wenz iD - Willemijne Schrijver by wenz iD

Genotyping and phenotyping of distant breast cancer metastases Willemijne A.M.E. Schrijver

Genotyping and phenotyping of distant breast cancer metastases © Willemijne A.M.E. Schrijver 2016

ISBN: 978-94-6233-487-8 Cover: wenz iD adjusted from a picture Willemijne Schrijver made Lay-out and Design: wenz iD | www.wenzid.nl Printed by: Gildeprint Drukkerijen, Enschede The work described in this thesis is part of a Dutch Cancer Society research program entitled ‘Genotyping and phenotyping of distant breast cancer metastases’ and was supported by Dutch Cancer Society grant UU 2011-5195, Philips Consumer Lifestyle and an A Sister’s Hope research grant. Financial support for the publication of this thesis was kindly provided by: ChipSoft, MRCHolland and the department of Pathology of the University Medical Center Utrecht

Genotyping and phenotyping of distant breast cancer metastases Genotyperen en fenotyperen van afstandsmetastasen van borstkanker (met een samenvatting in het Nederlands)

Proefschrift ter verkrijging van de graad van doctor aan de Universiteit Utrecht op gezag van de rector magnificus, Prof. dr. G.J. van der Zwaan, ingevolge het besluit van het college voor promoties in het openbaar te verdedigen op woensdag 14 december 2016 des middags te 2.30 uur door

Willemijne Alberta Maria Elisabeth Schrijver geboren op 14 april 1987 te Nijmegen

Promotoren:

Prof. dr. P.J. van Diest Prof. dr. E. van der Wall

Copromotor:

dr. C.B. Moelans

Voor mijn moeder

TABLE OF CONTENT List of abbreviations

Chapter 1

General introduction and thesis outline

Part one (Epi)genotyping of distant breast cancer metastases Chapter 2

Mutation profiling of key cancer genes in primary breast cancers and their distant metastases. Manuscript in preparation

Chapter 3

Progressive APOBEC3B mRNA expression in distant breast cancer metastases. Submitted

Chapter 4

Promoter hypermethylation profiling of distant breast cancer metastases. Breast Cancer Res Treat. 2015;151(1):41-55.

Chapter 5

Unravelling site-specific breast cancer metastasis: a microRNA expression profiling study. Oncotarget. 2016

123

Part two Phenotyping of distant breast cancer metastases Chapter 6

Receptor conversion in breast cancer metastases: a systematic review and meta-analysis. Manuscript in preparation

171

Chapter 7

Influence of decalcification procedures on immunohistochemistry and molecular pathology in breast cancer. Mod Pathol. 2016.

209

Chapter 8

Endocrine therapy imposes an evolutionary selection pressure on breast cancer metastases in malignant peritoneal and pleural effusions to lose steroid hormone receptors. Submitted

243

Chapter 9

Loss of FOXA1 expression in metastatic breast cancer drives acquired endocrine therapy resistance. Submitted

263

Chapter 10

Prospects of targeting the Gastrin Releasing Peptide Receptor, Chemokine C-X-C motif Receptor 4 and Somatostatin Receptor 2 for nuclear imaging and therapy in metastatic breast cancer. Submitted

299

Chapter 11

Summarizing discussion

319

Chapter 12

Future perspectives

331

Appendices Summary in Dutch/ Nederlandse samenvatting Acknowledgements/ Dankwoord Curriculum Vitae List of publications

338 344 350 351

LIST OF ABBREVIATIONS %AD percentage Added Dose APOBEC3B APOlipoprotein B Editing Catalytic subunit 3B AR Androgen Receptor ASCO/CAP American Society of Clinical Oncology/ College of American Pathologists ATG Translation Start Codon AUC Area Under the Curve Avg Average B&R Bloom & Richardson grading system BDT Big Dye Terminator Ber-EP4 EpCAM/Epithelial Specific Antigen Bp Base pairs C control tissue CT Cytosine to Thymine CB(M) Christensen’s Buffer (with Microwave) CBS Circular Binary Segmentation CDK12 Cyclin-Dependent Kinase 12 CEP17 Chromosome 17 centromere ChIP-seq Chromatin ImmunoPrecipitation massive parallel DNA sequencing CI Confidence Interval CMI Cumulative Methylation Index CNS Central Nervous System CpG Cytosine-phosphate-Guanine CT Computer Tomography CXCR4 chemokine C-X-C motif Receptor type 4 DMSO DiMethyl SulfOxide (c/ds)DNA (complementary/double-stranded) DeoxyriboNucleic Acid DNMT DNA Methyl-Transferase inhibitors DOT1L Disruptor Of Telomeric silencing-1 Like histone H3K79 methyltransferase DTC Disseminated Tumor Cell EDTA EthyleneDiamineTetraacetic Acid EMT Epithelial-Mesenchymal Transition EpCAM Epithelial Cell Adhesion Molecule ER(α) Estrogen Receptor (alpha) ERBB2 Erb-B2 Receptor Tyrosine Kinase 2 ESR1 EStrogen Receptor 1 EtOH Ethanol

F4 Formical-4 FAIRE Formaldehyde-Assisted Isolation of Regulatory Elements FATHMM Functional Analysis Through Hidden Markov Models FDG FluDeoxyGlucose FDR False Discovery Rate FF Fresh Frozen FFPE Formalin-Fixed Paraffin Embedded FOXA1 FOrkhead boX protein A1 GATA3 GATA Binding Protein 3 GI Gastro-Intestinal GO Gene Ontology GREB1 Growth Regulation by Estrogen in Breast cancer 1 GRIN2A Glutamate Receptor, Ionotropic, N-methyl d-aspartate 2A GRPR Gastrin Releasing Peptide Receptor H&E Haematoxylin and Eosin staining H3K27ac acetylation at the 27th lysine residue of the histone protein 3 H3K4me3 trimethylation of lysine 4 on histone H3 protein subunit HDAC Histone DeACetylase inhibitors HER2 Human Epidermal growth factor Receptor 2 hsa homo sapiens HSP90 Heat Shock Protein 90 HSPA8 Heat Shock 70 kDa Protein 8 IGFBP4 Insulin-like Growth Factor Binding Protein 4 IgG Immunoglobulin G IGO Integrated Genomics Operation IHC ImmunoHistoChemistry IP ImmunoPrecipitation IQR InterQuartile Range (F)ISH (Fluorescence) In Situ Hybridization KRT19 KeRaTin 19 LOH Loss Of Heterozygosity LR Logistic Regression M Metastasis MAF Mutant Allelic Fraction MAI Mitotic Activity Index MFS Metastasis-Free Survival miRs microRNAs MPS Massive Parallel Sequencing

MS-ARC Mass Spectrometry-Accurate Radioisotope Counting MS-MLPA Methylation-Specific Multiplex Ligand-dependent Probe Amplification MSK-IMPACT Memorial Sloan Kettering-Integrated Mutation Profiling of Actionable Cancer Targets MSP Methylation-Specific PCR MYC v-MYC avian myelocytomatosis Viral Oncogene Homolog N Normal tissue NaSCN Sodium thiocyanate P Primary tumor PCA Principal Component Analysis (qRT-)PCR (quantitative Real-Time) Polymerase Chain Reaction PET Positron Emission Tomography PIK3CA PhosphatidylInositol-4,5-bisphosphate 3-Kinase Catalytic subunit Alpha PR Progesterone Receptor PRISMA Preferred Reporting Items for Systematic Reviews and Meta-Analyses PSM Peptide Spectrum Match PTPRC Protein Tyrosine Phosphatase Receptor type C QC Quality Control QFI Quantitative Functional Index QM-MSP Quantitative Multiplex Methylation Specific PCR RB1 RetinoBlastoma 1 RIME Rapid Immunoprecipitation Mass spectrometry of Endogenous proteins RIN RNA Integrity Number (m)RNA (messenger) RiboNucleic Aid ROC Receiver Operating Characteristic RQ Relative Quantification RTU Ready-To-Use SE Standard Error Seq Sequencing SNV Single Nucleotide Variant SPECT Single Photon Emission Computed Tomography SSTR2 SomatoSTatin Receptor 2 TFF1 TreFoil Factor 1 TMA Tissue MicroArray TN(BC) Triple Negative (Breast Cancer) TP53 Tumor Protein p53 TRIM28 TRIpartite Motif-containing 28 TSS Transcription Start Site

Chapter 1

General introduction and thesis outline

CHAPTER 1

With 1.7 million new cases annually worldwide, breast cancer is the most common cancer in women and the second most common cancer overall (just behind lung cancer) 1. Since 1990, breast cancer related mortality rates have declined steadily, which is attributed to earlier detection through increased awareness and improved screening tools, and the introduction of new treatment modalities 2,3. Despite these advances, approximately 30% of breast cancer patients eventually develop recurrent or metastatic disease 4. Metastatic breast cancer remains essentially incurable and is responsible for the majority of breast cancer related deaths 5. Therefore, one of the most important challenges in breast cancer research is to elucidate underlying mechanisms by which cancer cells acquire their metastatic potential. The metastatic cascade is often represented as multiple successive, delineated steps that are followed by cells originating from the primary tumor. In succession, a cancer cell should pass through many steps: local invasion, intravasation, survival in the circulation, extravasation and colonization of the target organ 6. Actually, the fate of this process is determined by a complex series of interactions between metastatic cells and their organ microenvironment 7. These interactions were already postulated in 1889 by Paget. He examined autopsy reports of women diagnosed with metastatic breast cancer and noted that the organ distribution of metastases in these patients was non-random. Lung, liver, brain and bone seemed the most affected locations, while well circulated organs like the spleen and heart almost never harbored metastases. He therefore appointed the “seed and soil” analogy, where tumors are supposed to have a “seminal influence” on the metastatic microenvironment, and thereby act together with the distant organ to effect tumor metastases 8. The identity of these “seminal influences” remains elusive. Nowadays, breast tumors are often profoundly categorized into different molecular subtypes based on surrogate immunohistochemistry or gene expression profiling 9,10, because not only the preferred site of distant metastasis, but also the risk of relapse, prognosis and response to treatment are attributed to the type of breast cancer 11-13. This subtyping is classically based on ERα (estrogen receptor alpha), PR (progesterone receptor) and HER2 (human epidermal growth factor receptor 2) status. Especially ERα has been considered an important positive prognostic marker and a predictive marker of response to endocrine therapies 14. Although approximately 75% of breast cancers show ERα-positivity, their outcomes and responses to therapy vary extremely 15. Besides phenotypic markers, multiple genetic and epigenetic events have been identified that play a role in oncogenesis and progression, like mutations 16, copy number alterations17, methylation 18 and the presence or absence of certain microRNAs 19. These events lead to inactivation of tumor suppressor genes and cause alterations in proto-oncogenes (genes involved in the control of cell growth or division) that get activated and turn into oncogenes20. Only a small amount of these alterations are suspected to be driver events, playing an active role in carcinogenesis, but they are outnumbered by non-functional or passenger events 16. 14

General introduction

Which mechanisms underlie development of distant metastases remains a topic for debate. The two main but not necessarily mutually exclusive hypotheses are the linear and the parallel model of metastasis. According to the linear progression model, metastatic precursors leave the primary tumor at late stages of disease, after clonal evolution has given rise to a cell with metastatic ability. In this model, genetic modifications are thought to progressively accumulate in cancer cells of the primary tumor, whereby cells with advantageous mutations will survive and expand 21. Consequently, primary tumors and metastases are genetically closely related. The parallel progression model, on the other hand, assumes that cancer cells disseminate early during tumor progression at a stage when the primary lesion is small. Disseminated cells then evolve independently of the primary tumor to form metastases, resulting in genetic disparity between them 22. The direct comparison of primary breast tumors to paired macroscopic metastases is the most informative study design to evaluate these models of progression. Some studies have already compared immunohistochemical, epigenetic and molecular markers between primary breast tumors and lymph node metastases 23-25. However, there is still a relative scarcity of studies focusing on distant metastases, induced by difficulties in tissue acquisition and lack of faithful models of metastatic disease 26. Recent guidelines therefore agree that patients with accessible breast cancer metastases should be offered a biopsy or resection to confirm the diagnosis and to test for differences that would necessitate change of treatment strategy 27-29. A detailed (epi)genetic and phenotypic characterization and understanding of metastatic tumors is a prerequisite for the development of models that can predict which patient will develop metastatic disease, to personalize systemic treatments and to prevent such metastases from becoming clinically manifest. In this thesis we discuss the concordance and discordance of divergent genotypic and phenotypic characteristics in paired primary tumors and distant breast cancer metastases.

THESIS OUTLINE PART ONE (Epi)genotyping of distant breast cancer metastases In the first part of this thesis, we focus on the (epi)genetic similarities and differences between primary breast tumors and paired distant metastases. In Chapter 2, we compare the mutational profiles of actionable cancer targets between HER2-enriched and triple negative breast tumors and their matched distant metastases to brain and skin to find novel targetable drivers of metastatic progression. Next, we explore APOBEC3B (Apolipoprotein B editing catalytic subunit 3B) mRNA expression, a gain-of-function mutagenic enzyme, in primary breast cancers and paired metastases in Chapter 3, to gain more insight into 15

CHAPTER 1

the levels of expression during breast cancer progression. Subsequently, we concentrate on epigenetic markers. In Chapter 4, we perform promoter hypermethylation profiling of tumor suppressor genes by MS-MLPA (methylation specific multiplex ligation-dependent probe amplification). Furthermore, in Chapter 5, we study global miR (microRNA) expression patterns of multiple metastases per patient to pinpoint changes in miR expression during progression from the primary tumor to specific distant sites.

PART TWO Phenotyping of distant breast cancer metastases The second part describes divergent phenotypic biomarkers in primary breast tumors and paired distant metastases that have the potential to become markers for treatment response, therapeutic targets or prognostic or diagnostic factors. In Chapter 6, we review the frequency of receptor conversion (the change in hormone and HER2 receptor status) between primary breast tumors and distant breast cancer metastases, with special attention to metastasis location-specific differences. In Chapter 7, we study the influence of three routinely used decalcifying agents on assessment of hormone and HER2 receptor status and DNA/RNA quality in bone metastases, since bone is a frequent metastatic site among breast cancer patients. Thereafter, we evaluate the frequency of receptor conversion in primary breast carcinomas and their matched malignant peritoneal and/or pleural effusions in Chapter 8, aiming to optimize patient tailored therapy strategies that could improve quality of life and life expectancy in the later stages of the disease. In Chapter 9, we present a novel mechanism of acquired anti-estrogen resistance in metastatic breast cancer and highlight loss of FOXA1 and GATA3 expression as novel biomarkers for endocrine treatment resistance in ERÎą-positive metastatic breast cancer. Finally, in Chapter 10, we examine GRPR, CXCR4 and SSTR2 mRNA expression levels of primary tumors and paired metastases to evaluate whether nuclear imaging and therapy might be beneficial in metastatic breast cancer. The results of this thesis and future perspectives are discussed in Chapter 11.

General introduction

REFERENCES 1. Ferlay J, Soerjomataram I, Dikshit R, et al. Cancer incidence and mortality worldwide: Sources, methods and major patterns in GLOBOCAN 2012. Int J Cancer. 2015;136(5):E359-86. doi: 10.1002/ijc.29210 [doi]. 2. American Cancer Society. Cancer facts and figures 2005. Atlanta, GA: American Cancer Society, 2005:9–11. 3. DeSantis C, Siegel R, Bandi P, Jemal A. Breast cancer statistics, 2011. CA Cancer J Clin. 2011;61(6):409-418. doi: 10.3322/caac.20134 [doi]. 4. O’Shaughnessy J. Extending survival with chemotherapy in metastatic breast cancer. Oncologist. 2005;10 Suppl 3:20-29. doi: 10/suppl_3/20 [pii]. 5. Marsden CG, Wright MJ, Carrier L, Moroz K, Pochampally R, Rowan BG. “A novel in vivo model for the study of human breast cancer metastasis using primary breast tumor-initiating cells from patient biopsies”. BMC Cancer. 2012;12:10-2407-12-10. doi: 10.1186/1471-2407-12-10 [doi]. 6. Fidler IJ. The pathogenesis of cancer metastasis: The ‘seed and soil’ hypothesis revisited. Nat Rev Cancer. 2003;3(6):453-458. doi: 10.1038/nrc1098 [doi]. 7. Nguyen DX, Bos PD, Massague J. Metastasis: From dissemination to organ-specific colonization. Nat Rev Cancer. 2009;9(4):274-284. doi: 10.1038/nrc2622 [doi]. 8. Paget S. The distribution of secondary growths in cancer of the breast. 1889. Cancer Metastasis Rev. 1989;8(2):98101. 9. Perou CM, Sorlie T, Eisen MB, et al. Molecular portraits of human breast tumours. Nature. 2000;406(6797):747752. doi: 10.1038/35021093 [doi]. 10. Sorlie T, Perou CM, Tibshirani R, et al. Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications. Proc Natl Acad Sci U S A. 2001;98(19):10869-10874. doi: 10.1073/pnas.191367098 [doi]. 11. Kennecke H, Yerushalmi R, Woods R, et al. Metastatic behavior of breast cancer subtypes. J Clin Oncol. 2010;28(20):3271-3277. doi: 10.1200/JCO.2009.25.9820 [doi]. 12. Yersal O, Barutca S. Biological subtypes of breast cancer: Prognostic and therapeutic implications. World J Clin Oncol. 2014;5(3):412-424. doi: 10.5306/wjco.v5.i3.412 [doi]. 13. Savci-Heijink CD, Halfwerk H, Hooijer GK, Horlings HM, Wesseling J, van de Vijver MJ. Retrospective analysis of metastatic behaviour of breast cancer subtypes. Breast Cancer Res Treat. 2015;150(3):547-557. doi: 10.1007/ s10549-015-3352-0 [doi]. 14. Badve S, Nakshatri H. Oestrogen-receptor-positive breast cancer: Towards bridging histopathological and molecular classifications. J Clin Pathol. 2009;62(1):6-12. doi: 10.1136/ jcp.2008.059899 [doi]. 15. Blows FM, Driver KE, Schmidt MK, et al. Subtyping of breast cancer by immunohistochemistry to investigate a relationship between subtype and short and long term survival: A collaborative analysis of data for 10,159 cases from 12 studies. PLoS Med. 2010;7(5):e1000279. doi: 10.1371/journal.pmed.1000279 [doi].

16. Greenman C, Stephens P, Smith R, et al. Patterns of somatic mutation in human cancer genomes. Nature. 2007;446(7132):153-158. doi: nature05610 [pii]. 17. Li XC, Liu C, Huang T, Zhong Y. The occurrence of genetic alterations during the progression of breast carcinoma. Biomed Res Int. 2016;2016:5237827. doi: 10.1155/2016/5237827 [doi]. 18. Moarii M, Boeva V, Vert JP, Reyal F. Changes in correlation between promoter methylation and gene expression in cancer. BMC Genomics. 2015;16:873-015-1994-2. doi: 10.1186/s12864-015-1994-2 [doi]. 19. Kurozumi S, Yamaguchi Y, Kurosumi M, Ohira M, Matsumoto H, Horiguchi J. Recent trends in microRNA research into breast cancer with particular focus on the associations between microRNAs and intrinsic subtypes. J Hum Genet. 2016. doi: 10.1038/jhg.2016.89 [doi]. 20. Lodish H, Berk A, Zipursky SL, et al. Molecular Cell Biology. 4th Edition. New York: W. H. Freeman; 2000. Section 24.2, Proto-oncogenes and tumor-suppressor genes. Available from: http://www.ncbi.nlm.nih.gov/ books/NBK21662/. 21. Lorusso G, Ruegg C. New insights into the mechanisms of organ-specific breast cancer metastasis. Semin Cancer Biol. 2012;22(3):226-233. doi: 10.1016/j. semcancer.2012.03.007 [doi]. 22. Naxerova K, Jain RK. Using tumour phylogenetics to identify the roots of metastasis in humans. Nat Rev Clin Oncol. 2015;12(5):258-272. doi: 10.1038/ nrclinonc.2014.238 [doi]. 23. Barekati Z, Radpour R, Lu Q, et al. Methylation signature of lymph node metastases in breast cancer patients. BMC Cancer. 2012;12:244-2407-12-244. doi: 10.1186/14712407-12-244 [doi]. 24. Blighe K, Kenny L, Patel N, et al. Whole genome sequence analysis suggests intratumoral heterogeneity in dissemination of breast cancer to lymph nodes. PLoS One. 2014;9(12):e115346. doi: 10.1371/journal.pone.0115346 [doi]. 25. Zhao S, Xu L, Liu W, et al. Comparison of the expression of prognostic biomarkers between primary tumor and axillary lymph node metastases in breast cancer. Int J Clin Exp Pathol. 2015;8(5):5744-5748. 26. Almendro V, Kim HJ, Cheng YK, et al. Genetic and phenotypic diversity in breast tumor metastases. Cancer Res. 2014;74(5):1338-1348. doi: 10.1158/0008-5472.CAN13-2357-T [doi]. 27. Cardoso F, Costa A, Norton L, et al. 1st international consensus guidelines for advanced breast cancer (ABC 1). Breast. 2012;21(3):242-252. doi: 10.1016/j. breast.2012.03.003 [doi]. 28. Carlson RW, Allred DC, Anderson BO, et al. Metastatic breast cancer, version 1.2012: Featured updates to the NCCN guidelines. J Natl Compr Canc Netw. 2012;10(7):821-829. doi: 10/7/821 [pii]. 29. Van Poznak C, Somerfield MR, Bast RC, et al. Use of biomarkers to guide decisions on systemic therapy for women with metastatic breast cancer: American society of clinical oncology clinical practice guideline. J Clin Oncol. 2015;33(24):2695-2704. doi: 10.1200/ JCO.2015.61.1459 [doi].

Part one

(Epi)genotyping of distant breast cancer metastases

Chapter 2 Willemijne AME Schrijver, Charlotte K Y Ng, Kathleen A Burke, Salvatore Piscuoglio, Samuel H Berman, Jorge S Reis-Filho, Britta Weigelt, Paul J van Diest, Cathy B Moelans

Mutation profiling of key cancer genes in primary breast cancers and their distant metastases Manuscript in preparation

PART ONE | CHAPTER 2

ABSTRACT Although the repertoire of somatic genetic alterations of primary breast cancers have been extensively catalogued, the genetic differences between metastatic and primary tumors remain to be fully defined. Furthermore, detailed analyses of matched primary tumors and metastatic lesions have the potential to reveal novel targetable drivers of metastatic progression, in particular for patients with estrogen receptor (ER)-negative breast cancers, for whom treatment options are currently limited. DNA samples obtained from 17 formalin-fixed paraffin-embedded (FFPE) ER-negative/ HER2-positive (n=9) and ER-, progesterone receptor (PR)-, HER2-negative (i.e. triplenegative, n=8) primary breast cancers with paired brain or skin metastases and matched normal tissue were subjected to a hybridization capture-based massively parallel sequencing assay targeting all exons and selected introns of 341 of key cancer genes. Somatic single nucleotide variants, short insertions and deletions and copy number alterations were detected using state-of-the art bioinformatics algorithms. A large subset of the non-synonymous somatic mutations and gene copy number alterations (55%) identified were shared between the primary tumor and paired metastasis analyzed, however mutations restricted to either the primary tumor or metastases were also found. Although no metastasis location-specific alterations were detected, WNT-signaling was more often affected in the metastases of triple-negative breast cancers and cytokine mediated pathways in ER-negative/ HER2-positive metastases. In half of the patients, new potentially targetable alterations were found in the metastases relative to their matched primary tumors. The repertoire of somatic genetic alterations in metastatic breast cancer can differ from that of its primary tumor, even by the presence of driver and targetable somatic genetic alterations.

Mutational profiling of breast cancer metastases

INTRODUCTION Despite the advancements in our understanding of breast cancer and the introduction of novel therapies for breast cancer patients, metastatic breast cancer remains incurable. The survival of metastatic breast cancer patients has shown significant but only incremental increases over the last three decades 1,2. This is particularly a challenge for patients with estrogen-receptor (ER), progesterone receptor (PR) and HER2 negative (triple-negative (TN)) breast cancers (TNBCs) and HER2-positive breast cancers, who have been shown to have a higher incidence of brain metastases than ER-positive/ HER2-negative cancers 3. Massively parallel sequencing (MPS) analyses of breast cancer have revealed that breast cancers have a complex repertoire of somatic genetic alterations, and that these vary according to ER status. As a group, breast cancers have been shown to harbor few highly recurrently mutated genes, namely TP53, PIK3CA, GATA3, MAP3K1 and CDH1, and a large subset of genes mutated in <3% of breast cancers. In this subset of genes rarely mutated, targetable genetic alterations have been identified, including driver somatic genetic alterations affecting HER2, ESR1 and AKT1. Although some genotypic-phenotypic associations have been observed with the ER-status of breast cancers, there is no single highly recurrently mutated gene or highly recurrent mutation that defined ER-positive or ER-negative disease. In addition, recent studies have demonstrated that a large subset of breast cancers display intra-tumor genetic heterogeneity and are composed of mosaics of clones that, in addition to the truncal/ founder genetic events, harbor private mutations 4, and that the level and type of intra-tumor genetic heterogeneity may differ between ERpositive and ER-negative disease. Given the intra-tumor genetic heterogeneity described in breast cancers, it is plausible that in the progression from primary to metastatic disease, the metastatic process itself as well as the therapeutic interventions administered in the adjuvant setting may result in clonal selection. In this context, the analysis of primary vs. metastatic breast cancer would result in the identification of genes whose somatic genetic alterations are restricted to or enriched for in the metastatic lesions as compared to their respective primary tumors. Consistent with this notion, targeted massively parallel sequencing analysis of primary and metastatic breast cancers 5-7, as well as analysis of breast cancers, plasma DNA and metastatic lesions 8,9 have revealed that a varying proportion of somatic mutations, even those affecting known driver genes such as PIK3CA, SMAD4 and TP53, are either restricted to or enriched for in the metastatic lesion as compared to the primary tumor. These studies, however, either focused on hotspot mutations or on a limited number of breast cancers of different subtypes. In this study, we sought to define whether metastatic breast cancers would differ from their respective primary tumors in their repertoire of somatic genetic alterations. We have decided to focus on TNBCs and ER-/HER2-positive disease, and only brain and skin metastases to minimize the confounding variables stemming from the distinct biology of ER-positive vs. ER-negative breast cancer subtypes 10 and the notion that metastatic deposits to distinct anatomical sites may differ in terms of their biology and genetics. 23

PART ONE | CHAPTER 2

MATERIALS AND METHODS Patients Samples of primary and metastatic breast cancers were selected from an existing database entailing material from 400 patients from the Department of Pathology of the University Medical Center Utrecht in The Netherlands diagnosed between 1986 and 2014 11,12. For the purpose of this study, we selected TNBC and ER-negative/ PR-negative/ HER2-positive primary breast cancers from patients with brain or skin relapses for which sufficient primary and metastatic tumor tissue and normal tissues were available. Representative formalinfixed paraffin-embedded (FFPE) blocks from 17 primary-metastatic breast cancer pairs were retrieved, comprising eight triple-negative (4 with skin metastases, 4 with brain metastases) and nine ER-negative/HER2-positive breast cancers (4 with skin metastases, 5 with brain metastases). Normal tissues were obtained from blocks containing normal breast (n=4) or lymph node tissues (n=13). In the normal breast tissue, we have primarily retrieved DNA from stroma and adipose tissue, avoiding the terminal duct-lobular units. All cases were centrally reviewed by a pathologist with an interest and expertise in breast cancer (PJD), and graded following the Bloom and Richardson histological grading system (Table 1) 13. Immunohistochemical analysis of ER, PR and HER2 was initially performed for patient management, but centrally reviewed for the inclusion of cases in this study following the American Society of Clinical Oncology (ASCO)/ College of American Pathologists (CAP) guidelines 14,15. TNBCs were defined as tumors lacking ER (<1%), PR (<1%) and HER2 (0 or 1+ by immunohistochemistry, 2+ by immunohistochemistry, but lacking HER2 amplification by fluorescence (FISH) in situ hybridization, or non-amplified HER2 by FISH) expression and ER-negative/HER2-positive breast cancers were defined on the basis of lack of ER expression and HER2 overexpression (3+) and/or HER2 gene amplification. The use of leftover material requires no ethical approval according to Dutch legislation (“opt-out”). The use of anonymous or coded leftover material for scientific purposes is part of the standard treatment contract with patients; informed consent was sought and obtained where required by Dutch legislation 16. DNA extraction A 4µm-thick hematoxylin and eosin (H&E)-stained section from each FFPE tissue block was used to guide macro-dissection and to semi-quantitatively define the tumor cell content by a pathologist (PJD). Only tumors samples containing at least 10% neoplastic cells and normal tissue devoid of any tumor cells were cut (ten 10µm-thick sections) and subjected to macro-dissection using a scalpel, and areas with necrosis, dense lymphocytic infiltrates, and pre-invasive lesions were intentionally avoided. DNA extraction was performed using the QIAamp DNA FFPE tissue kit (Qiagen), and quantified using a Qubit fluorometer 24

Mutational profiling of breast cancer metastases

Table 1. Clinicopathological characteristics of the primary breast cancers included in this study Characteristics

Subgroup

All (n=17) n

Age at diagnosis (in years)

HER2 (n=9) %

TN (n=8) %

Difference %

Range Mean

31-61 48

37-61 50

31-55 46

Tumor cell estimate Range primary tumor (%) Mean

10-80 62

20-80 59

10-80 66

Histologic type

17 0

100 0

9 0

100 0

8 0

100 0

Histologic grade I (Bloom&Richardson) II III

0 7 10

0 41 59

0 4 5

0 44 56

0 2 6

0 25 75

MAI (Mitotic Activity Index)

Range Mean

4-68 31

4-56 21

12-68 45

Tumor diameter (cm)

Range Mean

1.3-6.5 3.1

1.4-6.5 3.2

1.3-4 2.9

Ductal Lobular

Lymph node status + Unknown

11 5 1

Time between primary tumor and metastasis~ (In days)

-15951919 605

-15951919 662

-15951496 468

-15951313 326

-26-1919 760

-21-1919 1083

- All Range Mean - Brain Range Mean - Skin Range Mean

65 29 6

8 1 0

89 11 0

3 4 1

0.016

ns 38 50 12

-26-1496 541

11-1496 645

-26-905 436

Location of metastases

Brain Skin

9 8

Tumor cell content metastasis (%)

Range Mean

30-80 66

Meta- or synchronous metastasis

M S

14 3

82 18

8 1

89 11

6 2

75 25

Treatment history*

naCT aCT aHT aTT unknown

2 10 2 3 4

12 59 12 18 24

0 5 1 3 4

0 56 11 33 44

2 5 1 0 4

25 63 12 0 50

53 47

5 4

56 44

30-80 59

4 4

50 50

50-80 71

ns ns

~ Negative numbers indicate that the metastasis was diagnosed before the primary tumor. * Numbers add up to more than 100% because of combination therapy. na: neoadjuvant; a: adjuvant CT: chemotherapy; HT: hormone therapy; TT: targeted therapy ns: not significant

PART ONE | CHAPTER 2

(Invitrogen, ThermoFisher Scientific). In addition, to determine DNA fragmentation, a size ladder PCR was performed using the specimen control size ladder kit (In Vivo Scribe) on a Veriti Thermal Cycler (Applied Biosystems) with a 35 cycle PCR reaction. All samples selected had â&#x2030;Ľ200 kB bands and were considered for MPS. Targeted capture massively parallel sequencing DNA extracted from the 51 FFPE tissue specimens (matched primary breast tumors, distant metastases to skin or brain, and normal tissue) from the seventeen patients included was subjected to targeted capture massively parallel sequencing on an Illumina HiSeq 2500 at the Memorial Sloan Kettering Cancer Center (MSKCC) Integrated Genomics Operation (IGO) using the Memorial Sloan Kettering-Integrated Mutation Profiling of Actionable Cancer Targets (MSK-IMPACT) assay, targeting all exons of 341 cancer genes harboring actionable mutations and non-coding regions of selected genes, following validated protocols as previously described 17,18. The complete list of genes targeted in this panel has been previously described 17,18. MSK-IMPACT sequencing data have been deposited into the NCBI Sequence Read Archive under the accession SRP078319. Analysis for the MSK-IMPACT sequencing data was performed as previously described 19. In brief, paired-ends were aligned to the human reference genome GRCh37 using the Burrows-Wheeler Aligner (BWA) 20. Local realignment, duplicate removal and quality score recalibration were performed using the Genome Analysis Toolkit (GATK) 21. Somatic mutations were defined using MuTect for single nucleotide variants (SNVs) 22, and Strelka and VarScan2 for small insertions and deletions (indels) 23,24. Variants with a mutant allelic fraction (MAF) of <1% and/ or variants supported by <5 reads and/ or covered by <10 reads at a given locus were disregarded 18. Additionally, variants whose MAF in the tumor was <5 times that in the matched normal sample were disregarded, as were variants, which were present at >5% minor allele frequency in dbSNP (Build 137) 18. Mpileup files generated from samtools mpileup (version 1.2 htslib 1.2.1) 25 for each sample were used to determine whether each mutation detected from the pipeline exists in the BAM file of the corresponding metastasis or primary tumor. Copy number alterations were identified using FACETS 26 as previously described 27. Copy number profiles of both primary tumor and metastases of patients 7 and 19 were excluded in the analyses performed due to insufficient sequencing depth, despite repeated rounds of sequencing of these samples. Identification of pathogenic mutations A combination of mutation function predictors, including MutationTaster 28, CHASM (breast) 29 and FATHMM 30, coupled with annotation of the genes mutated in the cancer gene lists described by Kandoth et al. (127 significantly mutated genes) 31, the Cancer Gene Census 32 and Lawrence et al. (Cancer5000-S gene set) 33 were employed to define the 26

Mutational profiling of breast cancer metastases

potential functional effect of each non-synonymous SNV. Missense SNVs defined as “nondeleterious”/”passenger” by both MutationTaster 28 and CHASM (breast) 29, a combination of predictors shown to have a high negative predictive value 34, were considered nonpathogenic alterations. Missense SNVs that were considered to be potentially nonpathogenic were further analyzed using CHASM (breast) and FATHMM 30; those alterations not considered to be “driver” and/ or “cancer” alterations, respectively, were classified as non-pathogenic alterations. The SNVs considered “driver” and/ or “cancer” alterations using this combination of mutation function predictors were defined as pathogenic if they affected genes in one of the cancer gene lists, or were otherwise classified as potentially pathogenic. Frameshift, splice-site and truncating mutations that were haploinsufficient or had LOH of the wild-type allele and affected genes included in at least one of the three cancer gene lists described above were considered pathogenic. The remaining frameshift, splice-site and truncating mutations were considered not to be pathogenic. For in-frame indels, those not defined as “deleterious” by both MutationTaster 28 and PROVEAN 35 were considered non-pathogenic. The remaining in-frame mutations were categorized as pathogenic based on their haploinsufficiency status, LOH status of the wild-type allele and whether they affected a gene included in the cancer gene lists as described above for the frameshift, splice-site and truncating mutations. Validation of TP53 mutations using Sanger sequencing The TP53 mutations identified by targeted capture sequencing were validated using Sanger sequencing in fourteen primary tumor – metastasis pairs for which sufficient DNA was available (for primer sequences see Supplementary Table S1). 70ng DNA of each matched primary breast tumor, metastasis and normal tissue sample were used for PCR analyses using standard protocols. PCR products were purified using Exonuclease I (E.coli; New England Biolabs) and shrimp alkaline phosphatase (Takara-ClonTech), followed by Big Dye Terminator (BDT) PCR with BDT buffer and ready reaction mix (ThermoFisher Scientific). Sequencing was performed on an ABI-3730 capillary sequencer (Applied Biosystems), and sequences were analyzed with Sequence Analysis Software 6 (Applied Biosystems) and mutations assessed with Mutation Surveyor Software (Soft Genetics). Statistical analyses Non-paired analyses between clinicopathological characteristics of patient groups were computed using the Mann-Whitney U test. Paired analyses between primary tumors and metastases were performed using the Wilcoxon signed-rank test. Kaplan Meier survival curves were computed for metastasis-free survival (MFS), defined as the time in days between the diagnosis of the primary breast tumor and the diagnosis of the metastasis. Most significantly up- or downregulated genes in the described pathways were found using ToppGene with Bonferroni-Holm correction for multiple comparisons (http://toppgene. 27

PART ONE | CHAPTER 2

cchmc.org). P-values <0.05 were considered statistically significant. All statistical calculations were using IBM SPSS Statistics 21 and visualized with GraphPad Prism 6 and R software (version 3.2.5).

RESULTS Sequencing statistics A mean target coverage of 301x (range 40-841x) was obtained across the tested samples with a mean average base coverage of 237x (range 20-866x) across the target panel (Supplementary Table S2). Total read and target coverage were significantly different between primary tumors, metastases and normal tissue (mean coverage primary tumors: 283x versus 451x for metastases and 170x for normal tissue). However, no differences were observed between molecular subtypes and metastasis locations, our target groups for statistical comparison (Supplementary Table S3). Across samples, a median of 90% of target bases were spanned by at least 50 sequence reads (Supplementary Figure S1A-C). High degree of similarity amongst somatic variants in matched primary tumors and distant metastases 195 non-synonymous SNVs were identified amongst the 34 tumor specimens, comprising seventeen primary and seventeen metastatic tumors (data available from the authors). 21 SNVs were found in the primary only (mean 1.23 per patient, range 0-5), 66 in the metastasis only (mean 3.88 per patient, range 0-22) and 54 were concordant between primary and matched metastasis (mean 3.18 per patient, range 1-22; Figure 1A). Thus, 55% of total variants were shared between primary tumor and metastasis, with a range of 15-100% per patient. Primary only or metastasis only variants could not be explained by copy number amplifications or homozygous deletions. However, a fraction of the low frequency variants may have been falsely called negative when a difference in tumor percentage was seen in combination with a gain or a loss (Supplementary Table S4). The number of concordant, primary only or metastasis only variants was similar in HER2enriched and triple negative tumors, and in brain and skin metastases (Figure 1B). When the MAF of the shared variants were corrected for tumor percentages, a significantly higher variant frequency was seen in metastases versus their paired primary tumors (p=0.001; Figure 1C). Furthermore, mutations found at higher allele frequency in the primary tumor were also more likely to be present in the corresponding distant metastasis (p<0.001; Figure 1D). Of the 195 variants found, 93% (n=182) were SNVs, 4% (n=8) deletions and 3% (n=5) insertions (Supplementary Figure S1D). In this cohort, 34% of the variants were detected as cancer drivers by FATHMM (Supplementary Figure S1E). These cancer drivers were more often shared between primary and metastasis, but this was not significant (p=0.099; 28

Mutational profiling of breast cancer metastases

Figure 1 Primary tumor Metastasis

HER2-enriched Brain

55%

11%

107/195

22/195

34% 66/195

Shared variants **

Metastasis only Shared Primary only

#1 #9 #19 #48 #52

#7 #8 #10 #188

#4 #5 #6 #14

Patients

#12 #13 #15 #16

Variant frequency in primary tumor ***

100

Variant fr eq uency

MAF corr ected for tum or%

Skin

1.1 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0

Brain

Total number of variants: 195

Triple negative

Skin

100

% of variants

Primary

Metastasis

80 60 40 20 0

Shared between primary and metastasis

Not shared

Figure 1. Somatic variants found with next generation sequencing in 341 cancer associates genes in 17 ERÎą-negative primary breast tumors and matched brain or skin metastases. A. Percentage of primary only, metastasis only and shared mutations. B. Frequency of somatic variants per patient, molecular subtype and metastasis location. C. Variant frequency in primary tumors that share variants with the metastasis or not. D. Comparison of tumor%-corrected variant frequency of shared variants in primary tumors versus paired metastases.

Figure 2A). However, detected variants of TP53 and PIK3CA, two well established cancer drivers in breast cancer and cancer in general, were shared in primaries and paired metastases in 13/14 patients (92.9%) and 4/5 patients (80%), respectively. PIK3CA mutations were more often seen in HER2-enriched primary tumors (p=0.162) whereas TP53 mutations were more frequent in triple negative primary tumors (p=0.036; Figure 2B). All NGSdetected TP53 variants in the 22 samples of fourteen patients were confirmed by Sanger sequencing (Supplementary Figure S2A; Supplementary Table S1). Copy number aberrations are largely shared between matched primary tumors and distant metastases All 341 interrogated genes showed copy number variations to some extent (data available from the authors). Sixty-eight (20% of interrogated genes) genes were amplified in one or more samples, 5 genes showed homozygous deletions and two genes showed both amplifications and homozygous deletions (NF2 and CDKN1B). Copy number patterns of primary tumors and metastases showed many similarities (Figure 3A-D). Thirteen amplified genes and one homozygously deleted gene were found in the primary tumor only 29

PART ONE | CHAPTER 2

(mean 0.82 aberrations per patient, range 0-4), 36 amplified genes and three homozygously deleted genes in the metastasis only (mean 2.29 aberrations per patient, range 0-15), and 60 amplified genes plus four homozygously deleted genes were concordant between primary tumors and matched metastases (mean 3.76 aberrations per patient, range 1-12). Comparable to the somatic variants, 55% (64/117) of amplifications and homozygous deletions were shared between primary tumors and metastases (Figure 3E). Amplifications and homozygous deletions were not significantly differentially distributed between primary tumors and metastases. When all copy number alterations, including gains and hemizygous losses, were compared between the primary tumors and metastases, eighteen genes showed significant differences: CCND1, FOXA1, GNAS, SF3B1, TRAF7 and PIK3CD were significantly more often gained or less often lost in the metastases and CDK8, FLT1, FLT3, NKX3-1, RASA1, DDR2, INPP4A, JAK1, TMEM127, VTCN1, WT1 and PAK7 were significantly more often lost or less often gained in the metastases (Supplementary Figure S2B). 7/9 samples with HER2 overexpression or amplification previously determined during routine diagnostic workup could be validated by NGS-based copy number analysis (samples of patients #7 and #19 were not validated due to insufficient DNA quality for proper copy number calling) and amplification in the primary tumor was always retained in the metastasis. HER2 amplification always co-occurred with CDK12 amplification (Supplementary Figure S2C). High degree of genetic disparity between patients Only one somatic variant was seen in more than one patient, namely PIK3CA p.His1047Arg, a known hotspot mutation with neutral impact (COSM775/COSM29325), in patient #16 and #48. Overall, 20 genes were recurrently affected (Supplementary Table S5), with the highest number of variants in TP53 (14 patients), PIK3CA (5 patients), RB1 (4 patients),

Figure 2

% of v ariants

Driver or passenger p=0,099

60 40 20 0

Driver

Passenger

Shared between primary and metastasis Not shared between primary and metastasis

B %of patients with a mutation

150

Driver mutations p=0,036

HER2-enriched Triple negative

100

p=0,162 50

PIK3CA

TP53

Figure 2. Driver and passenger mutations A. Driver and passenger mutations as assessed with FATHMM that are shared between primary tumors and metastases. B. Frequency of PIK3CA and TP53 mutations per molecular subtype of the primary tumor.

Mutational profiling of breast cancer metastases Figure 3 A

Percentage of samples

60 40 20 0 20 40 60 80 1

Chromosomes

20 22

Percentage of samples

40 20 0 20 40 60 80 2

Chromosomes

gains/amplifications

losses/homozygous deletions

0 20 40

20 22

60 80 1

Chromosomes

20 22

Gains/amplifications and losses/homozygous deletions in all metastasis samples

100

Amplifications and homozygous deletions in all metastasis samples

100

Gains/amplifications and losses/homozygous deletions in all primary tumor samples

100

Percentage of samples

Amplifications and homozygous deletions in all primary tumor samples

100

80 60 40 20 0 20 40 60 80

100

Chromosomes

20 22

E Primary tumor

12% 14/117

55% 64/117

Metastasis

33% 39/117

Total amplifications and homozygous deletions: 117

Figure 3. Copy number aberrations found with next generation sequencing in 341 cancer associated genes in 15 ERÎą-negative primary breast tumors and matched metastases to brain or skin. A. Amplifications and homozygous deletions in all primary tumor samples B. Gains, amplifications, losses and homozygous deletions in all primary tumor samples C. Amplifications and homozygous deletions in all metastasis samples D. Gains, amplifications, losses and homozygous deletions in all metastasis samples E. Percentage of primary only, metastasis only and shared amplifications and homozygous deletions.

DOT1L and GRIN2A (both 3 patients), respectively. Recurrently affected genes were not more commonly shared between the primary tumor and the metastasis than non-recurrently mutated genes (p=0.957). Supplementary Figure S3 depicts the mean allele frequency of each variant per individual patient, corrected for tumor percentage. A great variety was observed between number of shared, primary only and metastasis only variants. Recurrent amplifications (observed in more than one patient) were observed for 20 genes (Supplementary Table S6), with the most prevalent aberrations in MYC (8 patients), CDK12 and HER2 (7 patients), RECQL4 (4 patients) and ETV6 and FGF3 (both 3 patients). For homozygous deletions, only RB1 was affected more than once (two patients). Pathways Copy number aberrations and somatic mutations were combined to look for similarities between patients in the different analysis groups (HER2-enriched, triple negative, brain or skin metastases) for the three main driving pathways in our cohort: TP53, PIK3CA and HER2 (Figure 4). Large inter-patient differences were observed, but only few intra-patient variations. For all three pathways, HER2-driven primary tumors showed fewer aberrations than triple negative tumors (HER2-pathway: p=0.011; TP53-pathway: p=0.016; PIK3CA31

PART ONE | CHAPTER 2

pathway: p=0.009) and skin metastases showed fewer aberrations than brain metastases (HER2-pathway: p=0.001; TP53-pathway: p=0.002; PIK3CA-pathway: p=0.027). Next, we tried to assess which genetic alterations were important during the metastatic process of the two molecular subtypes and in the two different metastasis locations. To do so, we analyzed somatic variants and copy number aberrations that became more apparent in the metastases (metastasis only variants, significant increase in MAF (>15%) or significant gain in copy number) or less apparent in the metastases (primary only variants, significant decrease in MAF (>15%) or significant loss in copy number). The resulting genes are shown

Figure 4 A TP53-pathway

C 48

188

Location metastasis

Triple negative

Patient number

MDM4

188

HER2-enriched

Location metastasis

Triple negative

Patient number

PIK3CA

Chr.3

CDKN1A

PIK3CB

CDKN2A

FGFR3

PTEN

Chr.12 Chr.15

Chr.4 Chr.5

Chr.17

Chr.8 Chr.7 Chr.6

Chr.X Chr.22Chr.20

CDKN1A EGFR FGFR1 FGF3

Chr.11 Chr.12

ERBB3

amplification

deletion

loss

gain

missense

Triple negative

Brain

MYC CCND1 HRAS

KRAS MAP2K1 ERBB2 TP53 SRC AKT2 ARAF amplification

EP300

ABL1

FGF4 CDKN1B

MDM2

RICTOR

RPS6KB2

EGFR

ERBB3

PIK3R1

CHEK2

CDKN1A

CDKN1B

PIK3R2

Metastasis

PDGFRB

CCNE1

188

GSK3B

PDGFRA

AKT2

ERBB4

FGFR4

PIK3C3

BCL2L11

KIT

TP53

Primary

PIK3C2G

PDPK1

JUN

MDM2

CREBBP

Skin

ATM

AKT1

MTOR

IRS1 GSK3B

AKT3

ERBB4

PIK3R1

HER2-enriched 1

Sample type

Brain

Metastasis

PTEN Chr.2

Chr.3

Chr.11 Chr.10 Chr.9 Chr.7 Chr.6 Chr.5 Chr.12 Chr.14 Chr.16 Chr.18 Chr.17

PIK3CD PIK3R3

PIK3CG

Primary

PIK3CA TP63

Skin

ATR PIK3CB

MTOR

Chr.1

Metastasis

FGFR2

Primary

PIK3CD

AKT3

Skin

PIK3R3

Chr.19

ERBB2-pathway Molecular subtype

Sample type

Brain

AKT3 Chr.1

HER2-enriched

Chr.2

Chr.9 Chr.8 Chr.7 Chr.6 Chr.3

Chr.11

Sample type

Chr.22

PIK3CA-pathway Molecular subtype

Patient number

Chr.1

Molecular subtype Location metastasis

deletion

loss

gain

missense

truncation

MDM2

truncation

Chr.14 Chr.13

PTPN11 IRS2 AKT1

Chr.16

PDPK1

Chr.17

TSC2 ERBB2

Chr.19

AKT2 PIK3R2

20 15 10 5 0

Del Loss Gain Amp Miss Trunc

% of aberrations

40 30 20 10 0

HER2

Primary

Skin

Brain

Metastasis

15 10 5 0

Del Loss Gain Amp Miss Trunc

40 30 20 10

TN primary HER2 primary Skin metastasis Brain metastasis

TN brain

HER2 skin HER2 brain

40 30 20 10 HER2

Primary

Skin

Brain

Metastasis

Primary Metastasis

10 5

20 15 10 5

Cumulative aberrations per subgroup ***

***

40 30 20 10 0 TN brain

10 5 0

Del Loss Gain Amp Miss Trunc

HER2 skin HER2 brain

20 15 10 5 0

Del Loss Gain Amp Miss Trunc

TN primary HER2 primary Skin metastasis Brain metastasis

40 30 20 10 TN

HER2

Primary

Skin

Del Loss Gain Amp Miss Trunc

Cumulative aberrations per subgroup *

Primary Metastasis

Cumulative aberrations

0 TN skin

ERBB2: HER2 brain Primary Metastasis

Del Loss Gain Amp Miss Trunc

Primary Metastasis

25 20

ERBB2: HER2 skin

Del Loss Gain Amp Miss Trunc

% of aberrations

Del Loss Gain Amp Miss Trunc

0 TN skin

Cumulative aberrations

Primary Metastasis

ERBB2: TN brain Primary Metastasis

25 20

PIK3CA: HER2 brain

Cumulative aberrations per subgroup

Cumulative aberrations

Del Loss Gain Amp Miss Trunc

ERBB2: TN skin 30

Primary Metastasis

25 20

% of aberrations

Primary Metastasis

% of aberrations

PIK3CA: HER2 skin

% of aberrations

Primary Metastasis

TP53: HER2 brain

TP53: HER2 skin 30

Del Loss Gain Amp Miss Trunc

Primary Metastasis

25 20

% of aberrations

5 0

Del Loss Gain Amp Miss Trunc

% of aberrations

truncation

% of aberrations

missense

PIK3CA: TN brain

% of aberrations

Primary Metastasis

25 20

loss

gain

% of aberrations

deletion

PIK3CA: TN skin

% of aberrations

Primary Metastasis

25 20

% of aberrations

TP53: TN brain

% of aberrations

TP53: TN skin

% of aberrations

amplification

Brain

50 40 30

TN primary HER2 primary Skin metastasis Brain metastasis

20 10 0 TN skin

TN brain

HER2 skin HER2 brain

Metastasis

Figure 4. Copy number aberrations combined with somatic mutations in the three most affected pathways: TP53, PIK3CA and ERBB2. A. Genes of the TP53-pathway that were affected in our cohort. B. Percentage of genetic alterations within the TP53-pathway per molecular subtype and metastasis location. C. Genes of the PIK3CA-pathway that were affected in our cohort. D. Percentage of genetic alterations within the PIK3CA-pathway per molecular subtype and metastasis location. E. Genes of the ERBB2-pathway that were affected in our cohort. F. Percentage of genetic alterations within the ERBB2-pathway per molecular subtype and metastasis location.

Mutational profiling of breast cancer metastases

in Supplementary Table S7. Next to general cancer pathways, genes involved in NOTCHsignaling were less often and genes involved in WNT-signaling were more often affected in the metastases of triple negative tumors. In metastases of HER2-enriched tumors, cytokine mediated pathways were more often affected (IL-2, -4 and -7 and IFN gamma signaling). Tumors that disseminated to skin or brain showed no specific, but more divergent pathways that could be important in metastasis specificity. Low association between clinicopathological characteristics and somatic mutations and copy number aberrations Although only three synchronous metastases were included (#14, #15 and #188), no differences in percentage of concordant variants could be detected between primaries and synchronous metastases versus primaries and matched metachronous metastases. Also, the time between primary tumor and metastasis was not correlated to degree of concordance, although there was a visual trend towards higher discordance in tumor pairs with longer intervals in between (although the groups were too small to draw definite conclusions) (Supplemental Figure S4A). In this cohort, amount of discordance in genetic aberrations between primary and metastasis was not correlated to a shorter overall survival time. When treatment history was taken into account, a trend was seen towards fewer concordant mutations in primaries versus metastases (p=0.089) and more metastasis only mutations (p=0.075) in trastuzumab treated patients. For chemo- and hormonal therapy this trend was not observed (Supplemental Figure S4B). These results should however be interpreted with caution due to the low sample sizes. New targetable genetic alterations are apparent in metastases relative to paired primary tumors Finally, we evaluated the concordance of currently targetable mutations and amplifications between primary tumors and metastases. Genes were considered currently targetable when they were listed on the FDA website (http://www.fda.gov) as targets of approved drugs for breast cancer or other cancer types. Twelve currently targetable genes were present in our cohort. After exclusion of HER2 amplifications, 14/17 patients could have been stratified for targeted therapies and 13/23 of the aberrations remained expressed in the metastases. Furthermore, in 9/17 (53%) metastases new targetable options relative to the primary tumors were found (Supplemental Figure S4C; Supplementary Table S8).

PART ONE | CHAPTER 2

DISCUSSION To our knowledge, this work represents the first study performing next generation sequencing of a large set of actionable cancer targets in paired tumor tissue of patients with triple negative or HER2-driven primary breast tumors and matched distant metastases to skin or brain. In our search for novel targeted therapeutic options, we found extensive inter-patient differences and intra-patient similarities in somatic mutations and copy number alterations. We could validate the presence of well-known cancer drivers in the primary tumors that remained expressed in the metastases. Furthermore, distinct sets of genes were identified as being positively or negatively selected for during tumor progression. In a relatively large panel of 341 cancer-related genes, we detected a high degree of similarity amongst somatic variants in matched ERÎą-negative primary tumors and distant brain or skin metastases. 55% of variants were shared, 11% were only seen in the primary tumor and 28% only in the metastases. Likewise, Roy-Chowdhuri et al. found almost the same percentage of metastasis-only mutations (22%) and even a higher degree of concordant variants (77%) in 46 cancer-associated genes in 61 tumor pairs 36. cDNA microarray analyses also showed conserved gene expression profiles of primary breast tumors and their distant metastases, even after a long interval. In line with our results, metastasis site-specific differences were not found 37. In our cohort, the identified somatic variants were mainly found in known cancer driver genes like TP53, PIK3CA and RB1 36,38-40, but a large number of these and other variants were not previously associated with cancer (not found in COSMIC) and further studies are needed to explore whether these could constitute new cancer-associated variants. Although PIK3CA mutations were more apparent in HER2-driven and TP53 mutations in triple negative primary tumors, enrichment of their interacting pathways was not significantly different. HER2-driven tumors did show fewer genetic aberrations than triple negative tumors, as shown before 36. Two other frequently mutated genes were DOT1L and GRIN2A. DOT1L (Disruptor of telomeric silencing-1 Like Histone H3K79 Methyltransferase) has a role in chromatin organization and transcriptional misregulation in cancer. In MCF7, overexpression has been shown to increase cell migration and inhibition can downregulate proliferation, self-renewal and metastatic potential 41,42. GRIN2A (Glutamate Receptor, Ionotropic, N-Methyl D-Aspartate 2A) is involved in signal transmission, tumor growth and metastasis via G-protein-coupled receptors 43. The high mutation frequency of DOT1L and GRIN2A can therefore be explained by the enrichment for metastatic disease in our cohort, but further research should focus on the eventual prognostic or predictive potential of these genes. Besides the similarity in somatic variants between primary tumors and matched metastases, copy numbers profiles were largely preserved as well, particularly amplifications and homozygous deletions. This is in agreement with a previous study from our research group 34

Mutational profiling of breast cancer metastases

using MLPA 44. MYC, CDK12, HER2 and RECQL4 were most often amplified, and RB1 most frequently deleted. The functional importance of the co-amplification of CDK12 and HER2 that we observed in every HER2-enriched sample remains to be elucidated. Recently, Mertins et al. showed that the kinase CDK12 is a positive transcriptional regulator of homologous recombination repair genes in breast cancer and is found to be highly active in the majority of HER2-positive tumors 45. Also in ovarian carcinoma, CDK12 is suggested to be the key driver within the HER2 amplicon 46. In gastric tumors however, CDK12-HER2 fusion transcripts were described not to disrupt HER2 protein translation 47. MYC amplifications were more apparent in triple negative tumors, as also demonstrated by others 48,49. HoriuchiÂ et al. showed that aggressive breast tumors with elevated MYC levels are uniquely sensitive to CDK-inhibitors, making it a possible therapeutic target in triple negative tumors 48. We showed that MYC amplifications and other targetable mutations and amplifications in AKT2, BRCA1/2, BTK, CTLA4, ERBB2, FGFR, PIK3CA, PTCH1 and ROS1 were largely shared between primaries and metastases. However, large inter-patient differences were seen in the presence of these markers. This large genetic diversity was seen before, where no two tumors shared an identical genetic profile, emphasizing the importance of personalized medicine in daily clinical practice 38. Interestingly, 53% of patients harbored new potential actionable genetic aberrations in the metastases relative to the primary tumors, suggesting that mutational profiling of metastatic samples would be of added value. Unfortunately, most of these mutations and amplifications have not been shown to predict response to therapy in breast cancer yet, but continuous research is being performed in this field to assess the efficacy of these targeted therapies in different cancer types. Metastases often demonstrated an increase in variant frequency, copy number and novel mutations compared to the primaries, suggesting clonal evolution. However, primary-only mutations were seen as well, indicating clonal divergence within the primary tumor, either with other clones being selected for metastatic dissemination or with the metastasis branching off at an earlier time point. Importantly, intra-tumor heterogeneity should not be overlooked; differences between the primary tumor and the paired metastasis could be derived from a subclone not included due to sampling 5,50-52. On the other hand, the large concordance in genetic profiles and the resemblance in percentage of primary only/shared/ metastasis only mutations and copy number aberrations (12%, 55% and 33%, respectively) could hardly be caused solely by heterogeneity. Also, there was a trend towards conservation of important driver mutations (or evolving convergently), while probable passenger mutations diverged between primary tumors and metastases, which implies that metastases still need these driver mutations for dissemination to and/or maintenance at distant sites. This is emphasized by the finding that also in other studies, TP53 and PIK3CA were also the most frequently occurring mutations in metastases 5,36. Limitations of this study include the differences in total read and target coverage observed between primary tumors, metastases and normal tissue. The high concordance and increase 35

PART ONE | CHAPTER 2

in variant frequency in the metastases could not be explained solely by this observation. Secondly, the normal tissue included was mainly originating from axillary lymph nodes, so we did not include unaffected tissue of the specific locations of interest (breast, brain, skin). Therefore, we cannot exclude tissue-specific background mutational burden. Here, we have provided an insight into the presence of actionable cancer targets in estrogen receptor negative breast cancer metastases. In the battle against cancer we search for the common denominator; some kind of predictable pattern for therapeutic targeting. However, we showed that no patient and no tumor looks alike and the key in targeted treatment must be a tailored approach per tumor. As we found approximately half of the patients to harbor new potential actionable genetic aberrations in the metastases relative to the primary tumors, NGS of breast cancer metastases may be an important additional diagnostic approach in personalized treatment of metastatic breast cancer patients.

Acknowledgements This study is supported by Dutch Cancer Society grant UU 2011-5195 and Philips Consumer Lifestyle. JSR-F is funded in part by Breast Cancer Research Foundation (BCRF). Research reported in this publication was supported in part by a Cancer Center Support Grant of the National Institutes of Health/National Cancer Institute (Grant No. P30CA008748). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Mutational profiling of breast cancer metastases

REFERENCES 1. Cardoso F, Beishon M, Cardoso MJ, Corneliussen-James D, Gralow J, Mertz S, Mayer M, Paonessa D, Papadopoulos E, Peccatori FA, Sabelko K, Sakurai N, Spence D. Global status of advanced/metastatic breast cancer. 2005-2015 decade report. . 2016. 2. Dawood S, Broglio K, Gonzalez-Angulo AM, Buzdar AU, Hortobagyi GN, Giordano SH. Trends in survival over the past two decades among white and black patients with newly diagnosed stage IV breast cancer. J Clin Oncol. 2008;26(30):4891-4898. doi: 10.1200/JCO.2007.14.1168 [doi]. 3. Kennecke H, Yerushalmi R, Woods R, et al. Metastatic behavior of breast cancer subtypes. J Clin Oncol. 2010;28(20):3271-3277. doi: 10.1200/JCO.2009.25.9820 [doi]. 4. Ng CK, Martelotto LG, Gauthier A, et al. Intra-tumor genetic heterogeneity and alternative driver genetic alterations in breast cancers with heterogeneous HER2 gene amplification. Genome Biol. 2015;16:107-015-06576. doi: 10.1186/s13059-015-0657-6 [doi]. 5. Goswami RS, Patel KP, Singh RR, et al. Hotspot mutation panel testing reveals clonal evolution in a study of 265 paired primary and metastatic tumors. Clin Cancer Res. 2015;21(11):2644-2651. doi: 10.1158/1078-0432.CCR-142391 [doi]. 6. Pestrin M, Salvianti F, Galardi F, et al. Heterogeneity of PIK3CA mutational status at the single cell level in circulating tumor cells from metastatic breast cancer patients. Mol Oncol. 2015;9(4):749-757. doi: 10.1016/j. molonc.2014.12.001 [doi]. 7. Yates LR, Gerstung M, Knappskog S, et al. Subclonal diversification of primary breast cancer revealed by multiregion sequencing. Nat Med. 2015;21(7):751-759. doi: 10.1038/nm.3886 [doi]. 8. Murtaza M, Dawson SJ, Tsui DW, et al. Non-invasive analysis of acquired resistance to cancer therapy by sequencing of plasma DNA. Nature. 2013;497(7447):108112. doi: 10.1038/nature12065 [doi]. 9. Murtaza M, Dawson SJ, Pogrebniak K, et al. Multifocal clonal evolution characterized using circulating tumour DNA in a case of metastatic breast cancer. Nat Commun. 2015;6:8760. doi: 10.1038/ncomms9760 [doi]. 10. Ng CK, Schultheis AM, Bidard FC, Weigelt B, Reis-Filho JS. Breast cancer genomics from microarrays to massively parallel sequencing: Paradigms and new insights. J Natl Cancer Inst. 2015;107(5):10.1093/jnci/djv015. doi: 10.1093/jnci/djv015 [doi]. 11. Hoefnagel LDC, van de Vijver MJ, van Slooten H, et al. Receptor conversion in distant breast cancer metastases. Breast Cancer Research. 2010;12(5):R75. doi: 10.1186/ bcr2645. 12. Hoefnagel LDC, Moelans CB, Meijer SL, et al. Prognostic value of estrogen receptor alpha and progesterone receptor conversion in distant breast cancer metastases. Cancer. 2012;118(20):4929-4935. doi: 10.1002/cncr.27518. 13. Meyer JS, Alvarez C, Milikowski C, et al. Breast carcinoma malignancy grading by bloom-richardson system vs proliferation index: Reproducibility of grade and

advantages of proliferation index. Mod Pathol. 2005;18(8):1067-1078. doi: 3800388 [pii]. 14. Hammond ME, Hayes DF, Wolff AC, Mangu PB, Temin S. American society of clinical oncology/college of american pathologists guideline recommendations for immunohistochemical testing of estrogen and progesterone receptors in breast cancer. J Oncol Pract. 2010;6(4):195-197. doi: 10.1200/JOP.777003 [doi]. 15. Wolff AC, Hammond ME, Hicks DG, et al. Recommendations for human epidermal growth factor receptor 2 testing in breast cancer: American society of clinical oncology/college of american pathologists clinical practice guideline update. J Clin Oncol. 2013;31(31):39974013. doi: 10.1200/JCO.2013.50.9984 [doi]. 16. van Diest PJ. No consent should be needed for using leftover body material for scientific purposes. for. BMJ. 2002;325(7365):648-651. 17. Cheng DT, Mitchell TN, Zehir A, et al. Memorial sloan kettering-integrated mutation profiling of actionable cancer targets (MSK-IMPACT): A hybridization capturebased next-generation sequencing clinical assay for solid tumor molecular oncology. J Mol Diagn. 2015;17(3):251264. doi: 10.1016/j.jmoldx.2014.12.006 [doi]. 18. Schultheis AM, Ng CK, De Filippo MR, et al. Massively parallel sequencing-based clonality analysis of synchronous endometrioid endometrial and ovarian carcinomas. J Natl Cancer Inst. 2016;108(6):10.1093/jnci/ djv427. Print 2016 Jun. doi: 10.1093/jnci/djv427 [doi]. 19. Schultheis AM, Ng CK, De Filippo MR, et al. Massively parallel sequencing-based clonality analysis of synchronous endometrioid endometrial and ovarian carcinomas. J Natl Cancer Inst. 2016;108(6):djv427. doi: 10.1093/jnci/djv427 [doi]. 20. Li H, Durbin R. Fast and accurate short read alignment with burrows-wheeler transform. Bioinformatics. 2009;25(14):1754-1760. doi: 10.1093/bioinformatics/ btp324 [doi]. 21. McKenna A, Hanna M, Banks E, et al. The genome analysis toolkit: A MapReduce framework for analyzing next-generation DNA sequencing data. Genome Res. 2010;20(9):1297-1303. doi: 10.1101/gr.107524.110 [doi]. 22. Cibulskis K, Lawrence MS, Carter SL, et al. Sensitive detection of somatic point mutations in impure and heterogeneous cancer samples. Nat Biotechnol. 2013;31(3):213-219. doi: 10.1038/nbt.2514 [doi]. 23. Koboldt DC, Zhang Q, Larson DE, et al. VarScan 2: Somatic mutation and copy number alteration discovery in cancer by exome sequencing. Genome Res. 2012;22(3):568-576. doi: 10.1101/gr.129684.111 [doi]. 24. Saunders CT, Wong WS, Swamy S, Becq J, Murray LJ, Cheetham RK. Strelka: Accurate somatic small-variant calling from sequenced tumor-normal sample pairs. Bioinformatics. 2012;28(14):1811-1817. doi: 10.1093/ bioinformatics/bts271 [doi]. 25. Li H, Handsaker B, Wysoker A, et al. The sequence alignment/map format and SAMtools. Bioinformatics. 2009;25(16):2078-2079. doi: 10.1093/bioinformatics/ btp352 [doi]. 26. Shen R, Seshan VE. FACETS: Allele-specific copy number and clonal heterogeneity analysis tool for high-throughput

PART ONE | CHAPTER 2

DNA sequencing. Nucleic Acids Res. 2016. doi: gkw520 [pii]. 27. Piscuoglio S, Ng CK, Murray MP, et al. The genomic landscape of male breast cancers. Clin Cancer Res. 2016;22(16):4045-4056. doi: 10.1158/1078-0432.CCR-152840 [doi]. 28. Schwarz JM, Rodelsperger C, Schuelke M, Seelow D. MutationTaster evaluates disease-causing potential of sequence alterations. Nat Methods. 2010;7(8):575-576. doi: 10.1038/nmeth0810-575 [doi]. 29. Carter H, Chen S, Isik L, et al. Cancer-specific highthroughput annotation of somatic mutations: Computational prediction of driver missense mutations. Cancer Res. 2009;69(16):6660-6667. doi: 10.1158/00085472.CAN-09-1133 [doi]. 30. Shihab HA, Gough J, Cooper DN, et al. Predicting the functional, molecular, and phenotypic consequences of amino acid substitutions using hidden markov models. Hum Mutat. 2013;34(1):57-65. doi: 10.1002/humu.22225 [doi]. 31. Kandoth C, McLellan MD, Vandin F, et al. Mutational landscape and significance across 12 major cancer types. Nature. 2013;502(7471):333-339. doi: 10.1038/ nature12634 [doi]. 32. Futreal PA, Coin L, Marshall M, et al. A census of human cancer genes. Nat Rev Cancer. 2004;4(3):177-183. doi: 10.1038/nrc1299 [doi]. 33. Lawrence MS, Stojanov P, Mermel CH, et al. Discovery and saturation analysis of cancer genes across 21 tumour types. Nature. 2014;505(7484):495-501. doi: 10.1038/ nature12912 [doi]. 34. Winoto A, Baltimore D. Developmental regulation of the TCR alpha delta locus. Cold Spring Harb Symp Quant Biol. 1989;54 Pt 1:87-92. 35. Choi Y, Sims GE, Murphy S, Miller JR, Chan AP. Predicting the functional effect of amino acid substitutions and indels. PLoS One. 2012;7(10):e46688. doi: 10.1371/ journal.pone.0046688 [doi]. 36. Roy-Chowdhuri S, de Melo Gagliato D, Routbort MJ, et al. Multigene clinical mutational profiling of breast carcinoma using next-generation sequencing. Am J Clin Pathol. 2015;144(5):713-721. doi: 10.1309/ AJCPWDEQYCYC92JQ [doi]. 37. Weigelt B, Glas AM, Wessels LF, Witteveen AT, Peterse JL, vanâ&#x20AC;&#x2122;t Veer LJ. Gene expression profiles of primary breast tumors maintained in distant metastases. Proc Natl Acad Sci U S A. 2003;100(26):15901-15905. doi: 10.1073/ pnas.2634067100 [doi]. 38. Muller KE, Marotti JD, de Abreu FB, et al. Targeted nextgeneration sequencing detects a high frequency of potentially actionable mutations in metastatic breast cancers. Exp Mol Pathol. 2016;100(3):421-425. doi: S00144800(16)30037-5 [pii]. 39. Cancer Genome Atlas Network. Comprehensive molecular portraits of human breast tumours. Nature. 2012;490(7418):61-70. doi: 10.1038/nature11412 [doi].

40. Lee JY, Park K, Lim SH, et al. Mutational profiling of brain metastasis from breast cancer: Matched pair analysis of targeted sequencing between brain metastasis and primary breast cancer. Oncotarget. 2015;6(41):4373143742. doi: 10.18632/oncotarget.6192 [doi]. 41. Oktyabri D, Ishimura A, Tange S, Terashima M, Suzuki T. DOT1L histone methyltransferase regulates the expression of BCAT1 and is involved in sphere formation and cell migration of breast cancer cell lines. Biochimie. 2016;123:20-31. doi: 10.1016/j.biochi.2016.01.005 [doi]. 42. Zhang L, Deng L, Chen F, et al. Inhibition of histone H3K79 methylation selectively inhibits proliferation, selfrenewal and metastatic potential of breast cancer. Oncotarget. 2014;5(21):10665-10677. doi: 2496 [pii]. 43. Dorsam RT, Gutkind JS. G-protein-coupled receptors and cancer. Nat Rev Cancer. 2007;7(2):79-94. doi: nrc2069 [pii]. 44. Moelans CB, van der Groep P, Hoefnagel LD, et al. Genomic evolution from primary breast carcinoma to distant metastasis: Few copy number changes of breast cancer related genes. Cancer Lett. 2014;344(1):138-146. doi: 10.1016/j.canlet.2013.10.025 [doi]. 45. Mertins P, Mani DR, Ruggles KV, et al. Proteogenomics connects somatic mutations to signalling in breast cancer. Nature. 2016;534(7605):55-62. doi: 10.1038/nature18003 [doi]. 46. Marcotte R, Brown KR, Suarez F, et al. Essential gene profiles in breast, pancreatic, and ovarian cancer cells. Cancer Discov. 2012;2(2):172-189. doi: 10.1158/21598290.CD-11-0224 [doi]. 47. Zang ZJ, Ong CK, Cutcutache I, et al. Genetic and structural variation in the gastric cancer kinome revealed through targeted deep sequencing. Cancer Res. 2011;71(1):29-39. doi: 10.1158/0008-5472.CAN-10-1749 [doi]. 48. Horiuchi D, Kusdra L, Huskey NE, et al. MYC pathway activation in triple-negative breast cancer is synthetic lethal with CDK inhibition. J Exp Med. 2012;209(4):679696. doi: 10.1084/jem.20111512 [doi]. 49. Chandriani S, Frengen E, Cowling VH, et al. A core MYC gene expression signature is prominent in basal-like breast cancer but only partially overlaps the core serum response. PLoS One. 2009;4(8):e6693. doi: 10.1371/journal. pone.0006693 [doi]. 50. Kroigard AB, Larsen MJ, Laenkholm AV, et al. Clonal expansion and linear genome evolution through breast cancer progression from pre-invasive stages to asynchronous metastasis. Oncotarget. 2015;6(8):56345649. doi: 3111 [pii]. 51. Blighe K, Kenny L, Patel N, et al. Whole genome sequence analysis suggests intratumoral heterogeneity in dissemination of breast cancer to lymph nodes. PLoS One. 2014;9(12):e115346. doi: 10.1371/journal.pone.0115346 [doi]. 52. Ding L, Ellis MJ, Li S, et al. Genome remodelling in a basal-like breast cancer metastasis and xenograft. Nature. 2010;464(7291):999-1005. doi: 10.1038/nature08989 [doi].

Mutational profiling of breast cancer metastases

SUPPLEMENTALFigure S1 Supplementary A

Mean target coverage per sample mean 301x

800

Primary

Metastasis Normal tissue

600 400 200 0

1M 1C 1P 4M 4C 4P 5M 5C 5P 6M 6C 6P 7M 7C 7P 8M 8C 8P 9M 9C 9P 10M 10C 10P 12M 12C 12P 13M 13C 13P 14M 14C 14P 15M 15C 15P 16M 16C 16P 19M 19C 19P 48M 48C 48P 52M 52C 52P 188M 188C 188P

M ean target coverage

1000

Samples

% target bases spanned

Spanned target bases 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0

Primary Metastasis Normal

mean at 50x: 90%

100

# of reads

M ean base coverage

Mean target coverage per variant

1000 mean 237x

800 600 400 200 0 Variants

Found variants Del Frameshift variant Splice region variant

2 1

FATHMM

Inframe deletion

Ins

Total=8

Frameshift variant Inframe insertion

Unknown

89 66

Passenger/other Cancer driver

Total=5

Total=195

Del Ins SNV 182

Total=195

SNV 5 18

Missense variant Stop gained Splice region variant

159 Total=182

Supplementary Figure S1. Sequencing statistics. A. Mean target coverage per sample (primary tumor, metastasis and control per patient). B. Spanned target bases C. Mean target coverage per variant D. Types of found variants, including single nucleotide variants, deletions and insertions. E. FATHMM (Functional Analysis through Hidden Markov Models) results for assessment of potential driver and passenger mutations.

PART ONE | CHAPTER 2

Supplementary Figure S2 Patient #5 TP53 p.Cys238Phe: C>A

Primary

Metastasis 22.2%

12.5%

C T T T A A A G G A C

Significant different copy number aberrations between primary tumors and metastases

B Number of events

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0

homozygous deletion amplification

loss

* *

CCND1 FOXA1 GNAS SF3B1 TRAF7 PIK3CD

CDK8 FLT1

FLT3 NKX3-1 RASA1 DDR2 INPP4A JAK1 TMEM127 VTCN1 WT1

> GAINED, < LOST

gain

PAK7

>LOST, < GAINED

Copy number aberrations of individual patient ERBB2 CDK12

Primary tumor patient #9 log−ratio

4 2 0 −2

22 X

Chromosomes

Metastasis patient #9

13 14 15 16

−4

ERBB2 CDK12

log−ratio

4 2 0 −2

Chromosomes

13 14 15 16

−4

Supplementary Figure S2. A. Sanger sequencing validation of TP53 mutation p.Cys238Phe (Exon 7, Fw primers) in patient #5 (triple negative primary tumor with brain metastasis). B. Significantly different copy number alterations between primary tumors and metastases. C. Copy number alterations of patient #9 (HER2-enriched primary tumor with brain metastasis). CDK12 and ERBB2 are coamplified.

Mutational profiling of breast cancer metastases

Supplementary Figure S3

60 40 20 0

20 40 60 80 100 MAF Primary (corrected for tumor%)

80 60 40 20 0

60 40 20 0

20 0

40 20 0

40 20 0 0

20 40 60 80 100 MAF Primary (corrected for tumor%)

40 20 0

60 40 20 0 0

20 40 60 80 100 MAF Primary (corrected for tumor%)

20 0 0 20 40 60 80 100 MAF Primary (corrected for tumor%)

#14: TN brain 80 60 40 20 0

20 40 60 80 100 MAF Primary (corrected for tumor%)

0 20 40 60 80 100 MAF Primary (corrected for tumor%)

#16: TN skin

#15: TN skin 100

100

MAF Metastasis (corrected for tumor%)

#13: TN skin 100

100

20 40 60 80 100 MAF Primary (corrected for tumor%)

#12: TN skin 100

#188: HER2 skin 80

#6: TN brain

20 40 60 80 100 MAF Primary (corrected for tumor%)

0 20 40 60 80 100 MAF Primary (corrected for tumor%)

20 40 60 80 100 MAF Primary (corrected for tumor%)

MAF Metastasis (corrected for tumor%)

100

MAF Metastasis (corrected for tumor%)

#5: TN brain MAF Metastasis (corrected for tumor%)

MAF Metastasis (corrected for tumor%)

100

#10: HER2 skin 80

20 40 60 80 100 MAF Primary (corrected for tumor%)

#4: TN brain

20 40 60 80 100 MAF Primary (corrected for tumor%)

100

MAF Metastasis (corrected for tumor%)

#8 HER2 skin 100

100

20 40 60 80 100 MAF Primary (corrected for tumor%)

#7: HER2 skin Metastasis only Primary only Shared between primary and metastasis

MAF Metastasis (corrected for tumor%)

#52: HER2 brain 100

MAF Metastasis (corrected for tumor%)

#48: HER2 brain 100 MAF Metastasis (corrected for tumor%)

MAF Metastasis (corrected for tumor%)

#19: HER2 brain 100

MAF Metastasis (corrected for tumor%)

#8 HER2 brain 100

MAF Metastasis (corrected for tumor%)

#1: HER2 brain 100

80 60 40 20 0 0

20 40 60 80 100 MAF Primary (corrected for tumor%)

80 60 40 20 0 0 20 40 60 80 100 MAF Primary (corrected for tumor%)

Supplementary Figure S3. Mean allele frequency of all somatic variants per patient.

PART ONE | CHAPTER 2

Supplementary Figure S4 A

Time between primary and metastasis Low concordance between primary and metastasis

P e rc e n t s u rv iv al

100

High concordance between primary and metastasis 50

500

1000

1500

2000

2500

Time between primary tumor and metastasis (in days)

Shared mutations between primary tumors and paired metastases based on treatment history p=0.234

% shared mutations

80 60 40 20 0

Figure 5

Treatments

p=0.767

p=0.175

p=0.621

p=0.075

Neoadjuvant chemotherapy No neoadjuvant chemotherapy Adjuvant chemotherapy No adjuvant chemotherapy Adjuvant hormonal therapy No adjuvant hormonal therapy Adjuvant Herceptin No adjuvant Herceptin

60 40 20 0

Treatments

Targeted therapy options Amplifications

8 N u m b er o f p a tie n ts

p=0.089

% metastasis only mutations

p=0.174

p=0.427

100

Metastasis only mutations based on treatment history

Mutations

Primary tumors Metastases

6 4 2 0

AKT2

ERBB2

FGF PI3K pathway PIK3CA BRCA1/2

BTK

CTL4 PI3K pathway PIK3CA

PTCH1

ROS1

Potentially targetable genes

Supplementary Figure S4. The association of variants/aberrations with clinicopathological characteristics. A. Kaplan Meier curve depicting time between primary tumor and metastasis for highly concordant versus highly discordant samples B. Therapeutic regimens versus percentage of shared mutations and percentage of metastasis only mutations. C. Targeted therapy options for patients included in this study.

Mutational profiling of breast cancer metastases

Supplementary Table S1. Sanger sequencing primer sequences for TP53 mutations found by NGS. Exons

Primers

Sequence Fw

Sequence Rv

Samples

Exon 4.1

Short

5’ GGG GGC TGA GGA CCT GGT 3’

5’ GGA AGG GAC AGA AGA TGA C 3’

Internal

5’ CTG GTC CTC TGA CTG CTC 3’

5’ GAC AGA AGA TGA CAG GGG 3’

7M, 14P, 14M

Short

5’ GTC CAG ATG AAG CTC CCA G 3’

5’ ATA CGG CCA GGC ATT GAA GT 3’

Internal

5’ AGC TCC CAG AAT GCC AGA G 3’

5’ TGA AGT CTC ATG GAA GCC 3’

Short

5’ TGC TGC CGT GTT CCA GTT GC 3’

5’ AAC CTC CGT CAT GTG CTG T 3’

Internal

5’ CCG TGT TCC AGT TGC TTT ATC 3’

5’ GCT GTG ACT GCT TGT AGA TG 3’

Short

5’ CTG CCC TCA ACA AGA TGT TTT GCC 3’ 5’ GCC AGA CCT AAG AGC AAT CAG TG 3’

Internal

5’ ACA AGA TGT TTT GCC AAC TG 3’

5’ GAG CAA TCA GTG AGG AAT CAG 3’

Short

5’ GCG ACA GAG CGA GAT TCC ATC 3’

5’ GAA GAA ATC GGT AAG AGG TGG 3’

Internal

5’ CTT GCC ACA GGT CTC CCC AA 3’

5’ GCG GCA AGC AGA GGC TGG 3’

Short

5’ GGA CCT GAT TTC CTT ACT GC 3’

5’ TCT GAG GCA TAA CTG CAC CC 3’

Internal

5’ CCT TAC TGC CTC TTG CTT C 3’

5’ TAA CTG CAC CCT TGG TCT C 3’

Short

5’ AGG AGA CCA AGG GTG CAG T 3’

5’ CGG CAT TTT GAG TGT TAG A 3’

Internal

5’ GTT ATG CCT CAG ATT CAC T 3’

5’ TGA GTG TTA GAC TGG AAA C 3’

Exon 4.2

Exon 5.1

Exon 5.2

Exon 7

Exon 8

Exon 9

4P, 4M, 8M, 14P, 14M 11P, 52P, 52M 9P, 9M

5P, 5M, 12M, 15P 6P, 6M, 10P, 10M, 19M, 52M 16P, 16M

PART ONE | CHAPTER 2

Metastasis

106.02

98.94

98.17

96.54

92.86

87.25

80.20

37.44

Normal tissue

131.14

98.86

97.84

96.09

94.11

91.82

89.05

65.96

Primary

288.16

99.17

98.74

98.36

97.89

97.25

96.50

92.04

Metastasis

241.11

99.20

98.79

98.42

98.02

97.52

96.83

89.02

Normal tissue

292.41

99.19

98.70

98.18

97.56

96.75

95.87

89.95

Primary

282.89

99.28

98.98

98.71

98.43

98.13

97.75

94.09

Metastasis

840.65

99.41

99.28

99.14

98.99

98.85

98.72

98.06

Normal tissue

191.75

99.00

98.35

97.74

96.96

95.93

94.60

84.66

Primary

452.13

99.33

99.07

98.88

98.66

98.45

98.24

96.78

Metastasis

371.78

99.14

98.70

98.44

98.18

97.89

97.55

94.16

Normal tissue

66.18

98.68

96.05

90.41

81.91

71.00

58.64

16.23

Primary

477.67

99.16

98.77

98.51

98.27

98.06

97.83

95.62

Metastasis

194.31

99.13

98.58

98.02

97.39

96.32

94.68

80.03

Normal tissue

73.94

98.69

95.92

90.85

83.70

74.07

62.50

20.97

Primary

59.99

98.73

97.04

90.73

77.00

58.62

40.73

5.58

Metastasis

659.51

99.40

99.20

99.05

98.89

98.72

98.56

97.72

Normal tissue

236.69

99.10

98.61

98.05

97.18

96.08

94.89

86.62

Primary

291.52

99.25

98.87

98.54

98.25

97.87

97.42

93.73

Metastasis

395.86

99.34

99.11

98.88

98.64

98.37

98.09

96.16

Normal tissue

205.30

99.08

98.33

97.34

96.23

94.81

93.39

83.69

Primary

191.57

99.15

98.67

98.15

97.69

97.11

96.24

85.42

Metastasis

401.64

99.36

99.09

98.85

98.58

98.33

98.03

95.99

Normal tissue

256.71

99.17

98.69

98.11

97.31

96.43

95.42

90.03

Primary

313.79

99.21

98.78

98.43

98.04

97.58

97.01

93.32

Metastasis

626.99

99.31

99.06

98.85

98.63

98.45

98.24

97.01

Normal tissue

99.82

98.79

97.41

94.77

91.19

87.00

82.02

44.73

Primary

113.65

98.98

98.18

97.35

96.10

94.34

91.65

57.51

Metastasis

238.07

99.25

98.76

98.41

98.05

97.65

97.15

92.22

Normal tissue

127.27

99.02

97.84

96.09

93.94

91.47

88.39

63.67

Primary

192.26

99.05

98.53

98.10

97.58

96.78

95.72

84.48

Metastasis

404.47

99.10

98.70

98.40

98.13

97.84

97.47

94.64

Normal tissue

169.06

98.74

98.01

97.15

95.46

93.15

90.31

71.21

Primary

193.12

98.75

98.10

97.33

96.39

95.14

93.63

82.63

Mean target coverage

% of target bases >100x coverage

% of target bases >50x coverage

% of target bases >40x coverage

% of target bases >30x coverage

% of target bases >20x coverage

% of target bases >10x coverage

% of target bases >2x coverage

Sample type

Patient

Supplementary Table S2. Sequencing coverage

Mutational profiling of breast cancer metastases

188

% of target bases >100x coverage

% of target bases >50x coverage

% of target bases >30x coverage

% of target bases >20x coverage

% of target bases >10x coverage

% of target bases >2x coverage

Metastasis

442.91

99.37

98.97

98.71

98.45

98.24

98.02

96.13

Normal tissue

99.60

98.44

96.14

92.70

88.46

83.66

78.23

44.49

Mean target coverage

% of target bases >40x coverage

Sample type

Patient

Supplementary Table S2. Continued

Primary

236.13

99.10

98.64

98.27

97.91

97.45

96.85

90.74

Metastasis

558.19

99.35

99.07

98.83

98.60

98.35

98.10

96.28 82.53

Normal tissue

219.32

98.93

98.15

97.03

95.60

94.05

92.34

Primary

163.92

99.10

98.51

97.96

97.16

96.05

94.70

79.79

Metastasis

332.65

99.27

98.83

98.41

97.98

97.43

96.73

92.47

Normal tissue

40.04

98.43

92.86

76.34

56.06

38.76

25.89

3.48

Primary

96.65

98.78

97.99

95.84

91.60

85.27

77.29

35.25

Metastasis

796.45

99.31

99.07

98.88

98.69

98.51

98.33

96.62

Normal tissue

68.20

99.05

97.92

95.39

90.43

82.73

71.46

13.45

Primary

538.64

99.22

98.86

98.54

98.28

98.01

97.71

95.66

Metastasis

551.60

99.17

98.71

98.46

98.21

97.89

97.53

95.03

Normal tissue

226.56

98.96

98.22

97.15

95.82

94.41

92.93

83.57

Primary

482.16

99.13

98.71

98.41

98.13

97.76

97.29

94.72

Metastasis

503.98

99.38

98.93

97.66

95.97

94.12

91.89

75.37

Normal tissue

381.03

99.23

98.75

98.31

97.79

97.20

96.50

92.82

Primary

434.44

99.15

98.76

98.47

98.18

97.88

97.52

94.95

PART ONE | CHAPTER 2

Supplementary Table S3. Mean total reads and target coverage Mean

Total reads

Target coverage

All samples

22858101

301.17

Normal tissue

10546773

169.71

Primaries

24091019

282.86

HER2

24237706

299.66

23925995

263.97

33936510

450.95

brain

34445360

448.95

skin

33364054

453.20

Primaries vs. metastases

0.007*

0.008*

Primaries vs. normal tissue

0.000*

0.642

Metastases vs. normal tissue

0.000*

0.040*

HER2 vs. TN

0.773

Brain vs. skin

0.773

Metastases

Wilcoxon signed rank or Mann-Whitney test

* significant differences

Mutational profiling of breast cancer metastases

Supplementary Table S4. Primary and metastasis only mutations Primary only mutations Influence of tumor% and copy number aberrations patient

gene

Variant%

tumor%

copy number aberrations

primary

metastasis

primary

metastasis

primary

metastasis -1

DNMT1

p.Val3272Ile

4.1%

0.0%

POLE

p.Gln99Glu

3.6%

0.0%

NOTCH3

p.Pro800Leu

6.8%

0.0%

-1

GRIN2A

p.His677Gln

7.2%

0.0%

RAF1

p.Pro161Ser

4.0%

0.4%

WT1

p.Glu18Gln

4.9%

0.0%

DOT1L

p.Arg1199fs

15.0%

0.0%

-1

PTPRT

p.Arg1729His

6.4%

0.0%

No influence of tumor% and copy number aberrations patient

gene

Variant%

tumor%

copy number aberrations

primary

metastasis

primary

metastasis

primary

metastasis

RB1

p.Pro254Leu

19.5%

0.0%

POLE

p.Arg1111Trp

17.2%

0.0%

ARID1B

p.Pro308Leu

10.8%

0.0%

NOTCH1

p.Asp1880His

21.8%

0.0%

PRDM1

p.Pro2212Leu

23.7%

0.0%

ATRX

p.Ala46Thr

12.4%

0.0%

NOTCH4

p.Ala149Val

11.4%

0.0%

-1

TSC2

p.Glu2250Lys

6.2%

0.0%

DNMT3B

p.Pro320Ser

12.8%

0.0%

SPEN

p.Leu277Pro

6.6%

0.4%

RAD51B

p.Gln1087*

7.1%

0.0%

CTCF

p.Ala638Thr

9.1%

0.0%

ERBB2

p.Ser41del

6.1%

0.4%

-1

PART ONE | CHAPTER 2

Supplementary Table S4. Continued Metastasis only mutations Influence of tumor% and copy number aberrations patient

gene

Variant%

tumor%

copy number aberrations primary

metastasis

TSC2

p.Arg905Gln

0.5%

7.8%

NOTCH4

p.Ala1822Thr

0.0%

26.6%

BAP1

p.Ser460*

0.0%

3.3%

-1

p.Gln799Glu

0.0%

3.2%

-1

CDH1

p.Thr340Met

0.4%

8.1%

BTK

p.Gly12Ala

0.0%

14.0%

CDK12

p.Arg773Cys

0.4%

4.3%

MET

p.Pro563Ser

0.4%

7.6%

PIK3CG

p.His948Asn

0.0%

7.5%

NF1

p.Gly675Arg

0.0%

4.4%

RPTOR

p.Val829Met

0.0%

5.6%

CDH1

p.Asp854His

0.0%

4.3%

ERCC5

p.Met419Ile

0.0%

3.7%

FIP1L1

p.Glu334Gln

0.0%

4.0%

KDR

p.Gln1169Glu

0.0%

3.6%

MRE11A

p.Arg402Thr

0.0%

4.0%

TET2

p.Glu1411Gln

0.0%

3.5%

ATR

p.Val959Met

0.0%

3.3%

TP53

p.Arg273Cys

0.0%

3.9%

TP53

p.Arg273His

0.4%

4.9%

-1

primary

metastasis

primary

metastasis

Mutational profiling of breast cancer metastases

Supplementary Table S4. Continued Metastasis only mutations No influence of tumor% and copy number aberrations patient

gene

Variant%

tumor%

copy number aberrations

primary

metastasis

primary

metastasis

primary

metastasis

ATR

p.Ala572Gly

0.0%

19.2%

BTK

p.Asp435Gly

0.0%

75.0%

-1

KMT2C

p.Tyr4691*

0.0%

39.3%

PIK3R1

p.Arg386Gly

0.0%

24.6%

ARID1A

p.Glu1765Asp

0.0%

21.6%

ARID1A

p.Glu2098Asp

0.0%

20.1%

AXIN2

p.Val248Gly

0.4%

10.7%

-1

BARD1

p.Ala40Val

0.0%

8.0%

E2F3

p.His358fs

0.0%

12.5%

EP300

p.Asp1482Ala

0.0%

22.5%

MDC1

p.Glu609Gly

0.4%

8.0%

PIK3CA

p.Cys420Arg

1.5%

13.8%

SPEN

p.Glu2913Ala

0.0%

19.7%

XPO1

p.Phe973Val

0.0%

8.0%

APC

p.Glu853Gln

0.0%

26.4%

ARID2

p.Asp243His

0.0%

22.0%

BCOR

p.Ser560Leu

0.0%

5.5%

CBL

p.Leu380Val

0.0%

41.8%

-1

CTLA4

p.Pro63Ala

0.0%

18.6%

ERBB2

p.Glu975Gln

0.1%

6.1%

EWSR1

p.Arg469Thr

0.0%

6.0%

GRIN2A

p.Ser944*

0.0%

6.9%

MSH6

p.Trp628Cys

0.0%

24.0%

PTPRD

p.Glu206Lys

0.0%

7.2%

-1

RFWD2

p.Asp430Asn

0.0%

6.3%

SHQ1

p.Asp121His

0.0%

23.4%

ALOX12B

p.Ser23Leu

0.0%

8.3%

XIAP

p.Ala291Thr

0.0%

19.9%

-1

EPHA3

p.Asp17His

0.0%

20.6%

MAP3K13

p.Glu114*

0.0%

17.5%

FLT1

p.Arg508His

0.7%

32.4%

KMT2D

p.Glu5312Gln

1.3%

11.2%

BRCA1

p.Ser1507Gly

0.0%

27.1%

GSK3B

p.Pro404Leu

0.0%

23.6%

SMAD4

p.Gln450dup

0.0%

34.3%

-1

AXIN2

p.Ser304Thr

0.0%

20.6%

DOT1L

p.Asn589Ser

0.0%

6.3%

-1

RUNX1

p.Ser318Phe

0.0%

18.2%

CREBBP

p.Val1129Ile

0.0%

22.4%

HIST1H3B

p.Glu51Asp

0.0%

4.7%

ROS1

p.Ser2229Cys

0.0%

4.8%

-1

ROS1

p.Lys2228Gln

0.0%

4.5%

-1

GRIN2A

p.Gln1255Pro

0.0%

32.2%

PTPRT

p.Asn712Ser

0.0%

27.2%

DOT1L

p.Gln94*

0.0%

18.9%

-1

RB1

splice site

0.0%

53.4%

PART ONE | CHAPTER 2

Supplementary Table S5. Recurrent somatic variants TN-specific Gene

ARID1B

ATR BTK NOTCH4

CHROM 6

3 X 6

Tumor MAF

Sample ID

Mol. Subtype

Meta location

p.Ser41del

10.8%

13P

skin

p.Glu1564Gly

52.7%

brain

p.Glu1564Gly

13.8%

brain

p.Val959Met

3.3%

15M

skin

p.Ala572Gly

19.2%

brain

p.Gly12Ala

14.0%

14M

brain

p.Asp435Gly

75.0%

brain

p.Pro800Leu

11.4%

14P

brain

p.Ala1822Thr

26.6%

brain

p.Thr340Met

8.1%

10M

HER2

skin

p.Asp854His

4.3%

HER2

brain

p.Gln494*

39.5%

HER2

skin

p.Gln494*

40.0%

HER2

skin

p.Asp1482Ala

22.5%

HER2

skin

p.Lys460Glu

18.5%

HER2

skin

p.Lys460Glu

22.2%

HER2

skin

p.Glu206Lys

27.5%

HER2

brain

p.Glu206Lys

16.2%

HER2

brain

p.Pro2212Leu

12.4%

14P

brain

p.Glu1917*

20.6%

HER2

brain

p.Glu1917*

13.2%

HER2

brain

p.Gly12Ala

14.0%

14M

brain

p.Asp435Gly

75.0%

brain

p.Arg1111Trp

6.1%

14P

brain

p.Glu975Gln

6.1%

HER2

brain

p.Pro800Leu

11.4%

14P

brain

p.Ala1822Thr

26.6%

brain

p.Pro254Leu

3.6%

brain

p.Glu18Gln

17.2%

HER2

brain

p.Ser304Thr

20.6%

12M

skin

p.Val248Gly

10.7%

HER2

skin

p.Gln494*

39.5%

HER2

skin

p.Gln494*

40.0%

HER2

skin

p.Asp1482Ala

22.5%

HER2

skin

p.Val3272Ile

6.6%

12P

skin

p.Glu2913Ala

19.7%

HER2

skin

p.Val1501Leu

58.7%

12M

skin

p.Val1501Leu

65.1%

12P

skin

p.Arg1199fs

15.0%

12P

skin

p.Asn589Ser

6.3%

14M

brain

p.Gln94*

18.9%

52M

HER2

brain

p.Ala638Thr

7.2%

10P

HER2

skin

p.Gln1255Pro

32.2%

16M

skin

p.Ser944*

6.9%

HER2

brain

HER2-specific CDH1

EP300

SUFU

Brain-specific ATRX

BTK ERBB2 NOTCH4 POLE

X 17 6 12

Skin-specific AXIN2

EP300

SPEN

Non-specific

DOT1L

GRIN2A

Mutational profiling of breast cancer metastases

Supplementary Table S5. Continued Non-specific (continued) Gene

HIST1H3B

PIK3CA

PTPRT

RB1

TP53

TSC2

CHROM 6

Tumor MAF

Sample ID

Mol. Subtype

Meta location

p.Ala99Pro

10.0%

15M

skin

p.Ala99Pro

24.3%

15P

skin

p.Glu51Asp

4.7%

HER2

brain

p.His1047Arg

87.2%

16M

skin

p.His1047Arg

66.7%

16P

skin

p.His1047Arg

50.7%

48M

HER2

brain

p.His1047Arg

28.7%

48P

HER2

brain

p.Pro381Ala

23.1%

HER2

brain

p.Pro381Ala

11.3%

HER2

brain

p.Arg4Gln

17.7%

HER2

brain

p.Arg4Gln

17.3%

HER2

brain

p.Cys420Arg

13.8%

HER2

skin

p.Gln1087*

6.4%

16P

skin

p.Asn712Ser

27.2%

48M

HER2

brain

p.Asp109Tyr

39.5%

12M

skin

p.Asp109Tyr

41.5%

12P

skin

p.Leu277Pro

19.5%

brain

p.Arg787*

46.2%

HER2

brain

p.Arg787*

14.8%

HER2

brain

splice site

53.4%

52M

HER2

brain

p.Arg273Cys

3.9%

19M

HER2

brain

p.Arg273His

4.9%

52M

HER2

brain

p.Leu145Pro

77.1%

52M

HER2

brain

p.Leu145Pro

21.1%

52P

HER2

brain

p.Arg175His

72.8%

HER2

brain

p.Arg175His

35.0%

HER2

brain

p.Gly266Arg

29.5%

10M

HER2

skin

p.Gly266Arg

18.1%

10P

HER2

skin

p.Pro36fs

31.5%

HER2

skin

p.Pro36fs

30.3%

HER2

skin

p.Gln104*

51.5%

HER2

skin

p.Gln104*

7.8%

HER2

skin

p.Ser241Ala

57.3%

12M

skin

p.Ser241Ala

63.5%

12P

skin

p.Cys135Arg

17.0%

13M

skin

p.Cys135Arg

61.5%

13P

skin

p.Arg248Gln

37.3%

15M

skin

p.Arg248Gln

87.4%

15P

skin

p.Phe338fs

77.1%

16M

skin

p.Phe338fs

50.2%

16P

skin

p.Ala84fs

65.4%

14M

brain

p.Ala84fs

68.5%

14P

brain

p.Leu111fs

65.5%

brain

p.Leu111fs

49.0%

brain

p.Cys238Phe

22.2%

brain

p.Cys238Phe

12.5%

brain

p.Arg306*

66.2%

brain

p.Arg306*

57.5%

brain

p.Arg1729His

6.2%

16P

skin

p.Arg905Gln

7.8%

HER2

brain

PART ONE | CHAPTER 2

Supplementary Table S6. Recurrent amplifications and homozygous deletions TN-specific Gene

EPHB1

NF2

RB1

CCND1

ETV6

FGF19

FGF4

RECQL4

SOX9

CHROM

aberration

Sample ID

Mol. Subtype

Meta location

amp

brain

amp

brain

amp

14M

brain

amp

14P

brain

amp

brain

amp

brain

homdel

brain

homdel

brain

homdel

brain

homdel

14M

brain

homdel

14P

brain

amp

brain

amp

brain

amp

12M

skin

amp

brain

amp

brain

amp

15P

skin

amp

16M

skin

amp

16P

skin

amp

brain

amp

brain

amp

12M

skin

amp

brain

amp

brain

amp

12M

skin

amp

brain

amp

12P

skin

amp

13M

skin

amp

13P

skin

amp

15P

skin

amp

brain

amp

12M

skin

amp

12P

skin

amp

HER2

brain

amp

52M

HER2

brain

amp

52P

HER2

brain

amp

HER2

brain

amp

52M

HER2

brain

amp

52P

HER2

brain

amp

HER2

brain

amp

52M

HER2

brain

amp

52P

HER2

brain

amp

HER2

brain

amp

52M

HER2

brain

amp

52P

HER2

brain

amp

HER2

brain

amp

HER2

brain

amp

HER2

brain

amp

HER2

brain

HER2-specific HIST1H1C

HIST1H2BD

HIST1H3B

RARA

CDK12

Mutational profiling of breast cancer metastases

Supplementary Table S6. Continued HER2-specific (Continued) Gene

CDK12

ERBB2

CHROM

aberration

Sample ID

Mol. Subtype

Meta location

amp

48M

HER2

brain

amp

48P

HER2

brain

amp

52M

HER2

brain

amp

52P

HER2

brain

amp

HER2

skin

amp

HER2

skin

amp

10M

HER2

skin

amp

10P

HER2

skin

amp

188M

HER2

skin

amp

188P

HER2

skin

amp

HER2

brain

amp

HER2

brain

amp

HER2

brain

amp

HER2

brain

amp

48M

HER2

brain

amp

48P

HER2

brain

amp

52M

HER2

brain

amp

52P

HER2

brain

amp

HER2

skin

amp

HER2

skin

amp

10M

HER2

skin

amp

10P

HER2

skin

amp

188M

HER2

skin

amp

188P

HER2

skin

amp

48M

HER2

brain

amp

brain

amp

brain

amp

brain

amp

brain

amp

14M

brain

amp

14P

brain

amp

48M

HER2

brain

amp

14M

brain

amp

14P

brain

amp

HER2

brain

amp

52M

HER2

brain

amp

52P

HER2

brain

amp

HER2

brain

amp

52M

HER2

brain

amp

52P

HER2

brain

amp

HER2

brain

amp

52M

HER2

brain

amp

52P

HER2

brain

amp

brain

amp

brain

homdel

brain

amp

HER2

brain

amp

52M

HER2

brain

amp

52P

HER2

brain

Brain-specific AKT2

EPHB1

GATA1

HIST1H1C

HIST1H2BD

HIST1H3B

NF2

RARA

PART ONE | CHAPTER 2

Supplementary Table S6. Continued HER2-specific (Continued) Gene

RB1

CHROM

aberration

Sample ID

Mol. Subtype

Meta location

homdel

brain

homdel

brain

homdel

14M

brain

homdel

14P

brain

homdel

48M

HER2

brain

amp

HER2

brain

amp

15P

skin

amp

HER2

brain

amp

brain

amp

brain

amp

12M

skin

amp

HER2

brain

amp

HER2

brain

amp

52M

HER2

brain

amp

52P

HER2

brain

amp

HER2

skin

amp

HER2

skin

amp

12P

skin

amp

13M

skin

amp

13P

skin

amp

15M

skin

amp

15P

skin

amp

16M

skin

amp

16P

skin

amp

brain

amp

HER2

brain

amp

HER2

brain

amp

12M

skin

Non-specific CDKN1B

FGF3

MYC

PAK1

Mutational profiling of breast cancer metastases

Supplementary Table S7. Genes that are more or less often affected in primary breast tumors compared to paired metastases.

ARID1B ATRX CDKN1A CTCF ERBB2 HIST1H3B INPP4A NOTCH1 NOTCH4 PRDM1 RAD51B RB1 RET SPEN TSC2

ATR AXIN2 BRCA1 BTK CREBBP DOT1L GRIN2A GSK3B KMT2C MSH6 PIK3CD RBM10 ROS1 RUNX1 SMAD4

HER2-ENRICHED TUMOR

DDR2 DNMT3B IRS2 PIK3C2G POLE

Genes more affected in metastases

Genes less affected in metastases

Genes more affected in metastases

Genes less affected in metastases

TRIPLE NEGATIVE TUMOR

AKT1 ALOX12B APC ARID1A ARID2 ATRX AXIN2 BARD1 BRCA2 BRD4 CARD11 CBL CCND1 CTLA4 DOT1L E2F3 EP300 EPHA3 ERBB2 EWSR1 FAT1 FLT1 FOXA1 GNAQ GRIN2A HIST1H3B JAK1 KDM6A KLF4 KMT2D MAP3K13 MDC1 MDM4 NSD1 PIK3CA PIK3R1 PTPRD PTPRT RARA RB1 RECQL4 RFWD2 SHQ1 SPEN SUFU SUZ12 TET1 TP53 XIAP XPO1

PART ONE | CHAPTER 2

Supplementary Table S7. Continued Genes more affected in metastases

TUMORS THAT DISSEMINATED TO SKIN Genes less affected in metastases

Genes more affected in metastases

Genes less affected in metastases

TUMORS THAT DISSEMINATED TO BRAIN

ATRX

ALOX12B

ARID1B

ARID1A

CTCF

APC

AURKB

AXIN2

DNMT3B

ARID2

CDKN1A

BARD1

FLT3

ATR

HIST1H3B

BRAF

INPP4A

ATRX

NOTCH1

BRCA1

NOTCH4

BRCA2

PAX5

CCND1

PAK7

BRD4

PIK3C2G

E2F3

POLE

BTK

PRDM1

EP300

RAD51B

CARD11

TSC2

FLT1

RASA1

CBL

GRIN2A

WT1

CREBBP

GSK3B

CTLA4

IDH2

DOT1L

KMT2D

EPHA3

MDC1

EWSR1

MSH6

FAT1

PIK3CA

FOXA1

PIK3R1

GNAQ

ROS1

GRIN2A

SMAD4

HIST1H3B

XPO1

JAK1

KLF4

KMT2C MAP3K13 MDM4 PIK3CA PIK3CD PTPRD PTPRT RARA RECQL4 RFWD2 RUNX1 SDHA SHQ1 SUFU SUZ12 TET1 XIAP

Mutational profiling of breast cancer metastases

Supplementary Table S8. Targetable mutations Aberration ERBB2 amplifications

Primary

Metastasis

Agent

#1 #8 #9 #10 #48 #52 #188

Trastuzumab, lapatinib

PIK3CA amplifications

#52

PI3K pathway inhibitors

PIK3CA mutations

#1 #9 #16 #48

#1 #8 #9 #16 #48

Idelalisib

AKT2 amplifications

#6 #48

AKT inhibitors

BRCA1/2 mutations

#9 #12

PARP inhibitors

BTK mutations

#4 #14

Ibrutinib

CTLA4 mut

ROS1 mut

#15

#188 #14

#188 #6 #7 #14

PTCH1 mut

#19

Vismodegib

FGFR amp

#16 #1 #6

#16 #12 #6

FGFR inhibitors

PI3K pathway aberrations

Ipilimumab Crizotinib PI3K pathway inhibitors, mTOR inhibitors

This list is not complete. Its only provided to get an idea about the treatment possibilities.

Chapter 3 Willemijne AME Schrijver*, Anieta M Sieuwerts*, Simone U Dalm, Vanja de Weerd, Cathy B Moelans, Natalie ter Hoeve, Paul J van Diest, John WM Martens, Carolien HM van Deurzen * Both authors contributed equally to this study

Progressive APOBEC3B mRNA expression in distant breast cancer metastases

Submitted

PART ONE | CHAPTER 3

ABSTRACT APOBEC3B was recently identified as a gain-of-function enzymatic source of mutagenesis, which may offer novel therapeutic options with molecules that specifically target this enzyme. In primary breast cancer, APOBEC3B mRNA is deregulated in a substantial proportion of cases and its expression is associated with poor prognosis. However, its expression in breast cancer metastases, which are the main causes of breast cancer-related death, remained to be elucidated. RNA was isolated from 55 primary breast cancers and paired metastases, including regional lymph node (N = 20) and distant metastases (N = 35). APOBEC3B mRNA levels were measured by RT-qPCR. Expression levels of the primary tumors and corresponding metastases were compared, including subgroup analysis by estrogen receptor (ER/ESR1) status. Overall, APOBEC3B mRNA levels of distant metastases were significantly higher as compared to the corresponding primary breast tumor (P = 0.0015), an effect that was not seen for loco-regional lymph node metastases (P = 0.23). Subgroup analysis by ER-status showed that increased APOBEC3B levels in distant metastases were restricted to metastases arising from ER-positive primary breast cancers (P = 0.002). However, regarding ERnegative primary tumors, only loco-regional lymph node metastases showed increased APOBEC3B expression when compared to the corresponding primary tumor (P = 0.028). APOBEC3B mRNA levels are significantly higher in breast cancer metastases as compared to the corresponding ER-positive primary tumors. This suggests a potential role for APOBEC3B in luminal breast cancer progression, and consequently, a promising role for anti-APOBEC3B therapies in advanced stages of this frequent form of breast cancer.

APOBEC3B expression in breast cancer metastases

INTRODUCTION Breast cancer is the fifth cause of overall cancer related death 1 and this mortality is largely caused by progression of metastatic disease 2. Therefore, one of the most important challenges in breast cancer research includes the genetic changes and molecular mechanisms by which cancer cells acquire their metastatic ability. The generally accepted hypothesis is that metastases are caused by multiple intricate steps that arise in the primary tumor site 3 . Nevertheless, discordances between primary tumors and corresponding metastases are often encountered 4. However, therapies applied for disseminated disease are mainly based on primary tumor characteristics only. The study of molecular differences between matched primary tumors and metastatic lesions may improve our understanding of disease progression and has the potential to reveal novel, potentially targetable drivers of metastatic progression. Apolipoprotein B editing catalytic subunit 3B (APOBEC3B) is a member of the AID/ APOBEC family of deaminases, which is recognized for its ability to deaminate genomic DNA cytosines. APOBEC enzymes normally function in the innate immune system and in the protection against viral pathogens, but these enzymes can also generate Câ&#x2020;&#x2019;T mutations in the host genome 5. Recently, several studies showed that APOBEC3B is a common enzymatic mutagenic factor affecting the evolution of different cancer types, including breast cancer 5-19. In breast cancer, APOBEC3B mRNA is substantially upregulated in one third of cases and its expression is associated with mutational load, including certain driver mutations in PIK3CA and TP53 18,20. Besides, multiple studies have postulated that APOBEC3B influences the development of metastases and drug resistance, especially in estrogen receptor alpha (ERÎą)-positive breast cancer 5,21,22. In line with this, we previously reported an association between high APOBEC3B mRNA expression and poor outcome in a large cohort of patients with ERÎą-positive breast cancer 23. Since APOBEC3B is a gain-of-function mutagenic enzyme, it may be treatable with small molecules 5,24, which could have an important role in the management of metastatic disease. However, the expression of APOBEC3B in breast cancer metastases remained to be elucidated. In this study, we therefore quantified APOBEC3B mRNA in primary breast cancers and paired metastases to gain more insight into the levels of expression during breast cancer progression.

PART ONE | CHAPTER 3

METHODS Clinical pathological data In this study we adhered to the Code of Conduct of the Federation of Medical Scientific Societies in the Netherlands (http://www.fmwv.nl). The use of anonymous or coded left over material for scientific purposes is part of the standard treatment agreement with patients and therefore informed consent was not required according to Dutch law 25. We selected 73 formalin-fixed paraffin-embedded (FFPE) primary breast cancers and corresponding metastases from the pathology archives of the University Medical Center Utrecht and Erasmus Medical Center Rotterdam. Each specimen was reviewed by a pathologist to determine the percentage of invasive tumor cells. Inclusion criteria were: availability of clinical and pathological data, the presence of enough tumor tissue with the possibility to macro-dissect an area containing at least 50% tumor cells and good RNA quality and quantity to reliably determine expression levels by RT-qPCR (Supplementary methods). After applying these inclusion criteria, 55 paired primary tumors and metastases from different sites remained, including those from regional lymph nodes (N = 20), brain (N = 14), liver (N = 6), ovary (N = 4), lung (N = 4), bone (N = 4) and gastrointestinal tract (N = 3). Clinicopathological characteristics included age, primary tumor size, histological subtype, Bloom & Richardson score, ERα and HER2 expression and regional lymph node status. Furthermore, overall survival (death due to any cause) was reported. Detailed clinical information of this cohort is summarized in Table 1. RNA isolation and quantitative reverse transcriptase polymerase chain reaction (RT-qPCR) Ten 10 µm slides were cut from the primary tumors and paired metastases. The first and last sections (5 μm) were stained with hematoxylin and eosin to guide macro-dissection of the tumor cells for RNA extraction. Total RNA was isolated from the macro-dissected sections with the AllPrep DNA/RNA FFPE Kit (Qiagen) and resulting nucleic acid concentrations were measured with a Nanodrop 2000 system (ThermoFisher Scientific). cDNA was generated for 30 min at 48°C with RevertAid H minus (ThermoFisher Scientific) and gene-specific pre-amplified with Taqman PreAmp Master mix (ThermoFisher Scientific) for 15 cycles, followed by Taqman probe–based real time PCR according to the manufacturer’s instructions in a MX3000P Real-Time PCR System (Agilent). The following gene expression assays were evaluated (all from ThermoFisher Scientific): APOBEC3B, Hs00358981_m1; EPCAM, Hs00158980_m1; ESR1, Hs00174860_m1; ERBB2, Hs01001580_ m1, KRT19, Hs01051611_gH; PTPRC, Hs00236304_m1. mRNA levels were quantified relative to the average expression of GUSB, Hs9999908_m1 and HMBS, Hs00609297_m1 using the delta Cq (dCq = 2ˆ(average Cq reference genes – Cq target gene)) method.

APOBEC3B expression in breast cancer metastases

Table 1. Association of APOBEC3B mRNA expression with clinicopathological characteristics of the primary tumor.

All patients in this cohort

IQR

Median

PTPRC (CD45) mRNA log2

IQR

Median

AVG epithelial mRNA log2

IQR

Median

Clinical characteristics

APOBEC3B mRNA log2

No of patients*

Percentage of patients

100%

-6.14

-5.28

-3.11

-1.25

-2.98

-2.31

Age at surgery (years) ≤ 40

18%

-6.69

-4.62

-2.74

-1.40

-2.31

-2.89

41-55

38%

-5.80

-5.95

-3.23

-1.46

-3.14

-1.41

56-70

36%

-6.64

-4.16

-3.03

-1.25

-3.48

-7.68

> 70

-6.04

-2.08

-2.93

-0.54

-2.40

-1.09

P≠

0.76

0.52

1.00

Tumor size

≤ 2 cm

31%

-6.14

-6.31

-3.03

-0.94

-2.72

-2.28

2 ≤ 5 cm

49%

-5.92

-2.25

-2.93

-1.40

-3.62

-7.65

> 5 cm

13%

-6.91

-4.07

-3.17

-1.19

-3.27

-1.69

P≠

0.70

0.95

0.88

Histopathological subtypes† Ductal

78%

-5.80

-4.65

-2.93

-1.41

-3.27

-1.76

Lobular

15%

-8.50

-3.92

-3.14

-0.85

-2.31

-4.90

Other

-6.43

-3.19

-3.37

-0.58

-1.23

-2.24

0.07

0.41

0.31

Bloom & Richardson grade

I + II

18%

-6.39

-4.78

-3.11

-1.51

-6.94

-7.56

III

69%

-5.78

-4.65

-3.15

-1.33

-2.92

-2.37

0.43

0.52

0.08

ESR1 status

Negative

40%

-5.64

-3.79

-2.91

-1.24

-3.06

-2.00

Positive

60%

-6.40

-4.69

-3.17

-1.35

-2.98

-2.81

0.15

0.39

0.88

ERBB2 status

Negative

78%

-5.80

-5.01

-3.19

-1.06

-2.97

-2.90

Positive/ amplified

22%

-7.35

-3.70

-2.77

-1.59

-3.31

-1.67

0.042

0.11

0.96

PART ONE | CHAPTER 3

Table 1. Continued

All patients in this cohort

100%

IQR

Median

PTPRC (CD45) mRNA log2

IQR

Median

AVG epithelial mRNA log2

IQR

Median

Clinical characteristics

APOBEC3B mRNA log2

No of patients*

Percentage of patients

-6.14

-5.28

-3.11

-1.25

-2.98

-2.31

Regional lymph node status Negative

29%

-5.64

-4.40

-3.17

-1.14

-2.84

-2.63

Positive

60%

-6.40

-4.62

-3.00

-1.17

-2.98

-1.84

0.23

0.74

0.90

Time between primary tumor and studied metastasis

≤ 24 months

60%

-5.92

-5.45

-2.88

-1.16

-2.85

-2.09

> 24 months

40%

-6.43

-4.09

-3.24

-0.93

-3.20

-2.28

0.97

0.42

0.33

Overall survival status

Alive

40%

-5.92

-2.52

-3.11

-1.48

-3.67

-8.34

Deceased

60%

-6.65

-5.03

-3.11

-0.95

-2.81

-1.68

0.21

0.62

0.20

≠

* Due to missing values numbers do not always add up to 55. ≠ Spearman correlation significance (2-tailed). $ Mann-Whitney Test significance (2-tailed). † mRNA expression of ductal and lobular breast cancer was compared. AVG epithelial; average mRNA level of KRT19 and EPCAM. IQR; interquartile range.

Because data regarding ERα and HER2 expression at the protein level were incomplete for this cohort, we used ESR1 and ERBB2 mRNA expression levels to determine ESR1 and ERBB2 status (using a cut-off dCq for ESR1>1 and for ERBB2>3.5) (Supplementary methods, Supplementary Figure S1). Statistics SPSS version 23 was used for all statistical analyses. The Kolmogorov-Smirnov and ShapiroWilk tests were used to test for normality of the distributions. To compare mean values between two or more groups, the Mann-Whitney U Test or Kruskal-Wallis test was used, followed by a test for trend if appropriate. To compare values measured in primary cancers and paired metastases, the paired Wilcoxon Signed Ranks test was used. To correlate linear variables, the Spearman Rank Correlation test was used. P-values ≤ 0.05 were considered statistically significant. 64

APOBEC3B expression in breast cancer metastases

RESULTS APOBEC3B mRNA expression in primary breast cancer Since the Kolmogorov-Smirnov and Shapiro-Wilk tests showed that our data were not always normally distributed, we tested all our data non-parametrically. First, we correlated the levels of APOBEC3B mRNA measured in the primary tumors with traditional clinicopathological characteristics (Table 1). Besides a higher expression of APOBEC3B mRNA in ERBB2 negative tumors when compared to ERBB2 positive tumors (Spearmanâ&#x20AC;&#x2122;s Rho = -0.46, P = 0.002), APOBEC3B mRNA levels were not correlated to any of the studied parameters. To ensure that different levels of tumor cells or inflammatory cells in the primary tumors and their matched metastasis did not bias our data, we quantified the mRNA levels of KRT19 and EPCAM (as a measure for epithelial content) and PTPRC (the gene for the common leukocyte antigen CD45, as a measure for the presence of lymphocytes). Although a weak positive correlation was observed between APOBEC3B mRNA levels and epithelial content in the complete cohort (Spearmanâ&#x20AC;&#x2122;s Rho = 0.21, P = 0.031, N = 110), no significant correlations were observed between the mRNA levels of APOBEC3B and epithelial or infiltrate content when analyzed separately for the primary tumors and the metastases (Spearman correlation significance P > 0.05). In addition, epithelial and infiltrate content did not differ significantly between the primary tumors and matched metastases (paired Wilcoxon Signed Rank test P > 0.05). APOBEC3B mRNA expression in primary breast cancer and paired metastases Next, we correlated APOBEC3B mRNA expression in primary tumors and their matched metastases. This analysis revealed that APOBEC3B mRNA levels were significantly higher in the matched distant metastases as compared to the primary tumors (paired Wilcoxon Signed Rank test P = 0.0015, Figure 1A and Supplementary Table S1). In contrast, no difference was perceived between primary tumors and matched loco-regional lymph node metastases (paired Wilcoxon Signed Rank test P = 0.23). APOBEC3B mRNA levels measured in distant metastases (N = 35) showed a trend towards higher expression when compared to regional lymph node metastases (N = 20) of unmatched cases (Mann-Whitney U test P = 0.08, Figure 1A). No such trend was observed in primary tumors that disseminated either to loco-regional or distant locations (Mann-Whitney U test P = 0.42, Figure 1A). Subgroup analysis by distant metastatic site showed increased APOBEC3B expression for all locations, particularly for liver and ovary, although no significance was reached for any of the relatively small subgroups (paired Wilcoxon Signed Ranks test P > 0.05, Figure 1B and Supplementary Table S1). We also compared mRNA levels measured in primary tumors according to distant metastatic site. These analyses showed that APOBEC3B mRNA 65

PART ONE | CHAPTER 3

levels were lowest in primary tumors that metastasized to the ovaries and gastro-intestinal sites and highest in primary tumors that metastasized to lung, brain or bone (Kruskal-Wallis test P = 0.030), Figure 1B). APOBEC3B mRNA expression according to ESR1 and ERBB2 status of the primary tumor We observed no association between APOBEC3B mRNA levels and ESR1-status (Table 1). Several previous studies, however, showed higher APOBEC3B mRNA expression in ERαnegative tumors compared to ERα-positive tumors 23,26,27. Notably, in these studies, high APOBEC3B expression levels were only associated with poor prognosis for ERα-positive primary breast tumors. We therefore categorized our primary cohort into ESR1-positive and ESR1-negative primary tumors (Figure 1C-D). For the ESR1-positive primary tumors, a significantly higher expression was seen in paired distant metastases, but not in locoregional metastases (paired Wilcoxon Signed Ranks Test P = 0.002 and 0.53, respectively). In contrast, for the ESR1-negative primary tumors a significantly higher expression was seen in loco-regional metastases, but not in distant metastases (paired Wilcoxon Signed

Primary tumors vs. paired metastases

dCq_APOBEC3B

-5

-10

N = 20

N = 35

Locoregional metastases

Distant metastases

ESR1+ primary tumors p < 0.05

-10

N = 11

N = 22

Locoregional metastases

Distant metastases

-10

N=4

N=3

N=6

N=4

N = 14

N=4

Ovary

Liver

Bone

Brain

Lung

ESR1- primary tumors 0

dCq_APOBEC3B

-5

-15

p < 0.01

-15

Primary tumors vs. several metastasic subsites

p < 0.01

-15

dCq_APOBEC3B

p < 0.05

-5

-10

N=9 -15

Locoregional metastases

N = 14

Distant metastases

Figure 1. APOBEC3B mRNA expression differences between primary breast tumors and paired metastases. A APOBEC3B mRNA expression in primary breast tumors versus paired loco-regional and distant metastases. B APOBEC3B mRNA expression in primary breast tumors versus paired metastases, subdivided per location of metastasis (ovary (N = 4); gastro-intestinal tract (N = 3); liver (N = 6); bone (N = 4); brain (N = 14) and lung (N = 5). C APOBEC3B mRNA expression in ESR1-positive primary breast tumors versus paired distant and loco-regional metastases. D APOBEC3B mRNA expression in ESR1-negative primary breast tumors versus paired distant and loco-regional metastases. P-values obtained by paired Wilcoxon Signed Ranks test (2-tailed).

APOBEC3B expression in breast cancer metastases

Ranks Test P = 0.028 and 0.81, respectively). Receptor conversion from an ESR1-positive primary tumor to an ESR1-negative metastasis could not explain this finding (Supplementary Table S2). No such difference was seen after categorizing our patients according to ERBB2-status. Irrespective of ERBB2-status, APOBEC3B levels were only higher in the distant metastases and not in the loco-regional lymph nodes when compared to the paired primary tumor (paired Wilcoxon Signed Ranks Test P < 0.05 and > 0.05, respectively).

DISCUSSION APOBEC3B is thought to affect the evolution of breast cancer by somatically mutagenizing the cancer genome, which could potentially be abrogated by therapeutic intervention 5. Previous studies investigated APOBEC3B mRNA expression in primary breast tumors and paired normal tissue. These studies reported upregulation in primary tumors compared to normal tissue, especially in ERα-negative cases 17,26. However, metastatic disease remains the major cause of breast cancer related mortality and several studies reported discordances of (epi) genetic and immunohistochemical markers between primary tumors and matched metastases 4,28-32. To the best of our knowledge, no data is available regarding APOBEC3B expression in breast cancer metastases. Therefore, we set out to evaluate APOBEC3B mRNA expression in primary breast cancer and matched metastases. Importantly, we encountered a significant increase in APOBEC3B mRNA levels in the metastases compared to their corresponding primary tumor. Furthermore, distant metastases showed higher expression than loco-regional lymph node metastases. This implies a role for APOBEC3B not only at the stage of the primary tumor but also, and according to our data even more dominantly, during tumor evolution of metastatic breast cancer. Previous studies reported an association between APOBEC3B expression and aggressive characteristics of the primary breast cancer, including high histologic grade, genomic grade, advanced stage, negative ERα status and HER2 amplification 21,23,26,27,33. In this current, more concise study with a special focus on breast cancer metastases, we only observed a negative association between APOBEC3B and ERBB2 mRNA levels in both the primary tumor (Table 1) and the metastases (data not shown). However, our sample size was relatively small with a relatively high number of cases with loco-regional (36%) and brain metastases (25%), which could have biased our results. Overall, we did not find a correlation between APOBEC3B and ESR1-status of the primary tumor, while previous studies reported higher APOBEC3B mRNA expression in ERαnegative tumors compared to ERα-positive tumors 23,26,27. Interestingly, for our ESR1negative primary cases, a significantly higher expression of APOBEC3B was seen in paired loco-regional metastases only and not in paired distant metastases. For our ESR1-positive 67

PART ONE | CHAPTER 3

primary cases on the other hand, a significantly higher expression was seen in paired distant metastases and not in the loco-regional lymph nodes. This is especially noteworthy in view of our previous finding, that high levels of APOBEC3B were only associated with poor prognosis in ESR1-positive primary breast cancers, and not in ESR1-negative cases 23. Irrespective of ESR1-status, we showed that APOBEC3B expression was increased in distant metastases compared to the corresponding primary tumor, with highest expression in liver, lung, brain and bone metastases. APOBEC3B thus seems not only needed for breast cancer progression, but also for maintenance of the metastasis in distant environments. Since APOBEC3B is upregulated in numerous cancer types we wondered if these findings could be explained by the micro-environment of the distant site. In an article of Burns et al. 7, APOBEC3B expression levels determined by RNA-seq showed a lower expression in normal brain and ovarian tissue relative to normal breast tissue. Furthermore, brain tumors (lowgrade glioma, glioblastoma multiforme) and ovarian tumors (serous cystadenocarcinoma) also had lower APOBEC3B expression levels than breast carcinoma. This might imply that the higher APOBEC3B mRNA levels we found in breast cancer brain and ovarian metastases are independent of the micro-environment at these locations. Furthermore, since the pattern of APOBEC3B expression in primary tumors is retained and even increased in paired metastases, and shows a trend toward a possible metastatic location-specific pattern, one could envision that the primary tumor is already â&#x20AC;&#x2DC;primedâ&#x20AC;&#x2122; for an eventual site of dissemination. This should however be validated in a larger cohort. In theory, tumor heterogeneity could explain some of the observed differences between primary tumors and paired metastases. However, the reported distinct APOBEC3B mRNA expression levels of the primary tumor that were largely retained or increased in the paired metastases could not solely be explained by heterogeneity. In daily practice, the majority of metastases are not resected or biopsied. This likely resulted in a selection bias, since we only included primary tumors with available material of the paired metastasis. Another weakness of our study is the relatively small number of patients with distant metastases, which limited the reliability of subgroup analysis according to metastatic site. In conclusion, our findings add to the knowledge that APOBEC3B contributes to breast cancer progression and has now extended this to metastatic disease. Since APOBEC3B expression is at least retained and often even increased in distant metastases, our data suggest that it might also be an effective interventional candidate for disseminated breast cancer. Acknowledgements This study was supported by Cancer Genomics Netherlands (AMS and JWMM), Netherlands Organization of Scientific Research (NWO) (JWMM) and Dutch Cancer Society grant UU 2011-5195 and Philips Consumer Lifestyle (WAMES).

APOBEC3B expression in breast cancer metastases

REFERENCES 1. Ferlay J, Soerjomataram I, Dikshit R, et al. Cancer incidence and mortality worldwide: Sources, methods and major patterns in GLOBOCAN 2012. Int J Cancer. 2015;136(5):E359-86. doi: 10.1002/ijc.29210 [doi]. 2. Gupta GP, Massague J. Cancer metastasis: Building a framework. Cell. 2006;127(4):679-695. doi: S00928674(06)01414-0 [pii]. 3. Fidler IJ. The pathogenesis of cancer metastasis: The â&#x20AC;&#x2DC;seed and soilâ&#x20AC;&#x2122; hypothesis revisited. Nat Rev Cancer. 2003;3(6):453-458. doi: 10.1038/nrc1098 [doi]. 4. Moelans CB, van der Groep P, Hoefnagel LD, et al. Genomic evolution from primary breast carcinoma to distant metastasis: Few copy number changes of breast cancer related genes. Cancer Lett. 2014;344(1):138-146. doi: 10.1016/j.canlet.2013.10.025 [doi]. 5. Harris RS. Molecular mechanism and clinical impact of APOBEC3B-catalyzed mutagenesis in breast cancer. Breast Cancer Res. 2015;17:8-014-0498-3. doi: 10.1186/s13058014-0498-3 [doi]. 6. Burns MB, Lackey L, Carpenter MA, et al. APOBEC3B is an enzymatic source of mutation in breast cancer. Nature. 2013;494(7437):366-370. doi: 10.1038/nature11881 [doi]. 7. Burns MB, Temiz NA, Harris RS. Evidence for APOBEC3B mutagenesis in multiple human cancers. Nat Genet. 2013;45(9):977-983. doi: 10.1038/ng.2701 [doi]. 8. Kuong KJ, Loeb LA. APOBEC3B mutagenesis in cancer. Nat Genet. 2013;45(9):964-965. doi: 10.1038/ng.2736 [doi]. 9. Leonard B, Hart SN, Burns MB, et al. APOBEC3B upregulation and genomic mutation patterns in serous ovarian carcinoma. Cancer Res. 2013;73(24):7222-7231. doi: 10.1158/0008-5472.CAN-13-1753 [doi]. 10. Nowarski R, Kotler M. APOBEC3 cytidine deaminases in double-strand DNA break repair and cancer promotion. Cancer Res. 2013;73(12):3494-3498. doi: 10.1158/00085472.CAN-13-0728 [doi]. 11. Roberts SA, Lawrence MS, Klimczak LJ, et al. An APOBEC cytidine deaminase mutagenesis pattern is widespread in human cancers. Nat Genet. 2013;45(9):970-976. doi: 10.1038/ng.2702 [doi]. 12. Xuan D, Li G, Cai Q, et al. APOBEC3 deletion polymorphism is associated with breast cancer risk among women of european ancestry. Carcinogenesis. 2013;34(10):2240-2243. doi: 10.1093/carcin/bgt185 [doi]. 13. Gwak M, Choi YJ, Yoo NJ, Lee S. Expression of DNA cytosine deaminase APOBEC3 proteins, a potential source for producing mutations, in gastric, colorectal and prostate cancers. Tumori. 2014;100(4):112e-7e. doi: 10.1700/1636.17922 [doi]. 14. Jin Z, Han YX, Han XR. The role of APOBEC3B in chondrosarcoma. Oncol Rep. 2014;32(5):1867-1872. doi: 10.3892/or.2014.3437 [doi]. 15. Burns MB, Leonard B, Harris RS. APOBEC3B: Pathological consequences of an innate immune DNA mutator. Biomed J. 2015;38(2):102-110. doi: 10.4103/23194170.148904 [doi]. 16. Chan K, Roberts SA, Klimczak LJ, et al. An APOBEC3A hypermutation signature is distinguishable from the signature of background mutagenesis by APOBEC3B in human cancers. Nat Genet. 2015;47(9):1067-1072. doi: 10.1038/ng.3378 [doi]. 17. Zhang J, Wei W, Jin HC, Ying RC, Zhu AK, Zhang FJ. The roles of APOBEC3B in gastric cancer. Int J Clin Exp Pathol. 2015;8(5):5089-5096.

18. Kosumi K, Baba Y, Ishimoto T, et al. APOBEC3B is an enzymatic source of molecular alterations in esophageal squamous cell carcinoma. Med Oncol. 2016;33(3):26-0160739-7. Epub 2016 Feb 15. doi: 10.1007/s12032-0160739-7 [doi]. 19. Morganella S, Alexandrov LB, Glodzik D, et al. The topography of mutational processes in breast cancer genomes. Nat Commun. 2016;7:11383. doi: 10.1038/ ncomms11383 [doi]. 20. Henderson S, Chakravarthy A, Su X, Boshoff C, Fenton TR. APOBEC-mediated cytosine deamination links PIK3CA helical domain mutations to human papillomavirus-driven tumor development. Cell Rep. 2014;7(6):1833-1841. doi: 10.1016/j.celrep.2014.05.012 [doi]. 21. Periyasamy M, Patel H, Lai CF, et al. APOBEC3B-mediated cytidine deamination is required for estrogen receptor action in breast cancer. Cell Rep. 2015;13(1):108-121. doi: 10.1016/j.celrep.2015.08.066 [doi]. 22. Rebhandl S, Huemer M, Greil R, Geisberger R. AID/ APOBEC deaminases and cancer. Oncoscience. 2015;2(4):320-333. doi: 155 [pii]. 23. Sieuwerts AM, Willis S, Burns MB, et al. Elevated APOBEC3B correlates with poor outcomes for estrogenreceptor-positive breast cancers. Horm Cancer. 2014;5(6):405-413. doi: 10.1007/s12672-014-0196-8 [doi]. 24. Land AM, Wang J, Law EK, et al. Degradation of the cancer genomic DNA deaminase APOBEC3B by SIV vif. Oncotarget. 2015;6(37):39969-39979. doi: 10.18632/ oncotarget.5483 [doi]. 25. van Diest PJ. No consent should be needed for using leftover body material for scientific purposes. for. BMJ. 2002;325(7365):648-651. 26. Tsuboi M, Yamane A, Horiguchi J, et al. APOBEC3B high expression status is associated with aggressive phenotype in japanese breast cancers. Breast Cancer. 2016;23(5):780788. doi: 10.1007/s12282-015-0641-8 [doi]. 27. Zhang Y, Delahanty R, Guo X, Zheng W, Long J. Integrative genomic analysis reveals functional diversification of APOBEC gene family in breast cancer. Hum Genomics. 2015;9:34-015-0056-9. doi: 10.1186/s40246-015-0056-9 [doi]. 28. Hoefnagel LD, van de Vijver MJ, van Slooten HJ, et al. Receptor conversion in distant breast cancer metastases. Breast Cancer Res. 2010;12(5):R75. doi: 10.1186/bcr2645 [doi]. 29. Hoefnagel LD, Moelans CB, Meijer SL, et al. Prognostic value of estrogen receptor alpha and progesterone receptor conversion in distant breast cancer metastases. Cancer. 2012;118(20):4929-4935. doi: 10.1002/cncr.27518 [doi]. 30. Hoefnagel LD, van der Groep P, van de Vijver MJ, et al. Discordance in ERalpha, PR and HER2 receptor status across different distant breast cancer metastases within the same patient. Ann Oncol. 2013;24(12):3017-3023. doi: 10.1093/annonc/mdt390 [doi]. 31. Schrijver WA, Jiwa LS, van Diest PJ, Moelans CB. Promoter hypermethylation profiling of distant breast cancer metastases. Breast Cancer Res Treat. 2015;151(1):41-55. doi: 10.1007/s10549-015-3362-y [doi]. 32. Jiwa LS, van Diest PJ, Hoefnagel LD, et al. Upregulation of claudin-4, CAIX and GLUT-1 in distant breast cancer metastases. BMC Cancer. 2014;14:864-2407-14-864. doi: 10.1186/1471-2407-14-864 [doi]. 33. Cescon DW, Haibe-Kains B, Mak TW. APOBEC3B expression in breast cancer reflects cellular proliferation, while a deletion polymorphism is associated with immune activation. Proc Natl Acad Sci U S A. 2015;112(9):28412846. doi: 10.1073/pnas.1424869112 [doi].

PART ONE | CHAPTER 3

SUPPLEMENTAL Supplementary methods Quality and quantity control measurements for reliable quantitative reverse transcriptase polymerase chain reaction (RT-qPCR) For reliable RT-qPCR measurements, only samples that resulted in amplifiable products within 25 cycles for the used reference gene set at an input of 50 ng total RNA (92.9% of the samples) were considered to be of good quality to reliably determine RT-qPCR levels. Furthermore, a serially diluted FFPE breast tumor sample was included in each experiment to evaluate the linear amplification and efficiencies for all genes included in the panel and absence of amplification in the absence of reverse transcriptase. All gene transcripts were 100% efficient amplified (range 94%-102%) and were negative in the absence of reverse transcriptase. Estrogen receptor (ER/ESR1) and receptor tyrosine-protein kinase erbB-2 status/human epidermal growth factor 2 (HER2/ERBB2) status Because data regarding ER and HER2 expression on protein level of our data set was incomplete, ESR1 and ERBB2 mRNA expression was used to determine ESR1 and ERBB2 mRNA status (using a cut-off dCq for ESR1>1 and ERBB2>3.5 by optimal binning for n=92 and n=87 overlapping samples, respectively (Supplementary Figure S1)). Because ER and HER2 are determined on protein level in daily clinical practice (using a scoring system according to national and international guidelines 1,2), we investigated whether the ESR1 and ERBB2 mRNA status accurately reflected the ER and HER2 protein status as reported in the pathology reports in samples with known receptor protein status. These cut-offs resulted in a sensitivity of 0.88 and specificity of 0.85 for ESR1 and in a sensitivity of 0.89 and specificity of 0.97 for ERBB2. APOBEC3B mRNA expression versus conversion of ESR1 status Overall, APOBEC3B levels in distant metastases of ESR1-positive primary tumors showed the biggest increase when compared to their primary counterpart. For 13 of our cases we observed a conversion of ESR1 status; 8 cases (all with distant metastases) converted from an ESR1-positive primary tumor to an ESR1-negative distant metastasis and 5 cases (all with loco-regional lymph node metastases) from an ESR1-negative primary tumor to an ESR1-positive loco-regional metastasis, with for both conversion types an increased APOBEC3B mRNA level in the metastasis (Supplementary Table S2). As for both the converted and not converted cohorts APOBEC3B mRNA levels were higher in the metastasis, receptor conversion could thus not explain this finding of higher APOBEC3B levels in the metastases of ESR1-positive primary tumors. 70

APOBEC3B expression in breast cancer metastases

Supplementary Figure S1. Correlation of ER and HER2 protein status with ESR1 and ERBB2 mRNA levels. Arrows indicate used cut-off value.

Supplementary Table S1. APOBEC3B mRNA expression according to metastatic site.

Mean

PTPRC (CD45) mRNA log2

Mean

AVG epithelial mRNA log2

APOBEC3B mRNA log2

Mean

No of patients*

Percentage of patients

Tissue origin

Primary tumor

100%

-6.51

2.53

-3.04

1.07

-3.98

3.23

Paired Metastasis

100%

-5.36

2.39

-2.85

1.08

-4.48

2.84

Clinical parameter

0.0015

0.20

0.26

According metastatic site

Ovary

Primary tumor

-9.25

1.39

-3.31

1.25

-5.53

5.25

Paired Metastasis

-5.51

0.74

-3.03

1.20

-3.99

0.93

P†

0.07

0.72

0.47

GI tract

Primary tumor

-7.79

3.84

-3.32

0.20

-1.83

1.10

Paired Metastasis

-6.96

0.54

-3.93

1.88

-4.41

4.88

0.59

0.11

Liver

Primary tumor

11%

-7.59

2.65

-3.16

1.57

-2.66

1.35

Paired Metastasis

11%

-5.10

2.45

-2.20

1.17

-4.76

2.71

†

0.08

0.17

0.028

Loco-regional lymph node

Primary tumor

36%

-6.87

2.55

-2.82

0.91

-3.88

2.89

Paired Metastasis

36%

-6.12

2.68

-2.63

0.89

-3.42

3.03

P†

0.23

0.31

0.49

†

PART ONE | CHAPTER 3

Supplementary Table S1. Continued

Mean

PTPRC (CD45) mRNA log2

Mean

AVG epithelial mRNA log2

APOBEC3B mRNA log2

Mean

Percentage of patients

No of patients*

Bone

Primary tumor

-5.82

1.23

-3.01

1.28

-5.04

3.31

Paired Metastasis

-4.98

3.85

-2.63

0.96

-4.11

3.97

P†

0.47

0.27

0.47

Brain

Primary tumor

25%

-5.29

2.06

-3.10

1.03

-5.53

3.58

Paired Metastasis

25%

-4.46

2.00

-2.95

0.81

-5.67

2.82

P†

0.20

0.68

0.98

Lung

Primary tumor

-4.32

1.45

-3.29

1.69

-2.35

1.59

Paired Metastasis

-4.21

1.49

-3.82

1.47

-3.76

1.66

P†

0.72

0.47

0.14

Clinical parameter

* Due to missing values numbers do not always add up to 55. SD; standard deviation. AVG epithelial; average mRNA level of KRT19 and EPCAM. † Paired Wilcoxon Signed Ranks Test significance (2-tailed).

dCq APOBEC3B Primary (Average)

dCq APOBEC3B Metastasis (Average)

P-value*

loco-regional lymph node

-6.40

-5.97

0.61

0.032

Not converted

Metastasis type

ESR1 conversion primary to metastasis

Supplementary Table S2. ESR1 conversion from primary tumor to metastasis specified by site of metastasis.

distant metastasis

-5.86

-4.87

0.015

ESR1- primary to ESR1+ metastasis

loco-regional lymph node

-8.26

-6.57

0.11

ESR1+ primary to ESR1- metastasis

distant metastasis

-7.80

-5.16

0.07

* Wilcoxon Signed Ranks Test

0.041

APOBEC3B expression in breast cancer metastases

SUPPLEMENTAL REFERENCES 1. Wolff AC, Hammond ME, Hicks DG, et al. Recommendations for human epidermal growth factor receptor 2 testing in breast cancer: American society of clinical oncology/college of american pathologists clinical practice guideline update. Arch Pathol Lab Med. 2014;138(2):241-256. doi: 10.5858/arpa.2013-0953-SA [doi]. 2. NABON. Breast cancer dutch guideline, version 2.0. 2012.

Chapter 4

Willemijne AME Schrijver, Laura S Jiwa, Paul J van Diest and Cathy B Moelans

Promoter hypermethylation profiling of distant breast cancer metastases

Breast Cancer Res Treat. 2015;151(1):41-55

PART ONE | CHAPTER 4

ABSTRACT Promoter hypermethylation of tumor suppressor genes seems to be an early event in breast carcinogenesis and is, potentially reversible. This makes methylation a possible therapeutic target, a marker for treatment response and/or a prognostic factor. Methylation status of 40 tumor suppressor genes was compared between 53 primary breast tumors and their corresponding metastases to brain, lung, liver and skin. In paired analysis, a significant decrease in methylation values was seen in distant metastases compared to their primaries in 21/40 individual tumor suppressor genes. Further, primary tumors that metastasized to the liver clustered together, in line with the finding that primary breast carcinomas that metastasized to the brain, skin and lung showed higher methylation values in up to 27.5% of tumor suppressor genes than primary carcinomas that metastasized to the liver. Conversion in methylation status of several genes from the primary tumor to the metastasis had prognostic value, and methylation status of several genes in the metastases predicted survival after onset of metastases. Methylation levels for most of the analyzed tumor suppressor genes were lower in distant metastases compared to their primaries, pointing to the dynamic aspect of methylation of these tumor suppressor genes during cancer progression. Also, specific distant metastatic sites seem to show differences in methylation patterns, implying that hypermethylation profiles of the primaries may steer site-specific metastatic spread. Lastly, methylation status of the metastases seems to have prognostic value. These promising findings warrant further validation in larger patient cohorts and more tumor suppressor genes.

Hypermethylation in breast cancer metastases

INTRODUCTION With 1.7 million new cases causing 522,000 deaths worldwide per year, breast cancer is the leading cause of female cancer death 1. Early detection, optimal surgery and adjuvant therapy are the key strategies to improve prognosis. Although 5-year overall survival increased from 77% in the period 1978 to 1984 to 82% in the period 1995 to 2003, about 16% of patients will develop distant metastases and eventually die of the disease 2. Preferred site of distant metastases strongly depends on the subtype of breast cancer. Lobular type breast cancer preferentially metastasizes to bone, GI tract and ovaries, triple negative breast cancer to liver and brain, and luminal breast cancer to the bone and skin, while well circulated organs like the spleen and heart almost never harbor metastases 3-5. This “organotropism” was first described by Paget et al. about a century ago as the “seed and soil” analogy, where tumors are supposed to have a “seminal influence” on the metastatic microenvironment, and thereby act together with the distant organ to effect tumor metastases 6. The identity of these seminal influences remains elusive. Both genetic and epigenetic changes may play a role here. Epigenetic alterations are of pivotal interest since they cannot only influence tumor behaviour, but may also become important therapeutic targets as these processes are potentially reversible. Therapies that target DNA methylation (DNA methyl-transferase (DNMT) inhibitors) or histone modification (histone deacetylase (HDAC) inhibitors) already exist, but newer versions of these drugs need to be developed to improve future clinical management 7. Which mechanisms underlie development of distant metastases remains a topic for debate. The two main but not necessarily mutually exclusive hypotheses are the linear and the parallel model of metastasis. According to the linear model, genetic modifications progressively accumulate in cancer cells of the primary tumor, whereby cells with advantageous mutations will survive and expand through clonal evolution 8. If we translate this into epigenetic alterations such as promoter hypermethylation, one would expect that tumor suppressor genes in metastases show more methylation than primary carcinomas. An increase in methylation values during local tumor progression has already been shown9,10. In the parallel progression model cancer cells disseminate early during tumor progression at a stage when the primary lesion is small. Disseminated cells then evolve independently of the primary tumor to form metastases. According to this latter model, one would expect different methylation patterns in primaries and their matched metastases. Hypermethylation of tumor suppressor genes like APC, RASSF1A and FEZ1/LZTS1 in primary breast cancer has been reported to correlate with development of distant metastases11,12. However, little is known about the comparative methylation status of primary tumors and matched distant metastases, possibly related to the fact that metastatic material is rare. Rivenbark et al. compared the methylation status of CST6 in primary breast cancers to their lymph node metastases and showed that methylation-dependent silencing occurred 77

PART ONE | CHAPTER 4

more frequently in the lymph node metastases, possibly reflecting progression-related epigenetic events according to the linear model for metastasis 13. Here we report promoter hypermethylation profiling for 42 tumor suppressor genes by methylation specific multiplex ligation-dependent probe amplification (MS-MLPA) in 53 primary breast carcinomas and their matched non-bone distant metastases (skin, brain, lung and liver). This study is part of a project where we study genotype and phenotype of distant breast cancer metastases 14-16. Extensive knowledge of the hypermethylation status of tumor suppressor genes possibly involved in site-specific metastasis could lead to novel biomarkers predicting site of distant metastases and adjuvant targeted therapy strategies that could prevent such metastases from becoming clinically manifest.

MATERIALS AND METHODS Patients This study was performed on 53 formalin-fixed paraffin embedded (FFPE) samples of female primary breast carcinomas and 53 single corresponding metachronous non-bone distant metastases. The samples were selected randomly from an existing database entailing material from 300 patients from the departments of pathology of the University Medical Center Utrecht, the Meander Medical Center Amersfoort, the Deventer Hospital, the Rijnstate Hospital Arnhem, Tergooi Hospitals, the Academic Medical Center Amsterdam, the Radboud University Nijmegen Medical Center, the Canisius Wilhelmina Hospital Nijmegen, the Netherlands Cancer Institute Amsterdam, the Medical Center Alkmaar, the Medical Center Zaandam, the University Medical Center Groningen, the St. Antonius Hospital Nieuwegein, the Diakonessenhuis Utrecht, the Free University Medical Center Amsterdam, the Erasmus Medical Center Rotterdam, the Gelre hospital Apeldoorn, Isala clinics Zwolle, the Laboratory for Pathology Enschede, the Laboratory for Pathology Dordrecht and the Laboratory for Pathology Foundation Sazinon Hoogeveen, all in The Netherlands. This study was performed in accordance with the institutional medical ethical guidelines. The use of anonymous or coded left over material for scientific purposes is part of the standard treatment agreement with patients and therefore informed consent was not required according to Dutch law 17. Molecular subtypes of breast tumors were assigned as follows: Luminal A (ER+/PR+, HER2â&#x2C6;&#x2019;, low cellular proliferation), luminal B (ER+/PR+, HER2â&#x2C6;&#x2019;, low cellular proliferation or ER+/PR+, HER2+), triple negative or basal type (ER-/PR-, HER2-) and HER2 enriched (ER-/PR-, HER2+) as before 3. To set methylation cut-off values, non-paired normal breast tissue (n=25) was used from breast reduction specimens (mean age 39.4 years; n=15) and autopsy specimens (mean age 78

Hypermethylation in breast cancer metastases

48.9 years; n=10), with no significant difference in age compared to breast cancer patients (p=0.338). In addition, we analyzed normal non-paired tissue from brain (n=5), lung (n=5), liver (n=5) and skin (n=5) derived from our normal tissue biobank to exclude that methylation values in distant metastases would be influenced by admixture of normal surrounding tissue, with again no significant different in age (45.8 years) compared to patients with breast cancer (p=0.111). The mean patient age at diagnosis was 52.8 years and 84% of patients presented with invasive ductal carcinoma. Follow-up ranged between sixteen and 315 months and metastases were meanly diagnosed 55.4 months after the primary diagnosis. The localization of the metastases that were included was brain (n=11), lung (n=12), liver (n=10) and skin (n=20). Clinicopathological characteristics are shown in Table 1.

Table 1. Clinicopathological characteristics of the metastatic breast cancer patients (n=53) analyzed for methylation status of 40 tumor suppressor genes with MS-MLPA. Feature

Grouping

N or value

Age at diagnosis (in years)

Mean Range

52.8 27-88

Tumor size (in cm)

≤2 >2 and ≤5 >5 Not available

16 26 6 5

30 49 11 10

Histologic type

Invasive ductal Invasive lobular Metaplastic Micropapillary

45 4 3 1

84 8 6 2

Histologic grade (Bloom & Richardson)

I II III

1 12 40

2 22 76

MAI (per 2mm²)

Mean Range ≤12 ≥13

24.8 0-86 14 39

26 74

Lymph node status

Positive Negative Not available

25 24 4

47 45 8

Site of distant metastasis

Brain Lung Liver Skin

11 12 10 20

21 22 19 38

Molecular subtype

Luminal A Luminal B Triple negative HER2 enriched

11 28 12 2

21 53 22 4

Follow-up in months

Mean Range

94 16-315

PART ONE | CHAPTER 4

Table 1. Continued Feature

Grouping

N or value

Time between diagnosis of primary and metastasis (in months)

Mean Range

55.4 0.4-180.8

Time between diagnosis of metastasis and death (in months)

Mean Range

26.6 2.0-177.7

Treatment before resection of metastasis (adjuvant to surgery of primary breast tumor)

Chemotherapy Hormonal therapy Radiotherapy Combination of chemo-, hormonal and/or radiotherapy. Not available

19 17 26 22

36 32 49 42

Primary

Metastasis

+ -

35 18

66 34

Primary

Metastasis

+ -

22 31

42 58

Primary

0 1+ 2+ 3+

41 4 1 7

77 8 2 13

Metastasis

0 1+ 2+ 3+

38 5 4 6

72 9 8 11

ER-status*

PR-status*

HER2-status^

MAI: mitotic activity index. *: according to 10% threshold for positivity. ^: according to DAKO-scoring system

DNA extraction Four-micrometer sections were cut from each FFPE tissue block and stained with haematoxylin and eosin (HE). The HE-section was used to guide micro-dissection for DNA extraction and to estimate tumor percentage. Only samples containing 80 per cent tumor load or higher (both primary tumor and metastasis) were selected. For proteinase K-based DNA extraction, five 5-Âľm-thick slides were cut, and tumor areas were micro-dissected using a scalpel. Areas with necrosis, dense lymphocytic infiltrates, and pre-invasive lesions were intentionally avoided. The DNA concentration and absorbance at 260 and 280 nm were measured with a spectrophotometer (Nanodrop ND-1000, Thermo Scientific Wilmington, USA).

Hypermethylation in breast cancer metastases

MS-MLPA MS-MLPA was performed according to the manufacturerâ&#x20AC;&#x2122;s protocol using the SALSA MSMLPA probemixes ME001-C2 Tumor suppressor-1 and ME003-A1 Tumor suppressor-3 (Supplementary Table S1 and S2), each containing 15 internal control probes and in total 53 HhaI-sensitive probes against the following tumor suppressor genes: TP73, CASP8, VHL, RARB, MLH1 (2 loci), RASSF1A (2 loci), FHIT, APC, ESR1, CDKN2A/B, DAPK1, KLLN, CD44, GSTP1, ATM, CADM1, CDKN1B, CHFR, BRCA1/2, CDH13, HIC1, TIMP3 (2 loci), RDM2, RUNX3, HLTF (2 loci), SCGB3A1 (2 loci), ID4 (2 loci), TWIST1, SFR4 (2 loci), DLC1 (2 loci), SFR5 (2 loci), BNI3, H2AFX (2 loci), CCND2 (2 loci), CACNA1G, TGIF1, BCL2 and CACNA1A. Since MS-MLPA is based on the methylation-sensitive restriction enzyme HhaI, the choice of CpG site to be evaluated within the promoter region is highly dependent on the presence of the GCGC restriction site, and not so much based on correlation to expression in literature. At least 50 ng of DNA was used in each MS-MLPA reaction. DNA concentration control fragments, present in each MS-MLPA mix, were evaluated to check for sufficient DNA quantity. All reactions were performed according to the manufacturerâ&#x20AC;&#x2122;s instructions in a Veriti 96 Well Thermo Cycler (Applied Biosystems). A water sample, a 100% methylated (MCF-7 M.SssI methyltransferase treated) control and a negative control (human sperm DNA) were taken along in every MLPA run. Fragment separation was done by capillary electrophoresis on an ABI-3730 capillary sequencer (Applied Biosystems). Peak patterns derived by Genescan Analysis were evaluated using Genemapper (version 4.1) and Coffalyser. net software (version 9.4, MRC-Holland, Amsterdam, the Netherlands). The cumulative methylation index (CMI) was calculated as the sum of all quantitative methylation values per tumor. Raw methylation percentages of all genes are available from the authors. Correlation between mRNA expression and promoter methylation by TCGA To correlate methylation of the investigated tumor suppressor genes to mRNA expression, we used The Cancer Genome Atlas (https://tcga-data.nci.nih.gov/tcga/). TCGA Breast Invasive Carcinoma mRNA Expression z-Scores (RNA Seq V2 RSEM) data (n=1038) were downloaded via The cBioPortal for Cancer Genomics 18,19. Illumina Infinium Human DNA Methylation 27 level 3 data (calculated beta values (M/M+U), gene symbols, chromosomes and genomic coordinates) were downloaded via TCGA Data Portal (n=313). Statistical analyses were performed on data of all available CpG sites of the TCGA database compared to the CpG sites used for MS-MLPA. Statistics Unsupervised hierarchical clustering of log-transformed quantitative methylation values was performed using non-parametric Spearman correlation with R software (version 3.0.1), including all cases that were tested with both MLPA probemixes. Statistical analysis was 81

PART ONE | CHAPTER 4

executed on absolute methylation percentages as well as on dichotomized values, the latter were determined by ROC curve analyses of methylation values in normal breast tissue compared to primary breast tumor tissue. The Kolmogorov-Smirnov and Shapiro-Wilk test were used to test for normality of the distributions. Primary tumors and their paired metastases were compared per gene using the Wilcoxon signed-rank test. Non-paired analyses on patient differences and clinicopathological characteristics were computed using the Mann-Whitney test. The dichotomized values were analysed using McNemars test or chi-square test. Two-sided p-values < 0.05 were considered to be statistically significant. Correction for multiple comparisons was performed by the Bonferroni-Holm approach. Analysis of prognosis was performed using Kaplan Meier survival curves/logrank test for univariate analyses and Cox proportional hazard analysis for multivariate models (entry and remove limits 0.05), calculating hazard ratios (HR) with 95% confidence intervals (CI). TCGA mRNA z-scores were compared to percentages of DNA methylation by Pearsonâ&#x20AC;&#x2122;s r correlation. To evaluate whether site of distant metastasis is determined by specific methylation patterns of the primary tumors or rather by inherent molecular subtype, we performed logistic regression comparing the different metastatic sites one by one with quantitative methylation status of individual genes and molecular subtype as variables in the model. To evaluate whether adjuvant systemic treatment may influence conversion from low methylation in the primary to high methylation in the distant metastasis (or vice versa), we grouped patients according to conversion per individual gene and performed logistic regression for each individual gene including adjuvant chemotherapy (yes or no) and adjuvant hormonal therapy (yes or no) as variables in the model. All statistical calculations were done with IBM SPSS Statistics 21.

RESULTS Normal versus tumor tissue Appropriate cut-offs to dichotomize methylation values of tumor suppressor genes, derived from ROC curve analysis of MS-MLPA values in normal breast vs. primary breast tumor tissue, varied between 0.5 and 22.75% for the 40 genes (53 loci; Supplementary Table S7). Although we only included samples of breast cancer metastases that contained 80 per cent tumor load or higher, we wanted to further exclude that differences between primaries and metastases were due to the admixture of tumor micro-environment at distant sites. 17/40 genes showed significantly higher methylation values in normal lung, brain or liver than in normal breast (Supplementary Table S3; Figure 1a shows CASP8 as an example). Also the CMI values of normal liver and brain tissue were significantly higher than the CMI of normal lung, skin and breast tissue (Figure 1b). 82

Hypermethylation in breast cancer metastases

1000

600 400

B ra i

B re

as t

Sk in

200

Lu ng

Location normal tissue

brain: n=5 liver: n=5 lung: n=5 skin: n=5 breast: n=10

CMI

Li ve r

p < 0.01

800

B re a

Li ve r

p < 0.001

p < 0.01 p < 0.01 p < 0.01 p < 0.01

Sk in

p < 0.001

CMI in normal tissue

Lu ng

CASP8 in normal tissue

100

B ra in

Percentage methylation

Location normal tissue

Figure 1

Figure 1. Differences in quantitative methylation percentages of CASP8 (a) and the CMI (b) by MS-MLPA between various normal tissues. N=30 (brain n=5, liver n=5, lung n=5, skin n=5 and breast n=10). Small horizontal lines depict the median per group. The grey horizontal line depicts the cut-off for hypermethylation (4.5%). Chromosome location CASP8: chr2 (202122754-202152434), CpG site MS-MLPA probe: 202122649, #bp from probe to TSS: 104 and from probe to ATG: 104.

Unsupervised hierarchical clustering of the quantitative methylation values of primary breast tumors, paired distant metastases and normal tissues is shown in Figure 2. Normal liver and brain tissue seems to cluster together due to hypermethylation of some genes (APC, CDKN2B, CCND2 both loci, RASSF1A both loci and CASP8) as already mentioned above, and normal breast, lung and skin tissue showed a related pattern. Primary tumor versus metastasis Using quantitative methylation values, 52.5% (21/40) of genes were significantly less methylated in the metastases compared to their paired primary tumors : PRDM2 (p=0.036), RARB-2 (p=0.003), HLTF-2 (p=0.013), H2AFX-1 (p=0.001), CACNA1G (p=0.000), TGIF1 (p=0.029), TIMP3-1 (p=0.046), TP73 (p=0.019), FHIT (p=0.002), APC (p=0.048), CDKN2A (p=0.002), CDKN2B (p=0.012), PTEN (p=0.002), CD44 (p=0.011), ATM (p=0.000), CADM1 (p=0.006), CHFR (p=0.005), BRCA2 (p=0.001), HIC1 (p=0.001) and BRCA1 (p=0.002). After correction for multiple comparisons, H2AFX-1, CACNA1G, ATM, BRCA2 and HIC1 remained significant. CMI was not significantly different between primaries and metastases (p=0.454). Figure 3a shows quantitative methylation values of CACNA1G in primary tumors and their distant metastases as an example. Using dichotomized values, 55% (22/40) of the tested tumor suppressor genes, namely PRDM2 (p=0.049), RARB-1 (p=0.002), HLTF-2 (p=0.031), TWIST1 (p=0.012), H2AFX both loci (p=0.002 and p=0.049), CACNA1G (p=0.013), TGIF1 (p=0.002), TIMP3-3 (p=0.013), TP73 (p=0.007), FHIT (p=0.001), CDKN2A (p=0.029), DAPK1 (p=0.004), PTEN (p=0.008), CD44 (p=0.000), GSTP1 (p=0.013), ATM (p=0.000), CADM1 (p=0.000), CHFR 83

Value

TYPE

LOCATION

lung626 M10 P10 P15 P7 M7 M23 P23 M20 M27 P11 P27 M11 P59 M59 P9 M9 P1 M1 M15 M22 P24 P22 P20 M56 P19 M19 M49 P17 M17 P58 M58 M52 M5 P5 P52 P35 M35 P56 M32 P32 P44 P54 P48 M48 P21 M21 P18 M18 M38 breast107 P14 M14 P28 M44 M54 M47 P47 M57 P45 M45 P6 M6 P13 M13 P16 M16 P33 M33 P30 M30 M60 P43 M43 M24 M28 P49 P51 P60 P50 M51 P39 M39 P55 P57 M50 M55 P2 M2 P42 M42 P34 M34 M4 P4 P36 M36 P31 M31 P40 P38 M40 P37 M3 M37 P3 P41 M41 breast1245 skin1246 breast1028 breast6127 breast1018 skin57 breast821 breast114 breast1222 breast1039 breast1137 skin1014 skin4196 skin97 liver1043 liver636 liver1019 liver999 liver1296 brain681 brain788 brain749 brain758 brain766 lung6101 lung522 lung119 lung124

CHFR RARB_2 CDKN2B APC MLH1_1 CASP8 MLH1_2 TIMP3_3 HIC1 CDKN2A PTEN ATM BRCA1 CADM1 CDKN1B BRCA2 TP73 CD44 VHL HLTF_1 RDM2 ESR1 FHIT BCL2 H2AFX_2 RARB_1 TGIF DLC1_2 DLC1_1 TWIST1 CACNA1G BNIP3 TIMP3_2 H2AFX_1 TIMP3_1 HLTF_2 CDH13 CACNA1A CCND2_2 CCND2_1 SFRP4_1 ID4_2 SFRP4_2 RUNX3 DAPK1 SFRP5_1 SFRP5_2 ID4_1 SCGB3A1_2 SCGB3A1_1 RASSF1_2 RASSF1_1 GSTP1

brain lung liver skin breast

Primary Metastasis Normal

Figure 2. Unsupervised hierarchical clustering analysis of log transformed quantitative methylation percentages of 40 tumor suppressor genes (53 loci) in 53 primary breast tumors, 53 paired distant metastases and 30 normal tissues (breast n=10, brain n=5, lung n=5, liver n=5, skin n=5). The sidebars depict location of tissue and type (primary, metastasis or normal tissue).

Color Key

PART ONE | CHAPTER 4

Hypermethylation in breast cancer metastases

(p=0.031), BRCA2 (p=0.013), HIC1 (p=0.016) and BRCA1 (p=0.000) were significantly less methylated in the metastases than in the primaries. After correction for multiple comparisons, FHIT, CD44, ATM, CADM1 and BRCA1 stayed significant. PRDM2, HLTF-2, H2AFX-1, CACNA1G, TGIF1, TP73, FHIT, CDKN2A, PTEN, CD44, ATM, CADM1, CHFR, BRCA2, HIC1 and BRCA1 were significant in both quantitative and dichotomized analyses. Of these, PRDM2, H2AFX-1, TGIF1, TP73, CDKN2A and CD44 were more methylated in normal brain and/or liver tissues than in normal breast, which indicates that the generally lower methylation values in the distant metastases must be tumor cell specific and excludes the potential admixture of cells from the distant microenvironment being a confounder here. When comparing primaries and metastases for all investigated tumor suppressor genes per individual patient, significantly less methylation was seen in the metastases compared to the primary tumor in 30.2% (16/53; quantitative) or 41.5% (22/53; dichotomized) of patients (20.8 and 28.3% after correction for multiple comparisons, respectively). Only 15.1% (8/53; quantitative) or 3.8% (2/53; dichotomized) of patients showed significantly more methylation in the metastasis compared to the primary tumor (3.8 or 1.9%, respectively, if corrected for multiple comparisons). These higher methylation values cannot be explained by admixture of normal adjacent tissue in the metastases, since none of these patients had a metastasis in brain or liver, where high methylation values are found in normal tissue. In cluster analysis (Figure 2) 32/53 pairs of primaries and metastases clustered directly and another 9/53 pairs almost directly (within three positions), indicating that methylation patterns of the tested tumor suppressor genes show high patient specificity. Molecular subtype HER2 enriched tumors were excluded from statistical analyses, because of the small number. Triple negative tumors tended to cluster together, but the difference between luminal A and B was less distinct (Figure 4). PRDM2, RARB, CACNA1G (Figure 3b), SFRP4-2, H2AFX, CACNA1A, TIMP3-1/2 and DLC1-1 showed significantly less methylation in luminal A primary tumors compared to luminal B and/or triple negative primary tumors. Less methylation of SCGB3A1 was seen in triple negative tumors compared to the other subtypes. Further, more methylation of ID4-2 was seen in luminal B tumors compared to the other subtypes. When corrected for metastatic site these effects disappeared (Figure 3c), indicating that, although subgroups were small, molecular subtype is not a significant determinant of dissemination site in this group (Supplementary Table S4). No differences were seen between the CMI of the different molecular subtypes (p=0.199) (Supplementary Table S5). Concerning receptor status, 35% (14/40; quantitative) or 25% (10/40; dichotomized) of the tumor suppressor genes showed significantly higher methylation values in ER-positive tumors compared to ER-negative tumors. After correction for multiple comparisons, 85

CACNA1G in Primaries vs. Metastases

p < 0.01

50 40

n=53

30 20 10

ta se s

Figure 3

Sk in

Li ve r

Lu ng

B ra

he d ric en

H ER 2

50 p < 0.01

40 30

p < 0.05

brain: n=5 lung: n=5 liver: n=3 skin: n=15

20 10 0

Sk in

p < 0.01

p < 0.05

Li ve r

brain: n=11 lung: n=12 liver: n=10 skin: n=20

Percentage methylation

p < 0.01

0 A

en ric he d

p < 0.01

CACNA1G in Location E corrected for subtype (luminal B)

CACNA1G in Location of dissemination

Lu m in al

H ER 2

ne ga tiv e

Lu m in al

ne ga tiv e

luminal A: n=1 luminal B: n=5 triple negative: n=3 HER2 enriched: n=2

Lu ng

Tr ip le

luminal A: n=11 luminal B: n=28 triple negative: n=12 HER2 enriched: n=2

CACNA1G in Subtype corrected for location (brain)

Percentage methylation

p < 0.05

D Percentage methylation

Lu m in al

CACNA1G in Molecular subtype primary

Lu m in al

Percentage methylation

B ra i

M et as

im ar ie

Percentage methylation

PART ONE | CHAPTER 4

Figure 3. Quantitative methylation percentages of CACNA1G by MS-MLPA in primary breast tumors and their corresponding distant metastases (a). Methylation percentages in the primary tumor, divided per molecular subtype (b) and corrected for dissemination localisation (brain) (c) are shown thereunder. At the bottom, methylation percentages in the primary tumor, divided per dissemination location (d) and corrected for molecular subtype (luminal B) (e) are presented. Small horizontal lines depict the median per group. The grey horizontal line depicts the cut-off for hypermethylation (8.5%). Chromosome location CACNA1G: chr17:48638429-48704832, CpG site MS-MLPA probe: 48638728, #bp from probe to TSS: -300 and from probe to ATG: 92.

Hypermethylation in breast cancer metastases

5% of the tumor suppressor genes remained significant for both data types: SCGB3A1 (both loci), ID4-1, SFRP5-2, H2AFX-1 and FHIT. In PR-positive tumors this phenomenon was less distinct: 17.5% or 25% of genes (quantitative or dichotomized respectively) showed higher methylation values, but no significance remained after multiple comparisons correction. Further, in HER2-positive tumors more methylation was seen in 2.5% (quantitative) or 7.5% (dichotomized) of tumor suppressor genes, but again no significance remained when corrected for multiple comparisons. Metastatic site The following genes were significantly more methylated in primary tumors metastasizing to brain, lung or skin, than to liver: PRDM2 (quantitative and dichotomized), RARB-1 (quantitative and dichotomized), HLTF-1 (quantitative), ID4-2 (quantitative), TWIST1 (quantitative and dichotomized), SFRP4-2 (quantitative an dichotomized), DLC1 (both loci; quantitative), H2AFX-2 (quantitative and dichotomized), CACNA1G (quantitative and dichotomized) (Figure 3d), CACNA1A (quantitative) and TIMP3 (all three loci; quantitative, -b; dichotomized). Also in the heatmaps (Figure 4) a distinct cluster was formed by primary breast tumors that metastasized to liver. When corrected for molecular subtype by logistic regression, the largest differences in methylation of individual genes were seen between liver and skin (skin being more methylated), and also the CMI was significantly different here (p=0.039). Figure 3e shows significantly more methylation of CACNA1G in brain, lung and skin compared to liver (quantitative data) as an example. Association with clinicopathological characteristics Supplementary Table S5 shows the association between methylation in the primary tumor and classical clinicopathological characteristics. A higher CMI (quantitative values) significantly correlated with higher MAI (p=0.040), although there was no association to lymph node status, localization of metastases and molecular subtype. More aggressive tumor characteristics like higher grade and MAI showed a tendency to higher methylation values of individual genes. Logistic regression for methylation conversion between the primary cancers and their metastases did not show significance for chemo- or hormonal therapy for any of the genes, indicating that adjuvant systemic treatment is not a confounder in methylation conversion. No significant association was found (for both analysis methods) between methylation of individual tumor suppressor genes and age at diagnosis.

ER STATUS SUBTYPE META LOCATION

Value

P48

P21

P22

P2 P20

P23 P38

P1 P24

P31

P18

P14

P28

P54

P44

P32

P52

P59

P5 P56

P27

P11

P41

P34

P36

P3 P42

P37

P35

P40

P43

P30

P13

P17

P33

P45

P39

P19

P15

P10

P16

P60

P50

P51

P49

P57

P55

P58

P47

MLH1_2 TIMP3_3 HIC1 CDKN2A PTEN ATM RARB_2 APC CHFR VHL FHIT BCL2 ESR1 TP73 BRCA2 CADM1 CDKN1B CDKN2B CD44 CASP8 BRCA1 MLH1_1 CDH13 DLC1_2 DLC1_1 TIMP3_2 TIMP3_1 H2AFX_1 TGIF H2AFX_2 HLTF_1 RARB_1 HLTF_2 BNIP3 RUNX3 SFRP4_2 RDM2 CACNA1G TWIST1 GSTP1 SFRP5_1 SCGB3A1_2 SCGB3A1_1 SFRP4_1 ID4_2 DAPK1 SFRP5_2 ID4_1 RASSF1_2 RASSF1_1 CCND2_2 CCND2_1 CACNA1A

ER positive ER negative

luminal A luminal B triple negative HER2 enriched

brain lung liver skin

Figure 4. Unsupervised hierarchical clustering analysis of log transformed quantitative methylation percentages of 40 tumor suppressor genes (53 loci) in 53 primary breast tumors. The sidebars depict dissemination location, subtype (luminal A, luminal B, triple negative and HER2 enriched), and ER status (according to 10% threshold for positivity).

Color Key

PART ONE | CHAPTER 4

Hypermethylation in breast cancer metastases

Prognostic value Of the primary tumor characteristics, lymph node positivity, ER or PR negativity (10% cut-off for positivity) and HER2 positivity (DAKO score 3) were significantly correlated to worse survival (Table 2). When comparing survival curves of patients that showed methylation conversion from low to high or vice versa with those that did not, conversion of HLTF-2, ID4-2, SFRP4-1 and DAPK1 was correlated to worse overall survival (Figure 5a). Conversion for these genes was entered in Cox proportional hazard analyses together, where SFRP4-1 (HR 2.3, 95% CI 1.03-5.05) and HLTF-2 (HR 2.2, 95% CI 1.09-4.56) remained significant (Table 2). When analyzing prognostic value of methylation status of the individual genes in the metastases for survival time from biopsy of metastases to end of follow up, three out of the four aforementioned genes were again significant (ID4-2, SFRP4-1 and DAPK1) (Figure 5b).

Table 2. Cox proportional hazards modeling of tumor suppressor gene methylation. Clinicopathological characteristics are compared to time between resection of primary or metastasis and end of follow-up. Predictor

Bivariate model p value Time between resection of primary and end of follow-up

Methylation status in metastasis - ID4-2 - SFRP4-1 - DAPK1 - DLC1-1 - GSTP

N Time between resection of metastasis and end of follow-up 0.009 0.023 0.005 0.026 0.035

Conversion between primary* and metastasis - HLTF-2 - ID4-2 - SFRP4-1 - DAPK1

0.023 0.025 0.012 0.041

Molecular subtype - Luminal A - Luminal B - Triple negative - HER2 enriched

0.037 0.047 0.667

0.269 0.187 0.394

10 28 12 2

Location of metastasis - Brain - Lung - Liver - Skin

0.094 0.754 0.203

0.306 0.812 0.541

10 12 10 20

Tumor size <2 cm 2-5 cm >5 cm

0.699 0.756

0.039 0.330

16 25 6

PART ONE | CHAPTER 4

Table 2. Continued Histologic type - Ductal - Lobular - Metaplastic - Micropapillary

0.553 0.774 0.506

0.940 0.823 0.344

44 4 3 1

Histologic grade -I - II - III

0.963 0.026

0.663 0.303

1 11 40

MAI

0.359

0.712

Lymph node status

0.045

0.884

ER status

0.000

0.260

PR status

0.001

0.372

HER2 status

0.041

0.613

Age at diagnosis primary

0.385

0.202

Chemotherapy

0.118

0.998

Radiotherapy

0.064

0.024

Hormone therapy

0.236

0.907

Combination therapy - Chemoradiation - Radiohormonal therapy - Chemohormonal therapy - Chemoradiation + hormonal therapy

0.097 0.114 0.931

0.822 0.992 0.174

7 3 4 4

CMI primaries

0.642

0.876

CMI metastases

0.726

0.630

* Variables which are put in the multivariate model

Multivariate model of conversion between primary and metastasis in time between resection of metastasis and end of follow-up Parameter

Significance

Hazard ratio

95% Confidence interval Lower limit

Upper limit

SFRP4-1

0.042

2.279

1.028

5.052

HLTF-2

0.029

2.223

1.085

4.556

Bivariate analysis identified several significant (p<0.05) predictors of survival. Reference categories were set for those predictors with more than two categories. Based upon number of patients included, 4 predictors could be tested in multivariate Cox proportional hazard models. HLTF-2, ID4-2, SFRP4-1 and DAPK1 were used to generate composite models through forward conditional testing, with p<0.05 as the basis for retaining and removing variables.

2000

4000

6000

2000

4000

HLTF-2

6000

n=53 p=0.023

No conversion Negative conversion Positive conversion

Time between metastasis and end of follow-up

0.0

0.5

1.0

Time between metastasis and end of follow-up

0.0

Positive conversion

2000

4000

6000

2000

ID4-2

4000

6000

2000

4000

ID4-2 metastases

6000

n=53 p=0.009

0.0

0.5

1.0

2000

4000

ID4-2 metastases

6000

n=53 p=0.009

Unmethylated Methylated

Time between metastasis and end of follow-up

0.0

0.5

1.0

Unmethylated Methylated

n=53 p=0.025

No conversion Negative conversion Positive conversion

Time between metastasis and end of follow-up

0.0

0.5

1.0

Figure 5

Positive conversion

n=53 p=0.025

Time between metastasis and end of follow-up

0.0

0.5

2000

4000

Positive conversion

6000

n=53 p=0.012

2000

4000

SFRP4-1

6000

2000

4000

SFR4-1 metastases

6000

n=53 p=0.023

0.0

0.5

1.0

2000

4000

SFR4-1 metastases

6000

n=53 p=0.023

Unmethylated Methylated

Time between metastasis and end of follow-up

0.0

0.5

1.0

Unmethylated Methylated

n=53 p=0.012

No conversion Negative conversion Positive conversion

Time between metastasis and end of follow-up

0.0

0.5

1.0

Time between metastasis and end of follow-up

0.0

0.5

2000

4000

Positive conversion

6000

n=53 p=0.041

DAPK1

4000

6000

2000

4000

DAPK1 metastases

6000

n=53 p=0.005

0.0

0.5

1.0

2000

4000

DAPK1 metastases

6000

n=53 p=0.005

Unmethylated Methylated

Time between metastasis and end of follow-up

0.0

0.5

1.0

Unmethylated Methylated

n=53 p=0.041

No conversion Negative conversion Positive conversion

Time between metastasis and end of follow-up

0.0

0.5

1.0

Time between metastasis and end of follow-up

0.0

0.5

Time between metastasis and end of follow-up Time between metastasis and end of follow-up B Time between metastasis and end of follow-up Figure 5. Kaplan Meier survival curves of time between resection of metastasis to end of follow-up of HLTF-2, ID4-2, SFRP4-1 and DAPK1 of conversion of methylation status in the primary tumors compared toFigure paired5metastases (a). The dashed line depicts conversion from negative in the primary tumor to positive in the metastasis and the grey line depicts conversion from positive in the primary tumor to negative in the metastasis. Survival curves of ID4-2, SFRP4-1 and DAPK1 of methylation status of metastases are shown in (b). Chromosome location HLTF-2: chr3:148747904-148804341, CpG site MS-MLPA probe: 148804223, #bp from probe to TSS: -105 and from probe to ATG: -105. Chromosome location ID4-2: chr6:19837601-19842431, CpG site MS-MLPA probe: 19837620, #bp from probe to TSS: -20 and from probe to ATG: 365. Chromosome location SFRP4-1: chr7:37945535-37956525, CpG site MS-MLPA probe: 37956166, #bp from probe to TSS: -10632 and from probe to ATG: -9086. Chromosome location DAPK1: chr9:90113885-90323549, CpG site MS-MLPA probe: 90113281, #bp from probe to TSS: 603 and from probe to ATG: 711.

Fraction survi

Fraction survival

n=53 p=0.023

Fraction survi Fraction survival

Fraction survival Fraction survival

Fraction survi Fraction survival

Fraction survival Fraction survival

Fraction survi Fraction survival

Fraction survival Fraction survival

0.5

Hypermethylation in breast cancer metastases

PART ONE | CHAPTER 4

Correlation of methylation to mRNA expression by TCGA data extraction Despite possible heterogeneity in methylation between individual CpG sites within the same promoter region, we nevertheless tried to correlate methylation to mRNA expression by comparing the most closely located CpG sites between TCGA data and our MS-MLPA loci (criteria for matching: <1000 bp between CpG sites, significant inverse correlation, Pearsonâ&#x20AC;&#x2122;s r>-0.2; Supplementary Table S8). Note that these results thus need to be interpreted with caution. The evaluated CpG sites/regions of ATM, BCL2, BRCA1, BRCA2, CACNA1G, CADM1, CASP8, CCND2, CD44, CDKN2B, CHFR1, DAPK1, ESR1, GSTP1, HLTF, ID4, MLH1, PRDM2, PTEN, RARB, RASSF1, RUNX3, TIMP3, TP73 and TWIST1 (15/40 genes) showed a significant inverse correlation with mRNA expression when quantitative data were used Supplementary Table S6. Of these genes, fourteen showed higher methylation values in primaries compared to metastases in our cohort. For BNIP3, CACNA1A, CDH13, CDKN1B, FHIT, HIC1, SCGB3A1, SFRP4, SFRP5 and TGIF1 (10/40 genes) no correlation was found between CpG site methylation and mRNA expression.

DISCUSSION DNA methylation has a similar potential as genetic alterations in serving as a selectable driver during clonal expansion or metastatic dissemination, and could therefore yield valuable markers for cancer detection and prognosis as well as targets for new therapeutic strategies 20. Our study design allowed comparison of primary breast tumors to their paired distant metastases at different locations, enabling intra- and inter-individual comparison. Our results show a general tendency for lower methylation at primary tumor-methylated regions in the matched metastases of 21/40 tumor suppressor genes. It is unlikely that admixture of cells from the tumor micro-environment at distant sites have caused these lower methylation values. First, we only included metastatic samples that contained at least 80% tumor. Second, methylation values in normal breast were lower than in normal tissues from skin, lung, brain and liver, so admixture of such normal cells (especially from liver and brain) would have raised methylation values. Third, all normal tissues clustered together in unsupervised analysis, that also showed that primary tumors and their paired metastases cluster together. Therefore, most of these hypermethylation events are likely patient-specific and subject to specific selection across metastatic dissemination and expansion, emphasizing the need for personalized cancer treatment. Higher CMI correlated with higher MAI as did methylation values of individual genes, indicating that proliferation rate correlates with methylation, which is biologically plausible. Adjuvant chemo- or hormonal therapy did not seem to influence methylation conversion. To our knowledge, our study is the first that compared promoter methylation in a large 92

Hypermethylation in breast cancer metastases

group of multiple localizations of distant human breast cancer metastases to their matched primary breast carcinomas and we tried to apply the ‘reporting recommendations for tumor markers’ (REMARK criteria) as adequately as possible 21. Several studies have been performed describing just one metastatic site, just one tumor suppressor gene, non-matched pairs of primaries and metastases, methylation in the primary tumor only (compared to the metastasizing tendency), or have made use of mouse models instead of patient material 11,12,22-26 , but from those studies no conclusions can be drawn on intra-patient differences and site-specific markers. Rivenbark et al. demonstrated “epigenetic progression” by showing more methylation in lymph node metastases compared to the primary breast tumor 13, but Wu et al. showed no differences in methylation of 7 tumor suppressor genes in primary breast carcinomas compared to their matched distant metastases 27. The discrepant findings with our generally lower methylation values in distant metastases (largely in line with results in head and neck squamous cell carcinomas 28) are likely related to differences in distant metastasis localizations, differences in study populations and sample sizes, pairing of normal tissue, the inclusion of paired metastases, and variation in tumor suppressor genes and CpG regions studied, Further, methodologies for demonstration of methylation status (QM-MSP, methylation-specific PCR analysis, bisulfite sequencing, differential methylation hybridization, etc.) differ between studies. In our institute we have extensive experience using MS-MLPA 10,29-31, a restriction enzyme based assay that allows a multi-target approach on small amounts of DNA extracted from formalin-fixed paraffin embedded material. This technique shows a very good correlation with other techniques such as bisulfite pyrosequencing and (QM)MSP 32-37. Besides, a tumor or metastasisinitiating clone or sub-clone in each individual has a unique DNA methylation signature that is closely maintained across metastatic dissemination 20. However, for each tumor we chose one of many available tissue blocks (that contained the largest amount of tumor load), which could have led to sampling bias. A previous study from our group clearly demonstrated that, although most variation in methylation status is present between individual breast cancers, clonal epigenetic heterogeneity is seen within most primary breast carcinomas, indicating that methylation results from a single random sample may not be representative of the whole tumor29. In addition, for 12 genes two different CpG loci were analyzed separately, and exact results showed differences in methylation frequencies, indicating the presence of heterogeneous methylation. However, unsupervised hierarchical clustering showed an almost perfect correlation between 6-8 of the 12 genes of which different CpG sites were analyzed. These limitations could explain perhaps some but clearly not all of the differences in methylation values between primary and metastasis. To correct for the differences between location of dissemination, differences between molecular subtypes should be taken into account, since they are known to preferentially metastasize to specific distant sites 3,38. For instance, a general hypomethylation of basal-like tumors compared to differential methylation across non-basal-like subtypes is often 93

PART ONE | CHAPTER 4

reported 38,39. We indeed saw some clustering of triple negative tumors and one cluster almost entirely composed of ER-positive cancers, but no evident hypomethylation was seen compared to other subtypes. Distinct methylation patterns relative to breast cancer subtype and normal breast tissue as shown by Bardowell et al. (2013) were also not seen 38. Further, some of the chosen genes were significantly more methylated in tumors that metastasized to specific localizations (even when corrected for molecular subtype), which could lead to novel biomarkers predicting site of distant metastases and adjuvant targeted therapy strategies that could prevent such metastases from becoming clinically manifest. In a therapeutic setting, the correlation between methylation and mRNA/protein expression may become relevant, which is why we explored TCGA data. Generally, methylation at the investigated CpG sites by MS-MLPA, seemed inversely correlated to mRNA expression levels as demonstrated before 40 (despite possible heterogeneity in methylation between individual CpG sites used for MS-MLPA and TCGA test) indicating their relevance in gene silencing. Future studies should take into account actual protein expression of tumor suppressor genes in metastases in relation to methylation status. Theoretically, less methylation in metastases would prognostically be beneficial for the patient because of reactivation of these tumor suppressor genes. However, survival analysis showed that conversion of HLTF-2, ID4-2, SFRP4-1 and DAPK1 from positive in the primary tumor to negative in the metastasis was correlated to worse overall survival. Interestingly, methylation status of 3/4 of these genes (ID4-2, SFRP4-1 and DAPK1) predicted worse survival when hypermethylated in metastases. Most important independent predictors for shorter survival time over lymph node positivity and ER status were SFRP4 and HLTF, which are known predictors of worse survival. Hypermethylation of HLTF seems to predict poor outcome in colorectal 41,42 and lung cancer 43. SFRP4 is been shown to be an independent predictor for shorter survival in myelodysplastic syndrome 44 and invasive bladder cancer 45. However, these studies emphasize hypermethylation status in primary tumors and no studies were found on hypermethylation of these markers in paired metastases in relation to survival. Promoter hypermethylation of tumor suppressor genes is known to be an early event during carcinogenesis 9,10. There are several possible explanations for the trend that less promoter methylation of the investigated genes is seen in the metastases. First, the spread of tumor cells may take place even prior to methylation. It has been demonstrated before that, in breast, prostate and esophageal cancer, bone marrow DTCs (disseminated tumor cells: any tumor cell that has left the primary lesion and travelled to an ectopic environment, not necessarily forming a metastasis) display significantly fewer genetic aberrations than primary tumor cells 46-49. Dissemination of tumor cells that are still evolving may lead to allopatric selection and expansion of variant cells adapted to specific microenvironments 50. Second, it could be that methylation is a dynamic process and may even vary in different stages of the cell cycle. Graff et al. have 94

Hypermethylation in breast cancer metastases

shown that E-cadherin (a gene involved in homotypic cell-cell adhesion) in cell lines is hypermethylated when put in a culture model system for basement membrane invasion and hypomethylated in a tumor growth model 51. The reversibility of methylation of tumor suppressor genes could therefore be beneficial to tumor spread, whether it is a random process or a response to specific signals. In summary, we have shown that hypermethylation of tumor suppressor genes detected by MS-MLPA is generally lower in the distant metastases compared to the primary tumor. We already knew that hypermethylation, in contrast to DNA mutations, is reversible, but whether this is a random or controlled principle has not been fully elucidated. The question rises if the difference in methylation pattern between these primaries and metastases could be explained by the loss/ rearrangement of hypermethylation. Since we have shown that the 21/40 tested tumor suppressor genes show less methylation in metastases with respect to their matched primary carcinomas, methylation is probably not an epigenetic factor that could be used for therapy against metastatic tumor spread. However, since different metastasizing localizations show different methylation patterns, screening for a specific pattern that predicts most likely site of metastases could be a useful clinical tool. Further, methylation status of several genes seems to predict survival after metastases.Therefore, more tumor suppressor genes should be screened on larger databases and heterogeneity should be ruled out to include all tumor subclones.

Acknowledgements This study is supported by Dutch Cancer Society grant UU 2011-5195 and Philips Consumer Lifestyle. We especially would like to thank the University Medical Center Utrecht, the Meander Medical Center Amersfoort, the Deventer Hospital, the Rijnstate Hospital Arnhem, Tergooi Hospitals, the Academic Medical Center Amsterdam, the Radboud University Nijmegen Medical Center, the Canisius Wilhelmina Hospital Nijmegen, the Netherlands Cancer Institute Amsterdam, the Medical Center Alkmaar, the Medical Center Zaandam, the University Medical Center Groningen, the St. Antonius Hospital Nieuwegein, the Diakonessenhuis Utrecht, the Free University Medical Center Amsterdam, the Erasmus Medical Center Rotterdam, the Gelre hospital Apeldoorn, Isala clinics Zwolle, the Laboratory for Pathology Enschede, the Laboratory for Pathology Dordrecht and the Laboratory for Pathology Foundation Sazinon Hoogeveen for providing archival tissue for this study.

PART ONE | CHAPTER 4

REFERENCES 1. WHO, international agency for research on cancer. Latest world cancer statistics. global cancer burden rises to 14.1 million new cases in 2012:Â Marked increase in breast cancers must be addressed.Â . http://www.iarc.fr/en/mediacentre/pr/2013/pdfs/pr223_E.pdf. Updated 2013. Accessed 09/25, 2014. 2. van den Hurk CJ, Eckel R, van de Poll-Franse LV, et al. Unfavourable pattern of metastases in M0 breast cancer patients during 1978-2008: A population-based analysis of the munich cancer registry. Breast Cancer Res Treat. 2011;128(3):795-805. doi: 10.1007/s10549-011-1372-y [doi]. 3. Kennecke H, Yerushalmi R, Woods R, et al. Metastatic behavior of breast cancer subtypes. J Clin Oncol. 2010;28(20):3271-3277. doi: 10.1200/JCO.2009.25.9820 [doi]. 4. Ferlicot S, Vincent-Salomon A, Medioni J, et al. Wide metastatic spreading in infiltrating lobular carcinoma of the breast. Eur J Cancer. 2004;40(3):336-341. doi: S0959804903007251 [pii]. 5. Tseng LM, Hsu NC, Chen SC, et al. Distant metastasis in triple-negative breast cancer. Neoplasma. 2013;60(3):290294. doi: 10.4149/neo_2013_038 [doi]. 6. Paget S. The distribution of secondary growths in cancer of the breast. 1889. Cancer Metastasis Rev. 1989;8(2):98101. 7. Connolly R, Stearns V. Epigenetics as a therapeutic target in breast cancer. J Mammary Gland Biol Neoplasia. 2012;17(3-4):191-204. doi: 10.1007/s10911-012-9263-3 [doi]. 8. Lorusso G, Ruegg C. New insights into the mechanisms of organ-specific breast cancer metastasis. Semin Cancer Biol. 2012;22(3):226-233. doi: 10.1016/j. semcancer.2012.03.007 [doi]. 9. Park SY, Kwon HJ, Lee HE, et al. Promoter CpG island hypermethylation during breast cancer progression. Virchows Arch. 2011;458(1):73-84. doi: 10.1007/s00428010-1013-6 [doi]. 10. Verschuur-Maes AH, de Bruin PC, van Diest PJ. Epigenetic progression of columnar cell lesions of the breast to invasive breast cancer. Breast Cancer Res Treat. 2012;136(3):705-715. doi: 10.1007/s10549-012-2301-4 [doi]. 11. Swift-Scanlan T, Vang R, Blackford A, Fackler MJ, Sukumar S. Methylated genes in breast cancer: Associations with clinical and histopathological features in a familial breast cancer cohort. Cancer Biol Ther. 2011;11(10):853-865. doi: 15177 [pii]. 12. Chen L, Zhu Z, Sun X, et al. Down-regulation of tumor suppressor gene FEZ1/LZTS1 in breast carcinoma involves promoter methylation and associates with metastasis. Breast Cancer Res Treat. 2009;116(3):471-478. doi: 10.1007/s10549-008-0147-6 [doi]. 13. Rivenbark AG, Livasy CA, Boyd CE, Keppler D, Coleman WB. Methylation-dependent silencing of CST6 in primary human breast tumors and metastatic lesions. Exp Mol Pathol. 2007;83(2):188-197. doi: S0014-4800(07)00042-1 [pii].

14. Hoefnagel LDC, Beelen KJ, Opdam M, Vincent A, Linn SC, Van Diest PJ. Increased expression of phosporylated mTOR in metastatic breast tumors compared to primary tumors in patients who received adjuvant endocrine therapy. Cancer Res. 2012;72(24). 15. Hoefnagel LDC, van der Groep P, van de Vijver MJ, et al. Discordance in ER alpha, PR and HER2 receptor status across different distant breast cancer metastases within the same patient. Annals of Oncology. 2013;24(12):30173023. doi: 10.1093/annonc/mdt390. 16. Hoefnagel LDC, van de Vijver MJ, van Slooten H, et al. Receptor conversion in distant breast cancer metastases. Breast Cancer Research. 2010;12(5):R75. doi: 10.1186/ bcr2645. 17. van Diest PJ. No consent should be needed for using leftover body material for scientific purposes. for. BMJ. 2002;325(7365):648-651. 18. Cerami E, Gao J, Dogrusoz U, et al. The cBio cancer genomics portal: An open platform for exploring multidimensional cancer genomics data. Cancer Discov. 2012;2(5):401-404. doi: 10.1158/2159-8290.CD-12-0095 [doi]. 19. Gao J, Aksoy BA, Dogrusoz U, et al. Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal. Sci Signal. 2013;6(269):pl1. doi: 10.1126/ scisignal.2004088 [doi]. 20. Aryee MJ, Liu W, Engelmann JC, et al. DNA methylation alterations exhibit intraindividual stability and interindividual heterogeneity in prostate cancer metastases. Sci Transl Med. 2013;5(169):169ra10. doi: 10.1126/scitranslmed.3005211 [doi]. 21. McShane LM, Altman DG, Sauerbrei W, et al. REporting recommendations for tumor MARKer prognostic studies (REMARK). Breast Cancer Res Treat. 2006;100(2):229235. doi: 10.1007/s10549-006-9242-8 [doi]. 22. Acosta D, Suzuki M, Connolly D, et al. DNA methylation changes in murine breast adenocarcinomas allow the identification of candidate genes for human breast carcinogenesis. Mamm Genome. 2011;22(3-4):249-259. doi: 10.1007/s00335-011-9318-6 [doi]. 23. Noetzel E, Rose M, Sevinc E, et al. Intermediate filament dynamics and breast cancer: Aberrant promoter methylation of the synemin gene is associated with early tumor relapse. Oncogene. 2010;29(34):4814-4825. doi: 10.1038/onc.2010.229 [doi]. 24. Carraway HE, Wang S, Blackford A, et al. Promoter hypermethylation in sentinel lymph nodes as a marker for breast cancer recurrence. Breast Cancer Res Treat. 2009;114(2):315-325. doi: 10.1007/s10549-008-0004-7 [doi]. 25. Mehrotra J, Vali M, McVeigh M, et al. Very high frequency of hypermethylated genes in breast cancer metastasis to the bone, brain, and lung. Clin Cancer Res. 2004;10(9):3104-3109. 26. Salhia B, Kiefer J, Ross JT, et al. Integrated genomic and epigenomic analysis of breast cancer brain metastasis. PLoS One. 2014;9(1):e85448. doi: 10.1371/journal. pone.0085448 [doi]. 27. Wu JM, Fackler MJ, Halushka MK, et al. Heterogeneity of breast cancer metastases: Comparison of therapeutic

Hypermethylation in breast cancer metastases

target expression and promoter methylation between primary tumors and their multifocal metastases. Clin Cancer Res. 2008;14(7):1938-1946. doi: 10.1158/10780432.CCR-07-4082 [doi]. 28. Smiraglia DJ, Smith LT, Lang JC, et al. Differential targets of CpG island hypermethylation in primary and metastatic head and neck squamous cell carcinoma (HNSCC). J Med Genet. 2003;40(1):25-33. 29. Moelans CB, de Groot JS, Pan X, van der Wall E, van Diest PJ. Clonal intratumor heterogeneity of promoter hypermethylation in breast cancer by MS-MLPA. Mod Pathol. 2014;27(6):869-874. doi: 10.1038/ modpathol.2013.207 [doi]. 30. Suijkerbuijk KP, Fackler MJ, Sukumar S, et al. Methylation is less abundant in BRCA1-associated compared with sporadic breast cancer. Ann Oncol. 2008;19(11):18701874. doi: 10.1093/annonc/mdn409 [doi]. 31. Moelans CB, Verschuur-Maes AH, van Diest PJ. Frequent promoter hypermethylation of BRCA2, CDH13, MSH6, PAX5, PAX6 and WT1 in ductal carcinoma in situ and invasive breast cancer. J Pathol. 2011;225(2):222-231. doi: 10.1002/path.2930 [doi]. 32. Suijkerbuijk KP, Pan X, van der Wall E, van Diest PJ, Vooijs M. Comparison of different promoter methylation assays in breast cancer. Anal Cell Pathol (Amst). 2010;33(3):133-141. doi: 10.3233/ACP-CLO-2010-0542 [doi]. 33. Leong KJ, Wei W, Tannahill LA, et al. Methylation profiling of rectal cancer identifies novel markers of earlystage disease. Br J Surg. 2011;98(5):724-734. doi: 10.1002/ bjs.7422 [doi]. 34. Cardoso LC, Tenorio Castano JA, Pereira HS, et al. Constitutional and somatic methylation status of DMRH19 and KvDMR in wilms tumor patients. Genet Mol Biol. 2012;35(4):714-724. doi: 10.1590/S141547572012005000073 [doi]. 35. Lopez F, Sampedro T, Llorente JL, et al. Utility of MS-MLPA in DNA methylation profiling in primary laryngeal squamous cell carcinoma. Oral Oncol. 2014;50(4):291-297. doi: 10.1016/j. oraloncology.2014.01.003 [doi]. 36. Furlan D, Sahnane N, Mazzoni M, et al. Diagnostic utility of MS-MLPA in DNA methylation profiling of adenocarcinomas and neuroendocrine carcinomas of the colon-rectum. Virchows Arch. 2013;462(1):47-56. doi: 10.1007/s00428-012-1348-2 [doi]. 37. Pineda M, Mur P, Iniesta MD, et al. MLH1 methylation screening is effective in identifying epimutation carriers. Eur J Hum Genet. 2012;20(12):1256-1264. doi: 10.1038/ ejhg.2012.136 [doi]. 38. Bardowell SA, Parker J, Fan C, Crandell J, Perou CM, Swift-Scanlan T. Differential methylation relative to breast cancer subtype and matched normal tissue reveals distinct patterns. Breast Cancer Res Treat. 2013;142(2):365-380. doi: 10.1007/s10549-013-2738-0 [doi]. 39. Ulirsch J, Fan C, Knafl G, et al. Vimentin DNA methylation predicts survival in breast cancer. Breast Cancer Res Treat. 2013;137(2):383-396. doi: 10.1007/ s10549-012-2353-5 [doi].

40. Wang D, Yang PN, Chen J, et al. Promoter hypermethylation may be an important mechanism of the transcriptional inactivation of ARRDC3, GATA5, and ELP3 in invasive ductal breast carcinoma. Mol Cell Biochem. 2014;396(1-2):67-77. doi: 10.1007/s11010-0142143-y [doi]. 41. Philipp AB, Nagel D, Stieber P, et al. Circulating cell-free methylated DNA and lactate dehydrogenase release in colorectal cancer. BMC Cancer. 2014;14:245-2407-14-245. doi: 10.1186/1471-2407-14-245 [doi]. 42. Philipp AB, Stieber P, Nagel D, et al. Prognostic role of methylated free circulating DNA in colorectal cancer. Int J Cancer. 2012;131(10):2308-2319. doi: 10.1002/ijc.27505 [doi]. 43. Castro M, Grau L, Puerta P, et al. Multiplexed methylation profiles of tumor suppressor genes and clinical outcome in lung cancer. J Transl Med. 2010;8:86-5876-8-86. doi: 10.1186/1479-5876-8-86 [doi]. 44. Wang H, Fan R, Wang XQ, et al. Methylation of wnt antagonist genes: A useful prognostic marker for myelodysplastic syndrome. Ann Hematol. 2013;92(2):199209. doi: 10.1007/s00277-012-1595-y [doi]. 45. Marsit CJ, Karagas MR, Andrew A, et al. Epigenetic inactivation of SFRP genes and TP53 alteration act jointly as markers of invasive bladder cancer. Cancer Res. 2005;65(16):7081-7085. doi: 65/16/7081 [pii]. 46. Schmidt-Kittler O, Ragg T, Daskalakis A, et al. From latent disseminated cells to overt metastasis: Genetic analysis of systemic breast cancer progression. Proc Natl Acad Sci U S A. 2003;100(13):7737-7742. doi: 10.1073/ pnas.1331931100 [doi]. 47. Stoecklein NH, Hosch SB, Bezler M, et al. Direct genetic analysis of single disseminated cancer cells for prediction of outcome and therapy selection in esophageal cancer. Cancer Cell. 2008;13(5):441-453. doi: 10.1016/j. ccr.2008.04.005 [doi]. 48. Weckermann D, Polzer B, Ragg T, et al. Perioperative activation of disseminated tumor cells in bone marrow of patients with prostate cancer. J Clin Oncol. 2009;27(10):1549-1556. doi: 10.1200/JCO.2008.17.0563 [doi]. 49. Schardt JA, Meyer M, Hartmann CH, et al. Genomic analysis of single cytokeratin-positive cells from bone marrow reveals early mutational events in breast cancer. Cancer Cell. 2005;8(3):227-239. doi: S15356108(05)00261-8 [pii]. 50. Klein CA. Parallel progression of primary tumours and metastases. Nat Rev Cancer. 2009;9(4):302-312. doi: 10.1038/nrc2627 [doi]. 51. Graff JR, Gabrielson E, Fujii H, Baylin SB, Herman JG. Methylation patterns of the E-cadherin 5â&#x20AC;&#x2122; CpG island are unstable and reflect the dynamic, heterogeneous loss of E-cadherin expression during metastatic progression. J Biol Chem. 2000;275(4):2727-2732.

PART ONE | CHAPTER 4

SUPPLEMENTAL Supplementary Table S1. Description of the 24 tumor suppressor genes included in the ME001-C2 kit (MRC-Holland). Information about chromosome location, CpG site location and relative location near important gene regions is supplied in Supplementary Table S9. Gene

Function/description

TIMP3

Metalloprotease inhibitor

APC

Inhibits Wnt-signalling pathway and induces Î˛-catenine degradation

CDKN2A (p16)

Inhibitor of cyclin-dependent kinase 4/6

MLH1

DNA mismatch repair enzyme

ATM

Kinase responsive to DNA damage leading to p53 activation

RARB

Nuclear receptor for differentiation and growth arrest

CDKN2B (p15)

Inhibitor of cyclin-dependent kinase 4/6

HIC1

Downregulation of SIRT1, involved in p53 pathway

CHFR

Spindle assembly checkpoint protein

BRCA1

Maintenance and repair of DNA and chromosomal structure

15-17

CASP8

Apoptotic caspase

18,19

CDKN1B (p27)

Inhibitor of Cyclin-cdk complexes, thereby inducing cell-cycle arrest

PTEN

PIP3 phosphatase inhibiting AKT pathway

BRCA2

Maintenance and repair of DNA and chromosomal structure

19,22,23

CD44

Cell surface antigen involved in cell adhesion and migration

RASSF1

Inhibitor of cell cycle progression and proliferation

DAPK1

Kinase involved in cell growth control

VHL

Negative regulator of HIF-Îą, which mediates cellular response to hypoxia

ESR1

Estrogen receptor; transcription factor

TP73

Member of the p53 family of transcription factors, involved in cellular response to stress and development

FHIT

Hydrolase associated with limiting cell proliferation and inducing apoptosis

31,32

IGSF4

Cell adhesion molecule involved in cytoskeleton stability

CDH13

Cadherin located on the cell membrane, without intracellular domain. Its specific function is not clear, but it is among others associated with resistance to atherosclerosis and it protects endothelial cells from apoptosis due to oxidative stress.

GSTP1

Detoxifies electrophilic carcinogens

Hypermethylation in Breast cancer 1,2 3 4,5 6 7,8 9,10 11 12,13 14

20,21

9,25-27 28,29 1,30

23,33

1,23,33,34

Hypermethylation in breast cancer metastases

Supplementary Table S2. Description of the 18 tumor suppressor genes included in the ME003-A1 kit (MRC-Holland). Information about chromosome location, CpG site location and relative location near important gene regions is supplied in Supplementary Table S9. Gene

Function/Description

Hypermethylation in Breast cancer

CCND2

Cell cycle regulator by binding to cdk4 and cdk6

35,36

SCGB3A1 (HIN-1)

Secretoglobin that inhibits cell growth, migration and invasion by inhibition of the AKT signalling pathway

37-39

BNIP3

Mitochondrial pro-apoptotic protein by activation of Bax/Bak and interaction with BCL-2

DLC1

GTPase activating protein of the Rho family, involved in regulation of the cytoskeleton and cell migration

40,41

HLTF

Chromatin-remodelling factor belonging to the SWI/SNF family

SFRP5

Inhibitor of the Wnt-signalling pathway

H2AFX

Member of the histone H2A family, which facilitate wrapping of DNA into nucleosomes

CACNA1G

Belongs to the family of voltage-dependent calcium channels, which are associated with muscle contraction, hormone or neurotransmitter release and gene expression. This isoform is predominantly expressed in neuronal tissue

SFRP4

Inhibitor of the Wnt-signalling pathway

TWIST1

Basic helix-loop-helix domain-containing transcription factor essential for mesoderm specification and differentiation and organogenesis

BCL2

Anti-apoptotic protein by inhibition of caspase activity

CACNA1A

TIMP3

Metalloprotease inhibitor

ID4

Member of the inhibitor of DNA binding protein family, which inhibits DNA binding of basic helix-loop-helix transcription factors

48,49

RUNX3

Regulates cell proliferation, apoptosis, angiogenesis, cell adhesion and invasion

50-53

PRDM2 (RIZ)

Member of the nuclear histone/protein methyltransferase superfamily, that can bind to the retinoblastoma protein, the ER and the TPA-responsive element (MTE) of the hemeoxygenase-1 gene

TGIF1

Member of the TALE superclass, which inhibits retinoic aciddependent RXR-alpha transcription activation and acts as an active transcriptional co-repressor of SMAD2

RARB

Nuclear receptor for differentiation and growth arrest

42,43

38,44-46

47 -

1,2

9,10

PART ONE | CHAPTER 4

Supplementary Table S3. Significant differences seen in methylation status of normal tissue (derived from our normal tissue biobank) between various locations. Breast n=25, brain n=5, liver n=5, lung n=5. Lung, liver and brain tissue seems often more methylated than breast tissue. Gene

Lung>breast

Brain>breast

Liver>breast

PRDM

0.000

RUNX3

0.000

SCGB3A1-1

0.007

SFRP4-2

0.027

H2AFX-1

0.002

H2AFX-2

0.003

0.018

CCND2-1

0.025

0.000

CCND2-2

0.000

TGIF

0.042

TIMP3-3

0.039

TP73

0.049

CASP8

0.000

RASSF1A-1

0.000

RASSF1A-2

0.000

APC

0.000

ESR1

0.000

CDKN2A CDKN2B CD44 GSTP1

100

0.002 0.031

0.000 0.049 0.000

Hypermethylation in breast cancer metastases

Supplementary Table S4. Overview of molecular subtype and location of dissemination of the 53 selected patients. Luminal A tumors preferentially appear to disseminate to liver over lung, luminal B tumors to skin over liver and triple negative tumors to lung over skin. The group of HER2 enriched tumors is too small to perform statistics on.

Molecular subtype

Location of metastasis Brain

Lung

Liver

Skin

Luminal A

N=1

N=0

N=6

N=4

Luminal B

N=5

N=3

N=15

Triple negative

N=3

N=7

N=1

HER2 enriched

N=2

N=0

Total

Significant differences

N=11

Liver>lung

(20.8%)

(p=0.017)

N=28

Skin>liver

(52.8%)

(p=0.049)

N=12

Lung>skin

(22.6%)

(p=0.012)

N=2 (3.8%)

Total

N=11

N=12

N=10

N=20

N=53

(20.8%)

(22.6%)

(18.9%)

(37.7%)

(100%)

101

102 III>II

high>low

P(Q): 0.015

P(Q): 0.026

P(Q): 0.020 high>low P(dich): 0.021

P(Q): 0.049

+>-

->+

P(dich): 0.026 ->+

P(dich): 0.034 ->+

Lymph node status

ID4-1

P(Q): 0.034 II>I P(dich): 0.019

high>low

P(Q): 0.038

2-5> smaller than 2 P(Q): 0.019 ductal>lobular P(dich): 0.007

smaller than 2> bigger than 5

2-5>bigger than 5

P(Q): 0.010

P(Q): 0.037

P(Q): 0.040

MAI

H2AFX-1

FHIT

DLC1-1

P(dich)0.031

P(Q): 0.038

CDKN2A

CHFR

P(dich): 0.037

CCND2-2

CCND2-1

Q(dich): 0.019 II>I

2-5> smaller than 2

III>II

P(dich): 0.002 III>I

P(Q): 0.038P(dich): 0.014

Grade

CASP8

P(dich): 0.043 ductal>lobular

Type

P(Q): 0.016

P(dich): 0.006

Size

CACNA1G

CACNA1A

BRCA1

BCL2

APC

CMI

Supplementary Table S5. Association between methylation status of 53 patients for 40 tumor suppressor genes (53 loci) and classical clinicopathological characteristics. Only significant correlations (p<0.05) were shown.

PART ONE | CHAPTER 4

P(Q): 0.039 P(dich): 0.011

2-5>bigger than 5

P(Q): 0.037 high>low P(dich): 0.037

TIMP3-2

high>low

P(Q): 0.008 high>low P(dich): 0.000

P(dich): 0.030 high>low

P(Q): 0.007 high>low P(dich): 0.001

P(Q): 0.039 P(dich): 0.012 III>II

high>low

P(Q): 0.006 high>low P(dich): 0.005

P(Q): 0.015

TIMP3-1

SFR5-2

SFRP4-2

SFR4-1

SCGB3A1-2

SCGB3A1-1

P(Q): 0.018 III>II P(dich): 0.001

P(dich): 0.019 II>I

RASSF1-2 Ductal>lobular

P(dich): 0.001 II>I

RASSF1-1

P(Q): 0.030

P(dich): 0.000 III>II

RARB-1

RUNX3

P(dich): 0.006 III>II

P(Q): 0.017 III>II P(dich): 0.041

PRDM

MLH1-1

ID4-2 +>-

P(Q): 0.003

+>-

P(Q): 0.020 +>P(dich): 0.025

P(dich): 0.025 +>-

P(Q): 0.028 ->+ P(dich): 0.040

P(Q): 0.015

Hypermethylation in breast cancer metastases

103

PART ONE | CHAPTER 4

Supplementary Table S6. mRNA expression data of 1038 breast cancer patients from TCGA (RNA Seq V2 RSEM). Z-values of mRNA sequencing data for 40 tumor suppressor genes were compared to percentages of DNA methylation (beta values by the Illumina Infinium HumanMethylation27 platform). TCGA CpG-sites that matched closest to our MS-MLPA CpG-sites (criteria for matching: <1000 bp between CpG-sites, significant inverse correlation, Pearson’s r>-20), were shown when possible. Significant inverse correlations with Pearson’s r>-0.20 are depicted in bold. Quantitative data (Pearson’s r) Genes

Correlation

Significance

CpG-site MS-MLPA

CpG-site TCGA

#bp between TCGA and MS-MLPA CpG-sites

APC

0.167

p=0.004

112073489

112073502

-13

ATM

-0.315

p=0.000

108093863

108093680

183

BCL2

-0.204

p=0.000

60986763

60987808

-1045

BNIP3

BRCA1

-0.361

p=0.000

41277269

41277213

BRCA2

-0.266

p=0.000

32889752

32889023

729

CACNA1A

CACNA1G

-0.140

p=0.014

48638728

48637104

1624

CADM1

-0.129

p=0.024

115375372

115375226

146

CASP8

-0.418

p=0.000

202122649

202098951

23698

CCND2-2

-0.133

p=0.020

4382040

4382184

-144

CD44

-0.274

p=0.000

35160827

35160330

497

CDH13

CDKN1B

CDKN2A

0.137

p=0.017

21995276

21995312

-36

CDKN2B

-0.348

p=0.000

22008788

22005563

3225

CHFR1

-0.239

p=0.000

133464323

133463885

438

DAPK1

-0.460

p=0.000

90113281

90113813

-532

DLC1-2

0.139

p=0.015

12991070

13372132

-381062

ESR1

-0.264

p=0.000

152129218

152128338

880

FHIT

GSTP1

-0.200

p=0.000

67351212

67350976

236

H2AFX-2

0.115

p=0.045

118966473

118966382

HIC1

HLTF-1

-0.184

p=0.001

148804137

148803905

232

ID4-2

-0.302

p=0.000

19837620

19837350

270

MLH1-2

-0.245

p=0.000

37035032

37035399

-367

PRDM2

0.121

p=0.034

14026590

14031383

-4793

PTEN

-0.264

p=0.000

89622381

89622589

-208

RARB-2

-0.122

p=0.033

25469573

25469919

-346

RASSF1-2

-0.150

p=0.009

50378327

50375674

2653

104

Hypermethylation in breast cancer metastases

Supplementary Table S6. Continued Quantitative data (Pearsonâ&#x20AC;&#x2122;s r) Genes

Correlation

Significance

CpG-site MS-MLPA

CpG-site TCGA

#bp between TCGA and MS-MLPA CpG-sites

RUNX3

-0.243

p=0.000

25256377

25255838

539

SCGB3A1

SFRP4

SFRP5

TGIF1

TIMP3-1

-0.142

p=0.013

33197658

33197394

264

TP73

-0.204

p=0.000

3569155

3569624

-469

TWIST1

-0.113

p=0.048

19156724

19156086

638

VHL

0.169

p=0.003

10183424

10189941

-6517

105

PART ONE | CHAPTER 4

Supplementary Table S7. Cut-offs for dichotomization of methylation values for the 40 tested tumor suppressor genes (53 loci) to distinguish between hypermethylated and unmethylated. ROC curves were used to find the most appropriate cut-off between normal tissue (n=25 normal breast) and tumor tissue (n=53 primary breast tumors). The appropriate cut-offs varied between 0.5% and 22.75% for the 40 genes. Cut-offs for hypermethylation by ROC curves in %

Dynamic methylation range normal breast tissue in % (n=25)

Dynamic methylation range primary breast tumors in % (n=53)

2,50

0-21

0-48

11,50

0-24

7-100

8,50

0-17

0-70

10,75

5-17

3-65

0,50

0-4

0-7

1,50

0-6

0-8

3,50

0-8

0-97

16,50

0-43

3-76

12,50

0-18

7-100

4,50

0-12

0-77

1,50

0-8

0-72

1,50

0-8

0-56

10,50

0-14

5-100

1,50

0-16

0-59

3,50

0-11

0-59

0,50

0-4

0-55

2,50

0-9

0-72

10,50

4-37

2-42

9,50

6-21

4-42

0,50

0-2

0-6

7,25

0-38

3-83

7,50

0-35

5-100

8,50

5-27

4-53

1,00

0-3

0-10

0,50

0-3

0-5

2,50

0-5

0-65

0,50

0-7

0-95

6,50

0-19

0-75

2,75

1-21

0-72

10,75

4-18

6-64

4,50

2-95

3-77

1,75

0-5

0-6

0,50

0-11

0-6

4,50

0-12

3-19

106

Hypermethylation in breast cancer metastases

Supplementary Table S7. Continued Cut-offs for hypermethylation by ROC curves in %

Dynamic methylation range normal breast tissue in % (n=25)

Dynamic methylation range primary breast tumors in % (n=53)

10,50

2-53

7-100

10,50

0-66

0-100

1,75

0-5

0-7

22,75

4-75

4-86

8,75

3-90

4-16

8,75

3-16

4-18

4,50

1-77

0-38

3,50

0-14

2-60

6,88

3-16

4-21

7,50

2-16

5-20

7,50

3-28

4-77

3,50

2-14

0-13

4,75

0-19

0-83

3,75

1-6

0-8

0,75

0-7

0-26

3,75

1-11

2-9

17,50

5-32

14-100

5,50

2-17

0-13

3,50

1-9

2-100

107

PART ONE | CHAPTER 4

CpG-site MS-MLPA probe

Translation start site (ATG)

Transcription start site (TSS)

Gene position

Gene

Supplementary Table S8. List of 53 loci of 40 tumor suppressor genes tested in this study and their most important characteristics: including CpG site of MS-MLPA probe, chromosome location, transcription and translation start sites and #bp between MS-MLPA probe and TSS/ATG. Also, information used to correlate MS-MLPA data to TCGA data were depicted, including CpG site of Illumina TCGA methylation probe, #bp between TCGA methylation probe and TSS/ATG, #bp between TCGA methylation probe and MS-MLPA probe and correlation between TCGA mRNA expression z-values and TCGA methylation bÃ¨ta values.

APC

chr5:112073556-112181936

112073555

112090587

112073489

ATM

chr11:108093559-108239826

108093558

108098351

108093863

BCL2

chr18:60985282-60985899

60985281

60986763

BNIP3

chr10:133783191-133795435

133783190

133784137

133795284

BRCA1

chr17:41243452-41277468

41243451

41243452

41277269

BRCA2

chr13:32889617-32907524

32889616

32890597

32889752

CACNA1A

chr19:13317256-13617274

13317255

13318859

13616856

CACNA1G

chr17:48638429-48704832

48638428

48638820

48638728

CADM1

chr11:115044345-115375241

115270145

115375372

108

Hypermethylation in breast cancer metastases

-305

-1482

17098

4488

-1482

-12094

-11147

-33818

-33817

-136

845

-299601

-297997

-300

-105227

r-value

Direction of correlation

p-values of correlation between mRNA expression z-values and methylation bÃ¨ta values TCGA

#bp between Illumina probe TCGA and MS-MLPA probe

CpG-site Illumina probe TCGA

#bp between MS-MLPA probe and ATG

#bp between MS-MLPA probe and TSS

Supplementary Table S8. Continued

112073433

0,002

0,176

112073502

-13

0,004

0,167

112073570

-81

0,001

0,185

112073686

-197

0,004

0,166

112074043

-554

0,007

0,154

108093220

643

0,307

108093680

183

0,000

0,315

108093921

-58

0,001

0,187

60903834

82929

0,005

0,160

60904237

82526

0,001

0,188

60904328

82435

0,003

0,170

60904418

82345

0,655

60985380

1383

0,000

0,309

60985645

1118

0,000

0,244

60986026

737

0,017

0,136

60986622

141

0,006

0,156

60986679

0,865

60986911

-148

0,184

60987429

-666

0,271

60987808

-1045

0,000

0,204

133794911

373

0,091

133796476

-1192

0,760

41272508

4761

0,000

0,217

41277059

210

0,000

0,268

41277213

0,000

0,361

41277322

-53

0,000

0,327

41277444

-175

0,000

0,319

41278622

-1353

0,654

32889023

729

0,000

0,266

32889752

0,018

0,135

13616871

-15

0,959

13617366

-510

0,175

48637104

1624

0,014

0,140

48638187

541

0,666

48638703

0,546

48639239

-511

0,535

48639585

-857

0,858

48639836

-1108

0,866

115374969

403

0,104

115375226

146

0,024

0,129

109

PART ONE | CHAPTER 4

CpG-site MS-MLPA probe

Translation start site (ATG)

Transcription start site (TSS)

Gene position

Gene

Supplementary Table S8. Continued

CASP8

chr2:202122754-202152434

202122753

202122649

CCND2-1

chr12:4382902-4414522

4382901

4383206

4381850 4381850 4381850 4381850 4381850 4381850 4381850 4381850 4381850 4381850 4381850

CCND2-2

chr12:4382902-4414522

4382901

4383206

4382040

CD44

chr11:35160417-35253949

35160416

35160850

35160827

CDH13

chr16:82660399-83214630

82660398

82660697

82660765

CDKN1B

chr12:12870302-12875305

12870301

12870773

12870608

CDKN2A

chr9:21967751-21994490

21967750

21974475

21995276

110

Hypermethylation in breast cancer metastases

1051

861

-411

-367

-307 -27526

1356

1166

-68

165 -20801

r-value

Direction of correlation

p-values of correlation between mRNA expression z-values and methylation bÃ¨ta values TCGA

104

#bp between Illumina probe TCGA and MS-MLPA probe

#bp between MS-MLPA probe and ATG

104

CpG-site Illumina probe TCGA

#bp between MS-MLPA probe and TSS

Supplementary Table S8. Continued

202098257

24392

0,000

0,448

202098951

23698

0,000

0,418

4381777

0,170

4382184

-334

0,020

0,133

4382985

-1135

0,306

4383117

-1267

0,164

4383507

-1657

0,110

4383619

-1769

0,102

4383699

-1849

0,037

0,120

4383724

-1874

0,180

4383770

-1920

0,035

0,121

4384022

-2172

0,001

0,182

4384890

-3040

0,002

0,176

4381777

263

0,170

4382184

-144

0,020

0,133

4382985

-945

0,306

4383117

-1077

0,164

4383507

-1467

0,110

4383619

-1579

0,102

4383699

-1659

0,037

0,120

4383724

-1684

0,180

4383770

-1730

0,035

0,121

4384022

-1982

0,001

0,182

4384890

-2850

0,002

0,176

35160330

497

0,000

0,274

35160400

427

0,000

0,231

35160462

365

0,000

0,231

35160571

256

0,002

0,177

35160892

-65

0,020

0,133

35161229

-402

0,024

0,130

82660328

437

0,517

82660490

275

0,391

82660670

0,167

82660873

-108

0,161

82661725

-960

0,225

82671450

-10685

0,089

12870119

489

0,480

12870463

145

0,205

21968233

27043

0,003

0,170

21995312

-36

0,017

0,137

111

PART ONE | CHAPTER 4

CpG-site MS-MLPA probe

Translation start site (ATG)

Transcription start site (TSS)

Gene position

Gene

Supplementary Table S8. Continued

CDKN2B

chr9:22008716-22008952

22008715

22008788

CHFR

chr12:133416938-133464204

133416937

133418139

133464323

DAPK1

chr9:90113885-90323549

90113884

90113992

90113281

DLC1-1

chr8:12940872-12990809

12940871

12943319

12990544

DLC1-2

chr8:12940872-12990809

12940871

12943319

12991070

ESR1

chr6:151977830-152163922

152128813

152129047

152129218

FHIT

chr3:59735036-61237133

61068514

61236889

12990544

112

Hypermethylation in breast cancer metastases

-73

-47386

603

-49673 -50199 -405

-168375

-46184

711

-47225 -47751 -171

-168375

r-value

Direction of correlation

p-values of correlation between mRNA expression z-values and methylation bÃ¨ta values TCGA

#bp between Illumina probe TCGA and MS-MLPA probe

CpG-site Illumina probe TCGA

#bp between MS-MLPA probe and ATG

#bp between MS-MLPA probe and TSS

Supplementary Table S8. Continued

22005288

3500

0,000

0,364

22005563

3225

0,000

0,348

22008970

-182

0,349

22009355

-567

0,885

133420801

43522

0,001

0,197

133424221

40102

0,015

0,139

133424709

39614

0,281

133429949

34374

0,514

133430226

34097

0,141

133430387

33936

0,535

133430564

33759

0,000

0,212

133435649

28674

0,644

133435770

28553

0,167

133463607

716

0,033

0,122

133463694

629

0,000

0,269

133463885

438

0,000

0,239

133464462

-139

0,000

0,218

133464857

-534

0,000

0,233

90112101

1180

0,000

0,396

90112515

766

0,005

0,160

90112519

762

0,014

0,140

90113120

161

0,008

0,152

90113813

-532

0,000

0,460

90114156

-875

0,000

0,393

90123546

-10265

0,006

0,156

13372132

-381588

0,015

0,139

13373090

-382546

0,000

0,239

13372132

-381062

0,015

0,139

13373090

-382020

0,000

0,239

152128328

890

0,000

0,223

152128338

880

0,000

0,264

152128743

475

0,030

0,125

152129036

182

0,116

152129400

-182

0,369

152129791

-573

0,567

152130207

-989

0,000

0,324

60363505

873384

0,181

61236652

237

0,113

61236909

-20

0,962

61236911

-22

0,953

61237063

-174

0,890

61237172

-283

0,670

61237270

-381

0,556

113

PART ONE | CHAPTER 4

CpG-site MS-MLPA probe

Translation start site (ATG)

Transcription start site (TSS)

Gene position

Gene

Supplementary Table S8. Continued

GSTP1

chr11:67351066-67354124

67351065

67351314

67351212

H2AFX-1

chr11:118964585-118966177

118964584

118965672

118966217

H2AFX-2

chr11:118964585-118966177

118964584

118965672

118966473

HIC1

chr17:1958263-1962981

1958262

1959984

1958385

HLTF-1

chr3:148747904-148804341

148804118

148804137

HLTF-2

chr3:148747904-148804341

148804118

148804223

ID4-1

chr6:19837601-19842431

19837600

19837985

19837032

ID4-2

chr6:19837601-19842431

19837600

19837985

19837620

MLH1-1

chr3:37034841-37092337

37034840

37035038

37034698

MLH1-2

chr3:37034841-37092337

37034840

37035038

37035032 37035032 37035032 37035032 37035032

PRDM2

chr1:14026735-14151574

14026734

14042100

14026590

PTEN

chr10:89623195-89728532

89623194

89624226

89622381

RARB-1

chr3:25469834-25639422

25469753

25542702

25469336

114

Hypermethylation in breast cancer metastases

-1633 -1889 -123

-19

-105

568 -20 142

-192

144

813

417

-545 -801 1599

-19

-105

953 365 340

15510

1845

73366

r-value

Direction of correlation

p-values of correlation between mRNA expression z-values and methylation bÃ¨ta values TCGA

102

#bp between Illumina probe TCGA and MS-MLPA probe

#bp between MS-MLPA probe and ATG

-147

CpG-site Illumina probe TCGA

#bp between MS-MLPA probe and TSS

Supplementary Table S8. Continued

67350976

236

0,000

0,200

67351608

-396

0,000

0,267

67351786

-574

0,000

0,273

118966186

0,375

118966382

-165

0,045

0,115

118966186

287

0,375

118966382

0,045

0,115

1958117

268

0,947

1958162

223

0,878

1958412

-27

0,899

148803905

232

0,001

0,184

148805262

-1125

0,367

148805294

-1157

0,267

148803905

318

0,001

0,184

148805262

-1039

0,367

148805294

-1071

0,267

19837350

-318

0,000

0,302

19838211

-1179

0,003

0,171

19837350

270

0,000

0,302

19838211

-591

0,003

0,171

37033625

1073

0,864

37033980

718

0,330

37034154

544

0,157

37034840

-142

0,643

37035399

-701

0,000

0,245

37033625

1407

0,864

37033980

1052

0,330

37034154

878

0,157

37034840

192

0,643

37035399

-367

0,000

0,245

14029875

-3285

0,862

14030611

-4021

0,116

14031383

-4793

0,034

0,121

89621773

608

0,982

89622589

-208

0,000

0,264

89623138

-757

0,016

0,138

89623336

-955

0,001

0,190

89623432

-1051

0,043

0,116

89624102

-1721

0,068

25469402

-66

0,296

25469577

-241

0,374

25469720

-384

0,872

25469919

-583

0,033

0,122

115

PART ONE | CHAPTER 4

CpG-site MS-MLPA probe

Translation start site (ATG)

Transcription start site (TSS)

Gene position

Gene

Supplementary Table S8. Continued

RARB-2

chr3:25469834-25639422

25469753

25542702

25469573

RASSF1-1

chr3:50374001-50378367

50374000

50374970

50378321 50378321 50378321 50378321 50378321 50378321 50378321 50378321 50378321

RASSF1-2

chr3:50374001-50378367

50374000

50374970

50378327

RUNX3

chr1:25256078-25291612

25256077

25256377

SCGB3A1-1

chr5:180017105-180018487

180017104

180017230

180018465

SCGB3A1-2

chr5:180017105-180018487

180017104

180017230

180018693

SFRP4-1

chr7:37945535-37956525

37945534

37947080

37956166

SFRP4-2

chr7:37945535-37956525

37945534

37947080

37956392

116

Hypermethylation in breast cancer metastases

180

-4321

-4327

-300

73129

-3351

-3357

-300

25469402

171

0,296

25469577

-4

0,374

25469720

-147

0,872

25469919

-346

0,033

50374304

4017

0,926

50374670

3651

0,071

50374929

3392

0,443

50375252

3069

0,376

50375459

2862

0,068

50375674

2647

0,009

50378191

130

0,128

50378413

-92

0,090

50378425

-104

0,154

50374304

4023

0,926

50374670

3657

0,071

50374929

3398

0,443

50375252

3075

0,376

50375459

2868

0,068

50375674

2653

0,009

50378191

136

0,128

50378413

-86

0,090

50378425

-98

0,154

25228582

27795

0,111

25228687

27690

0,910

25229099

27278

0,312

25229192

27185

0,086

25255838

539

25256369

25257566

-1189

0,232

25258332

-1955

0,494

25290947

-34570

0,402

25291385

-35008

0,000

25292072

-35695

0,101

0,122

0,150

0,000

0,243

0,012

0,144

0,329

0,183

25292215

-35838

0,001

-1235

180017623

842

-1589

-1463

180017623

1070

-10632

-9086

37955598

568

0,505

37955824

342

0,965

37956018

148

0,483

37956276

-110

0,250

37956554

-388

0,441

37955598

794

0,505

37955824

568

0,965

37956018

374

0,483

37956276

116

0,250

37956554

-162

0,441

-9312

r-value

-1361

-10858

Direction of correlation

p-values of correlation between mRNA expression z-values and methylation bÃ¨ta values TCGA

#bp between Illumina probe TCGA and MS-MLPA probe

CpG-site Illumina probe TCGA

#bp between MS-MLPA probe and ATG

#bp between MS-MLPA probe and TSS

Supplementary Table S8. Continued

117

PART ONE | CHAPTER 4

CpG-site MS-MLPA probe

Translation start site (ATG)

Transcription start site (TSS)

Gene position

Gene

Supplementary Table S8. Continued

SFRP5-1

chr10:99526508-99531756

99526507

99527270

99531516

SFRP5-2

chr10:99526508-99531756

99526507

99527270

99531918 99531918 99531918

TGIF1

chr18:3451591-3458406

3451590

3451977

3451691

TIMP3

chr22:33196802-33259028

33196801

33197987

33197658

TIMP3

chr22:33196802-33259028

33196801

33197987

33197819

TIMP3

chr22:33196802-33259028

33196801

33197987

33198047 33198047 33198047

TP73

chr1:3569129-3652765

3569128

3598929

3569155

TWIST1

chr7:19156336-19156944

19156335

19156724

VHL

chr3:10183319-10195354

10183318

10183531

10183424

118

Hypermethylation in breast cancer metastases

-5411

-101 -857

-1018

-1246

-27

-389

-106

-4648

286 329

168

-60

29774

-389

107

99531164

352

0,175

99531309

207

0,243

99532398

-882

0,063

99531164

754

0,175

99531309

609

0,243

99532398

-480

0,063

3411743

39948

0,190

3412088

39603

0,103

33197034

624

0,203

33197394

264

0,013

33198055

-397

0,354

33197034

785

0,203

33197394

425

0,013

33198055

-236

0,354

33197034

1013

0,203

33197394

653

0,013

33198055

-8

0,354

3568004

1151

3568212

943

3568669

486

0,426

3569386

-231

3569624

-469

3579978

r-value

Direction of correlation

p-values of correlation between mRNA expression z-values and methylation bÃ¨ta values TCGA

-4246

#bp between Illumina probe TCGA and MS-MLPA probe

#bp between MS-MLPA probe and ATG

-5009

CpG-site Illumina probe TCGA

#bp between MS-MLPA probe and TSS

Supplementary Table S8. Continued

0,142

0,000

0,242

0,005

0,159

0,032

0,123

0,000

0,204

-10823

0,000

0,231

3607110

-37955

0,004

0,162

3607222

-38067

0,000

0,235

3607425

-38270

0,000

0,269

19156086

638

0,048

0,113

19156550

174

0,197

19156902

-178

0,322

19157263

-539

0,337

19157710

-986

0,148

19158134

-1410

0,132

10184319

-895

0,347

10184584

-1160

0,520

10189941

-6517

0,003

0,169

119

PART ONE | CHAPTER 4

SUPPLEMENTAL REFERENCES 1. Hoque MO, Prencipe M, Poeta ML, et al. Changes in CpG islands promoter methylation patterns during ductal breast carcinoma progression. Cancer Epidemiol Biomarkers Prev. 2009;18(10):2694-2700. doi: 10.1158/1055-9965.EPI-08-0821 [doi]. 2. Lui EL, Loo WT, Zhu L, Cheung MN, Chow LW. DNA hypermethylation of TIMP3 gene in invasive breast ductal carcinoma. Biomed Pharmacother. 2005;59 Suppl 2:S3635. doi: S0753-3322(05)80079-4 [pii]. 3. Liu Z, Yang L, Cui DX, et al. Methylation status and protein expression of adenomatous polyposis coli (APC) gene in breast cancer. Ai Zheng. 2007;26(6):586-590. doi: 1000-467X200706586 [pii]. 4. Wang L, Tang L, Xie R, Nie W, Chen L, Guan X. P16 promoter hypermethylation is associated with increased breast cancer risk. Mol Med Rep. 2012;6(4):904-908. doi: 10.3892/mmr.2012.1001 [doi]. 5. Celebiler Cavusoglu A, Sevinc AI, Saydam S, et al. Promoter methylation and expression changes of CDH1 and P16 genes in invasive breast cancer and adjacent normal breast tissue. Neoplasma. 2010;57(5):465-472. 6. Murata H, Khattar NH, Kang Y, Gu L, Li GM. Genetic and epigenetic modification of mismatch repair genes hMSH2 and hMLH1 in sporadic breast cancer with microsatellite instability. Oncogene. 2002;21(37):56965703. doi: 10.1038/sj.onc.1205683 [doi]. 7. Bhattacharya N, Mukherjee N, Singh RK, et al. Frequent alterations of MCPH1 and ATM are associated with primary breast carcinoma: Clinical and prognostic implications. Ann Surg Oncol. 2013;20 Suppl 3:S424-32. doi: 10.1245/s10434-012-2715-6 [doi]. 8. Vo QN, Kim WJ, Cvitanovic L, Boudreau DA, Ginzinger DG, Brown KD. The ATM gene is a target for epigenetic silencing in locally advanced breast cancer. Oncogene. 2004;23(58):9432-9437. doi: 1208092 [pii]. 9. Shukla S, Mirza S, Sharma G, Parshad R, Gupta SD, Ralhan R. Detection of RASSF1A and RARbeta hypermethylation in serum DNA from breast cancer patients. Epigenetics. 2006;1(2):88-93. doi: 2679 [pii]. 10. Marzese DM, Hoon DS, Chong KK, et al. DNA methylation index and methylation profile of invasive ductal breast tumors. J Mol Diagn. 2012;14(6):613-622. doi: 10.1016/j.jmoldx.2012.07.001 [doi]. 11. Buyru N, Altinisik J, Ozdemir F, Demokan S, Dalay N. Methylation profiles in breast cancer. Cancer Invest. 2009;27(3):307-312. doi: 10.1080/07357900802350814 [doi]. 12. Fujii H, Biel MA, Zhou W, Weitzman SA, Baylin SB, Gabrielson E. Methylation of the HIC-1 candidate tumor suppressor gene in human breast cancer. Oncogene. 1998;16(16):2159-2164. doi: 10.1038/sj.onc.1201976 [doi]. 13. Parrella P, Scintu M, Prencipe M, et al. HIC1 promoter methylation and 17p13.3 allelic loss in invasive ductal carcinoma of the breast. Cancer Lett. 2005;222(1):75-81. doi: S0304-3835(04)00678-0 [pii]. 14. Privette LM, Gonzalez ME, Ding L, Kleer CG, Petty EM. Altered expression of the early mitotic checkpoint protein, CHFR, in breast cancers: Implications for tumor

120

suppression. Cancer Res. 2007;67(13):6064-6074. doi: 0008-5472.CAN-06-4109 [pii]. 15. Catteau X, Simon P, Noel JC. Stromal expression of matrix metalloproteinase 2 in cancer-associated fibroblasts is strongly related to human epidermal growth factor receptor 2 status in invasive breast carcinoma. Mol Clin Oncol. 2016;4(3):375-378. doi: 10.3892/mco.2015.721 [doi]. 16. Esteller M, Silva JM, Dominguez G, et al. Promoter hypermethylation and BRCA1 inactivation in sporadic breast and ovarian tumors. J Natl Cancer Inst. 2000;92(7):564-569. 17. Bianco T, Chenevix-Trench G, Walsh DC, Cooper JE, Dobrovic A. Tumour-specific distribution of BRCA1 promoter region methylation supports a pathogenetic role in breast and ovarian cancer. Carcinogenesis. 2000;21(2):147-151. 18. Wu Y, Alvarez M, Slamon DJ, Koeffler P, Vadgama JV. Caspase 8 and maspin are downregulated in breast cancer cells due to CpG site promoter methylation. BMC Cancer. 2010;10:32-2407-10-32. doi: 10.1186/1471-2407-10-32 [doi]. 19. Pal R, Srivastava N, Chopra R, et al. Investigation of DNA damage response and apoptotic gene methylation pattern in sporadic breast tumors using high throughput quantitative DNA methylation analysis technology. Mol Cancer. 2010;9:303-4598-9-303. doi: 10.1186/1476-45989-303 [doi]. 20. Zhang HY, Liang F, Jia ZL, Song ST, Jiang ZF. Mutation, methylation and expression in breast cancer patients. Oncol Lett. 2013;6(1):161-168. doi: 10.3892/ol.2013.1331 [doi]. 21. Garcia JM, Silva J, Pena C, et al. Promoter methylation of the PTEN gene is a common molecular change in breast cancer. Genes Chromosomes Cancer. 2004;41(2):117-124. doi: 10.1002/gcc.20062 [doi]. 22. Bosviel R, Durif J, Guo J, et al. BRCA2 promoter hypermethylation in sporadic breast cancer. OMICS. 2012;16(12):707-710. doi: 10.1089/omi.2012.0060 [doi]. 23. Moelans CB, Verschuur-Maes AH, van Diest PJ. Frequent promoter hypermethylation of BRCA2, CDH13, MSH6, PAX5, PAX6 and WT1 in ductal carcinoma in situ and invasive breast cancer. J Pathol. 2011;225(2):222-231. doi: 10.1002/path.2930 [doi]. 24. Branham MT, Marzese DM, Laurito SR, et al. Methylation profile of triple-negative breast carcinomas. Oncogenesis. 2012;1:e17. doi: 10.1038/oncsis.2012.17 [doi]. 25. Sebova K, Zmetakova I, Bella V, et al. RASSF1A and CDH1 hypermethylation as potential epimarkers in breast cancer. Cancer Biomark. 2011;10(1):13-26. doi: 10.3233/ CBM-2012-0230 [doi]. 26. Buhmeida A, Merdad A, Al-Maghrabi J, et al. RASSF1A methylation is predictive of poor prognosis in female breast cancer in a background of overall low methylation frequency. Anticancer Res. 2011;31(9):2975-2981. doi: 31/9/2975 [pii]. 27. Kajabova V, Smolkova B, Zmetakova I, et al. RASSF1A promoter methylation levels positively correlate with estrogen receptor expression in breast cancer patients. Transl Oncol. 2013;6(3):297-304.

Hypermethylation in breast cancer metastases

28. Ahmed IA, Pusch CM, Hamed T, et al. Epigenetic alterations by methylation of RASSF1A and DAPK1 promoter sequences in mammary carcinoma detected in extracellular tumor DNA. Cancer Genet Cytogenet. 2010;199(2):96-100. doi: 10.1016/j. cancergencyto.2010.02.007 [doi]. 29. Tserga A, Michalopoulos NV, Levidou G, et al. Association of aberrant DNA methylation with clinicopathological features in breast cancer. Oncol Rep. 2012;27(5):16301638. doi: 10.3892/or.2011.1576 [doi]. 30. Ramos EA, Camargo AA, Braun K, et al. Simultaneous CXCL12 and ESR1 CpG island hypermethylation correlates with poor prognosis in sporadic breast cancer. BMC Cancer. 2010;10:23-2407-10-23. doi: 10.1186/14712407-10-23 [doi]. 31. Ingvarsson S, Sigbjornsdottir BI, Huiping C, Jonasson JG, Agnarsson BA. Alterations of the FHIT gene in breast cancer: Association with tumor progression and patient survival. Cancer Detect Prev. 2001;25(3):292-298. 32. Yang Q, Yoshimura G, Sakurai T, Kakudo K. The fragile histidine triad gene and breast cancer. Med Sci Monit. 2002;8(7):RA140-4. doi: 2628 [pii]. 33. Saxena A, Dhillon VS, Shahid M, et al. GSTP1 methylation and polymorphism increase the risk of breast cancer and the effects of diet and lifestyle in breast cancer patients. Exp Ther Med. 2012;4(6):1097-1103. doi: 10.3892/ etm.2012.710 [doi]. 34. Lee JS. GSTP1 promoter hypermethylation is an early event in breast carcinogenesis. Virchows Arch. 2007;450(6):637-642. doi: 10.1007/s00428-007-0421-8 [doi]. 35. Zhu W, Qin W, Hewett JE, Sauter ER. Quantitative evaluation of DNA hypermethylation in malignant and benign breast tissue and fluids. Int J Cancer. 2010;126(2):474-482. doi: 10.1002/ijc.24728 [doi]. 36. Verschuur-Maes AH, de Bruin PC, van Diest PJ. Epigenetic progression of columnar cell lesions of the breast to invasive breast cancer. Breast Cancer Res Treat. 2012;136(3):705-715. doi: 10.1007/s10549-012-2301-4 [doi]. 37. Krop IE, Sgroi D, Porter DA, et al. HIN-1, a putative cytokine highly expressed in normal but not cancerous mammary epithelial cells. Proc Natl Acad Sci U S A. 2001;98(17):9796-9801. doi: 10.1073/pnas.171138398 [doi]. 38. Fackler MJ, McVeigh M, Evron E, et al. DNA methylation of RASSF1A, HIN-1, RAR-beta, cyclin D2 and twist in in situ and invasive lobular breast carcinoma. Int J Cancer. 2003;107(6):970-975. doi: 10.1002/ijc.11508 [doi]. 39. Feng W, Orlandi R, Zhao N, et al. Tumor suppressor genes are frequently methylated in lymph node metastases of breast cancers. BMC Cancer. 2010;10:378-2407-10-378. doi: 10.1186/1471-2407-10-378 [doi]. 40. Teramoto A, Tsukuda K, Yano M, et al. Less frequent promoter hypermethylation of DLC-1 gene in primary breast cancers. Oncol Rep. 2004;12(1):141-144. 41. Yuan BZ, Durkin ME, Popescu NC. Promoter hypermethylation of DLC-1, a candidate tumor suppressor gene, in several common human cancers. Cancer Genet Cytogenet. 2003;140(2):113-117. doi:

S016546080200674X [pii]. 42. Takagi H, Sasaki S, Suzuki H, et al. Frequent epigenetic inactivation of SFRP genes in hepatocellular carcinoma. J Gastroenterol. 2008;43(5):378-389. doi: 10.1007/s00535008-2170-0 [doi]. 43. Veeck J, Geisler C, Noetzel E, et al. Epigenetic inactivation of the secreted frizzled-related protein-5 (SFRP5) gene in human breast cancer is associated with unfavorable prognosis. Carcinogenesis. 2008;29(5):991-998. doi: 10.1093/carcin/bgn076 [doi]. 44. Sturgeon SR, Balasubramanian R, Schairer C, Muss HB, Ziegler RG, Arcaro KF. Detection of promoter methylation of tumor suppressor genes in serum DNA of breast cancer cases and benign breast disease controls. Epigenetics. 2012;7(11):1258-1267. doi: 10.4161/epi.22220 [doi]. 45. Cheol Kim D, Thorat MA, Lee MR, et al. Quantitative DNA methylation and recurrence of breast cancer: A study of 30 candidate genes. Cancer Biomark. 2012;11(23):75-88. doi: B17771525002132H [pii]. 46. Swift-Scanlan T, Vang R, Blackford A, Fackler MJ, Sukumar S. Methylated genes in breast cancer: Associations with clinical and histopathological features in a familial breast cancer cohort. Cancer Biol Ther. 2011;11(10):853-865. doi: 15177 [pii]. 47. Stone A, Cowley MJ, Valdes-Mora F, et al. BCL-2 hypermethylation is a potential biomarker of sensitivity to antimitotic chemotherapy in endocrine-resistant breast cancer. Mol Cancer Ther. 2013;12(9):1874-1885. doi: 10.1158/1535-7163.MCT-13-0012 [doi]. 48. Umetani N, Mori T, Koyanagi K, et al. Aberrant hypermethylation of ID4 gene promoter region increases risk of lymph node metastasis in T1 breast cancer. Oncogene. 2005;24(29):4721-4727. doi: 1208538 [pii]. 49. Noetzel E, Veeck J, Horn F, Hartmann A, Knuchel R, Dahl E. Promoter methylation of ID4. A marker for recurrencefree survival in human breast cancer. Pathologe. 2008;29 Suppl 2:319-327. doi: 10.1007/s00292-008-1038-7 [doi]. 50. Chen LF. Tumor suppressor function of RUNX3 in breast cancer. J Cell Biochem. 2012;113(5):1470-1477. doi: 10.1002/jcb.24074 [doi]. 51. Park SY, Kwon HJ, Lee HE, et al. Promoter CpG island hypermethylation during breast cancer progression. Virchows Arch. 2011;458(1):73-84. doi: 10.1007/s00428010-1013-6 [doi]. 52. Kang HF, Dai ZJ, Bai HP, et al. RUNX3 gene promoter demethylation by 5-aza-CdR induces apoptosis in breast cancer MCF-7 cell line. Onco Targets Ther. 2013;6:411417. doi: 10.2147/OTT.S43744 [doi]. 53. Subramaniam MM, Chan JY, Soong R, et al. RUNX3 inactivation by frequent promoter hypermethylation and protein mislocalization constitute an early event in breast cancer progression. Breast Cancer Res Treat. 2009;113(1):113-121. doi: 10.1007/s10549-008-9917-4 [doi]. 54. Du Y, Carling T, Fang W, Piao Z, Sheu JC, Huang S. Hypermethylation in human cancers of the RIZ1 tumor suppressor gene, a member of a histone/protein methyltransferase superfamily. Cancer Res. 2001;61(22):8094-8099.

121

Chapter 5 Willemijne AME Schrijver, Paul J van Diest, Dutch Distant Breast Cancer Metastases Consortium*, Cathy B Moelans * Members are listed in the acknowledgement

Unravelling site-specific breast cancer metastasis: a microRNA expression profiling study Oncotarget. 2016

PART ONE | CHAPTER 5

ABSTRACT Distant metastasis is still the main cause of death from breast cancer. MicroRNAs (miRs) are important regulators of many physiological and pathological processes, including metastasis. Molecular breast cancer subtypes are known to show a site-specific pattern of metastases formation. In this study, we set out to determine the underlying molecular mechanisms of site-specific breast cancer metastasis by microRNA expression profiling. To identify a miR signature for metastatic breast carcinoma that could predict metastatic localization, we compared global miR expression in 23 primary breast cancer specimens with their corresponding multiple distant metastases to ovary (n=9), skin (n=12), lung (n=10), brain (n=4) and gastrointestinal tract (n=10) by miRCURY microRNA expression arrays. For validation, we performed quantitative real-time (qRT) PCR on the discovery cohort and on an independent validation cohort of 29 primary breast cancer specimens and their matched metastases. miR expression was highly patient specific and miR signatures in the primary tumor were largely retained in the metastases, with the exception of several differentially expressed, location specific miRs. Validation with qPCR demonstrated that hsa-miR-106b-5p was predictive for the development of lung metastases. In time, the second metastasis often showed a miR upregulation compared to the first metastasis. This study discovered a metastatic site-specific miR and found miR expression to be highly patient specific. This may lead to novel biomarkers predicting site of distant metastases, and to adjuvant, personalized targeted therapy strategies that could prevent such metastases from becoming clinically manifest.

124

MicroRNAs in breast cancer metastases

INTRODUCTION With a worldwide incidence of 1.67 million and a mortality of 522,000 patients, breast cancer is the leading cause of female cancer and the fifth cause of overall cancer death 1. The majority of solid tumor related mortality is caused by metastatic progression 2, rendering the genetic changes and molecular mechanisms by which cancer cells acquire their metastatic ability one of the most important challenges in breast cancer research. MicroRNAs (miRs) may be involved here, as critical regulators of global mRNA expression in both physiological and pathological processes, including cancer 3. miRs are a group of small non-coding RNAs able to regulate gene expression at the posttranscriptional level by binding to target mRNAs 4. Dysregulation of miRs occurs in various types of cancer and is associated with tumor initiation, drug resistance, and metastasis. Therefore, therapeutic strategies based on modulating the expression levels of miRs and identifying their targets are promising approaches for cancer treatment 5. Even though there have been several studies investigating the role of individual miRs in breast cancer metastasis, often only focussing on their presence in primary tumors 6-8, few global miR expression profiling studies have yet been performed in paired primary breast tumors and their solid distant metastases 9. For lymph node metastases, a metastatic miR signature has already been identified comprising over- and underexpressed miRs 10,11. Extensive knowledge of the miRs involved in distant metastasis could lead to novel biomarkers predicting site of distant metastases and adjuvant targeted therapy strategies that could prevent such metastases from becoming clinically manifest. Intrinsic (molecular) subtypes of breast cancer have been shown to preferentially metastasize to specific sites. E.g., while luminal ERÎą-positive cases prefer to seed to the bone, triple negative and HER2-driven cancer metastases often go to the brain 12. We therefore set out to study global miR expression patterns of 23 primary breast cancer specimens and their corresponding multiple solid distant metastases on selected locations, to pinpoint changes in miR expression during progression from the primary tumor to specific distant sites.

MATERIALS AND METHODS Patients From a series of 481 patients gathered at the department of pathology of the University Medical Center Utrecht in The Netherlands within the framework of a Dutch Cancer Society project on the genotype and phenotype of distant breast cancer metastases 16-19, we selected 25 formalin-fixed paraffin embedded (FFPE) tissue specimens of female primary breast carcinomas and per patient two corresponding distant metastases to lung (n=10), brain (n=4), skin (n=12), ovary (n=10) and gastro-intestinal sub sites (GI; n=10) (cohort 1). 125

PART ONE | CHAPTER 5

Per patient, the metastatic locations could be subdivided into lung-skin (n=3), lung-ovary (n=3), lung-brain (n=4), skin-ovary (n=3), skin-skin (n=3), GI-GI (n=3) and ovary-GI (n=4). Independent validation was performed in 29 matched patients (cohort 2; matched according to age at diagnosis of the primary, molecular subtype, location and time to metastasis) with single metastases to ovary, skin, lung, brain and gastro-intestinal subsites. Clinicopathological characteristics of both cohorts are shown in Table 1. To correct for tissue specific differences in miR expression, four independent normal tissues were selected per tumor location except brain (Supplementary Table S1). Molecular IHC-surrogate subtypes of breast tumors were assigned as follows: Luminal A-like (ER+/PR+, HER2−, Ki-67<15), luminal B-like (ER+/PR+, HER2−, Ki-67>15 or ER+/PR+, HER2+), triple negative or basal-like (ER-/PR-, HER2-) and HER2 enriched (ER-/PR-, HER2+), as before12. The experiments were performed in accordance with the institutional medical ethical guidelines. The use of anonymous or coded left over material for scientific purposes is part of the standard treatment agreement with patients and therefore informed consent was not required according to Dutch law 39. RNA extraction Four-µm thick sections were cut from each FFPE tissue block and stained with haematoxylin and eosin (H&E). The H&E-section was used to guide macro-dissection and to estimate tumor percentage. Only samples containing 80 per cent tumor load or higher (both primary tumor and metastases) were selected. Four 10-µm-thick slides were cut and deparaffinized in xylene. Tumor areas were macro-dissected using a scalpel and areas with necrosis, dense lymphocytic infiltrates, and pre-invasive lesions were intentionally avoided. RNA extraction was carried out with the miRNeasy FFPE kit (Qiagen) according to the manufacturer’s instructions and samples were eluted in 25µL RNAse free water. Total RNA concentration was measured spectrophotometrically (Nanodrop ND-1000, Thermo Scientific Wilmington, DE, USA). Only samples with a concentration of >50 ng/µL and a total amount of 500ng RNA were included for microarray analysis, resulting in 23 matched primary tumor and multiple metastases pairs. miR array profiling All experiments were conducted at Exiqon Services, Denmark. The quality of the total RNA was verified by an Agilent 2100 Bioanalyzer profile. 400 ng total RNA from sample and reference was labeled with Hy3™ and Hy5™ fluorescent label, respectively, using the miRCURY LNA™ microRNA Hi-Power Labeling Kit, Hy3™/Hy5™ (Exiqon, Denmark) following the procedure described by the manufacturer. The Hy3™-labeled samples and a Hy5™-labeled reference RNA sample were mixed pair-wise and hybridized to the miRCURY LNA™ microRNA Array 7th Gen (Exiqon, Denmark), which contains capture probes targeting all miRs for human, mouse or rat registered in the miRBASE 18.0. 126

MicroRNAs in breast cancer metastases

Table 1. Clinicopathological characteristics of 23 primary tumors with two paired metastases included for microRNA expression profiling (cohort 1: microarray profiling and qPCR validation) and a validation cohort of 29 primary tumors and single paired metastases (cohort 2: qPCR validation). Characteristics

Subgroup

Cohort 1 (n=23)

Number of samples

Primaries Metastases

N=23 N=46

N=29 N=30

Age at diagnosis (in years)

Range Mean

29-72 49

31-88 53

0.151

Tumor diameter (in cm)

Range Mean

0.5-4 2.4

1,5-10 3,6

0.057

Histologic type

Ductal Lobular Other

N=16 N=5 N=2

69.6% 21.7% 8.7%

N=19 N=8 N=2

65.5% 27.6% 6.9%

0.814

Histologic grade (Bloom & Richardson)

I II III

N=2 N=11 N=10

8.7% 47.8% 43.5%

N=4 N=13 N=12

13.8% 44.8% 41.4%

0.745

MAI (per 2mmÂ˛)

Range Mean

0-50 15

Molecular subtype

Luminal A Luminal B Triple negative HER2-enriched

N=12 N=7 N=3 N=1

52.2% 30.4% 13.1% 4.3%

N=16 N=4 N=8 N=1

55.2% 13.8% 27.6% 3.4%

0.823

Lymph node status

+ Unknown

N=10 N=4 N=9

43.5% 17.4% 39.1%

N=15 N=10 N=4

51.7% 34.5% 13.8%

0.037*

Metastasis location

Lung Brain Skin Ovary GI

N=10 N=4 N=12 N=10 N=10

21.7% 8.7% 26.2% 21.7% 21.7%

N=6 N=7 N=5 N=6 N=5

20.7% 24.2% 17.2% 20.7% 17.2%

0.491

Time between primary tumor and metastasis (in days)

Range Mean

0-8965 1752

225-3296 1340

Lung

Range Mean Range Mean Range Mean Range Mean Range Mean

467-5502 1615 631-1224 896 0-3458 1682 0-8965 2525 -17193944 1603

367-2480 1250 371-2970 1379 605-1872 1161 225-2345 1278 714-3296 1687

Lung-skin Lung-ovary Lung-brain Skin-ovary Skin-skin GI-GI Ovary-GI

N=3 N=3 N=4 N=3 N=3 N=3 N=4

Brain Skin Ovary GI

Metastasis subgroups

Cohort 2 (n=29) N

0-102 20

0.719

0.222 0.828 0.850 0.225 0.346 0.758

13.0% 13.0% 17.5% 13.0% 13.0% 13.0% 17.5%

*: significant difference between the two cohorts.

127

PART ONE | CHAPTER 5

Hybridization was performed according to the miRCURY LNA™ microRNA Array Instruction manual using a Tecan HS4800™ hybridization station (Tecan, Austria). After hybridization the microarray slides were scanned and stored in an ozone free environment (ozone level below 2.0 ppb) in order to prevent potential bleaching of the fluorescent dyes. The miRCURY LNA™ microRNA Array slides were scanned using the Agilent G2565BA Microarray Scanner System (Agilent Technologies, Inc., USA) and the image analysis was carried out using the ImaGene 9.0 software (BioDiscovery, Inc., USA). The quantified signals were background corrected (Normexp with offset value 10, see 40) and normalized using quantile normalization method, to enable good between-slide normalization to minimize the intensity-dependent differences between the samples. qPCR validation Reverse transcription was performed with the Universal cDNA Synthesis Kit II (miRCURY LNA™ microRNA PCR, Polyadenylation and cDNA synthesis, Exiqon, Denmark). qPCR was performed in duplicate on the ViiA™ 7 Real-Time PCR System (Applied Biosystems) with the ExiLENT SYBR® Green master mix (Exiqon) and ROX as a passive reference. Each run included non-template controls and a calibrator sample. An appropriate endogenous control miR was selected based on the threshold-filtered data from the array profiling. MiRs were included in the analysis if i) they were found in all samples, ii) they had a probe signal of at least 7 in the lowest sample and iii) they had average signal of at least 7.5 across all samples. This data set was run through the NormFinder algorithm 41 to get a stability value for the expression of each miR in the data set. Filtering the most stable hits by assay availability resulted in hsa-miR-483-3p as the best candidate. For the qPCR validation figures, the test and validation cohorts were combined. Prediction of metastasis location by assessing specific miRs in the primary tumor Mann-Whitney U test analyses were performed to compare miR expression in primaries that disseminated to a specific site versus primaries that disseminated to all other tested sites (the rest). For these analyses we also searched the anonymised medical histories of these patients, to find out if they also had metastases in the selected organs (brain, GI, lung, skin, ovary) of which no tumor material was present (Supplementary Table S2). With this information we corrected for selection bias. To ascertain the role of the candidate location-specific miRs in the metastatic cascade, we used miRTarBase, the experimentally validated microRNA-target interactions database 14. The mRNA targets with strong validation evidence (obtained by reporter assay, western blot or qPCR) were imported in ToppGene Suite 15 for pathway enrichment analysis.

128

MicroRNAs in breast cancer metastases

Statistical analyses The Threshold filter data obtained by miR array profiling of almost 2098 miRs was manually checked and non-human miRs were excluded. miRs with no signal or a signal <7 in >25% of samples were excluded as well. Roughly 700 miRs were expressed above background in every sample (Supplementary Figure S1). Unsupervised hierarchical clustering of Threshold filtered data was performed using non-parametric Spearman correlation with R (version 3.2.5). Non-paired analyses on patient differences and clinicopathological characteristics were computed using the Kruskal-Wallis and Mann-Whitney U test. Paired analyses between primary tumors and metastases were done using the Wilcoxon signed-rank test. P-values below 0.05 were considered significant. Thereafter, correction for multiple comparisons was performed by the Benjamini Hochberg procedure. Univariable and multivariable relationships were tested by logistic regression (method: Forward LR) with 95% confidence intervals (CI). All statistical calculations were done with IBM SPSS Statistics 21 and visualized with GraphPad Prism 6 and R.

RESULTS miRs differentially expressed in primary breast cancer versus corresponding multiple distant metastases First, the samples of cohort 1 were subjected to miRCURY microRNA expression array profiling. A principal component analysis (PCA) of the samples showed only small variances between the common reference channels, indicating that the observed variances of the tissue samples were likely related to biological differences between the tumor samples and not technical variances (Supplementary Figure S2). When PCA was performed for the most varying components (PC1 and PC2), the primary tumors and some but not all of the metastases locations (GI, skin and lung) clustered only vaguely together (Figure 1). Molecular subtype and patient number seemed to be the most important clustering variables. An unsupervised cluster analysis of all detected miRs did not readily separate the samples into groups, but some agreement in miR expression was seen in ERÎą+ primary tumors and GI and ovarian metastases. No clear distinction was observed based on the samples being either primary tumors or distant metastases, but for 9/23 (39%) of the patients, the primary tumor and (one of) the corresponding distant metastasis clustered together, indicating very similar miR expression patterns (Figure 2). For this group, the time between the primary and the metastasis was significantly shorter (p=0.021; mean 947 days; range 0-3867 days) than for the group that did not cluster together (mean 1988 days; range 0-8965). However, when individual miR expression was correlated to time between primary and metastases, 129

PART ONE | CHAPTER 5

no significant relation was found. For another 9/23 of the patients, the two metastases clustered roughly together, but here no significant differences were seen in time span when compared to patient samples that did not cluster.

SampleType

metastasis primary

PC2

Site GI

brain skin lung breast ovary

-2

-4

-10

-5

PC1

Figure 1. Principal Component Analysis (PCA) of most varying components (PC1 and PC2) between all primary tumors (n=23) and paired multiple distant metastases (n=46) of cohort 1, subjected to miRCURY microRNA expression array profiling.

In non-paired analyses, 48 miRs were identified that were differentially expressed between primary tumors and metastases (21 upregulated and 27 downregulated in the metastases versus the primary tumors). In paired analyses per metastasis location, 101 miRs were identified that exhibited a significantly altered expression between paired primary tumors and metastases (Supplementary Table S3). Almost no overlap was seen in differentially expressed miRs per metastatic location, only hsa-miR-3201 was dysregulated in lung and ovarian metastases. Interestingly, compared to other metastatic localizations, ovaries generally demonstrated more differentially expressed miRs (n=86 versus n=4 for skin, n=11 for lung, n=2 for GI and n=0 for brain).

130

3.5

SUBTYPE

META LOCATION

PRIM LOCATION

Value

15M1

15P

4P 19M2

8M1 16M2

8M2 16P

25M2

4M1

21P

17M2

17M1 17P

4M2

1M1

6M1 19M1

2M1 19P

15M2

25P

20P

20M2

8P 11P

18P

16M1

2P 12P

9P 24M1

24P

2M2

13P

25M1

23M1

24M2

5M1

14M1

6M2

11M1

14P

3M1

10M2

23P

14M2

10M1

5M2

13M1

20M1

9M2

18M2

23M2

1M2

12M1

21M1

10P

18M1

13M2

12M2

21M2

9M1

3M2

X3656 X4459 X4636 X5523p X4473 X665 X4472 X47323p X4533 X46533p X31585p X47955p X3149 X8773p X4468 X1493p X5745p X3611 X243p X23a3p X5701 X165p X4500 X125a5p X1413p X200c3p X1433p X23b3p X223p X29b3p let7b5p X125b5p X4451 X4698 X4478 X35915p X2053p X47123p X519e5p X921 X1273e X6755p X663a X47325p X103a3p X1264 X4780 X4788 X2113 X51873p X711 X46943p X46335p X3621 X4488 X47253p X5572 X3202 X3687 X4484 X1833p X1184 X3915 X2143p X4455 X7445p X31563p X36675p X1827 X323p X1973 X46953p X47093p X4421 X638 X4530 X3646 X5704 X5681b X1505p X302a3p X47283p X47475p X548ap5p X47953p X4301 X4286 X3182 X47145p X215p X3178 X4507 X46675p X498 X943 X3941 X1260a X55813p X4644 X4657 X4235p X50025p X1255a X3353p let7e5p X548as3p X4306 X1469 X47075p X4335p X255p X3196 X1255b23p X47645p X36065p X99b5p X47423p X548an X44235p X4447 X8775p X4903p X21153p X378a3p X5845p X4431 X3660 X1555p X1273f X4654 X47625p X5423p X12475p X4329 X1290 X934 X4450 X1284 X1976 X3924 X4955p X451b X4233p X1273a X4297 X46455p let7d5p let7i5p X4791 X107 X4245p X106b5p X3133 X200a3p let7c5p X1955p X29a3p X1013p X1423p X199a3p X199a5p X29c3p X15a5p X30b5p X200b3p X1263p X27a3p X30c5p X26b5p let7g5p let7a5p X10a5p X1455p X10b5p X451a X2055p X47255p X4853p X1343p X46773p X1321 X50063p X5743p X4540 X3148 X664a3p X46463p X1299 X3935 X3654 X1248 X4324 X8743p X4425 X4682 X4258 X551a let7a23p X12363p X518e5p X5853p X147b X4835p X47235p X4513 X4449 X46873p X13045p X4436b5p X12853p X4429 X4784 X47775p X21165p X4739 X4893p X44455p X630 X55843p X5193 X23553p let7d3p X516b5p X12855p X4426 X642b5p X4279 X4290 X4913p X39405p X1246 X4443 X1260b X31243p X4284 X6603p X19085p X47505p X4268 X4419b X4456 X4508 X3685 X4833p X4285 X3976 X4475 X47265p X4516 X46393p X513a5p X31363p X5684 X1587 X1275 X4531 X4532 X4299 X1273g3p X47643p X4497 X4505 X2043p X664b3p X48003p X4454 X5100 X47875p X4467 X371b5p X3960 X47083p

GI lung skin brain ovary primary

skin & ovary skin & lung skin & skin lung & ovary lung & brain ovary & GI GI & GI metastasis

Luminal A Luminal B Triple Negative HER2 driven

Figure 2. Hierarchical cluster analysis of Threshold filtered data of all primary tumors (n=23) and paired multiple distant metastases (n=46) of cohort 1, subjected to the miRCURY microRNA expression array.

Color Key

MicroRNAs in breast cancer metastases

131

PART ONE | CHAPTER 5

Validation of miRs differentially expressed in primary breast tumors versus corresponding distant metastasis as analyzed by quantitative PCR analysis The following analyses were performed on cohort 1: i) primary tumors versus metastases (paired and unpaired), ii) primary tumors that disseminated to a specific site (brain, lung, GI, ovary or skin) and iii) differences between molecular subtypes. The miRs with the highest fold changes in these analyses were validated using real-time PCR. Moderate to good correlations were seen between microarray data and qPCR validation (Supplementary Table S4). Subsequently, these miRs were validated in the independent cohort 2. Only hsa-miR-16-5p was excluded from further validation due to low correlations. qPCR validation of hsa-miR-200a-3p, hsa-miR-29b-3p, hsa-miR-451a, hsa-miR-125b-5p, hsamiR-143-3p and hsa-miR-3182 in both cohorts are shown in Supplementary Figure S3. Comparison between multiple metastases per patient A general tendency was observed towards a higher expression in the second metastasis compared to the first. Especially in the significantly upregulated miRs, an increase in fold change was shown in the second metastasis (compared to the primary tumor) relative to the first metastasis (Figure 3). However, no significant correlation was seen between miR expression and time between primary and metastasis. As an example, the miR with the highest upregulation (miR-451a; FC M1: 2.55 and FC M2: 2.91) was plotted against

A Fold Change of upregulated miRs miR-451a

2.5 2.0 1.8 1.6 1.4 1.2 Metastasis 1 Metastasis 2

p < 0.01

Fold Change compared to primary

3.0

B Fold Change of downregulated miRs

p < 0.01

0.9

0.8

0.7

0.60 0.55 Metastasis 1 Metastasis 2

Figure 3. Fold change of significantly upregulated (a) and downregulated (b) miRs of both metastases (compared to the primary tumor). Results of the Threshold filtered data of the miRCURY microRNA expression array of cohort 1. miR-451a shows the highest upregulation. Wilcoxon signed-rank test was used with * p < 0.05, ** p < 0.01, *** p < 0.001

132

MicroRNAs in breast cancer metastases

timespan between primary and metastasis and metastatic location (Supplementary Figure S4). Location of metastasis and patient specific differences were the largest contributors to the differences in miR expression. Time to distant recurrence was not significantly different (p=0.637) between the metastatic locations of cohort 1 (Supplementary Figure S5). Expression differences in primary tumors and metastases of miRs known to have an oncogenic and tumor suppressive potential in primary breast tumors A literature search resulted in 38 oncogenic and tumor suppressive miRs that play a role in the metastatic cascade in primary tumors (Supplementary Table S5) 4,8,13. Expression of these miRs was evaluated in the primary tumors compared to the metastases of cohort 1 and 26 oncogenic (n=14) and tumor suppressive (n=12) miRs were expressed in all tested samples. Of the fourteen selected oncogenic miRs, nine miRs with a role in EMT, invasion and angiogenesis were significantly upregulated in the metastases compared to the primaries (Figure 4a). Of the twelve selected tumor suppressive miRs, five miRs were significantly downregulated in metastases (Figure 4b). Prediction of site-specific metastasis by miR expression in the primary tumor miRs predicting metastasis location based on expression levels in the primary tumor are listed in Supplementary Table S3. In a multivariable regression model corrected for molecular subtype, histologic type, histologic grade, tumor diameter, lymph node status, age at diagnosis and MAI, miR-106b-5p was an independent predictor of lung and GI metastases, miR-7-5p of skin metastases and miR-1273g-3p of ovarian metastases (Figure 5a). These findings were validated by qPCR in cohorts 1 and 2. Only miR-106b-5p remained an independent predictor of metastases to the lung. ROC-curve analysis showed an AUC of 0.828 (95% CI 0.701-0.955; SE 0.07; Supplementary Figure S6) with an RQ value of â&#x2030;Ľ1.208 (sensitivity 0.94; 1-specificity 0.34). Although not significant, in all three tested miRs (hsa-miR-106b-5p, hsa-miR-1273g-3p and hsa-miR-7-5p) the same expression trend was observed in the metastases. In (independent) normal tissue, a significantly lower expression was seen compared to primary tumors and metastases (Figure 5b). To ascertain the role of hsa-miR-106b-5p, hsa-miR-1273g-3p and hsa-miR-7-5p in the metastatic process, we used miRTarBase, the experimentally validated microRNA-target interactions database 14. The mRNA targets with strong validation evidence (obtained by reporter assay, western blot or qPCR) are listed in Supplementary Table S6. These target genes were subsequently imported in ToppGene Suite 15 to find enriched pathways. Unfortunately, for hsa-miR-1273g-3p, there are no known targets with strong evidence. Hsa-miR-106b-5p appears to have an important role in both lung and breast cancer. Furthermore, this miR is a key player in cell cycle control and regulation and the cellular response to stress. Hsa-miR-7-5p plays a role in breast cancer, melanoma and bacterial invasion of epithelial cells. Regarding metastatic properties, focal adhesion, apoptosis and angiogenesis are significantly enriched pathways. 133

PART ONE | CHAPTER 5

Oncogenic miRs 13.5 13.0

Threshold filter data (expression)

12.5 12.0

p < 0.01

11.5 11.0

p < 0.05

p < 0.01

10.5

p < 0.01

10.0 9.5

p < 0.05

9.0

p < 0.05

p < 0.01

8.5

p < 0.01

8.0 7.5 7.0 6.5

EMT

P-1908-5p

M-1908-5p

P-199a-5p

M-199a-5p

P-93-5p

M-93-5p

P-23b-3p

M-23b-3p

P-21-5p

M-21-5p

P-24-3p

M-24-3p

P-10b-5p

M-10b-5p

P-221-3p

M-221-3p

P-222-3p

M-222-3p

P-107

M-107

P-103a-3p

M-103a-3p

P-182-5p

M-182-5p

P-29a-3p

M-29a-3p

P-150-5p

5.5

M-150-5p

6.0

Angiogenesis Invasion

Tumor suppressive miRs 11.5 p < 0.05

Threshold filter data (expression)

11.0 10.5 10.0 9.5

p < 0.05

9.0 8.5 p < 0.05

8.0 7.5 7.0 6.5

EMT

M-335-3p

P-335-3p

P-146b-5p

M-146b-5p

M-22-5p

P-22-5p

M-let-7i-5p

P-let-7i-5p

M-let-7a-2-3p

P-let-7a-2-3p

M-let-7d-3p

P-let-7d-3p

P-let-7e-5p

M-let-7e-5p

P-205-3p

M-205-3p

M-7-5p

P-7-5p

M-7-2-3p

P-7-2-3p

M-194-3p

P-194-3p

M-155-5p

5.5

P-155-5p

6.0

Migration Invasion

Figure 4. Expression differences in primary tumors and metastases of miRs known to have an oncogenic and tumor suppressive potential in primary breast tumors. Data from cohort 1. A) Expression of 15 oncogenic miRs with a role in EMT, invasion and angiogenesis in primary tumors versus metastases. B) Expression of 12 tumor suppressive miRs with a role in EMT, invasion and migration in primary tumors versus metastases. Wilcoxon signed-rank test was used with * p < 0.05, ** p < 0.01, *** p < 0.001. Purple: miRs with a role in EMT, orange: miRs with a role in EMT and invasion, green: miRs with a role in invasion, blue: miRs with a role in invasion and angiogenesis, red: miRs with a role in angiogenesis, pink: miRs with a role in invasion and migration, turquoise: miRs with a role in migration.

134

MicroRNAs in breast cancer metastases

qPCR hsa-miR-106b-5p in lung

p < 0.05

10 9

p < 0.05

p < 0.001

8 7

Primaries

Metastases

Primaries

primaries that disseminated to GI primaries that did not disseminate to GI GI metastases non GI metastases GI normal tissue non GI normal tissue

p < 0.01

8 7

10 5

6 5

p < 0.001

p < 0.05

0 Primaries

Primaries

Metastases

Normal tissue

qPCR hsa-miR-7-5p in skin

Microarray hsa-miR-7-5p in skin

p < 0.05 p < 0.05

9 p < 0.05

p < 0.01 p < 0.05

p < 0.01

primaries that disseminated to skin primaries that did not disseminate to skin skin metastases non skin metastases skin normal tissue non skin normal tissue

45 20

15 10

5 6

0 Primaries

Primaries

Metastases

p < 0.001

p < 0.05

Normal tissue

qPCR hsa-miR-1273g-3p in ovary

Microarray hsa-miR-1273g-3p in ovary

p < 0.05 p < 0.01

p < 0.001 p < 0.05

primaries that disseminated to ovary primaires that did not disseminate to ovary ovarian metastases non ovarian metastases ovarian normal tissue non ovarian normal tissue

30 15

Threshold filter data (expression)

Normal tissue

p < 0.05

Metastases

qPCR hsa-miR-106b-5p in GI

Threshold filter data (expression)

primaries that disseminated to lung primaries that did not disseminate to lung lung metastases non lung metastases lung normal tissue non lung normal tissue

Microarray hsa-miR-106b-5p in GI

Threshold filter data (expression)

p < 0.01

Threshold filter data (expression)

Microarray hsa-miR-106b-5p in lung

10 5

9 Primaries

Metastases

Primaries

Metastases

Normal tissue

Figure 5. MiRs shown to be predictive in the primary tumor for metastasis location. A) with microarray profiling of cohort 1; miR-106b-5p was an independent predictor for lung and GI metastases, miR-7-5p for skin metastases and miR-1273g-3p for ovarian metastases. B) qPCR validation on the same cohort plus an independent cohort revealed the same trend as in A), but only miR-106b-5p remained an independent predictor for lung metastases. Mann-Whitney U test was used with * p < 0.05, ** p < 0.01, *** p < 0.001

135

PART ONE | CHAPTER 5

DISCUSSION Systemic therapies are still largely guided by the characteristics of the primary tumor, while discordance between primary tumors and metastases are often encountered 16-19. Molecular differences between matched primary tumors and metastatic lesions have the potential to reveal novel, potentially targetable drivers of metastatic progression. In this study we performed miR expression profiling in primary breast tumors and matched multiple metastases. We demonstrated that the expression of known â&#x20AC;&#x2DC;metastamiRsâ&#x20AC;&#x2122; was generally higher in the metastases compared to the primary tumors. Also, the abundance of specific miRs in the primary tumor seemed to be metastasis location-specific, which could potentially be exploited to gain more knowledge about the metastatic cascade. Especially hsa-miR-106b-5p seems to be a predictor of lung metastases. Several studies have examined the role of individual miRs in primary metastatic breast cancer. For example, miR-148a, miR-33a, miR-34a and miR-199a/b-3p are thought to suppress metastasis 20 and to inhibit tumor cell migration and invasion 21-23. In contrast, miR-762 and miR-1228, amongst others, are thought to promote breast cancer cell proliferation and invasion 24 and metastasis 25. However, little is known about the full miR profile of primary breast tumors compared to paired distant metastases. Gravgaard et al. already performed expression profiling of primary breast tumors and matched distant metastases to liver (n=5) and brain (n=9) 9. In line with Gravgaard et al., we observed a higher miR expression similarity between primary tumors and metastases when the recurrence interval was shorter. However, a time dependent effect on the individual miR level was not found. Furthermore, they reported 97 altered miRs between primary tumors and brain metastases, while we found none, possibly due to our smaller sample size (n=4). Baffa et al. compared thirteen primary breast tumors and matched lymph node metastases and found five upregulated and six downregulated miRs 10. Only upregulation of hsa-miR30b and downregulation of hsa-miR-125b corresponded to our findings, suggesting that some miRs have an influence on the metastatic cascade in general, while others could correlate to location specificity. In another study that compared miR expression in primary breast tumors and lymph node metastases (n=97) a downregulation of hsa-miR-151-5p was seen in metastases, while we detected an upregulation 26. This may be explained by the fact that our metastases were distant, with different microenvironments and progression routes playing a role. This is further supported by the fact that specific miRs can have an oncogenic potential in one cancer type and a tumor suppressive effect in the other 27. Several gastric and colorectal cancer studies discovered the same possible pro-metastatic miRs as we did, suggesting that these miRs can influence cancer progression in general (miR-10a: 28; miR-335: 29; miR-143: 30). However, there is virtually no overlap in candidate miRs across prior studies. Whether this arises from dissimilarities in tumor types, metastasis locations, patient characteristics or the use of varying techniques remains unclear. Findings 136

MicroRNAs in breast cancer metastases

with qPCR and microarray did not always correlate well 31, which was in this study overcome by selecting only miRs with high fold changes for validation. Overall, there was a tendency for higher expression of certain oncogenic miRs in the metastases compared to the primaries. This observation is in line with Huang et al., who reported a higher expression of hsa-miR-373 in paired lymph node metastases of eleven patients 32. In addition, they reported higher levels of hsa-miR-373 in the primary tumors that disseminated to lymph nodes compared to lymph node negative samples. We saw a similar tendency for hsa-miR-106b-5p, but this should be validated in a larger group. Also Korpal et al. found a higher expression of hsa-miR-200s in paired lung metastases, stressing its potential role in metastatic colonization 33. Overall, only few differentially expressed miRs were found between primary tumors and matched metastases (with an exception for ovarian metastases), suggesting that miR expression is largely retained in metastases. Ferracin et al. already showed that primary tumors of different origin display a distinct miR expression profile and that metastases retain a large part of these miRs 34. The latter was visualized by unsupervised hierarchical clustering, where primary tumors and metastases clustered together. We also observed some clustering of primaries and paired metastases, but the metastatic locations hardly clustered. This could be due to patient specific differences or to too much variation within subsites (different GI and skin locations). Why the ovarian metastases differ so clearly from the other metastatic sites remains to be elucidated. The H&E slides were reinspected by an experienced breast pathologist, who was convinced about their origin from the breast. However, there is still a small chance that a part of these tumors are primary ovarian cancers since metastatic carcinomas may mimic primary ovarian carcinomas. Also, mucinous ovarian carcinomas are difficult to distinguish from metastatic adenocarcinomas 35. The biggest microRNA expression differences were seen between normal tissue and tumor tissue (primaries and metastases). Our findings agree with Neerincx et al. who described that miR expression in primaries and metastases is similar, but differs largely from normal tissue 36. Other studies frequently use â&#x20AC;&#x2DC;normal tissueâ&#x20AC;&#x2122; of regions near the tumor, which may introduce bias when presumably normal tissue has already been affected. By comparing to reference material of all metastasis locations we tried to rule out tissue-specific background as much as possible. Because of limited availability, we used (unpaired) normal tissue that did not originate from the same patients, which may have introduced patient-specific background differences. Smeets et al. developed a predictor of lymph node metastases based on miR expression profiling of the primary tumor 11. Here, we demonstrated that high expression of hsa-miR106b-5p in the primary tumor can predict lung metastases. A review about the influence of miR-106 in cancer showed a moderate accuracy in identifying gastric and colorectal cancer and lymphoma patients 37. In breast cancers, miR-106b was found to be associated with a high risk of recurrence, and was mentioned as a putative plasma marker for risk assessment 38. 137

PART ONE | CHAPTER 5

Certain limitations to our study include the small samples sizes per metastatic location and potential tumor heterogeneity, which may explain some of the observed differences between primary tumors and paired metastases. Furthermore, by making use of microarray technology we may have underestimated downregulated miRs, since the applied threshold prevents the detection of lowly expressed miRs. In summary, we have shown that primary tumor miR expression patterns are largely retained in metastases, except from some location specific miRs. miR-106b-5p expression in the primary tumor seems to be an independent predictor of lung metastases. miR-7-5p, miR-1273g-3p and miR-106b-5p could be predictors for skin, ovarian and GI metastases as well, respectively, but these results require validation in a larger and independent cohort. This miR expression profiling study thereby identified possible therapeutic targets and predictive markers of site-specific metastasis. The large patient specific differences further stress the uniqueness of individual tumors and thereby the need for individualized treatment.

Acknowledgements The Dutch Distant Breast Cancer Metastases Consortium included the Departments of Pathology from the University Medical Center Utrecht, Meander Medical Center Amersfoort, Hospital Gelderse Vallei Ede, Academic Medical Center Amsterdam, Medical Center Alkmaar, Radboud University Nijmegen Medical Center, Canisius Wilhelmina Hospital Nijmegen, VU University Medical Center Amsterdam, The Netherlands Cancer Institute Amsterdam, Groene Hart Hospital Gouda, University Medical Center Groningen, St Antonius Hospital Nieuwegein, Diakonessenhuis Utrecht, Isala klinieken Zwolle, Erasmus Medical Center Rotterdam, Gelre Hospital Apeldoorn and the Laboratories for Pathology Dordrecht, ‘s Hertogenbosch, Terneuzen, Symbiant Zaandam, Sazinon, Hoogeveen, Oost Nederland Enschede (LabPON), all in The Netherlands. We thank Stichting PALGA for the national query for cases. This publication was realized with support of ‘Stichting Vrienden UMC Utrecht’ and a Sister’s Hope. We would like to thank Annelisa Cornel for helping with some qPCR assays and Antonio Sorrentino for advice and contact with Exiqon Services, Vedbæk, Denmark.

138

MicroRNAs in breast cancer metastases

REFERENCES 1. Ferlay J, Soerjomataram I, Dikshit R, et al. Cancer incidence and mortality worldwide: Sources, methods and major patterns in GLOBOCAN 2012. Int J Cancer. 2015;136(5):E359-86. doi: 10.1002/ijc.29210 [doi]. 2. Gupta GP, Massague J. Cancer metastasis: Building a framework. Cell. 2006;127(4):679-695. doi: S00928674(06)01414-0 [pii]. 3. Pillai RS. MicroRNA function: Multiple mechanisms for a tiny RNA? RNA. 2005;11(12):1753-1761. doi: 11/12/1753 [pii]. 4. van Schooneveld E, Wildiers H, Vergote I, Vermeulen PB, Dirix LY, Van Laere SJ. Dysregulation of microRNAs in breast cancer and their potential role as prognostic and predictive biomarkers in patient management. Breast Cancer Res. 2015;17:21-015-0526-y. doi: 10.1186/s13058015-0526-y [doi]. 5. Takahashi RU, Miyazaki H, Ochiya T. The roles of MicroRNAs in breast cancer. Cancers (Basel). 2015;7(2):598-616. doi: 10.3390/cancers7020598 [doi]. 6. Ma L, Weinberg RA. Micromanagers of malignancy: Role of microRNAs in regulating metastasis. Trends Genet. 2008;24(9):448-456. doi: 10.1016/j.tig.2008.06.004 [doi]. 7. Nicoloso MS, Spizzo R, Shimizu M, Rossi S, Calin GA. MicroRNAs--the micro steering wheel of tumour metastases. Nat Rev Cancer. 2009;9(4):293-302. doi: 10.1038/nrc2619 [doi]. 8. Wang L, Wang J. MicroRNA-mediated breast cancer metastasis: From primary site to distant organs. Oncogene. 2012;31(20):2499-2511. doi: 10.1038/onc.2011.444 [doi]. 9. Gravgaard KH, Lyng MB, Laenkholm AV, et al. The miRNA-200 family and miRNA-9 exhibit differential expression in primary versus corresponding metastatic tissue in breast cancer. Breast Cancer Res Treat. 2012;134(1):207-217. doi: 10.1007/s10549-012-1969-9 [doi]. 10. Baffa R, Fassan M, Volinia S, et al. MicroRNA expression profiling of human metastatic cancers identifies cancer gene targets. J Pathol. 2009;219(2):214-221. doi: 10.1002/ path.2586 [doi]. 11. Smeets A, Daemen A, Vanden Bempt I, et al. Prediction of lymph node involvement in breast cancer from primary tumor tissue using gene expression profiling and miRNAs. Breast Cancer Res Treat. 2011;129(3):767-776. doi: 10.1007/s10549-010-1265-5 [doi]. 12. Kennecke H, Yerushalmi R, Woods R, et al. Metastatic behavior of breast cancer subtypes. J Clin Oncol. 2010;28(20):3271-3277. doi: 10.1200/JCO.2009.25.9820 [doi]. 13. Bouyssou JM, Manier S, Huynh D, Issa S, Roccaro AM, Ghobrial IM. Regulation of microRNAs in cancer metastasis. Biochim Biophys Acta. 2014;1845(2):255-265. doi: 10.1016/j.bbcan.2014.02.002 [doi]. 14. Chou CH, Chang NW, Shrestha S, et al. miRTarBase 2016: Updates to the experimentally validated miRNA-target interactions database. Nucleic Acids Res. 2016;44(D1):D239-47. doi: 10.1093/nar/gkv1258 [doi]. 15. Chen J, Bardes EE, Aronow BJ, Jegga AG. ToppGene suite for gene list enrichment analysis and candidate gene

prioritization. Nucleic Acids Res. 2009;37(Web Server issue):W305-11. doi: 10.1093/nar/gkp427 [doi]. 16. Hoefnagel LDC, Beelen KJ, Opdam M, Vincent A, Linn SC, Van Diest PJ. Increased expression of phosporylated mTOR in metastatic breast tumors compared to primary tumors in patients who received adjuvant endocrine therapy. Cancer Res. 2012;72(24). 17. Hoefnagel LDC, van der Groep P, van de Vijver MJ, et al. Discordance in ER alpha, PR and HER2 receptor status across different distant breast cancer metastases within the same patient. Annals of Oncology. 2013;24(12):30173023. doi: 10.1093/annonc/mdt390. 18. Hoefnagel LDC, van de Vijver MJ, van Slooten H, et al. Receptor conversion in distant breast cancer metastases. Breast Cancer Research. 2010;12(5):R75. doi: 10.1186/ bcr2645. 19. Hoefnagel LDC, Moelans CB, Meijer SL, et al. Prognostic value of estrogen receptor alpha and progesterone receptor conversion in distant breast cancer metastases. Cancer. 2012;118(20):4929-4935. doi: 10.1002/cncr.27518. 20. Xu X, Zhang Y, Jasper J, et al. MiR-148a functions to suppress metastasis and serves as a prognostic indicator in triple-negative breast cancer. Oncotarget. 2016;7(15):20381-20394. doi: 10.18632/oncotarget.7953 [doi]. 21. Zhang C, Zhang Y, Ding W, Lin Y, Huang Z, Luo Q. MiR33a suppresses breast cancer cell proliferation and metastasis by targeting ADAM9 and ROS1. Protein Cell. 2015;6(12):881-889. doi: 10.1007/s13238-015-0223-8 [doi]. 22. Li SQ, Wang ZH, Mi XG, Liu L, Tan Y. MiR-199a/b-3p suppresses migration and invasion of breast cancer cells by downregulating PAK4/MEK/ERK signaling pathway. IUBMB Life. 2015;67(10):768-777. doi: 10.1002/iub.1433 [doi]. 23. Li G, Yao L, Zhang J, et al. Tumor-suppressive microRNA34a inhibits breast cancer cell migration and invasion via targeting oncogenic TPD52. Tumour Biol. 2016;37(6):74817491. doi: 10.1007/s13277-015-4623-4 [doi]. 24. Li Y, Huang R, Wang L, et al. microRNA-762 promotes breast cancer cell proliferation and invasion by targeting IRF7 expression. Cell Prolif. 2015;48(6):643-649. doi: 10.1111/cpr.12223 [doi]. 25. Lin L, Liu D, Liang H, Xue L, Su C, Liu M. MiR-1228 promotes breast cancer cell growth and metastasis through targeting SCAI protein. Int J Clin Exp Pathol. 2015;8(6):6646-6655. 26. Krell J, Frampton AE, Jacob J, et al. The clinico-pathologic role of microRNAs miR-9 and miR-151-5p in breast cancer metastasis. Mol Diagn Ther. 2012;16(3):167-172. doi: 10.2165/11632740-000000000-00000 [doi]. 27. Calin GA, Croce CM. MicroRNA signatures in human cancers. Nat Rev Cancer. 2006;6(11):857-866. doi: nrc1997 [pii]. 28. Chen W, Tang Z, Sun Y, et al. miRNA expression profile in primary gastric cancers and paired lymph node metastases indicates that miR-10a plays a role in metastasis from primary gastric cancer to lymph nodes. Exp Ther Med. 2012;3(2):351-356. doi: 10.3892/ etm.2011.411 [doi].

139

PART ONE | CHAPTER 5

29. Vickers MM, Bar J, Gorn-Hondermann I, et al. Stagedependent differential expression of microRNAs in colorectal cancer: Potential role as markers of metastatic disease. Clin Exp Metastasis. 2012;29(2):123-132. doi: 10.1007/s10585-011-9435-3 [doi]. 30. Kulda V, Pesta M, Topolcan O, et al. Relevance of miR-21 and miR-143 expression in tissue samples of colorectal carcinoma and its liver metastases. Cancer Genet Cytogenet. 2010;200(2):154-160. doi: 10.1016/j. cancergencyto.2010.04.015 [doi]. 31. Koshiol J, Wang E, Zhao Y, Marincola F, Landi MT. Strengths and limitations of laboratory procedures for microRNA detection. Cancer Epidemiol Biomarkers Prev. 2010;19(4):907-911. doi: 10.1158/1055-9965.EPI-10-0071 [doi]. 32. Huang Q, Gumireddy K, Schrier M, et al. The microRNAs miR-373 and miR-520c promote tumour invasion and metastasis. Nat Cell Biol. 2008;10(2):202-210. doi: 10.1038/ncb1681 [doi]. 33. Korpal M, Ell BJ, Buffa FM, et al. Direct targeting of Sec23a by miR-200s influences cancer cell secretome and promotes metastatic colonization. Nat Med. 2011;17(9):1101-1108. doi: 10.1038/nm.2401 [doi]. 34. Ferracin M, Pedriali M, Veronese A, et al. MicroRNA profiling for the identification of cancers with unknown primary tissue-of-origin. J Pathol. 2011;225(1):43-53. doi: 10.1002/path.2915 [doi]. 35. Bruls J, Simons M, Overbeek LI, Bulten J, Massuger LF, Nagtegaal ID. A national population-based study provides

140

insight in the origin of malignancies metastatic to the ovary. Virchows Arch. 2015;467(1):79-86. doi: 10.1007/ s00428-015-1771-2 [doi]. 36. Neerincx M, Sie DL, van de Wiel MA, et al. MiR expression profiles of paired primary colorectal cancer and metastases by next-generation sequencing. Oncogenesis. 2015;4:e170. doi: 10.1038/oncsis.2015.29 [doi]. 37. Zhang B, Xu CW, Shao Y, et al. Evaluation and identification of microRNA-106 in the diagnosis of cancer: A meta-analysis. Int J Clin Exp Med. 2014;7(10):3746-3756. 38. Zheng R, Pan L, Gao J, et al. Prognostic value of miR-106b expression in breast cancer patients. J Surg Res. 2015;195(1):158-165. doi: 10.1016/j.jss.2014.12.035 [doi]. 39. van Diest PJ. No consent should be needed for using leftover body material for scientific purposes. for. BMJ. 2002;325(7365):648-651. 40. Ritchie ME, Silver J, Oshlack A, et al. A comparison of background correction methods for two-colour microarrays. Bioinformatics. 2007;23(20):2700-2707. doi: btm412 [pii]. 41. Andersen CL, Jensen JL, Orntoft TF. Normalization of real-time quantitative reverse transcription-PCR data: A model-based variance estimation approach to identify genes suited for normalization, applied to bladder and colon cancer data sets. Cancer Res. 2004;64(15):52455250. doi: 10.1158/0008-5472.CAN-04-0496 [doi].

MicroRNAs in breast cancer metastases

SUPPLEMENTAL

Supplementary Figure S1. Plot illustrating the number of microRNAs detectable in the microarray above background threshold (1.2 times the 25th percentile of the overall signal intensity of the slide) for each sample of cohort 1(out of a total of 2098 possible microRNAs).

141

PART ONE | CHAPTER 5

Supplementary Figure S2. Matrix PCA plot of the common reference channels of the microRNAs detectable with the microarray of cohort 1. Analyses were performed on all tumor samples. The observed variances of the tissue samples were likely related to biological differences between the tumor samples and not technical variances, because most variation was seen in PC1 and PC2.

142

MicroRNAs in breast cancer metastases

A 25

Primaries vs. metastases (unpaired) p<0.05

Primary tumors Metastases

Primaries vs. ovarian metastases (paired) p<0.01

Primary tumors Metastases

p<0.001

10 p<0.01

p<0.01

hsa-miR-200a-3p

hsa-miR-29b-3p

hsa-miR-125b-5p

hsa-miR-451a

hsa-miR-143-3p

qPCR hsa-miR-3182 in brain p<0.05

Primaries that disseminated to brain Primaries that did not disseminate to brain Brain metastases Non brain metastases

Primaries

Metastases

Supplementary Figure S3. Validation with qPCR of significantly deregulated miRs in primary tumors versus metastases of cohort 1 and 2. A) Unpaired comparison for hsa-miR-200a-3p, hsa-miR-29b-3p and hsa-miR451a. b) Paired assessment for hsa-miR-125b-5p and hsa-miR-143-3p. C) Unpaired comparison of primary tumors that did and did not disseminate to brain and brain versus non brain metastases for hsa-miR-3182. Mann-Whitney U test was used for unpaired and Wilcoxon signed-rank test for paired analyses with p < 0.05, ** p < 0.01, *** p < 0.001.

143

PART ONE | CHAPTER 5

B microarray hsa-miR-451a

microarray hsa-miR-451a

GI brain skin lung ovary

Threshold filtered data

Threshold filter data

11 10

Spearman's rho: 0.172 p=0.282

6 Primary

Metastasis 1

Metastasis 2

-1000

1000

2000

3000

4000

10000

time in days between primary and metastasis

Supplementary Figure S4. Threshold filtered data of the miRCURY microRNA expression array of miR-451a (highest fold change of metastases compared to the primary tumor) in cohort 1. A) Expression of all metastases 1 and 2 compared to their paired primary tumor. B) Expression of all metastases compared to the time interval between the primary tumor. No significant correlation was seen.

Time between primary and metastasis (divided per location)

metastasis location

skin

ovary

lung

brain -2000 -1000

1000

2000

3000

4000

5000

6000

7000

8000

9000

Time in days between primary and metastasis

Supplementary Figure S5. Time in days between paired primary tumors and metastases per location for cohort 1. No significant differences were seen between locations.

144

MicroRNAs in breast cancer metastases

Supplementary Figure S6. ROC curve of hsa-miR-106b-5p qPCR expression in primary tumors that disseminated to lung vs. primary tumors that not disseminate to lung (cohort 1 and 2 together).

145

PART ONE | CHAPTER 5

Supplementary Table S1. Characteristics of normal tissue included for determination of background microRNA levels using qPCR. Normal tissue

Subgroup

Age

Breast (n=4)

Range Mean

22-55 43

Lung (n=4)

Range Mean

17-59 44

Skin (n=4)

Range Mean

42-65 50

Ovary (n=4)

Range Mean

52-83 70

GI (n=4)

Range Mean

25-79 63

Supplementary Table S2. Location of metastases per patient in both cohorts. Cohort

Patient #

Metastases included in this study

Metastases to other organs (no tumor material available)

1st

skin-ovary

bone

1st

skin-ovary

lung, bone

1st

skin-ovary

liver, bone, brain

1st

skin-lung

bone, brain, liver

1st

skin-lung

bone, liver

1st

skin-lung

1st

skin-skin

1st

skin-skin

bone

1st

skin-skin

bone

1st

lung-ovary

1st

lung-ovary

liver

1st

lung-ovary

skin

1st

lung-brain

1st

lung-brain

1st

lung-brain

bone

1st

lung-brain

1st

ovary-GI

1st

ovary-GI

bone

1st

ovary-GI

1st

ovary-GI

bone, skin, liver, uterus

1st

GI-GI

bone, liver

1st

GI-GI

skin, liver, bone

1st

GI-GI

2nd

101

brain

lung

146

MicroRNAs in breast cancer metastases

Supplementary Table S2. Continued 2nd

102

brain

2nd

103

brain

lung

2nd

104

brain

2nd

105

brain

2nd

106

brain

bone

2nd

107

brain

2nd

108

lung

bone

2nd

109

lung

bone

2nd

110

lung

2nd

111

lung

2nd

112

lung

2nd

113

lung

2nd

114

skin

bone, liver, lung

2nd

115

skin-GI

2nd

116

skin

brain

2nd

117

skin

bone

2nd

118

skin

2nd

120

ovary

bone, liver

2nd

121

ovary

brain

2nd

122

ovary

liver

2nd

123

ovary

bone, uterus, GI

2nd

124

ovary

2nd

125

ovary

2nd

128

liver

2nd

129

2nd

130

2nd

131

2nd

132

ovary

147

PART ONE | CHAPTER 5

Supplementary Table S3. Analyses on microarray data of cohort 1. A. primary tumors versus metastases (unpaired). Upregulated

Downregulated

All miRs

ER+ FC

All FC

miR

ER+ FC

hsa-miR-125a-5p

0,020

1,38

hsa-miR-125b-2-3p

0,002

0,82

0,042

0,84

hsa-miR-126-3p

0,042

1,59

hsa-miR-195-3p

0,030

0,92

hsa-miR-148b-3p

0,020

1,27

0,042

1,29

hsa-miR-302a-3p

0,030

0,84

0,030

0,83

hsa-miR-151a-5p

0,012

1,28

0,012

1,35

hsa-miR-3162-3p

0,020

0,87

0,030

0,86

hsa-miR-15b-5p

0,020

1,39

0,020

1,46

hsa-miR-3196

0,030

0,73

hsa-miR-16-5p

0,012

1,61

0,012

1,75

hsa-miR-3606-5p

0,012

0,82

hsa-miR-197-3p

0,030

1,19

hsa-miR-3940-5p

0,000

0,68

0,002

0,69

hsa-miR-200a-3p

0,030

1,73

hsa-miR-422a

0,020

0,91

hsa-miR-203a

0,030

1,51

hsa-miR-4255

0,002

0,91

hsa-miR-223-3p

0,000

1,44

0,002

1,42

hsa-miR-4258

0,020

0,77

0,030

0,76

hsa-miR-23c

0,002

1,26

0,000

1,33

hsa-miR-4426

0,020

0,85

0,030

0,84

hsa-miR-25-3p

0,042

1,25

hsa-miR-4445-5p

0,042

0,80

hsa-miR-29b-3p

0,030

1,74

0,012

1,95

hsa-miR-4456

0,002

0,75

0,000

0,72

hsa-miR-30b-5p

0,042

1,53

0,042

1,66

hsa-miR-4486

0,030

0,81

hsa-miR-320a

0,042

1,41

0,042

1,48

hsa-miR-4488

0,006

0,68

0,030

0,69

hsa-miR-320b

0,012

1,41

0,012

1,47

hsa-miR-4507

0,030

0,77

hsa-miR-32-5p

0,006

1,47

hsa-miR-4513

0,012

0,69

0,042

0,67

hsa-miR-3591-3p

0,030

1,15

hsa-miR-4632-3p

0,002

0,78

0,002

0,76

hsa-miR-3654

0,020

1,27

hsa-miR-4653-3p

0,000

0,55

0,002

0,55

hsa-miR-4284

0,030

1,49

0,020

1,56

hsa-miR-506-5p

0,006

0,86

0,006

0,84

2,91

hsa-miR-548ap-5p/ hsa-miR-548j-5p

0,006

0,68

0,020

0,68

hsa-miR-628-3p

0,020

0,88

0,030

0,88

hsa-miR-639

0,042

0,84

0,012

0,81

hsa-miR-654-5p

0,042

0,86

hsa-miR-761

0,006

0,86

0,000

0,85

hsa-miR-874-3p

0,012

0,71

hsa-miR-920

0,042

0,91

hsa-miR-451a

148

0,006

2,55

0,012

MicroRNAs in breast cancer metastases

Supplementary Table S3. Analyses on microarray data of cohort 1. B. primary tumors versus metastases (paired per location of metastasis). No differentially expressed miRs for brain metastases were detected.

P vs M skin (n=9)

All

miRs

P vs M lung (n=10)

ER+ FC

All FC

P vs M ovary (n=10)

ER+ FC

All FC

P vs M GI (n=7)

ER+ FC

All FC

ER+ FC

hsa-let-7g-5p

0,018 1,55

hsa-let-7i-5p

0,018 1,30

hsa-miR-101-3p

0,039 2,13

hsa-miR-103a-3p

0,039 1,24

hsa-miR-106b-5p

0,040 1,26

hsa-miR-10b-5p

0,040 2,18

hsa-miR-1248

0,034 0,82 0,034 0,82

hsa-miR-1252-5p

0,041 1,02

hsa-miR-125a-5p

0,018 1,41

0,021 0,75

hsa-miR-125b-2-3p

hsa-miR-125b-5p

0,019 1,56

hsa-miR-1297

0,041 1,47

hsa-miR-136-5p

0,044 1,28

0,042 1,25

hsa-miR-141-3p

0,018 1,36

hsa-miR-143-3p

0,019 1,74

hsa-miR-145-5p

0,019 1,69

hsa-miR-146b-5p

0,042 1,24

hsa-miR-147b

0,043 0,73

hsa-miR-16-5p

0,043 2,09

hsa-miR-195-5p

0,019 2,03

0,019 1,43

hsa-miR-199a3p/199b-3p

hsa-miR-199a-5p

0,044 0,91

hsa-miR-19a-3p

0,044 1,42

hsa-miR-19b-3p

0,045 1,65

hsa-miR-200b-3p

0,005 1,60 0,018 1,55

hsa-miR-200c-3p

0,018 1,30

hsa-miR-205-3p

0,019 0,91

hsa-miR-2116-5p

0,045 0,80

hsa-miR-214-3p

0,019 0,95

hsa-miR-23b-3p

0,020 1,33

hsa-miR-23c

0,046 1,16

hsa-miR-24-3p

0,046 1,35

hsa-miR-25-5p

0,020 0,77

hsa-miR-26a-5p

0,020 1,56

hsa-miR-26b-5p

0,020 1,93

hsa-miR-27b-3p

0,020 1,62

149

PART ONE | CHAPTER 5

Supplementary Table S3B. Continued

P vs M skin (n=9)

All

miRs

P vs M lung (n=10)

ER+ FC

All FC

P vs M ovary (n=10)

ER+ FC

All FC

P vs M GI (n=7)

ER+ FC

All FC

ER+ FC

hsa-miR-29a-3p

0,020 2,07

hsa-miR-29b-3p

0,021 2,23

hsa-miR-29c-3p

0,047 2,44

hsa-miR-302a-3p

0,021 0,91

hsa-miR-30b-5p

0,047 1,73

hsa-miR-30c-5p

0,021 1,71

hsa-miR-30d-5p

0,021 1,56

hsa-miR-30e-3p

0,048 1,36

hsa-miR-30e-5p

0,021 1,86

hsa-miR-3135a

0,049 0,87

hsa-miR-3175

0,018 1,60

hsa-miR-3178

0,010 0,76 0,018 0,82

hsa-miR-3201

0,018 1,17

0,049 0,93

hsa-miR-320a

0,050 1,53

hsa-miR-320b

0,022 1,63

hsa-miR-335-3p

0,022 0,97

hsa-miR-34a-5p

0,022 1,41

hsa-miR-3611

0,051 1,02

hsa-miR-361-5p

0,022 1,16

hsa-miR-3654

0,028 1,52 0,018 1,70

hsa-miR-3656

0,051 0,54

hsa-miR-378a-3p

0,013 1,22 0,013 1,22

hsa-miR-424-5p

0,022 1,54

hsa-miR-4285

0,052 0,68

hsa-miR-4286

0,023 1,42

hsa-miR-4288

0,023 1,16

hsa-miR-4301

0,019 1,12

hsa-miR-4324

0,023 1,28

hsa-miR-4328

0,053 1,47

hsa-miR-4421

0,054 0,95

hsa-miR-4443

0,018 0,62

hsa-miR-4445-5p

0,054 0,79

hsa-miR-4456

0,018 0,59

hsa-miR-4459

0,055 0,60

hsa-miR-4488

0,056 0,60

hsa-miR-451b

0,018

150

MicroRNAs in breast cancer metastases

Supplementary Table S3B. Continued

P vs M skin (n=9)

All

miRs

P vs M lung (n=10)

ER+ FC

All FC

P vs M ovary (n=10)

ER+ p

All FC

P vs M GI (n=7)

ER+ FC

All FC

ER+ FC

hsa-miR-4531

0,057 0,85

hsa-miR-4532

0,035 0,71

hsa-miR-4533

0,058 0,74

hsa-miR-4658

0,059 0,97

hsa-miR-4687-3p

0,060 0,64

hsa-miR-4708-3p

0,061 0,82

hsa-miR-4712-3p

0,023 0,87

hsa-miR-4726-5p

0,023 0,85

hsa-miR-4739

0,062 0,73

hsa-miR-4765

0,063 0,82

hsa-miR-4777-5p

0,024 0,88

hsa-miR-4795-5p

0,064 0,57

hsa-miR-4800-5p

0,065 0,91

hsa-miR-483-5p

0,066 0,90

hsa-miR-487b-3p

0,067 1,21

hsa-miR-5004-5p

0,068 0,93

hsa-miR-516b-5p

0,069 0,85

hsa-miR-550b-3p

0,071 1,01

hsa-miR-5572

0,072 0,47

hsa-miR-620

0,074 0,84

hsa-miR-622

0,075 0,79

hsa-miR-635

0,024 0,982 0,024 0,982

hsa-miR-636

0,077 0,963 0,077 0,963

hsa-miR-664b-3p

0,025 0,897

hsa-miR-675-3p

0,078 1,05

hsa-miR-761

0,024 0,86

hsa-miR-877-5p

0,024 0,69

hsa-miR-891a-5p

0,025 0,86

hsa-miR-920

0,025 0,94

151

152

lung vs rest

brain vs rest

GI vs rest

0,006

hsa-miR-5096

0,032

hsa-miR-4268

hsa-miR-4284

0,01

0,018

hsa-miR-548t-5p

hsa-miR-505-5p

0,01

hsa-miR-451b

0,01

0,004

hsa-miR-302e

hsa-miR-106b-5p

0,004

hsa-miR-210-3p

0,039

hsa-miR-5002-5p

hsa-miR-223-3p

0,001

hsa-miR-4780

0,022

hsa-miR-4646-3p

hsa-miR-210-3p

0,032

hsa-miR-4284

0,008

0,018

hsa-miR-1273g-3p

hsa-miR-106b-5p

0,018

0,032

hsa-miR-4764-3p

hsa-miR-302e

hsa-miR-146b-5p

ovary vs rest

0,006

hsa-miR-7-5p

skin vs rest

miRs

Subgroups

C. site of metastasis predictive miRs in primary tumor.

1,15

1,16

1,46

1,20

1,33

1,20

1,49

1,25

1,35

1,55

1,15

1,34

1,68

1,40

1,21

1,26

1,31

All

lung>rest

rest>lung

lung>rest

rest>brain

brain>rest

rest>brain

brain>rest

GI>rest

rest>GI

ovary>rest

rest>ovary

ovary>rest

skin>rest

0,003

0,03

0,01

0,011

0,022

0,004

0,022

0,042

1,56

1,71

1,74

1,53

1,22

1,29

1,41

1,37

1,31

Luminal

0,021

0,005

Univariate analyses

direction

Supplementary Table S3. Analyses on microarray data of cohort 1.

lung>rest

brain>rest

rest>GI

rest>ovary

ovary>rest

rest>ovary

rest>skin

skin>rest

direction

0,031

0,018

0,037

1,432-1964,930

0,002-0,560

0,000-0,0487

1,23-792,37

95% CI

All

1.843

1.457

1.782

1.650

Multivariate analyses

PART ONE | CHAPTER 5

MicroRNAs in breast cancer metastases

Supplementary Table S4. Correlations of Threshold filtered data from the Exiqon microarray versus RQ data from SYBR Green qPCR assays of the most deregulated miRs in cohort 1. miRs

Sig. (2-tailed)

Spearmanâ&#x20AC;&#x2122;s rho

Analysis group

miR-106b-5p

0.017

0.294

Predictive in primary tumors for lung and GI metastases

miR-125b-5p

0.008

0.435

Upregulated in ovarian metastases versus primary tumors (paired)

miR-1273g-3p

0.020

0.285

Predictive in primary tumors for ovarian metastases

miR-143-3p

0.050

0.328

Upregulated in ovarian metastases versus primary tumors (paired)

miR-16-5p

0.204*

0.217

miR-200a-3p

0.008

0.445

Significantly upregulated in metastases versus primary tumors (unpaired)

miR-29b-3p

0.017

0.400

Significantly upregulated in metastases versus primary tumors (unpaired)

miR-3182

0.000

0.720

Significantly upregulated in brain metastases versus non brain metastases

miR-451a

0.000

0.777

Significantly upregulated in metastases versus primary tumors (unpaired)

miR-7-5p

0.015

0.300

Predictive in primary tumors for skin metastases

* non-significant correlations

153

PART ONE | CHAPTER 5

Supplementary Table S5. 38 oncogenic and tumor suppressive miRs that play a role in the metastatic cascade and are expressed in primary breast tumors. In Figure 4

miR

Type

Role

Ref

Tumor suppressive

EMT, anoikis

1-4

Yes

Oncogenic

EMT, invasion, angiogenesis, metastasis

No*

10b

Oncogenic

Migration, invasion, metastasis

6-12

Yes

17/20

Tumor suppressive

Proliferation

13,14

No*

17/92

Tumor suppressive and oncogenic

NK antitumoral activity, metastasis, invasion

15-17

No$*

Oncogenic

Invasion, metastasis, migration, EMT, apoptosis

18-28

Yes

Tumor suppressive

Growth, senescence, metastasis, invasion

29-31

Yes

Oncogenic

Invasion, angiogenesis

Yes

Oncogenic

Metastasis, invasion

Yes

29a/b

Oncogenic

EMT, metastasis, angiogenesis, invasion

34,35

Yes

Tumor suppressive

Apoptosis, tumorigenesis, metastasis

36,37

No*

Tumor suppressive

Invasion, metastasis, progression, intraand extravasation

38,39

No*

34a/b/c

Tumor suppressive

Metastasis, invasion

40-42

No*

Oncogenic

Metastasis, angiogenesis, invasion

Yes

103/107

Oncogenic

EMT, migration, invasion, metastasis

Yes

124

Tumor suppressive

Metastasis, EMT, invasion

No*

125b

Tumor suppressive and oncogenic

Proliferation, differentiation, migration, invasion

46-49

No$

126

Tumor suppressive

Metastasis, angiogenesis, tumorigenesis, proliferation, cell adhesion

50-56

No*

127/197/ 222/223

Tumor suppressive

Quiescence, proliferation

No*

132

Oncogenic

Neovascularization

No*

145

Tumor suppressive

Apoptosis, cell motility, invasion, metastasis, EMT, angiogenesis

59-62

No*

146a/b

Tumor suppressive

Survival, metastasis, invasion, migration

63-65

Yes

154

MicroRNAs in breast cancer metastases

Supplementary Table S5. Continued In Figure 4

miR

Type

Role

Ref

150

Oncogenic

EMT

Yes

155

Tumor suppressive

Growth, proliferation, survival, EMT, migration, invasion

67-71

Yes

182

Oncogenic

EMT

72,73

Yes

193b

Tumor suppressive

Migration, invasion, metastasis

No*

194

Tumor suppressive

Invasion, EMT

Yes

199

Oncogenic

Angiogenesis

Yes

200 family

Tumor suppressive and oncogenic

Growth, metastasis, EMT, MET, migration, invasion, proliferation, apoptosis

77-92

No$

205

Tumor suppressive

Proliferation, invasion, EMT, metastasis

84,93-96

Yes

206

Tumor suppressive

Migration, invasion, metastasis, apoptosis

97-102

No*

221/222

Oncogenic

EMT, invasion

335

Tumor suppressive

Metastasis, migration, invasion

10,51,107

Yes

373/520c

Oncogenic

Migration, invasion, metastasis

10,108,109

No*

448

Tumor suppressive

EMT, metastasis

110

No*

661

Tumor suppressive

Cell motility, invasion, EMT

111,112

No*

1908

Oncogenic

Angiogenesis

Yes

Let-7 family

Tumor suppressive

Metastasis, invasion

113-115

Yes

99,103106

Yes

* not present in all tested samples $ both tumor suppressive and oncogenic properties

155

PART ONE | CHAPTER 5

Supplementary Table S6. hsa-miR-106b-5p, hsa-miR-1273g-3p and hsa-miR-7-5p were inserted in miRTarBase to find experimentally validated microRNA-target interactions. The mRNA targets with strong validation evidence (obtained by reporter assay, western blot or qPCR) are listed. These target genes were subsequently imported in ToppGene Suite to show enriched pathways. For hsa-miR-1273g-3p no known targets with strong evidence were reported. hsa-miR-7-5p

hsa-miR-106b-5p

Bonferonni

ToppGene

Significant pathways

Targets with strong evidence

miRTarBase

Bonferonni

ToppGene

Significant pathways

Targets with strong evidence

miRTarBase

ABCC1

Small cell lung cancer

0.0000

APC

G1 to S cell cycle control

0.0000

BCL2

Pathways in cancer

0.0000

APP

Hepatitis B

0.0000

CCNE1

Prostate cancer

0.0000

ATG16L1

Integrated Pancreatic Cancer Pathway 0.0000

CUL5

Alpha6-Beta4 Integrin Signaling Pathway

0.0000

BCL2L11

Pathways in cancer

0.0000

EGFR

Prolactin Signaling Pathway

0.0000

CASP7

Cell cycle

0.0000

GFI1

PI3K-Akt signaling pathway

0.0000

CASP8

Bladder cancer

0.0000

HELLS

ErbB signaling pathway

0.0000

CCND1

Cyclin D associated events in G1

0.0000

HOXB3

Focal adhesion

0.0000

CCND2

G1 Phase

0.0000

HOXB5

Insulin/IGF pathway-protein kinase B signaling cascade

0.0000

CDKN1A

E2F transcription factor network

0.0000

IGF1R

Signaling Pathways in Glioblastoma

0.0000

E2F1

DNA damage response (only ATM dependent)

0.0000

IRS1

Hepatitis B

0.0000

E2F3

Mitotic G1-G1/S phases

0.0000

IRS2

Proteoglycans in cancer

0.0000

E2F5

Pancreatic cancer

0.0000

JARID2

Endometrial cancer

0.0000

EOMES

DNA damage response

0.0000

KLF4

Acute myeloid leukemia

0.0000

FAS

Integrated Breast Cancer Pathway

0.0000

MYC

Genes related to IL4 rceptor signaling in B lymphocytes

0.0000

ITCH

Viral carcinogenesis

0.0000

PAK1

HIF-1 signaling pathway

0.0000

JAK1

Cell cycle

0.0000

PAX6

Colorectal cancer

0.0000

KAT2B

TGF-beta Receptor Signaling Pathway 0.0000

PIK3CD

EGFR1 Signaling Pathway

0.0000

MAPK9

Prostate Cancer

0.0000

PIK3CG

Glioma

0.0000

MFN2

MicroRNAs in cancer

0.0000

PIK3R3

Focal Adhesion

0.0000

MMP2

Cyclins and Cell Cycle Regulation

0.0000

PSME3

Pancreatic cancer

0.0000

MYC

RB in Cancer

0.0000

PTK2

Genes related to PIP3 signaling in cardiac myocytes

0.0000

PKD2

HTLV-I infection

0.0000

RAF1

Neurotrophin signaling pathway

0.0000

PTEN

Integrated Cancer pathway

0.0000

RELA

Fc epsilon receptor (FCERI) signaling

0.0000

PURA

Glioma

0.0000

SETD8

Melanoma

0.0000

RB1

p53 Signaling Pathway

0.0000

SKP2

Chronic myeloid leukemia

0.0000

RBL1

p53 signaling pathway

0.0000

SLC7A5

Hepatitis C

0.0000

RBL2

Tumor Suppressor Arf Inhibits Ribosomal Biogenesis

0.0000

SNCA

IL-2 Receptor Beta Chain in T cell Activation

0.0000

RUNX3

Melanoma

0.0000

156

MicroRNAs in breast cancer metastases

Supplementary Table S6. Continued hsa-miR-7-5p

hsa-miR-106b-5p

Bonferonni

ToppGene

Significant pathways

Targets with strong evidence

miRTarBase

Bonferonni

ToppGene

Significant pathways

Targets with strong evidence

miRTarBase

SRSF1

Apoptosis

0.0000

SMAD7

Chronic myeloid leukemia

0.0000

TET2

Regulation of Actin Cytoskeleton

0.0001

STAT3

Signaling Pathways in Glioblastoma

0.0000

XIAP

Apoptosis

0.0001

TCEAL1

p53 pathway feedback loops 2

0.0000

XRCC2

Ras signaling pathway

0.0001

TP53

Adipogenesis

0.0000

YY1

p53 pathway feedback loops 2

0.0001

TWIST1

Small cell lung cancer

0.0000

Telomeres, Telomerase, Cellular Aging, and Immortality

0.0001

VEGFA

Prostate cancer

0.0000

Signaling by ERBB2

0.0001

WEE1

Proteoglycans in cancer

0.0000

Type II diabetes mellitus

0.0001

ZBTB4

Cell Cycle

0.0000

Insulin Signaling

0.0001

Cell Cycle: G1/S Check Point

0.0000

IL2-mediated signaling events

0.0001

Cell Cycle, Mitotic

0.0000

Downstream Signaling Events Of B Cell Receptor (BCR)

0.0001

Pre-NOTCH Transcription and Translation

0.0000

IGF-1 Signaling Pathway

0.0002

TSH signaling pathway

0.0000

T cell receptor signaling pathway

0.0002

Signaling by NOTCH

0.0000

Adaptive Immune System

0.0002

PI3K-Akt signaling pathway

0.0000

Non-small cell lung cancer

0.0002

Cdk2, 4, and 6 bind cyclin D in G1, 0.0001 while cdk2/cyclin E promotes the G1/S transition.

VEGF signaling pathway

0.0003

Direct p53 effectors

0.0001

Chemokine signaling pathway

0.0003

Intrinsic Pathway for Apoptosis

0.0001

VEGF signaling pathway

0.0003

Apoptosis

0.0001

Influence of Ras and Rho proteins on G1 to S Transition

0.0004

IL-7 Signaling Pathway

0.0001

EPO Receptor Signaling

0.0004

Androgen receptor signaling pathway

0.0002

MicroRNAs in cancer

0.0004

Pre-NOTCH Expression and Processing

0.0002

IL-4 signaling Pathway

0.0004

Apoptosis

0.0003

Toxoplasmosis

0.0004

Endometrial cancer

0.0003

Epstein-Barr virus infection

0.0005

Wnt Signaling Pathway and Pluripotency

0.0003

Renal cell carcinoma

0.0005

extrinsic apoptotic

0.0004

Erk1/Erk2 Mapk Signaling pathway

0.0005

Apoptosis signaling pathway

0.0004

IGF1 pathway

0.0005

TGF Beta Signaling Pathway

0.0004

157

PART ONE | CHAPTER 5

Supplementary Table S6. Continued hsa-miR-7-5p

158

hsa-miR-106b-5p

Bonferonni

ToppGene

Significant pathways

Targets with strong evidence

miRTarBase

Bonferonni

ToppGene

Significant pathways

Targets with strong evidence

miRTarBase

Insulin/IGF pathway-mitogen activated 0.0006 protein kinase kinase/MAP kinase cascade

ErbB signaling pathway

0.0004

Regulation of actin cytoskeleton

0.0007

Non-small cell lung cancer

0.0004

Signaling by the B Cell Receptor (BCR)

0.0007

Senescence and Autophagy

0.0005

B cell receptor signaling pathway

0.0008

Wnt Signaling Pathway

0.0006

Prolactin signaling pathway

0.0008

Spinal Cord Injury

0.0006

Signaling by SCF-KIT

0.0009

p75(NTR)-mediated signaling

0.0007

T cell activation

0.0010

Influence of Ras and Rho proteins on G1 to S Transition

0.0007

Insulin signaling pathway

0.0010

Colorectal cancer

0.0007

Downstream signaling of activated FGFR

0.0011

Oncostatin M Signaling Pathway

0.0008

Angiotensin II mediated activation of JNK Pathway via Pyk2 dependent signaling

0.0012

G0 and Early G1

0.0008

Signaling by ERBB4

0.0014

FAS signaling pathway ( CD95 )

0.0013

Angiogenesis

0.0014

Cellular responses to stress

0.0014

Non-alcoholic fatty liver disease (NAFLD)

0.0015

Epstein-Barr virus infection

0.0014

Signaling of Hepatocyte Growth Factor 0.0017 Receptor

FAS signaling pathway

0.0015

Downstream signal transduction

0.0018

SMAD2/SMAD3:SMAD4 heterotrimer regulates transcription

0.0015

Progesterone-mediated oocyte maturation

0.0018

miRNAs involved in DDR

0.0016

Signaling by FGFR

0.0019

Measles

0.0018

Constitutive PI3K/AKT Signaling in Cancer

0.0020

Wnt signaling pathway

0.0023

DAP12 signaling

0.0020

Regulation of nuclear SMAD2/3 signaling

0.0024

B Cell Receptor Signaling Pathway

0.0020

Alzheimers Disease

0.0025

Aldosterone-regulated sodium reabsorption

0.0021

Regulation of Wnt-mediated beta catenin signaling and target gene transcription

0.0025

Fc gamma R-mediated phagocytosis

0.0024

p53 pathway

0.0025

Signaling by FGFR in disease

0.0033

FAS pathway and Stress induction of HSP regulation

0.0034

Ceramide signaling pathway

0.0035

IL-2 Receptor Beta Chain in T cell Activation

0.0034

MicroRNAs in breast cancer metastases

Supplementary Table S6. Continued hsa-miR-7-5p

hsa-miR-106b-5p

Bonferonni

ToppGene

Significant pathways

Targets with strong evidence

miRTarBase

Bonferonni

ToppGene

Significant pathways

Targets with strong evidence

miRTarBase

Signaling by EGFR

0.0036

TSLP Signaling Pathway

0.0034

PI3K events in ERBB4 signaling

0.0037

Delta-Notch Signaling Pathway

0.0036

PIP3 activates AKT signaling

0.0037

Non-alcoholic fatty liver disease (NAFLD)

0.0037

PI3K/AKT Signaling in Cancer

0.0037

FOXM1 transcription factor network

0.0038

PI-3K cascade

0.0037

RB Tumor Suppressor/Checkpoint Signaling in response to DNA damage

0.0050

PI3K events in ERBB2 signaling

0.0037

IL-2 Signaling pathway

0.0051

Signaling by EGFR in Cancer

0.0039

Transcriptional activity of SMAD2/ SMAD3:SMAD4 heterotrimer

0.0056

Estrogen signaling pathway

0.0039

Signal Transduction

0.0062

Signaling by PDGF

0.0040

Transcriptional activation of cell cycle inhibitor p21

0.0063

DAP12 interactions

0.0040

Transcriptional activation of p53 responsive genes

0.0063

Keratinocyte Differentiation

0.0042

MAPK signaling pathway

0.0063

Apoptosis signaling pathway

0.0043

FoxO family signaling

0.0080

PI3K/AKT activation

0.0043

Notch-mediated HES/HEY network

0.0087

SOS-mediated signalling

0.0043

IL-3 Signaling Pathway

0.0097

Signalling by NGF

0.0044

Cellular Senescence

0.0107

GAB1 signalosome

0.0045

Regulation of transcriptional activity by 0.0118 PML

IL-3 Signaling Pathway

0.0047

IL2-mediated signaling events

0.0120

Amoebiasis

0.0059

Id Signaling Pathway

0.0120

MicroRNAs in cardiomyocyte hypertrophy

0.0059

TNF signaling pathway

0.0128

Integrated Pancreatic Cancer Pathway 0.0064

G1/S Transition

0.0139

Genes related to the insulin receptor pathway

0.0064

METS affect on Macrophage Differentiation

0.0141

Cadmium induces DNA synthesis and proliferation in macrophages

0.0066

Telomeres, Telomerase, Cellular Aging, and Immortality

0.0141

Id Signaling Pathway

0.0069

Apoptosis is mediated by caspases, cysteine proteases arranged in a proteolytic cascade.

0.0168

NGF signalling via TRKA from the plasma membrane

0.0078

TP53 network

0.0168

Viral carcinogenesis

0.0090

Apoptosis Modulation by HSP70

0.0168

Role of EGF Receptor Transactivation by GPCRs in Cardiac Hypertrophy

0.0096

HIV-I Nef: negative effector of Fas and 0.0186 TNF

159

PART ONE | CHAPTER 5

160

Significant pathways

Targets with strong evidence

Bonferonni

miRTarBase ToppGene

Significant pathways

hsa-miR-106b-5p

miRTarBase ToppGene Targets with strong evidence

hsa-miR-7-5p

Bonferonni

Supplementary Table S6. Continued

Rap1 signaling pathway

0.0108

G1 phase

0.0188

Role of LAT2/NTAL/LAB on calcium mobilization

0.0111

S Phase

0.0228

mTOR signaling pathway

0.0123

Validated targets of C-MYC transcriptional repression

0.0242

PDGF signaling pathway

0.0125

Fas Signaling Pathway

0.0292

mTOR signaling pathway

0.0131

CTCF: First Multivalent Nuclear Factor 0.0304

Trefoil Factors Initiate Mucosal Healing 0.0156

AGE/RAGE pathway

0.0310

TSH signaling pathway

0.0159

Cyclin E associated events during G1/S transition

0.0310

Measles

0.0162

Hepatitis C

0.0319

Natural killer cell mediated cytotoxicity

0.0168

Cyclin A:Cdk2-associated events at S phase entry

0.0329

AGE/RAGE pathway

0.0179

Cell Cycle: G2/M Checkpoint

0.0346

Cyclin E associated events during G1/S transition

0.0179

mitogen activated protein kinase signaling

0.0346

Role of ERBB2 in Signal Transduction and Oncology

0.0180

FasL mediated signaling pathway

0.0376

Signaling by Leptin

0.0180

Loss of Function of SMAD2/3 in Cancer

0.0437

Insulin Signaling Pathway

0.0180

TGFBR1 LBD Mutants in Cancer

0.0437

Ceramide Signaling Pathway

0.0180

Loss of Function of SMAD4 in Cancer

0.0437

Cyclin A:Cdk2-associated events at S phase entry

0.0190

SMAD2/3 Phosphorylation Motif Mutants in Cancer

0.0437

Ras Pathway

0.0202

Loss of Function of TGFBR1 in Cancer 0.0437

Multiple antiapoptotic pathways from IGF-1R signaling lead to BAD phosphorylation

0.0207

Signaling by TGF-beta Receptor Complex

0.0437

Ras Signaling Pathway

0.0207

Prolactin signaling pathway

0.0437

Inhibition of Cellular Proliferation by Gleevec

0.0207

TGFBR2 Kinase Domain Mutants in Cancer

0.0437

Fc epsilon RI signaling pathway

0.0226

Signaling by TGF-beta Receptor Complex in Cancer

0.0437

Erk and PI-3 Kinase Are Necessary for 0.0236 Collagen Binding in Corneal Epithelia

Loss of Function of TGFBR2 in Cancer 0.0437

CXCR4 Signaling Pathway

0.0236

TGFBR2 MSI Frameshift Mutants in Cancer

0.0437

Endothelin signaling pathway

0.0267

SMAD4 MH2 Domain Mutants in Cancer

0.0437

MicroRNAs in breast cancer metastases

Significant pathways

Targets with strong evidence

Bonferonni

miRTarBase ToppGene

Significant pathways

hsa-miR-106b-5p

miRTarBase ToppGene Targets with strong evidence

hsa-miR-7-5p

Bonferonni

Supplementary Table S6. Continued

Hypoxia response via HIF activation

0.0303

TGFBR1 KD Mutants in Cancer

0.0437

Bacterial invasion of epithelial cells

0.0313

WNT Signaling Pathway

0.0443

Integrated Breast Cancer Pathway

0.0317

Disease

0.0319

3-phosphoinositide biosynthesis

0.0340

HTLV-I infection

0.0357

p53 pathway

0.0365

Growth Hormone Signaling Pathway

0.0380

Integrin signalling pathway

0.0467

Axon guidance mediated by netrin

0.0469

Delta-Notch Signaling Pathway

0.0487

161

PART ONE | CHAPTER 5

SUPPLEMENTAL REFERENCES 1. Foekens JA, Sieuwerts AM, Smid M, et al. Four miRNAs associated with aggressiveness of lymph node-negative, estrogen receptor-positive human breast cancer. Proc Natl Acad Sci U S A. 2008;105(35):13021-13026. doi: 10.1073/ pnas.0803304105 [doi]. 2. Webster RJ, Giles KM, Price KJ, Zhang PM, Mattick JS, Leedman PJ. Regulation of epidermal growth factor receptor signaling in human cancer cells by microRNA-7. J Biol Chem. 2009;284(9):5731-5741. doi: 10.1074/jbc. M804280200 [doi]. 3. Reddy SD, Ohshiro K, Rayala SK, Kumar R. MicroRNA-7, a homeobox D10 target, inhibits p21-activated kinase 1 and regulates its functions. Cancer Res. 2008;68(20):81958200. doi: 10.1158/0008-5472.CAN-08-2103 [doi]. 4. Kong X, Li G, Yuan Y, et al. MicroRNA-7 inhibits epithelial-to-mesenchymal transition and metastasis of breast cancer cells via targeting FAK expression. PLoS One. 2012;7(8):e41523. doi: 10.1371/journal. pone.0041523 [doi]. 5. Ma L, Young J, Prabhala H, et al. miR-9, a MYC/MYCNactivated microRNA, regulates E-cadherin and cancer metastasis. Nat Cell Biol. 2010;12(3):247-256. doi: 10.1038/ncb2024 [doi]. 6. Tang J, Ahmad A, Sarkar FH. The role of microRNAs in breast cancer migration, invasion and metastasis. Int J Mol Sci. 2012;13(10):13414-13437. doi: 10.3390/ ijms131013414 [doi]. 7. Ma L, Teruya-Feldstein J, Weinberg RA. Tumour invasion and metastasis initiated by microRNA-10b in breast cancer. Nature. 2007;449(7163):682-688. doi: nature06174 [pii]. 8. Gabriely G, Teplyuk NM, Krichevsky AM. Context effect: microRNA-10b in cancer cell proliferation, spread and death. Autophagy. 2011;7(11):1384-1386. doi: 10.4161/ auto.7.11.17371 [doi]. 9. Bourguignon LY, Wong G, Earle C, Krueger K, Spevak CC. Hyaluronan-CD44 interaction promotes c-srcmediated twist signaling, microRNA-10b expression, and RhoA/RhoC up-regulation, leading to rho-kinaseassociated cytoskeleton activation and breast tumor cell invasion. J Biol Chem. 2010;285(47):36721-36735. doi: 10.1074/jbc.M110.162305 [doi]. 10. Edmonds MD, Hurst DR, Vaidya KS, Stafford LJ, Chen D, Welch DR. Breast cancer metastasis suppressor 1 coordinately regulates metastasis-associated microRNA expression. Int J Cancer. 2009;125(8):1778-1785. doi: 10.1002/ijc.24616 [doi]. 11. Moriarty CH, Pursell B, Mercurio AM. miR-10b targets Tiam1: Implications for rac activation and carcinoma migration. J Biol Chem. 2010;285(27):20541-20546. doi: 10.1074/jbc.M110.121012 [doi]. 12. Ma L, Reinhardt F, Pan E, et al. Therapeutic silencing of miR-10b inhibits metastasis in a mouse mammary tumor model. Nat Biotechnol. 2010;28(4):341-347. doi: 10.1038/ nbt.1618 [doi]. 13. Yu Z, Wang C, Wang M, et al. A cyclin D1/microRNA 17/20 regulatory feedback loop in control of breast cancer cell proliferation. J Cell Biol. 2008;182(3):509-517. doi: 10.1083/jcb.200801079 [doi].

162

14. Yu Z, Willmarth NE, Zhou J, et al. microRNA 17/20 inhibits cellular invasion and tumor metastasis in breast cancer by heterotypic signaling. Proc Natl Acad Sci U S A. 2010;107(18):8231-8236. doi: 10.1073/pnas.1002080107 [doi]. 15. Jiang H, Wang P, Li X, et al. Restoration of miR17/20a in solid tumor cells enhances the natural killer cell antitumor activity by targeting Mekk2. Cancer Immunol Res. 2014;2(8):789-799. doi: 10.1158/2326-6066.CIR-13-0162 [doi]. 16. Fan M, Sethuraman A, Brown M, Sun W, Pfeffer LM. Systematic analysis of metastasis-associated genes identifies miR-17-5p as a metastatic suppressor of basal-like breast cancer. Breast Cancer Res Treat. 2014;146(3):487-502. doi: 10.1007/s10549-014-3040-5 [doi]. 17. Liu S, Goldstein RH, Scepansky EM, Rosenblatt M. Inhibition of rho-associated kinase signaling prevents breast cancer metastasis to human bone. Cancer Res. 2009;69(22):8742-8751. doi: 10.1158/0008-5472.CAN-091541 [doi]. 18. Huang TH, Wu F, Loeb GB, et al. Up-regulation of miR-21 by HER2/neu signaling promotes cell invasion. J Biol Chem. 2009;284(27):18515-18524. doi: 10.1074/jbc. M109.006676 [doi]. 19. Han M, Wang Y, Liu M, et al. MiR-21 regulates epithelialmesenchymal transition phenotype and hypoxiainducible factor-1alpha expression in third-sphere forming breast cancer stem cell-like cells. Cancer Sci. 2012;103(6):1058-1064. doi: 10.1111/j.1349-7006.Â 2012.02281.x [doi]. 20. Lu Z, Liu M, Stribinskis V, et al. MicroRNA-21 promotes cell transformation by targeting the programmed cell death 4 gene. Oncogene. 2008;27(31):4373-4379. doi: 10.1038/onc.2008.72 [doi]. 21. Qi L, Bart J, Tan LP, et al. Expression of miR-21 and its targets (PTEN, PDCD4, TM1) in flat epithelial atypia of the breast in relation to ductal carcinoma in situ and invasive carcinoma. BMC Cancer. 2009;9:163-2407-9-163. doi: 10.1186/1471-2407-9-163 [doi]. 22. Song B, Wang C, Liu J, et al. MicroRNA-21 regulates breast cancer invasion partly by targeting tissue inhibitor of metalloproteinase 3 expression. J Exp Clin Cancer Res. 2010;29:29-9966-29-29. doi: 10.1186/1756-9966-29-29 [doi]. 23. Baffa R, Fassan M, Volinia S, et al. MicroRNA expression profiling of human metastatic cancers identifies cancer gene targets. J Pathol. 2009;219(2):214-221. doi: 10.1002/ path.2586 [doi]. 24. Connolly EC, Van Doorslaer K, Rogler LE, Rogler CE. Overexpression of miR-21 promotes an in vitro metastatic phenotype by targeting the tumor suppressor RHOB. Mol Cancer Res. 2010;8(5):691-700. doi: 10.1158/1541-7786. MCR-09-0465 [doi]. 25. Du J, Yang S, An D, et al. BMP-6 inhibits microRNA-21 expression in breast cancer through repressing deltaEF1 and AP-1. Cell Res. 2009;19(4):487-496. doi: 10.1038/ cr.2009.34 [doi]. 26. Yan LX, Huang XF, Shao Q, et al. MicroRNA miR-21 overexpression in human breast cancer is associated with advanced clinical stage, lymph node metastasis and

MicroRNAs in breast cancer metastases

patient poor prognosis. RNA. 2008;14(11):2348-2360. doi: 10.1261/rna.1034808 [doi]. 27. Zhu S, Si ML, Wu H, Mo YY. MicroRNA-21 targets the tumor suppressor gene tropomyosin 1 (TPM1). J Biol Chem. 2007;282(19):14328-14336. doi: M611393200 [pii]. 28. Zhu S, Wu H, Wu F, Nie D, Sheng S, Mo YY. MicroRNA-21 targets tumor suppressor genes in invasion and metastasis. Cell Res. 2008;18(3):350-359. doi: 10.1038/cr.2008.24 [doi]. 29. Pandey DP, Picard D. miR-22 inhibits estrogen signaling by directly targeting the estrogen receptor alpha mRNA. Mol Cell Biol. 2009;29(13):3783-3790. doi: 10.1128/ MCB.01875-08 [doi]. 30. Xu D, Takeshita F, Hino Y, et al. miR-22 represses cancer progression by inducing cellular senescence. J Cell Biol. 2011;193(2):409-424. doi: 10.1083/jcb.201010100 [doi]. 31. Patel JB, Appaiah HN, Burnett RM, et al. Control of EVI-1 oncogene expression in metastatic breast cancer cells through microRNA miR-22. Oncogene. 2011;30(11):12901301. doi: 10.1038/onc.2010.510 [doi]. 32. Zhang H, Hao Y, Yang J, et al. Genome-wide functional screening of miR-23b as a pleiotropic modulator suppressing cancer metastasis. Nat Commun. 2011;2:554. doi: 10.1038/ncomms1555 [doi]. 33. Du WW, Fang L, Li M, et al. MicroRNA miR-24 enhances tumor invasion and metastasis by targeting PTPN9 and PTPRF to promote EGF signaling. J Cell Sci. 2013;126(Pt 6):1440-1453. doi: 10.1242/jcs.118299 [doi]. 34. Gebeshuber CA, Zatloukal K, Martinez J. miR-29a suppresses tristetraprolin, which is a regulator of epithelial polarity and metastasis. EMBO Rep. 2009;10(4):400-405. doi: 10.1038/embor.2009.9 [doi]. 35. Chou J, Lin JH, Brenot A, Kim JW, Provot S, Werb Z. GATA3 suppresses metastasis and modulates the tumour microenvironment by regulating microRNA-29b expression. Nat Cell Biol. 2013;15(2):201-213. doi: 10.1038/ncb2672 [doi]. 36. Yu F, Deng H, Yao H, Liu Q, Su F, Song E. Mir-30 reduction maintains self-renewal and inhibits apoptosis in breast tumor-initiating cells. Oncogene. 2010;29(29):4194-4204. doi: 10.1038/onc.2010.167 [doi]. 37. Wu F, Zhu S, Ding Y, Beck WT, Mo YY. MicroRNAmediated regulation of Ubc9 expression in cancer cells. Clin Cancer Res. 2009;15(5):1550-1557. doi: 10.1158/10780432.CCR-08-0820 [doi]. 38. Augoff K, McCue B, Plow EF, Sossey-Alaoui K. miR-31 and its host gene lncRNA LOC554202 are regulated by promoter hypermethylation in triple-negative breast cancer. Mol Cancer. 2012;11:5-4598-11-5. doi: 10.1186/1476-4598-11-5 [doi]. 39. Sossey-Alaoui K, Downs-Kelly E, Das M, Izem L, Tubbs R, Plow EF. WAVE3, an actin remodeling protein, is regulated by the metastasis suppressor microRNA, miR31, during the invasion-metastasis cascade. Int J Cancer. 2011;129(6):1331-1343. doi: 10.1002/ijc.25793 [doi]. 40. Lujambio A, Calin GA, Villanueva A, et al. A microRNA DNA methylation signature for human cancer metastasis. Proc Natl Acad Sci U S A. 2008;105(36):13556-13561. doi: 10.1073/pnas.0803055105 [doi]. 41. Ahn YH, Gibbons DL, Chakravarti D, et al. ZEB1 drives

prometastatic actin cytoskeletal remodeling by downregulating miR-34a expression. J Clin Invest. 2012;122(9):3170-3183. doi: 10.1172/JCI63608 [doi]. 42. Yang S, Li Y, Gao J, et al. MicroRNA-34 suppresses breast cancer invasion and metastasis by directly targeting fra-1. Oncogene. 2013;32(36):4294-4303. doi: 10.1038/ onc.2012.432 [doi]. 43. Fang L, Du WW, Yang W, et al. MiR-93 enhances angiogenesis and metastasis by targeting LATS2. Cell Cycle. 2012;11(23):4352-4365. doi: 10.4161/cc.22670 [doi]. 44. Martello G, Rosato A, Ferrari F, et al. A MicroRNA targeting dicer for metastasis control. Cell. 2010;141(7):1195-1207. doi: 10.1016/j.cell.2010.05.017 [doi]. 45. Liang YJ, Wang QY, Zhou CX, et al. MiR-124 targets slug to regulate epithelial-mesenchymal transition and metastasis of breast cancer. Carcinogenesis. 2013;34(3):713-722. doi: 10.1093/carcin/bgs383 [doi]. 46. Feliciano A, Castellvi J, Artero-Castro A, et al. miR-125b acts as a tumor suppressor in breast tumorigenesis via its novel direct targets ENPEP, CK2-alpha, CCNJ, and MEGF9. PLoS One. 2013;8(10):e76247. doi: 10.1371/ journal.pone.0076247 [doi]. 47. Scott GK, Goga A, Bhaumik D, Berger CE, Sullivan CS, Benz CC. Coordinate suppression of ERBB2 and ERBB3 by enforced expression of micro-RNA miR-125a or miR125b. J Biol Chem. 2007;282(2):1479-1486. doi: M609383200 [pii]. 48. Ferracin M, Bassi C, Pedriali M, et al. miR-125b targets erythropoietin and its receptor and their expression correlates with metastatic potential and ERBB2/HER2 expression. Mol Cancer. 2013;12(1):130-4598-12-130. doi: 10.1186/1476-4598-12-130 [doi]. 49. Tang F, Zhang R, He Y, Zou M, Guo L, Xi T. MicroRNA125b induces metastasis by targeting STARD13 in MCF-7 and MDA-MB-231 breast cancer cells. PLoS One. 2012;7(5):e35435. doi: 10.1371/journal.pone.0035435 [doi]. 50. Png KJ, Halberg N, Yoshida M, Tavazoie SF. A microRNA regulon that mediates endothelial recruitment and metastasis by cancer cells. Nature. 2011;481(7380):190194. doi: 10.1038/nature10661 [doi]. 51. Tavazoie SF, Alarcon C, Oskarsson T, et al. Endogenous human microRNAs that suppress breast cancer metastasis. Nature. 2008;451(7175):147-152. doi: 10.1038/ nature06487 [doi]. 52. Fish JE, Santoro MM, Morton SU, et al. miR-126 regulates angiogenic signaling and vascular integrity. Dev Cell. 2008;15(2):272-284. doi: 10.1016/j.devcel.2008.07.008 [doi]. 53. Harris TA, Yamakuchi M, Ferlito M, Mendell JT, Lowenstein CJ. MicroRNA-126 regulates endothelial expression of vascular cell adhesion molecule 1. Proc Natl Acad Sci U S A. 2008;105(5):1516-1521. doi: 10.1073/ pnas.0707493105 [doi]. 54. Sempere LF, Preis M, Yezefski T, et al. Fluorescence-based codetection with protein markers reveals distinct cellular compartments for altered MicroRNA expression in solid tumors. Clin Cancer Res. 2010;16(16):4246-4255. doi: 10.1158/1078-0432.CCR-10-1152 [doi].

163

PART ONE | CHAPTER 5

55. Li X, Shen Y, Ichikawa H, Antes T, Goldberg GS. Regulation of miRNA expression by src and contact normalization: Effects on nonanchored cell growth and migration. Oncogene. 2009;28(48):4272-4283. doi: 10.1038/onc.2009.278 [doi]. 56. Zhang Y, Yang P, Sun T, et al. miR-126 and miR-126* repress recruitment of mesenchymal stem cells and inflammatory monocytes to inhibit breast cancer metastasis. Nat Cell Biol. 2013;15(3):284-294. doi: 10.1038/ncb2690 [doi]. 57. Lim PK, Bliss SA, Patel SA, et al. Gap junction-mediated import of microRNA from bone marrow stromal cells can elicit cell cycle quiescence in breast cancer cells. Cancer Res. 2011;71(5):1550-1560. doi: 10.1158/0008-5472.CAN10-2372 [doi]. 58. Anand S, Majeti BK, Acevedo LM, et al. MicroRNA-132mediated loss of p120RasGAP activates the endothelium to facilitate pathological angiogenesis. Nat Med. 2010;16(8):909-914. doi: 10.1038/nm.2186 [doi]. 59. Sachdeva M, Zhu S, Wu F, et al. p53 represses c-myc through induction of the tumor suppressor miR-145. Proc Natl Acad Sci U S A. 2009;106(9):3207-3212. doi: 10.1073/ pnas.0808042106 [doi]. 60. Sachdeva M, Mo YY. MicroRNA-145 suppresses cell invasion and metastasis by directly targeting mucin 1. Cancer Res. 2010;70(1):378-387. doi: 10.1158/0008-5472. CAN-09-2021 [doi]. 61. Spizzo R, Nicoloso MS, Lupini L, et al. miR-145 participates with TP53 in a death-promoting regulatory loop and targets estrogen receptor-alpha in human breast cancer cells. Cell Death Differ. 2010;17(2):246-254. doi: 10.1038/cdd.2009.117 [doi]. 62. Gotte M, Mohr C, Koo CY, et al. miR-145-dependent targeting of junctional adhesion molecule A and modulation of fascin expression are associated with reduced breast cancer cell motility and invasiveness. Oncogene. 2010;29(50):6569-6580. doi: 10.1038/ onc.2010.386 [doi]. 63. Xiang M, Birkbak NJ, Vafaizadeh V, et al. STAT3 induction of miR-146b forms a feedback loop to inhibit the NF-kappaB to IL-6 signaling axis and STAT3-driven cancer phenotypes. Sci Signal. 2014;7(310):ra11. doi: 10.1126/scisignal.2004497 [doi]. 64. Bhaumik D, Scott GK, Schokrpur S, Patil CK, Campisi J, Benz CC. Expression of microRNA-146 suppresses NF-kappaB activity with reduction of metastatic potential in breast cancer cells. Oncogene. 2008;27(42):5643-5647. doi: 10.1038/onc.2008.171 [doi]. 65. Hurst DR, Edmonds MD, Scott GK, Benz CC, Vaidya KS, Welch DR. Breast cancer metastasis suppressor 1 up-regulates miR-146, which suppresses breast cancer metastasis. Cancer Res. 2009;69(4):1279-1283. doi: 10.1158/0008-5472.CAN-08-3559 [doi]. 66. Yokobori T, Suzuki S, Tanaka N, et al. MiR-150 is associated with poor prognosis in esophageal squamous cell carcinoma via targeting the EMT inducer ZEB1. Cancer Sci. 2013;104(1):48-54. doi: 10.1111/cas.12030 [doi]. 67. Jiang S, Zhang HW, Lu MH, et al. MicroRNA-155 functions as an OncomiR in breast cancer by targeting the suppressor of cytokine signaling 1 gene. Cancer Res.

164

2010;70(8):3119-3127. doi: 10.1158/0008-5472.CAN-094250 [doi]. 68. Zhang CM, Zhao J, Deng HY. MiR-155 promotes proliferation of human breast cancer MCF-7 cells through targeting tumor protein 53-induced nuclear protein 1. J Biomed Sci. 2013;20:79-0127-20-79. doi: 10.1186/14230127-20-79 [doi]. 69. Kong W, Yang H, He L, et al. MicroRNA-155 is regulated by the transforming growth factor beta/smad pathway and contributes to epithelial cell plasticity by targeting RhoA. Mol Cell Biol. 2008;28(22):6773-6784. doi: 10.1128/ MCB.00941-08 [doi]. 70. Kong W, He L, Coppola M, et al. MicroRNA-155 regulates cell survival, growth, and chemosensitivity by targeting FOXO3a in breast cancer. J Biol Chem. 2010;285(23):1786917879. doi: 10.1074/jbc.M110.101055 [doi]. 71. Liang Z, Wu H, Reddy S, et al. Blockade of invasion and metastasis of breast cancer cells via targeting CXCR4 with an artificial microRNA. Biochem Biophys Res Commun. 2007;363(3):542-546. doi: S0006-291X(07)01935-3 [pii]. 72. Chiang CH, Chu PY, Hou MF, Hung WC. MiR-182 promotes proliferation and invasion and elevates the HIF1alpha-VEGF-A axis in breast cancer cells by targeting FBXW7. Am J Cancer Res. 2016;6(8):1785-1798. 73. Fite K, Gomez-Cambronero J. Down-regulation of MicroRNAs (MiRs) 203, 887, 3619 and 182 prevents vimentin-triggered, phospholipase D (PLD)-mediated cancer cell invasion. J Biol Chem. 2016;291(2):719-730. doi: 10.1074/jbc.M115.686006 [doi]. 74. Li XF, Yan PJ, Shao ZM. Downregulation of miR-193b contributes to enhance urokinase-type plasminogen activator (uPA) expression and tumor progression and invasion in human breast cancer. Oncogene. 2009;28(44):3937-3948. doi: 10.1038/onc.2009.245 [doi]. 75. Meng Z, Fu X, Chen X, et al. miR-194 is a marker of hepatic epithelial cells and suppresses metastasis of liver cancer cells in mice. Hepatology. 2010;52(6):2148-2157. doi: 10.1002/hep.23915 [doi]. 76. Pencheva N, Tran H, Buss C, et al. Convergent multimiRNA targeting of ApoE drives LRP1/LRP8-dependent melanoma metastasis and angiogenesis. Cell. 2012;151(5):1068-1082. doi: 10.1016/j.cell.2012.10.028 [doi]. 77. Zhang HF, Xu LY, Li EM. A family of pleiotropically acting microRNAs in cancer progression, miR-200: Potential cancer therapeutic targets. Curr Pharm Des. 2014;20(11):1896-1903. doi: CPD-EPUB-54314 [pii]. 78. Castilla MA, Diaz-Martin J, Sarrio D, et al. MicroRNA-200 family modulation in distinct breast cancer phenotypes. PLoS One. 2012;7(10):e47709. doi: 10.1371/journal. pone.0047709 [doi]. 79. Roy SS, Gonugunta VK, Bandyopadhyay A, et al. Significance of PELP1/HDAC2/miR-200 regulatory network in EMT and metastasis of breast cancer. Oncogene. 2014;33(28):3707-3716. doi: 10.1038/ onc.2013.332 [doi]. 80. Dykxhoorn DM, Wu Y, Xie H, et al. miR-200 enhances mouse breast cancer cell colonization to form distant metastases. PLoS One. 2009;4(9):e7181. doi: 10.1371/ journal.pone.0007181 [doi].

MicroRNAs in breast cancer metastases

81. Bracken CP, Gregory PA, Kolesnikoff N, et al. A doublenegative feedback loop between ZEB1-SIP1 and the microRNA-200 family regulates epithelial-mesenchymal transition. Cancer Res. 2008;68(19):7846-7854. doi: 10.1158/0008-5472.CAN-08-1942 [doi]. 82. Burk U, Schubert J, Wellner U, et al. A reciprocal repression between ZEB1 and members of the miR-200 family promotes EMT and invasion in cancer cells. EMBO Rep. 2008;9(6):582-589. doi: 10.1038/embor.2008.74 [doi]. 83. Gibbons DL, Lin W, Creighton CJ, et al. Contextual extracellular cues promote tumor cell EMT and metastasis by regulating miR-200 family expression. Genes Dev. 2009;23(18):2140-2151. doi: 10.1101/gad.1820209 [doi]. 84. Gregory PA, Bert AG, Paterson EL, et al. The miR-200 family and miR-205 regulate epithelial to mesenchymal transition by targeting ZEB1 and SIP1. Nat Cell Biol. 2008;10(5):593-601. doi: 10.1038/ncb1722 [doi]. 85. Iliopoulos D, Polytarchou C, Hatziapostolou M, et al. MicroRNAs differentially regulated by akt isoforms control EMT and stem cell renewal in cancer cells. Sci Signal. 2009;2(92):ra62. doi: 10.1126/scisignal.2000356 [doi]. 86. Iliopoulos D, Lindahl-Allen M, Polytarchou C, Hirsch HA, Tsichlis PN, Struhl K. Loss of miR-200 inhibition of Suz12 leads to polycomb-mediated repression required for the formation and maintenance of cancer stem cells. Mol Cell. 2010;39(5):761-772. doi: 10.1016/j. molcel.2010.08.013 [doi]. 87. Li S, Wang Q, Wang Y, Chen X, Wang Z. PLC-gamma1 and Rac1 coregulate EGF-induced cytoskeleton remodeling and cell migration. Mol Endocrinol. 2009;23(6):901-913. doi: 10.1210/me.2008-0368 [doi]. 88. Park SM, Gaur AB, Lengyel E, Peter ME. The miR-200 family determines the epithelial phenotype of cancer cells by targeting the E-cadherin repressors ZEB1 and ZEB2. Genes Dev. 2008;22(7):894-907. doi: 10.1101/gad.1640608 [doi]. 89. Schickel R, Park SM, Murmann AE, Peter ME. miR-200c regulates induction of apoptosis through CD95 by targeting FAP-1. Mol Cell. 2010;38(6):908-915. doi: 10.1016/j.molcel.2010.05.018 [doi]. 90. Uhlmann S, Zhang JD, Schwager A, et al. miR-200bc/429 cluster targets PLCgamma1 and differentially regulates proliferation and EGF-driven invasion than miR-200a/141 in breast cancer. Oncogene. 2010;29(30):4297-4306. doi: 10.1038/onc.2010.201 [doi]. 91. Korpal M, Ell BJ, Buffa FM, et al. Direct targeting of Sec23a by miR-200s influences cancer cell secretome and promotes metastatic colonization. Nat Med. 2011;17(9):1101-1108. doi: 10.1038/nm.2401 [doi]. 92. Yu SJ, Hu JY, Kuang XY, et al. MicroRNA-200a promotes anoikis resistance and metastasis by targeting YAP1 in human breast cancer. Clin Cancer Res. 2013;19(6):13891399. doi: 10.1158/1078-0432.CCR-12-1959 [doi]. 93. Liu J, Mao Q, Liu Y, Hao X, Zhang S, Zhang J. Analysis of miR-205 and miR-155 expression in the blood of breast cancer patients. Chin J Cancer Res. 2013;25(1):46-54. doi: 10.3978/j.issn.1000-9604.2012.11.04 [doi]. 94. Elgamal OA, Park JK, Gusev Y, et al. Tumor suppressive function of mir-205 in breast cancer is linked to HMGB3

regulation. PLoS One. 2013;8(10):e76402. doi: 10.1371/ journal.pone.0076402 [doi]. 95. Iorio MV, Casalini P, Piovan C, et al. microRNA-205 regulates HER3 in human breast cancer. Cancer Res. 2009;69(6):2195-2200. doi: 10.1158/0008-5472.CAN-082920 [doi]. 96. Wu H, Zhu S, Mo YY. Suppression of cell growth and invasion by miR-205 in breast cancer. Cell Res. 2009;19(4):439-448. doi: 10.1038/cr.2009.18 [doi]. 97. Li Y, Hong F, Yu Z. Decreased expression of microRNA-206 in breast cancer and its association with disease characteristics and patient survival. J Int Med Res. 2013;41(3):596-602. doi: 10.1177/0300060513485856 [doi]. 98. Song G, Zhang Y, Wang L. MicroRNA-206 targets notch3, activates apoptosis, and inhibits tumor cell migration and focus formation. J Biol Chem. 2009;284(46):31921-31927. doi: 10.1074/jbc.M109.046862 [doi]. 99. Di Leva G, Gasparini P, Piovan C, et al. MicroRNA cluster 221-222 and estrogen receptor alpha interactions in breast cancer. J Natl Cancer Inst. 2010;102(10):706-721. doi: 10.1093/jnci/djq102 [doi]. 100. Leivonen SK, Makela R, Ostling P, et al. Protein lysate microarray analysis to identify microRNAs regulating estrogen receptor signaling in breast cancer cell lines. Oncogene. 2009;28(44):3926-3936. doi: 10.1038/ onc.2009.241 [doi]. 101. Adams BD, Cowee DM, White BA. The role of miR-206 in the epidermal growth factor (EGF) induced repression of estrogen receptor-alpha (ERalpha) signaling and a luminal phenotype in MCF-7 breast cancer cells. Mol Endocrinol. 2009;23(8):1215-1230. doi: 10.1210/me.20090062 [doi]. 102. Kondo N, Toyama T, Sugiura H, Fujii Y, Yamashita H. miR-206 expression is down-regulated in estrogen receptor alpha-positive human breast cancer. Cancer Res. 2008;68(13):5004-5008. doi: 10.1158/0008-5472.CAN-080180 [doi]. 103. Stinson S, Lackner MR, Adai AT, et al. TRPS1 targeting by miR-221/222 promotes the epithelial-to-mesenchymal transition in breast cancer. Sci Signal. 2011;4(177):ra41. doi: 10.1126/scisignal.2001538 [doi]. 104. Hwang MS, Yu N, Stinson SY, et al. miR-221/222 targets adiponectin receptor 1 to promote the epithelial-tomesenchymal transition in breast cancer. PLoS One. 2013;8(6):e66502. doi: 10.1371/journal.pone.0066502 [doi]. 105. Nassirpour R, Mehta PP, Baxi SM, Yin MJ. miR-221 promotes tumorigenesis in human triple negative breast cancer cells. PLoS One. 2013;8(4):e62170. doi: 10.1371/ journal.pone.0062170 [doi]. 106. Pichiorri F, Palmieri D, De Luca L, et al. In vivo NCL targeting affects breast cancer aggressiveness through miRNA regulation. J Exp Med. 2013;210(5):951-968. doi: 10.1084/jem.20120950 [doi]. 107. Zhang S, Kim K, Jin UH, et al. Aryl hydrocarbon receptor agonists induce microRNA-335 expression and inhibit lung metastasis of estrogen receptor negative breast cancer cells. Mol Cancer Ther. 2012;11(1):108-118. doi: 10.1158/1535-7163.MCT-11-0548 [doi].

165

PART ONE | CHAPTER 5

108. Huang Q, Gumireddy K, Schrier M, et al. The microRNAs miR-373 and miR-520c promote tumour invasion and metastasis. Nat Cell Biol. 2008;10(2):202-210. doi: 10.1038/ncb1681 [doi]. 109. Yan GR, Xu SH, Tan ZL, Liu L, He QY. Global identification of miR-373-regulated genes in breast cancer by quantitative proteomics. Proteomics. 2011;11(5):912920. doi: 10.1002/pmic.201000539 [doi]. 110. Li QQ, Chen ZQ, Cao XX, et al. Involvement of NF-kappaB/miR-448 regulatory feedback loop in chemotherapy-induced epithelial-mesenchymal transition of breast cancer cells. Cell Death Differ. 2011;18(1):16-25. doi: 10.1038/cdd.2010.103 [doi]. 111. Reddy SD, Pakala SB, Ohshiro K, Rayala SK, Kumar R. MicroRNA-661, a c/EBPalpha target, inhibits metastatic tumor antigen 1 and regulates its functions. Cancer Res. 2009;69(14):5639-5642. doi: 10.1158/0008-5472.CAN-090898 [doi].

166

112. Vetter G, Saumet A, Moes M, et al. miR-661 expression in SNAI1-induced epithelial to mesenchymal transition contributes to breast cancer cell invasion by targeting nectin-1 and StarD10 messengers. Oncogene. 2010;29(31):4436-4448. doi: 10.1038/onc.2010.181 [doi]. 113. Yu F, Yao H, Zhu P, et al. Let-7 regulates self renewal and tumorigenicity of breast cancer cells. Cell. 2007;131(6):1109-1123. doi: S0092-8674(07)01417-1 [pii]. 114. Qian P, Zuo Z, Wu Z, et al. Pivotal role of reduced let-7g expression in breast cancer invasion and metastasis. Cancer Res. 2011;71(20):6463-6474. doi: 10.1158/00085472.CAN-11-1322 [doi]. 115. Lamouille S, Subramanyam D, Blelloch R, Derynck R. Regulation of epithelial-mesenchymal and mesenchymalepithelial transitions by microRNAs. Curr Opin Cell Biol. 2013;25(2):200-207. doi: 10.1016/j.ceb.2013.01.008 [doi].

MicroRNAs in breast cancer metastases

167

Part Two

Phenotyping of distant breast cancer metastases

Chapter 6 Willemijne AME Schrijver, Karijn PM Suijkerbuijk, Carla H van Gils, Elsken van der Wall, Paul J van Diest, Cathy B Moelans

Receptor conversion in breast cancer metastases: a systematic review and meta-analysis Manuscript in preparation

PART TWO | CHAPTER 6

ABSTRACT In metastatic breast cancer, hormone and/or HER2-targeted therapy decision making is still largely based on tissue characteristics of the primary tumor. However, a change of ERα, PR and HER2 status in distant metastases has frequently been reported. The actual incidence of this phenomenon has been debated. We performed a meta-analysis including 39 studies assessing receptor conversion from paired primary breast tumors to distant breast cancer metastases. We noted the direction of the change (positive to negative or vice versa) and performed subgroup analyses for different thresholds for positivity, the type of test used to assess HER2 receptor status and metastasis location-specific differences. Overall, the incidence of receptor conversion varied significantly between studies. For ERα, PR and HER2 we found random effects pooled discordance proportions of 19%, 31% and 10%, respectively. For ERα and PR, a switch from positive to negative receptor status occurred significantly more often than from negative to positive. Furthermore, ERα discordance was significantly higher in central nervous system compared to skin metastases (21% versus 5%, p=0.049) and PR discordance was higher in bone (43%, p=0.041) and liver metastases (47%. p=0.03), compared to central nervous system metastases (23%). Receptor conversion in breast cancer is frequent, stressing the need for personalized cancer care. Large prospective studies assessing the impact of receptor conversion on treatment efficacy and survival are needed. Meanwhile, reassessing receptor status in metastases is strongly encouraged.

172

Receptor conversion in breast cancer metastases

INTRODUCTION Despite advances in breast cancer treatment during the last decades, most metastatic breast cancer patients still have poor life expectancy. Therefore, acquiring more profound insights into the phenotypic and molecular composition of metastatic tumors, paving the way for more effective therapeutic regiments, is of utmost importance. The estrogen receptor alpha (ERα), progesterone receptor (PR) and human epidermal growth factor receptor 2 (HER2) status have proven their clinical utility in guiding therapeutic decision-making for metastatic breast cancer 1. However, prescription of endocrine or HER2-targeted therapies is mainly directed on the biomarker status of the primary tumor. However, increasing evidence shows extensive differences between immunohistochemically assessed tissue characteristics of primary breast tumors and their paired metastases 2-6. For ERα, PR and HER2, widely varying discordance rates have been reported so far: 3-54% for ERα, 5-78% for PR and 0-34% for HER2 7-9. This change of hormone receptor and/or HER2 status between primary tumor and paired metastasis within a patient is usually denoted ‘receptor conversion’. Several guidelines have now consented that patients with accessible breast cancer metastases should be offered a biopsy or resection to confirm the diagnosis of metastases and to reassess ERα, PR and HER2 status 10-12. Obtaining metastatic material is however not without risk, introducing potential hemorrhage and infection. Furthermore, to date, there is no evidence supporting improved survival outcomes when treatment regimens are based on the receptor status of the metastasis instead of the primary tumor 10. Although previous studies have summarized available data and literature, a solid systematic review addressing receptor conversion in distant metastases including meta-analysis to date is lacking. Other studies only included lymph node metastases 13 or assessed receptor conversion in pooled loco-regional and distant metastases 10, despite large differences between these two groups 9. In this study we focus on distant metastases, since they are the largest contributor to breast cancer related mortality 14. Furthermore, tissue characteristics of these distant sites are not always reassessed due to difficulty in obtaining a biopsy, leading to potential suboptimal treatment. Therefore we set out to systematically evaluate the frequency of receptor conversion between primary breast tumors and distant breast cancer metastases (excluding regional lymph nodes), with special attention to thresholds for positivity (1% versus 10% for ERα and PR), the type of test used to assess HER2 status (immunohistochemistry, fluorescence in situ hybridization or a combination of both) and metastasis location-specific differences.

173

PART TWO | CHAPTER 6

MATERIALS AND METHODS Selection of studies The databases Embase, Cochrane and PubMed were searched on July 11th 2016 for relevant studies, covering a time period from 1986 until 2016. The literature search used the following terms (with synonyms, MeSH-terms and closely related words): “breast cancer” and “metastasis”, combined with “estrogen receptor/ERα”, “progesterone receptor/PR”, “HER2/neu”, “immunohistochemistry/IHC” or “in-situ hybridization/ISH” and “receptor conversion/dis- or concordance” . Duplicates were eliminated using RefWorks. The search strategy is listed in Appendix 1. All articles were screened for relevance. Original full text research articles directly describing immunohistochemically assessed ERα, PR or HER2status in primary breast tumors compared to paired distant metastases were included. Exclusion criteria were: case reports, meta-analyses and reviews, cytology specimens (or circulating tumor cells or tissue collected by fine needle aspiration), male patients, less than ten described cases, axillary lymph node or locoregional metastases, other methodology than IHC or ISH, other receptors than ERα, PR and HER2 and languages other than English. Data collection 5521 unique articles were identified and screened. When no full text was available online, printed copies of these articles were requested. Titles and abstracts were screened for relevance. 3733 articles were excluded, because they did not meet the selection criteria (Figure 1). Reference lists of the papers of interest were screened manually and using Scopus to ensure sensitivity of the search strategy and to identify additional relevant studies, which identified 11 additional articles. 51 selected publications were independently reviewed by two of the authors (WS and KS) to determine the eligibility of each article in the meta-analysis. Quality assessment of included studies was performed by critical appraisal, based on standardized criteria for diagnostic research using the QUADAS-2 tool for quality assessment of diagnostic accuracy studies (Table 1) 15. This tool consists of four key domains covering patient selection, index test, reference standard, and flow of patients through the study (timing of the index test and reference standard). Each domain was assessed in terms of the risk of bias and the first three were also assessed in terms of concerns regarding applicability. Risk of bias and concern of applicability for each domain was rated as low, high or unclear. Studies with two or more high or unclear ratings were excluded for this meta-analysis. In case of disagreement the study was discussed until consensus was reached among the two investigators. The following details were extracted: total number of patients evaluated, clinicopathologic characteristics of the primary tumor (if reported), site of and time to relapse, ERα, PR and HER2 discordance rates with direction of conversion (positive to negative or vice versa) and information about treatment and survival. The technique used 174

Receptor conversion in breast cancer metastases

Figure 1. Flow diagram of study selection for this meta-analysis

to define endocrine receptor or HER2 status (IHC and/or ISH) and the specific antibodies or probes were also registered. According to the ESMO and ASCO clinical practice guidelines, using a standardized assessment methodology (1% or 10% cut-off, Allred or H-score) for defining ERÎą and PR positivity 16 is a prerequisite. HER2 ISH should be used on all samples or in case of an ambiguous (2+) IHC score 17. Therefore we focused on studies that met these criteria, but due to low numbers we also included studies that did not perform ISH. To correct for this bias, we included subanalyses to check for conversion differences between used techniques. For complete and transparent reporting of the results of this review, we used the preferred reporting items for systematic reviews and meta-analyses (PRISMA) statement checklist18. 175

PART TWO | CHAPTER 6

Table 1. Critical Appraisal according to Quadas-2 Reference standard

Flow and timing

Patient selection

Index test

Reference standard

Concerns regarding applicability

Index test

Risk of bias

Patient selection

Study

Amir et al., 2012b 38,49,53 - BRITS study Thompson et al., 2010 - DESTINY study Amir et al., 2012a

Low

Applicable for review

Bogina et al., 2011 37

Low

Brogi et al., 2011 66

Low

Applicable for review

Chan et al., 2012 67

Low

Cummings et al., 2014 68

Low

Duchnowska et al., 2012 32

Low

Edgerton et al., 2003 45

Low

Fabi et al., 2011 39

Low

Fuchs et al., 2006 69

Low

Gaedcke et al., 2007 70

Low

Gonzalez-Angulo et al., 2011 71

Low

Hilton et al., 2010 61

Low

Hoefnagel et al., 2013 4

Low

Hoefnagel et al., 2010 & 2012 2,3

Low

Jensen et al., 2010 34

Low

Karagoz Ozen et al., 2014 47

Low

Kulka et al., 2016 62

Low

Nakamura et al., 2013 35

Low

Regitnig et al., 2004 63

Low

Santinelli et al., 2008 72

Low

Simmons et al., 2009 52

Low

Thomson et al., 2016 73

Low

Yonemori et al., 2008 36

Low

Zidan et al., 2005 74

Low

Aurilio et al., 2013 50

Moderate

176

Applicable for review Applicable for review Applicable for review Applicable for review Applicable for review Applicable for review Applicable for review Applicable for review Applicable for review Applicable for review Applicable for review Applicable for review Applicable for review Applicable for review Applicable for review Applicable for review Applicable for review Applicable for review Applicable for review Applicable for review Applicable for review Applicable for review Applicable for review

Receptor conversion in breast cancer metastases

Reference standard

Index test

Concerns regarding applicability Flow and timing

Reference standard

Index test

Risk of bias Patient selection

Study

Patient selection

Table 1. Continued

Bachmann + + ? + Moderate + + + Applicable et al., 2013 75 for review Botteri ? + ? + Moderate + + + Applicable et al., 2012 76 for review Cabioglu + + ? + Moderate + + + Applicable 77 et al., 2009 for review Chang + + ? + Moderate + + + Applicable et al., 2011 44 for review Curigliano + + ? + Moderate + + + Applicable et al., 2011 51 for review Curtit + + ? + Moderate + + + Applicable et al., 2013 31 for review Gancberg + + ? + Moderate + + + Applicable et al., 2002 78 for review Idirisinghe + + ? + Moderate + + + Applicable 33 et al., 2010 for review Lorincz + + ? + Moderate + + + Applicable 79 et al., 2006 for review Lower + + ? + Moderate + + + Applicable et al., 2009 46 for review Omoto + + ? + Moderate + + + Applicable et al., 2010 80 for review Shen + + + Moderate + + + Applicable et al., 2015 48 for review St. Romain + + ? + Moderate + + + Applicable 81 et al., 2012 for review Vincent-Salomon + + ? + Moderate + + Applicable 82 et al., 2002 for review Shao + + + + Low + ? ? Not applicable et al., 2011 83 for review Wu + + + + Low Not applicable et al., 2008 84 for review Lower + + + Moderate + + Not applicable et al., 2005 85 for review Amir + High + + Not applicable et al., 2008 86 for review Gullo ? ? ? High Not applicable 87 et al., 2013 for review Kalinsky + ? High Not applicable et al., 2015 88 for review Kamby + ? ? High Not applicable et al., 1989 89 for review Koo ? ? ? High ? Not applicable et al., 2010 90 for review Lear-Kaul + ? ? + High + Not applicable et al., 2003 91 for review Nogami + ? ? + High ? + Not applicable 92 et al., 2014 for review Schwarz + ? ? + High + Not applicable 93 et al., 2004 for review Welter + ? High + Not applicable et al., 2008 94 for review +: Low risk, -: High risk, ?: Unclear In case of disagreement the study was discussed, resulting in consensus about inclusion among the two investigators. Amir et al., 2012 53 describe two clinical studies, namely the BRITS study 38 and the DESTINY study 49. For total conversion the data of both clinical studies was used. However, for the direction of conversion (positive to negative and vice versa) discordance percentages were only presented in the pooled study 53. Hoefnagel et al., 2012 and 2010 2,3 describe the same cohort and therefore only the most recent study (Hoefnagel et al., 2012 3) was included.

177

PART TWO | CHAPTER 6

Statistical analysis The proportion of ERα, PR and HER2 changes and 95% confidence intervals (CIs) were calculated for each study. Subanalyses were performed for thresholds for positivity (1% versus 10% for ERα and PR), the type of test used to assess receptor status (IHC, FISH or a combination of both) and location of metastasis. For HER2 immunohistochemistry, 0 and 1+ were considered negative, 2+ as equivocal and 3+ as positive. For meta-analysis, the conservative random effects model was used to calculate the pooled proportion, and statistical significance was determined using the Z-test 19. Heterogeneity across studies was assessed using both Q test and the Higgins I2 20,21. I2 values of 0%, 25%, 50%, and 75% are indicated as ‘no’, ‘low’, ‘moderate’ and ‘high’ heterogeneity, respectively. To test publications bias, funnel plots were made and the Egger’s test was used 22. Comparison of subgroups was performed using Mann-Whitney U or Kruskal-Wallis tests. All statistical tests for meta-analyses were performed using Comprehensive Meta-Analysis 2.0 software (Biostat, Englewood, NJ, USA) and IBM SPSS Statistics 23. P-values <0.05 (two-sided) were considered statistically significant.

RESULTS Funnel plots showed high publication bias when including studies describing less than twenty patients. We therefore excluded 12 such articles. The main characteristics of the remaining thirty-nine studies are reported in Supplementary Table 1. Data on ERα, PR and HER2 status in the primary tumor and corresponding distant metastasis were available in 27, 24 and 35 studies, respectively. The discordance rate was assessed in 1948 patients for ERα, in 1730 patients for PR and in 2440 patients for HER2. Mean age at diagnosis of the primary tumors was 51 years (26 studies; range 22-93 years), 86% of tumors were of the ductal type (14 studies; 1038/1204 tumors) and the mean time between primary tumor and matched distant metastasis was 51 months (28 studies; range 0-432 months). Heterogeneity and publication bias Overall, variation between studies was high. For ERα, the heterogeneity for total conversion and conversion from positive to negative was moderate to high (I²: 73% and 66% respectively), but for conversion from negative to positive low to moderate (I²: 37%). This barely changed when studies were more specifically subdivided per threshold for positivity (Table 2). A similar trend was seen for PR conversion, but less heterogeneity was perceived for the 10% threshold of positivity (I²: 41%, 51% and 18% for total, positive to negative and negative to positive conversion respectively). For HER2, the smallest variation between studies was seen when FISH was used assess receptor status (I²: 69%, 35% and 45% for total, positive to negative and negative to positive conversion respectively). 178

0.010

0.000

-/+

0.05

0.11

0.05

0.08

0.18

0.47

10%

0.55

0.39

0.60

0.52

0.90

0.48

IHC

0.07

0.013

0.07

% 11 4 9

H 74 36 52

FISH

69 35 45

0.034 0.034

0.06

%: percentage of conversion, PB: publication bias calculated with Egger’s test, H: heterogeneity according to the I2%. Bold < 0.05

+/-

-/+

0.010

0.005

0.10

0.28

+/-

-/+

+/-

0.63

total

All

total

HER2 total

Table 2. Overview of subanalyses with conversion percentages, publication bias and heterogeneity

0.000

82 0.000

0.000

IHC + FISH

Receptor conversion in breast cancer metastases

179

PART TWO | CHAPTER 6

Pooled proportion of ERα discordance between primary breast tumor and distant metastasis The total discordance proportion for ERα varied between studies from 7% to 51%, with a pooled random effects proportion of 19% (95% CI: 16-23%). The proportion of conversion from positive to negative was 13% (CI: 10-16%) and from negative to positive 7% (95% CI: 5-8%) (Figure 2). We divided studies into two groups using the 1% or 10% threshold for positivity, showing a total pooled ERα conversion proportion of 18% (95% CI: 14-23%) and 19% (95% CI: 15-25%) respectively (Supplementary Figure S1). No significant difference between both cut-offs was perceived for total discordance proportions (p=0.54). The amount of conversion from positive to negative was 11% (95% CI: 8-14) for the 1% threshold and 13% (95% CI: 9-19%) for the 10% threshold. Conversion from negative to positive occurred in 6% of tumors for both thresholds (95% CI 4-8% for the 1% and 95% CI: 4-10% for the 10% threshold). Conversion from positive to negative occurred significantly more often than from negative to positive (p=0.002, p=0.007 and p=0.018 for the total group, 1% and 10% thresholds for positivity, respectively). Pooled proportion of PR discordance between primary breast tumor and distant metastasis The meta-analytic pooled proportion for PR was 31% (95% CI: 27-36%), with a change from positive to negative of 24% (95% CI: 20-29%) and 8% (95% CI: 6-9%) vice versa (Figure 3). For the 1% threshold for positivity, the total proportion of discordance was 33% (95% CI: 27-40%), the conversion from positive to negative was 25% (95% CI: 19-33%) and the conversion from negative to positive was 7% (95% CI: 6-10). For the 10% cut-off, these values were 31% (95% CI: 26-36%), 24% (95% CI: 19-29%) and 8% (95% CI: 6-10%), respectively (Supplementary Figure S2). Also for PR, no significant difference in total discordance proportions was seen between the two cut-offs for positivity (p=0.54) and conversion from positive to negative occurred significantly more often than from negative to positive (p=0.000, p=0.000 and p=0.001 for the total group, 1% and 10% thresholds for positivity, respectively). Pooled proportion of HER2 discordance between primary breast tumor and distant metastasis The pooled proportion of HER2 conversion was 10% (95% CI: 8-14%). Positive to negative conversion occurred in 5% (95% CI: 3-7%) and negative to positive in 7% (95% CI: 5-8%) of all cases in this meta-analysis (Figure 4). We subdivided studies into three groups, studies using FISH only, studies using IHC only and studies using a combination of IHC and FISH (in case of 2+/equivocal IHC) to assess receptor status. The total discordance proportion for FISH, IHC and FISH + IHC was 11% (95% CI: 6-21%), 13% (95% CI: 7-22%) and 10% (95% CI: 7-14%) respectively. Conversion from positive to negative was 4% (95% CI: 2-9%), 6% (95% CI: 3-11%) and 5% (95% CI: 3-8%) and from negative to positive 9% (95% CI: 5-15%), 8% (95% CI: 5-14%) and 6% (95% CI: 5-8%) for the three groups (Supplementary Figure S3). 180

Receptor conversion in breast cancer metastases

Figure 2. Study-specific and pooled estimate for ERÎą discordance proportions for studies reporting ERÎą IHC in primary breast tumors and paired distant metastases. Study-specific data are ordered by date of publication. Discordance percentages are shown for total conversion (a), conversion from positive to negative (b) and conversion from negative to positive (c).

181

PART TWO | CHAPTER 6

Figure 3. Study-specific and pooled estimate for PR discordance proportions for studies reporting PR IHC in primary breast tumors and paired distant metastases. Study-specific data are ordered by date of publication. Discordance percentages are shown for total conversion (a), conversion from positive to negative (b) and conversion from negative to positive (c).

182

Receptor conversion in breast cancer metastases

Figure 4. Study-specific and pooled estimate for HER2 discordance proportions for studies reporting HER2 IHC in primary breast tumors and paired distant metastases. Study-specific data are ordered by date of publication. Discordance percentages are shown for total conversion (a), conversion from positive to negative (b) and conversion from negative to positive (c). uuu (see next page for Figure 4C)

183

PART TWO | CHAPTER 6

ttt

Figure 4. Continued

No significant difference was seen between total discordance proportions of these groups (p=0.46), and the pooled proportions of conversion from positive to negative and vice versa did not significantly differ (p=0.46, p=0.80, p=0.30 and p=0.17 for the total group, IHC+FISH, IHC only and FISH only, respectively). Some FISH-based studies used a HER2/CEP17 ratio of >2.0 and others of >2.2 to assess positivity. When comparing total discordance proportions between these two groups, the 2.2 threshold showed a significantly lower total pooled proportion than the 2.0 threshold (7% versus 11%; p=0.044). Pooled proportion of location-specific discordance between primary breast tumor and distant metastasis Additionally, discordance analyses were performed within subgroups representing the most frequent distant metastatic sites. CNS, bone, liver, skin and lung metastases were described in ten, four, six, two and two studies, respectively. Total discordance proportions for ERÎą, PR and HER2 are shown in Figure 5. ERÎą discordance in primary tumors with CNS metastases was borderline significantly higher than in primary tumors with skin metastases (p=0.049, 21% versus 5%). PR discordance was significantly higher in primary tumors with bone (p=0.041; 43%) or liver metastases (p=0.03; 47%), compared to primary tumors with CNS metastases (23%). For HER2, no significant differences were observed for pooled discordance proportions between metastatic sub sites (p=0.84).

184

Receptor conversion in breast cancer metastases

Figure 5. Study-specific and pooled estimates for metastasis location-specific discordance proportions for studies repor ting ERα, PR and/or HER2 IHC in primary breast tumors and paired distant metastases. Study-specific data are ordered by date of publication. Locations of metastasis are grouped in CNS (central nervous system), bone, liver, lung and skin. The Y-axis shows the total pooled discordance proportion of all locations. Discordance percentages are shown for total conversion of ERα (a), total conversion of PR (b) and total conversion of HER2 (c).

185

PART TWO | CHAPTER 6

DISCUSSION At present, discordance of ERα, PR and HER2 status between primary breast tumors and paired metastases is well recognized 10. However, some notable questions remained unanswered: how frequent is this phenomenon, which factors influence its occurrence and does a treatment switch based on the characteristics of the metastasis promote survival? With this meta-analysis we aimed to answer especially the first question and reviewed the literature about the other two. For ERα, PR and HER2 we found random effects pooled discordance proportions of 19%, 31% and 10%, respectively. For ERα and PR, a switch from positive to negative receptor status occurred significantly more often than from negative to positive. Furthermore, metastasis location-specific differences were found. Together, these findings confirm the idea that breast cancer is a very heterogeneous disease, and stress the importance of personalized cancer care. In this meta-analysis we focused on immunohistochemically assessed receptor status of ERα, PR and HER2 with a predefined threshold for positivity, as recommended by clinical practice guidelines to enable reliable treatment decision making 16,17. Risk of relapse, prognosis and response to treatment are attributed to the type of breast cancer determined by these markers 23-25. Especially ERα has been considered an important positive prognostic marker and a predictive marker of response to endocrine therapies 26. Although approximately 75% of breast cancers show ERα-positivity, their outcome and response to therapy vary extremely 27. Receptor conversion is thought to be the result of clonal selection or selective pressure of therapy 28-30. Some studies in this meta-analysis indeed reported an effect of chemotherapy exposure on ERα or PR receptor conversion and of previous trastuzumab therapy on HER2 conversion 31-37. Other articles, however, could not demonstrate a correlation between receptor discordance and previous systemic therapies 38,39. In primary breast cancer, sequential biopsies have shown that ERα levels are reduced slightly with intervening endocrine therapy, while PR levels decrease more dramatically with up to half of tumors completely losing PR expression when resistance develops 40. Therefore, PR loss in the metastasis may be an important predictive marker for endocrine therapy response failure41-43. In recent years, clinical guidelines have increasingly started advising to re-assess metastatic tissue characteristics whenever possible 1,11,12. However, solid clinical evidence supporting these guidelines is currently lacking. One study reported responses to trastuzumab in 2 out of 5 patients with positive HER2 status after conversion 44. Other articles do show survival differences between concordant and discordant tumors, but the relation between therapy administration and discordance is poorly reported. For example, Chang et al., and Edgerton 186

Receptor conversion in breast cancer metastases

et al., showed a significantly better overall survival in patients without HER2 conversion compared to patients with conversion 44,45 and Lower et al. reported that patients with negative to positive conversion performed better compared to conversion from positive to negative 46. Regarding ERÎą and/or PR, conversion from a positive primary tumor to a negative metastasis was associated with significantly worse survival compared to patients remaining receptor positive. In contrast, no significant survival difference was seen between patients showing conversion from negative to positive compared to patients remaining negative 3. Consensus is however not reached about the influence of receptor conversion on survival 32,47,48. Change of therapeutic plan after biopsy of the metastasis has been reported in 14% to 62% for ERÎą and PR and 67% for HER2 38,49-52, but the long-term effect of this therapy switch has not been reported. Change in therapy was more often seen when there was apparent gain of receptor status 53. Regarding the data available, randomized controlled trials in this setting no longer seem to be ethical. Moreover, large prospective studies with sufficient follow-up on survival and therapy response are very much needed to gain more insight in the real clinical significance for breast cancer patients. A major limitation of describing immunohistochemical receptor conversion is the heterogeneity of the studies included. This could be attributed to the fact that many studies used retrospectively assessed data, potentially leading to differences in staining protocols, inter-observer bias and analytical errors 7. For example, analysis of HER2 IHC and FISH on the same primary breast tumors in different labs already showed discordance proportions of 18% and 12% respectively 54. Other factors that could cause heterogenic findings are differences in primary tumor characteristics (Supplementary Table S1; for example tumor type, grading and nodal status) and time interval between primary tumor and metastasis. We did extract these variables from the included studies whenever possible, but they were often reported independently from the conversion statistics. Sub analyses based on these data were therefore impossible. Several other limitations are different study cohorts (age and race), small sample size, use of different techniques interchangeably (ligand-binding assay versus immunohistochemistry), influence of divergent systemic therapies and use of different cut-offs for positivity (1% versus 10% for ERÎą and PR) 2. Therefore, we tried to select a group of studies that is as homogenous as possible. First, we excluded material obtained with fine needle aspiration, since insufficient sampling may potentially cause false negative results 55-57. Furthermore, we chose to focus on solid metastases only, given that a significant difference has been perceived between axillary lymph node metastases and distant metastases 8,9. We also tried to primarily include studies that reassessed receptor status of the primary tumor together with the metastasis to exclude potential technical differences, but this was rarely reported. Still, variation between studies remained high, which decreased by focusing on specific subgroups with matching thresholds for positivity (Table 2) or location of metastasis 187

PART TWO | CHAPTER 6

(Figure 5). For ERα, the lowest amount of publication bias and variability was seen for the 1% threshold for positivity. For PR, publication bias was smaller for the 1% threshold, but variability was smaller in the 10% threshold group. For both ERα and PR, no significant difference was observed between receptor conversion levels between both thresholds. For HER2, the least publication bias and variability were seen when only FISH had been used to assess positivity, and again no significant difference in receptor conversion was seen between techniques (Table 2). Since discordance percentages did not significantly differ between techniques, receptor conversion can be seen as a true biological phenomenon and is not solely the result of limited accuracy of receptor assays, as sometimes thought 12,58. Furthermore, the finding of metastatic subsite-specific differences in discordance frequency underlines its true biology. We demonstrated that ERα discordance was significantly higher in CNS compared to skin metastases and PR discordance was higher in bone and liver metastases compared to CNS metastases. Sometimes, it is thought that bone receptor status cannot be reliably assessed since antigenicity may be altered by decalcifying agents that enable sectioning of bone 59,60. However, bone was not the only site with high discordance rates. Moreover, we only included studies on bone metastases that were not decalcified 52,61 or decalcified with EDTA 50,62,63, since EDTA was shown not to alter ERα, PR and HER2 immunohistochemistry 64. Discordance percentages per metastatic subsite were assessed before by Yeung et al. 65. For CNS/brain, bone, liver and lung metastases they found total discordance proportion for ERα of 17%, 47.5%, 15% and 28% and for PR of 22%, 36%, 45% and 30.5%, respectively. For the same locations discordance, we report proportions of 21%, 29%, 14% and 9% for ERα and 23%, 41%, 45%, 19% for PR. Except for lung metastases, the pattern of discordance is roughly similar, with few overlapping studies between both analyses. These similarities suggest that discordance in distant metastases shows a locationspecific pattern, potentially adjusting to micro-environmental needs of the target organ. This systematic review confirms the frequent occurrence of ERα, PR and HER2 receptor conversion. High heterogeneity was however seen between patients, receptors, techniques used and subsites of metastasis leading to dispersed discordance percentages. Therefore, personalized cancer care is of the utmost importance. Although not yet prospectively examined, multiple studies report survival differences between patients with concordant and discordant receptor statuses, where the effect of treatment may be a confounder. Based on the present meta-analysis, we advise to biopsy and re-assess receptor status in distant metastases whenever possible. Prospective studies with sufficient (post-treatment) followup are needed to assess the clinical implications of receptor conversion for breast cancer treatment.

188

Receptor conversion in breast cancer metastases

REFERENCES 1. Van Poznak C, Somerfield MR, Bast RC, et al. Use of biomarkers to guide decisions on systemic therapy for women with metastatic breast cancer: American society of clinical oncology clinical practice guideline. J Clin Oncol. 2015;33(24):2695-2704. doi: 10.1200/ JCO.2015.61.1459 [doi]. 2. Hoefnagel LDC, van de Vijver MJ, van Slooten H, et al. Receptor conversion in distant breast cancer metastases. Breast Cancer Research. 2010;12(5):R75. doi: 10.1186/ bcr2645. 3. Hoefnagel LDC, Moelans CB, Meijer SL, et al. Prognostic value of estrogen receptor alpha and progesterone receptor conversion in distant breast cancer metastases. Cancer. 2012;118(20):4929-4935. doi: 10.1002/cncr.27518. 4. Hoefnagel LDC, van der Groep P, van de Vijver MJ, et al. Discordance in ER alpha, PR and HER2 receptor status across different distant breast cancer metastases within the same patient. Annals of Oncology. 2013;24(12):30173023. doi: 10.1093/annonc/mdt390. 5. Jiwa LS, van Diest PJ, Hoefnagel LD, et al. Upregulation of claudin-4, CAIX and GLUT-1 in distant breast cancer metastases. BMC Cancer. 2014;14:864-2407-14-864. doi: 10.1186/1471-2407-14-864 [doi]. 6. Kroigard AB, Larsen MJ, Thomassen M, Kruse TA. Molecular concordance between primary breast cancer and matched metastases. Breast J. 2016;22(4):420-430. doi: 10.1111/tbj.12596 [doi]. 7. Turner NH, Di Leo A. HER2 discordance between primary and metastatic breast cancer: Assessing the clinical impact. Cancer Treat Rev. 2013;39(8):947-957. doi: 10.1016/j.ctrv.2013.05.003. 8. Houssami N, Macaskill P, Balleine RL, Bilous M, Pegram MD. HER2 discordance between primary breast cancer and its paired metastasis: Tumor biology or test artefact? insights through meta-analysis. Breast Cancer Res Treat. 2011;129(3):659-674. doi: 10.1007/s10549-011-1632-x. 9. Aurilio G, Disalvatore D, Pruneri G, et al. A meta-analysis of oestrogen receptor, progesterone receptor and human epidermal growth factor receptor 2 discordance between primary breast cancer and metastases. Eur J Cancer. 2014;50(2):277-289. doi: 10.1016/j.ejca.2013.10.004. 10. Van Poznak C, Somerfield MR, Bast RC, et al. Use of biomarkers to guide decisions on systemic therapy for women with metastatic breast cancer: American society of clinical oncology clinical practice guideline. Journal of clinical oncology : official journal of the American Society of Clinical Oncology. 2015;33(24):2695-704. doi: 10.1200/ JCO.2015.61.1459. 11. Cardoso F, Costa A, Norton L, et al. 1st international consensus guidelines for advanced breast cancer (ABC 1). Breast. 2012;21(3):242-252. doi: 10.1016/j. breast.2012.03.003 [doi]. 12. Carlson RW, Allred DC, Anderson BO, et al. Metastatic breast cancer, version 1.2012: Featured updates to the NCCN guidelines. J Natl Compr Canc Netw. 2012;10(7):821-829. doi: 10/7/821 [pii]. 13. Carlsson J, Nordgren H, Sjostrom J, et al. HER2 expression in breast cancer primary tumours and corresponding

metastases. original data and literature review. Br J Cancer. 2004;90(12):2344-2348. doi: 10.1038/sj.bjc.6601881. 14. Marsden CG, Wright MJ, Carrier L, Moroz K, Pochampally R, Rowan BG. â&#x20AC;&#x153;A novel in vivo model for the study of human breast cancer metastasis using primary breast tumor-initiating cells from patient biopsiesâ&#x20AC;?. BMC Cancer. 2012;12:10-2407-12-10. doi: 10.1186/1471-2407-12-10 [doi]. 15. Whiting PF, Rutjes AW, Westwood ME, et al. QUADAS-2: A revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med. 2011;155(8):529-536. doi: 10.7326/0003-4819-155-8-201110180-00009 [doi]. 16. Senkus E, Kyriakides S, Ohno S, et al. Primary breast cancer: ESMO clinical practice guidelines for diagnosis, treatment and follow-up. Ann Oncol. 2015;26 Suppl 5:v830. doi: 10.1093/annonc/mdv298 [doi]. 17. Hammond ME, Hayes DF, Wolff AC, Mangu PB, Temin S. American society of clinical oncology/college of american pathologists guideline recommendations for immunohistochemical testing of estrogen and progesterone receptors in breast cancer. J Oncol Pract. 2010;6(4):195-197. doi: 10.1200/JOP.777003 [doi]. 18. Moher D, Liberati A, Tetzlaff J, Altman DG, PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. Int J Surg. 2010;8(5):336-341. doi: 10.1016/j.ijsu.2010.02.007 [doi]. 19. Borenstein M, Hedges LV, Higgins JP, Rothstein HR. A basic introduction to fixed-effect and random-effects models for meta-analysis. Res Synth Methods. 2010;1(2):97-111. doi: 10.1002/jrsm.12 [doi]. 20. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med. 2002;21(11):1539-1558. doi: 10.1002/sim.1186 [doi]. 21. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ. 2003;327(7414):557-560. doi: 10.1136/bmj.327.7414.557 [doi]. 22. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997;315(7109):629-634. 23. Yersal O, Barutca S. Biological subtypes of breast cancer: Prognostic and therapeutic implications. World J Clin Oncol. 2014;5(3):412-424. doi: 10.5306/wjco.v5.i3.412 [doi]. 24. Savci-Heijink CD, Halfwerk H, Hooijer GK, Horlings HM, Wesseling J, van de Vijver MJ. Retrospective analysis of metastatic behaviour of breast cancer subtypes. Breast Cancer Res Treat. 2015;150(3):547-557. doi: 10.1007/ s10549-015-3352-0 [doi]. 25. Kennecke H, Yerushalmi R, Woods R, et al. Metastatic behavior of breast cancer subtypes. J Clin Oncol. 2010;28(20):3271-3277. doi: 10.1200/JCO.2009.25.9820 [doi]. 26. Badve S, Nakshatri H. Oestrogen-receptor-positive breast cancer: Towards bridging histopathological and molecular classifications. J Clin Pathol. 2009;62(1):6-12. doi: 10.1136/ jcp.2008.059899 [doi]. 27. Blows FM, Driver KE, Schmidt MK, et al. Subtyping of breast cancer by immunohistochemistry to investigate a relationship between subtype and short and long term

189

PART TWO | CHAPTER 6

survival: A collaborative analysis of data for 10,159 cases from 12 studies. PLoS Med. 2010;7(5):e1000279. doi: 10.1371/journal.pmed.1000279 [doi]. 28. Pusztai L, Viale G, Kelly CM, Hudis CA. Estrogen and HER-2 receptor discordance between primary breast cancer and metastasis. Oncologist. 2010;15(11):1164-1168. doi: 10.1634/theoncologist.2010-0059 [doi]. 29. Rhodes A, Jasani B, Barnes DM, Bobrow LG, Miller KD. Reliability of immunohistochemical demonstration of oestrogen receptors in routine practice: Interlaboratory variance in the sensitivity of detection and evaluation of scoring systems. J Clin Pathol. 2000;53(2):125-130. 30. Perez EA, Suman VJ, Davidson NE, et al. HER2 testing by local, central, and reference laboratories in specimens from the north central cancer treatment group N9831 intergroup adjuvant trial. J Clin Oncol. 2006;24(19):30323038. doi: 24/19/3032 [pii]. 31. Curtit E, Nerich V, Mansi L, et al. Discordances in estrogen receptor status, progesterone receptor status, and HER2 status between primary breast cancer and metastasis. Oncologist. 2013;18(6):667-674. doi: 10.1634/ theoncologist.2012-0350. 32. Duchnowska R, Dziadziuszko R, Trojanowski T, et al. Conversion of epidermal growth factor receptor 2 and hormone receptor expression in breast cancer metastases to the brain. Breast Cancer Research. 2012;14(4):R119. doi: 10.1186/bcr3244. 33. Idirisinghe PKA, Thike AA, Cheok PY, et al. Hormone receptor and C-ERBB2 status in distant metastatic and locally recurrent breast cancer pathologic correlations and clinical significance. Am J Clin Pathol. 2010;133(3):416-429. 34. Jensen JD, Knoop A, Ewertz M, Laenkholm A. ER, HER2, and TOP2A expression in primary tumor, synchronous axillary nodes, and asynchronous metastases in breast cancer. Breast Cancer Res Treat. 2012;132(2):511-521. doi: 10.1007/s10549-011-1610-3. 35. Nakamura R, Yamamoto N, Onai Y, Watanabe Y, Kawana H, Miyazaki M. Importance of confirming HER2 overexpression of recurrence lesion in breast cancer patients. Breast Cancer. 2013;20(4):336-341. doi: 10.1007/ s12282-012-0341-6. 36. Yonemori K, Tsuta K, Shimizu C, et al. Immunohistochemical profiles of brain metastases from breast cancer. J Neurooncol. 2008;90(2):223-228. doi: 10.1007/s11060-008-9654-x. 37. Bogina G, Bortesi L, Marconi M, et al. Comparison of hormonal receptor and HER-2 status between breast primary tumours and relapsing tumours: Clinical implications of progesterone receptor loss. Virchows Archiv. 2011;459(1):1-10. doi: 10.1007/s00428-011-1097-7. 38. Thompson AM, Jordan LB, Quinlan P, et al. Prospective comparison of switches in biomarker status between primary and recurrent breast cancer: The breast recurrence in tissues study (BRITS). Breast Cancer Research. 2010;12(6):R92. doi: 10.1186/bcr2771. 39. Fabi A, Di Benedetto A, Metro G, et al. HER2 protein and gene variation between primary and metastatic breast cancer: Significance and impact on patient care. Clinical Cancer Research. 2011;17(7):2055-2064. doi: 10.1158/1078-0432.CCR-10-1920.

190

40. Cui X, Schiff R, Arpino G, Osborne CK, Lee AV. Biology of progesterone receptor loss in breast cancer and its implications for endocrine therapy. J Clin Oncol. 2005;23(30):7721-7735. doi: 23/30/7721 [pii]. 41. Bardou VJ, Arpino G, Elledge RM, Osborne CK, Clark GM. Progesterone receptor status significantly improves outcome prediction over estrogen receptor status alone for adjuvant endocrine therapy in two large breast cancer databases. J Clin Oncol. 2003;21(10):1973-1979. doi: 10.1200/JCO.2003.09.099 [doi]. 42. Bartlett JM, Brookes CL, Robson T, et al. Estrogen receptor and progesterone receptor as predictive biomarkers of response to endocrine therapy: A prospectively powered pathology study in the tamoxifen and exemestane adjuvant multinational trial. J Clin Oncol. 2011;29(12):1531-1538. doi: 10.1200/JCO.2010.30.3677 [doi]. 43. Brankovic-Magic M, Jankovic R, Neskovic-Konstantinovic Z, Nikolic-Vukosavljevic D. Progesterone receptor status of breast cancer metastases. J Cancer Res Clin Oncol. 2002;128(1):55-60. doi: 10.1007/s00432-001-0299-9 [doi]. 44. Chang HJ, Han S, Oh D, et al. Discordant human epidermal growth factor receptor 2 and hormone receptor status in primary and metastatic breast cancer and response to trastuzumab. Jpn J Clin Oncol. 2011;41(5):593599. doi: 10.1093/jjco/hyr020. 45. Edgerton SM, Moore II D, Merkel D, Thor AD. erbB-2 (HER-2) and breast cancer progression. Appl Immunohistochem Mol Morphol. 2003;11(3):214-221. 46. Lower EE, Glass E, Blau R, Harman S. HER-2/neu expression in primary and metastatic breast cancer. Breast Cancer Res Treat. 2009;113(2):301-306. doi: 10.1007/ s10549-008-9931-6. 47. Karagoz Ozen DS, Ozturk MA, Aydin O, Turna ZH, Ilvan S, Ozguroglu M. Receptor expression discrepancy between primary and metastatic breast cancer lesions. Oncology research and treatment. 2014;37(11):622-6. doi: 10.1159/000368312. 48. Shen Q, Sahin AA, Hess KR, et al. Breast cancer with brain metastases: Clinicopathologic features, survival, and paired biomarker analysis. Oncologist. 2015;20(5):466473. doi: 10.1634/theoncologist.2014-0107. 49. Amir E, Miller N, Geddie W, et al. Prospective study evaluating the impact of tissue confirmation of metastatic disease in patients with breast cancer. J Clin Oncol. 2012;30(6):587-592. 50. Aurilio G, Monfardini L, Rizzo S, et al. Discordant hormone receptor and human epidermal growth factor receptor 2 status in bone metastases compared to primary breast cancer. Acta Oncol. 2013;52(8):1649-1656. doi: 10.3109/0284186X.2012.754990. 51. Curigliano G, Bagnardi V, Viale G, et al. Should liver metastases of breast cancer be biopsied to improve treatment choice? Ann Oncol. 2011;22(10):2227-2233. doi: 10.1093/annonc/mdq751 [doi]. 52. Simmons C, Miller N, Geddie W, et al. Does confirmatory tumor biopsy alter the management of breast cancer patients with distant metastases? Ann Oncol. 2009;20(9):1499-1504. doi: 10.1093/annonc/mdp028 [doi].

Receptor conversion in breast cancer metastases

53. Amir E, Clemons M, Purdie CA, et al. Tissue confirmation of disease recurrence in breast cancer patients: Pooled analysis of multi-centre, multi-disciplinary prospective studies. Cancer Treat Rev. 2012;38(6):708-714. 54. Perez EA, Suman VJ, Davidson NE, et al. HER2 testing by local, central, and reference laboratories in specimens from the north central cancer treatment group N9831 intergroup adjuvant trial. J Clin Oncol. 2006;24(19):30323038. doi: 24/19/3032 [pii]. 55. Rao R, Lilley L, Andrews V, Radford L, Ulissey M. Axillary staging by percutaneous biopsy: Sensitivity of fine-needle aspiration versus core needle biopsy. Ann Surg Oncol. 2009;16(5):1170-1175. doi: 10.1245/s10434-009-0421-9 [doi]. 56. Ganott MA, Zuley ML, Abrams GS, et al. Ultrasound guided core biopsy versus fine needle aspiration for evaluation of axillary lymphadenopathy in patients with breast cancer. ISRN Oncol. 2014;2014:703160. doi: 10.1155/2014/703160 [doi]. 57. Meunier M, Clough K. Fine needle aspiration cytology versus percutaneous biopsy of nonpalpable breast lesions. Eur J Radiol. 2002;42(1):10-16. doi: S0720048X01004806 [pii]. 58. Pusztai L, Viale G, Kelly CM, Hudis CA. Estrogen and HER-2 receptor discordance between primary breast cancer and metastasis. Oncologist. 2010;15(11):1164-1168. doi: 10.1634/theoncologist.2010-0059. 59. Gruchy JR, Barnes PJ, Dakin Hache KA. CytoLyt(R) fixation and decalcification pretreatments alter antigenicity in normal tissues compared with standard formalin fixation. Appl Immunohistochem Mol Morphol. 2015;23(4):297-302. doi: 10.1097/PAI.0000000000000082 [doi]. 60. Bussolati G, Leonardo E. Technical pitfalls potentially affecting diagnoses in immunohistochemistry. J Clin Pathol. 2008;61(11):1184-1192. doi: 10.1136/ jcp.2007.047720 [doi]. 61. Hilton JF, Amir E, Hopkins S, et al. Acquisition of metastatic tissue from patients with bone metastases from breast cancer. Breast Cancer Res Treat. 2011;129(3):761765. doi: 10.1007/s10549-010-1264-6 [doi]. 62. Kulka J, Székely B, Lukács LV, et al. Comparison of predictive immunohistochemical marker expression of primary breast cancer and paired distant metastasis using surgical material: A practice-based study. J Histochem Cytochem. 2016;64(4):256-267. 63. Regitnig P, Schippinger W, Lindbauer M, Samonigg H, Lax SF. Change of HER-2/neu status in a subset of distant metastases from breast carcinomas. J Pathol. 2004;203(4):918-926. doi: 10.1002/path.1592 [doi]. 64. Schrijver WA, van der Groep P, Hoefnagel LD, et al. Influence of decalcification procedures on immunohistochemistry and molecular pathology in breast cancer. Mod Pathol. 2016. doi: 10.1038/ modpathol.2016.116 [doi]. 65. Yeung C, Hilton J, Clemons M, et al. Estrogen, progesterone, and HER2/neu receptor discordance between primary and metastatic breast tumours-a review. Cancer Metastasis Rev. 2016;35(3):427-437. doi: 10.1007/ s10555-016-9631-3 [doi].

66. Brogi E, Murphy CG, Johnson ML, et al. Breast carcinoma with brain metastases: Clinical analysis and immunoprofile on tissue microarrays. Ann Oncol. 2011;22(12):2597-2603. doi: 10.1093/annonc/mdr022 [doi]. 67. Chan A, Morey A, Brown B, Hastrich D, Willsher P, Ingram D. A retrospective study investigating the rate of HER2 discordance between primary breast carcinoma and locoregional or metastatic disease. BMC Cancer. 2012;12:555. doi: 10.1186/1471-2407-12-555. 68. Cummings MC, Simpson PT, Reid LE, et al. Metastatic progression of breast cancer: Insights from 50 years of autopsies. J Pathol. 2014;232(1):23-31. doi: 10.1002/ path.4288. 69. Fuchs IB, Siemer I, Bühler H, et al. Epidermal growth factor receptor changes during breast cancer metastasis. Anticancer Res. 2006;26(6 B):4397-4401. 70. Gaedcke J, Traub F, Milde S, et al. Predominance of the basal type and HER-2/neu type in brain metastasis from breast cancer. Mod Pathol. 2007;20(8):864-870. doi: 3800830 [pii]. 71. Gonzalez-Angulo AM, Ferrer-Lozano J, Stemke-Hale K, et al. PI3K pathway mutations and PTEN levels in primary and metastatic breast cancer. Molecular Cancer Therapeutics. 2011;10(6):1093-1101. doi: 10.1158/15357163.MCT-10-1089. 72. Santinelli A, Pisa E, Stramazzotti D, Fabris G. HER-2 status discrepancy between primary breast cancer and metastatic sites. impact on target therapy. International Journal of Cancer. 2008;122(5):999-1004. doi: 10.1002/ijc.23051. 73. Thomson AH, Mcgrane J, Mathew J, et al. Changing molecular profile of brain metastases compared with matched breast primary cancers and impact on clinical outcomes. Br J Cancer. 2016;114(7):793-800. 74. Zidan J, Dashkovsky I, Stayerman C, Basher W, Cozacov C, Hadary A. Comparison of HER-2 overexpression in primary breast cancer and metastatic sites and its effect on biological targeting therapy of metastatic disease. Br J Cancer. 2005;93(5):552-556. doi: 6602738 [pii]. 75. Bachmann C, Grischke EM, Staebler A, Schittenhelm J, Wallwiener D. Receptor change-clinicopathologic analysis of matched pairs of primary and cerebral metastatic breast cancer. J Cancer Res Clin Oncol. 2013;139(11):1909-1916. 76. Botteri E, Disalvatore D, Curigliano G, et al. Biopsy of liver metastasis for women with breast cancer: Impact on survival. Breast. 2012;21(3):284-288. doi: 10.1016/j. breast.2011.12.014. 77. Cabioglu N, Sahin AA, Morandi P, et al. Chemokine receptors in advanced breast cancer: Differential expression in metastatic disease sites with diagnostic and therapeutic implications. Ann Oncol. 2009;20(6):10131019. 78. Gancberg D, Di Leo A, Cardoso F, et al. Comparison of HER-2 status between primary breast cancer and corresponding distant metastatic sites. Ann Oncol. 2002;13(7):1036-1043. 79. Lorincz T, Toth J, Badalian G, Timar J, Szendroi M. HER-2/neu genotype of breast cancer may change in bone metastasis. Pathology & Oncology Research. 2006;12(3):149-152.

191

PART TWO | CHAPTER 6

80. Omoto Y, Kurosumi M, Hozumi Y, et al. Immunohistochemical assessment of primary breast tumors and metachronous brain metastases, with particular regard to differences in the expression of biological markers and prognosis. Exp Ther Med. 2010;1(4):561-567. doi: 10.3892/etm_00000088 [doi]. 81. St Romain P, Madan R, Tawfik OW, Damjanov I, Fan F. Organotropism and prognostic marker discordance in distant metastases of breast carcinoma: Fact or fiction? A clinicopathologic analysis. Hum Pathol. 2012;43(3):398404. doi: 10.1016/j.humpath.2011.05.009. 82. Vincent-Salomon A, Pierga J-, Couturier J, et al. HER2 status of bone marrow micrometastasis and their corresponding primary tumours in a pilot study of 27 cases: A possible tool for anti-HER2 therapy management? Br J Cancer. 2007;96(4):654-659. 83. Shao J, Pan C-, Wang M-, Wu X-, Ma B. Discordance of aldehyde dehydrogenase 1 and estrogen receptor expression between primary and metastatic focuses of breast cancer. Acta Anat Sin. 2013;44(2):224-228. 84. Wu JM, Fackler MJ, Halushka MK, et al. Heterogeneity of breast cancer metastases: Comparison of therapeutic target expression and promoter methylation between primary tumors and their multifocal metastases. Clin Cancer Res. 2008;14(7):1938-1946. doi: 10.1158/10780432.CCR-07-4082 [doi]. 85. Lower EE, Glass EL, Bradley DA, Blau R, Heffelfinger S. Impact of metastatic estrogen receptor and progesterone receptor status on survival. Breast Cancer Res Treat. 2005;90(1):65-70. doi: 10.1007/s10549-004-2756-z [doi]. 86. Amir E, Ooi WS, Simmons C, et al. Discordance between receptor status in primary and metastatic breast cancer: An exploratory study of bone and bone marrow biopsies. Clin Oncol. 2008;20(10):763-768. doi: 10.1016/j. clon.2008.08.005.

192

87. Gullo G, Bettio D, Zuradelli M, et al. Level of HER2/neu amplification in primary tumours and metastases in HER2-positive breast cancer and survival after trastuzumab therapy. Breast. 2013;22(2):190-193. doi: 10.1016/j.breast.2013.01.005 [doi]. 88. Kalinsky K, Mayer JA, Xu X, et al. Correlation of hormone receptor status between circulating tumor cells, primary tumor, and metastasis in breast cancer patients. Clin Transl Oncol. 2015;17(7):539-546. doi: 10.1007/s12094015-1275-1 [doi]. 89. Kamby C, Rasmussen BB, Kristensen B. Estrogen-receptor status of primary breast carcinomas and their metastases - relation to pattern of spread and survival after recurrence. Br J Cancer. 1989;60(2):252-257. doi: 10.1038/ bjc.1989.264. 90. Koo JS, Jung W, Jeong J. Metastatic breast cancer shows different immunohistochemical phenotype according to metastatic site. Tumori. 2010;96(3):424-432. 91. Lear-Kaul KC, Yoon HR, Kleinschmidt-DeMasters BK, McGavran L, Singh M. Her-2/neu status in breast cancer metastases to the central nervous system. Arch Pathol Lab Med. 2003;127(11):1451-1457. doi: OA3091 [pii]. 92. Nogami T, Shien T, Tanaka T, et al. The discordance between primary breast cancer lesions and pulmonary metastatic lesions in expression of aldehyde dehydrogenase 1-positive cancer cells. Breast Cancer. 2014;21(6):698-702. doi: 10.1007/s12282-013-0445-7. 93. Schwarz C, Lubbert H, Rahn W, Schonfeld N, Serke M, Loddenkemper R. Medical thoracoscopy: Hormone receptor content in pleural metastases due to breast cancer. Eur Respir J. 2004;24(5):728-730. doi: 24/5/728 [pii]. 94. Welter S, Jacobs J, Krbek T, Totsch M, Stamatis G. Pulmonary metastases of breast cancer. when is resection indicated? Eur J Cardiothorac Surg. 2008;34(6):1228-1234. doi: 10.1016/j.ejcts.2008.07.063 [doi].

Receptor conversion in breast cancer metastases

SUPPLEMENTAL

6 Supplementary Figure S1. Study-specific and pooled estimates for ERÎą discordance proportions for studies reporting ERÎą IHC in primary breast tumors and paired distant metastases. Study-specific data are ordered by date of publication. Analyses are subdivided for studies using a 1% (upper) and studies using a 10% threshold for positivity (lower). Discordance percentages are shown for total conversion (a), conversion from positive to negative (b) and conversion from negative to positive (c). uuu (see next page for Supplementary Figure S1C)

193

PART TWO | CHAPTER 6

ttt Supplementary Figure S1. Continued

194

Receptor conversion in breast cancer metastases

Supplementary Figure S2. Study-specific and pooled estimates for PR discordance proportions for studies reporting PR IHC in primary breast tumors and paired distant metastases. Study-specific data are ordered by date of publication. Analyses are subdivided for studies using a 1% (upper) and studies using a 10% threshold for positivity (lower). Discordance percentages are shown for total conversion (a), conversion from positive to negative (b) and conversion from negative to positive (c).

195

PART TWO | CHAPTER 6

Supplementary Figure S3. Study-specific and pooled estimates for HER2 discordance proportions for studies reporting HER2 IHC, FISH and/or a combination of both in primary breast tumors and paired distant metastases. Study-specific data are ordered by date of publication. Analyses are subdivided for studies using a combination of IHC and FISH (upper), studies using IHC only (middle) and studies using FISH only (lower). Discordance percentages are shown for total conversion (a), conversion from positive to negative (b) and conversion from negative to positive (c). uuu (see next page for Supplementary Figure S3C)

196

Receptor conversion in breast cancer metastases

6 ttt Supplementary Figure S3. Continued

197

198 R

Italy

USA

Botteri et al., 2012 5

Italy

France

Poland

USA

Curigliano et al., 2011 11

Curtit et al., 2013 12

Duchnowska et al., 2012 13

USA

Canada

the Netherlands

Gonzalez-Angulo et al., 2011 19

Hilton et al., 2010 20

Hoefnagel et al., 2013 21

Singapore

Denmark

Idirisinghe et al., 2010 23

Jensen et al., 2012 24

the Netherlands

Belgium

Gancberg et al., 2002 18

Germany

Gaedcke et al., 2007 17

Hoefnagel et al., 2012

Germany

Fuchs et al., 2006 16

Italy

Fabi et al., 2011 15

Edgerton et al., 2003

Australia

Cummings et al., 2014 10

South Korea

Chang et al., 2011 9

Australia

Chan et al., 2012 8

USA

Cabioglu et al., 2009 7

Brogi et al., 2011

Italy

Bogina et al., 2011 4

Germany

Bachmann et al., 2013 3

Italy

Aurilio et al., 2013 2

Design

UK/Canada

Country

Amir et al., 2012 1 *

Study

A. Study characteristics

233

100

120

235

255

100

107

120

Number of eligible subjects

yes

ISH

≥1%

≥10%

≥1% and ≥10%

≥10%

Allred score ≥3 or ≥1%

RIS

DAKO: 2+ for CISH

DAKO

DAKO: 2+ for SISH

DAKO: 2+ for FISH

DAKO: 3+ positive

DAKO: 2+ for FISH

DAKO: ≥1+ for FISH

≥10%

DAKO: 2+ for FISH

DAKO: 3+ is positive

DAKO: 2+ for FISH

DAKO: 2+ for SISH

DAKO: 2+ for FISH

DAKO: 2+/3+ for FISH

HER2 IHC scoring

Allred score ≥3 or ≥1%

≥10%

≥1%

≥10%

≥1%

≥10%

≥1%

Allred score ≥3 or ≥1%

ER/PR scoring

Study characteristics

Supplementary Table S1. The main characteristics of the thirty-nine studies included in this meta-analysis.

CISH: ratio ≥2.2

SISH: ≥6 copies

FISH: ratio ≥2.2

FISH: ratio ≥2.0

FISH: ≥ 4 copies

FISH: ratio ≥2.2, SISH: ≥6 copies

FISH: ratio ≥2.0

FISH: ratio ≥2.2. SISH: ≥6 copies

FISH: ratio ≥2.2

SISH: ratio ≥2.2

FISH: ratio ≥ 2.2

FISH: ratio ≥2.2

HER2 ISH scoring

PART TWO | CHAPTER 6

Austria

Italy

USA

Canada

USA

Regitnig et al., 2004 31

Santinelli et al., 2008 32

Shen et al., 2015 33

Simmons et al., 2009 34

France

Japan

Israel

Thomson et al., 2016 36

Vincent-Salomon et al., 2002 37

Yonemori et al., 2008 38

Zidan et al., 2005 39

St Romain et al., 2012

Japan

Omoto et al., 2010 30

Japan

Nakamura et al., 2013 29

USA

Hungary

Lower et al., 2009 28

Lorincz et al., 2006

Hungary

Kulka et al., 2016 26

Turkey

Karagoz Ozen et al., 2014 25

382

yes

≥10%

Allred score ≥3

≥1%

≥10%

≥1%

≥10%

Allred score ≥3

≥1%

FISH: ratio ≥2.2 FISH: ratio ≥4.0 NR FISH: ratio ≥2.2 FISH: ratio ≥2.0 FISH: ratio ≥2.2 NR NR FISH: ratio ≥2.2 NR NR NR

DAKO: 2+ for FISH DAKO: 2+/3+ for FISH DAKO DAKO: 2+ for FISH DAKO: 2+/3+ positive DAKO: 2+ for FISH DAKO: 2+ for FISH DAKO: 2+ for FISH DAKO: 2+ for FISH DAKO: 2+ for FISH DAKO: 2+/3+ positive DAKO: 2+ for FISH DAKO: 2+/3+ for FISH

Receptor conversion in breast cancer metastases

199

200

28-87; 61

40-52; 47

29-67; 49

34-93; 62

26-83; 45

22-79; 44

31-85; 50

32-73; 48

43-61,2; 52

26-75; 45

29-87; 57

26-80; 49

NR; 50

26-92; 56

30-83; 48

34-76; 55

25-93; 54

30-77; 49

Aurilio et al., 2013 2

Bachmann et al., 2013 3

Bogina et al., 2011 4

Botteri et al., 2012 5

Brogi et al., 2011 6

Cabioglu et al., 2009 7

Chan et al., 2012 8

Chang et al., 2011 9

Cummings et al., 2014 10

Curigliano et al., 2011 11

Curtit et al., 2013 12

Duchnowska et al., 2012 13

Edgerton et al., 2003 14

Fabi et al., 2011 15

Fuchs et al., 2006 16

Gaedcke et al., 2007 17

Gancberg et al., 2002 18

Gonzalez-Angulo et al., 2011 19

Hilton et al., 2010 20

Hoefnagel et al., 2013 21

Hoefnagel et al., 2012 22 ~

Idirisinghe et al., 2010 23

Jensen et al., 2012 24

Karagoz Ozen et al., 2014 25

Age at diagnosis (range; median)

ductal: 46, lobular: 4, other: 8

ductal: 69, lobular: 1, other: 2

ductal: 192, lobular: 20, other: 20

ductal: 40, lobular: 8, other: 3

ductal: 42, lobular: 6

ductal: 88, lobular: 5

ductal: 98, lobular: 10, ductolobular: 2, other: 5

ductal: 225, lobular: 14, other: 7

ductal: 28, lobular: 11, ductolobular: 8, other: 8

ductal: 89, lobular: 7, other: 4

ductal: 21, lobular: 2, other: 1

Type

Primary tumor characteristics

Amir et al., 2012 1 *

Study

B. Primary tumor characteristics

I: 1, II: 30, III: 40

I: 8, II: 61, III: 161

I: 4, II: 20, III: 24

I: 18, II: 27, III: 43

I: 6, II: 41, III: 56

I: 1, II: 21, III: 31

I: 1, II: 24, III: 27

I/II: 16, III: 8

Grade

Supplementary Table S1. The main characteristics of the thirty-nine studies included in this meta-analysis.

T1: 6, T2: 20, T3: 19, T4: 11

T1: 14, T2: 20, T3: 5, T4: 9

T1: 20, T2: 8, T3: 1, T4: 1

T1: 112, T2: 102, T3/4: 26

T1: 10, T2: 41, T3: 4, T4: 1

T1: 45, T2: 45, T3/4: 5

T1: 17, T2: 6, T3: 1

T1: 51, T2: 11, T3: 12

Stage

+: 110, -: 81

+: 35, -: 13

+: 16, -: 14

+: 49, -: 37

+: 147, -: 99

+: 92, -: 5

+:11, -: 13

Nodal status

PART TWO | CHAPTER 6

NR; 59

24-83; NR

33-69; 47

33-78; 54

31-76; 50

24-73; 46

NR; 51

31-74; 49

29-82; 56

Kulka et al., 2016 26

Lorincz et al., 2006 27

Lower et al., 2009 28

Nakamura et al., 2013 29

Omoto et al., 2010 30

Regitnig et al., 2004 31

Santinelli et al., 2008 32

Shen et al., 2015 33

Simmons et al., 2009 34

St Romain et al., 2012 35

Thomson et al., 2016 36

Vincent-Salomon et al., 2002 37

Yonemori et al., 2008 38

Zidan et al., 2005 39

ductal: 40, lobular: 3, other: 1

ductal: 33, lobular: 2

ductal: 27, ductolobular: 3, other: 1

NR NR NR

T1: 74, T2: 170, T3: 45, T4: 35

NR # NR

T1: 11, T2: 9, T3: 4, T4: 7 T1: 24, T2: 8, T3: 3 # #

I: 1, II: 14, III: 16 I: 3, II: 14, III: 18

NR NR NR NR

NR T2/T3: 41 NR NR

NR I: 11, II: 25, III: 7 NR

+: 11, -: 20

I: 1, II: 10, III: 12

Receptor conversion in breast cancer metastases

201

202 3 S, 53 M

2-191; 36 0-432; 37

26-135; 36

0-69; 24

Botteri et al., 2012 5

Brogi et al., 2011 6

Cabioglu et al., 2009 7

Chan et al., 2012 8

NR; 36

Edgerton et al., 2003 14

Fabi et al., 2011 15

Fuchs et al., 2006 16

1-216; NR

5-210; 90

NR -18-182; NR

1-175; 46

Gancberg et al., 2002 18

Gonzalez-Angulo et al., 2011 19

Hilton et al., 2010 20

Hoefnagel et al., 2013 21

Hoefnagel et al., 2012 22 ~

Idirisinghe et al., 2010 23

Gaedcke et al., 2007

6-176; 97

Duchnowska et al., 2012 13

0-277; 35

Curtit et al., 2013 12

3 S, 117 M

18 S, 217 M

16 S, 239 M

0-220; 41

Curigliano et al., 2011 11

4 S, 46 M

Cummings et al., 2014 10

Chang et al., 2011

6-216;Â 74

Bogina et al., 2011 4

7-73; 36

Bachmann et al., 2013 3

0-227; 50

0-332; 86

Timing

Aurilio et al., 2013 2

Time between primary and metastasis in months (range; median)

Comparison primary to metastasis

Amir et al., 2012 1 *

Study

C. Comparison between primary tumors and paired metastases

Supplementary Table S1. The main characteristics of the thirty-nine studies included in this meta-analysis.

yes

Same tests, same time

yes, 1 observer

no, 2 observers

yes, 3 observers

NR, 2 observers

yes, 2 observers

yes, 2/3 observers

no, 2 observers

yes, 1 observer

NR, 2 observers

yes, 1 observer

no, 1 observer

Interpretation masked

PART TWO | CHAPTER 6

NR -16-103; 35

12-228; 67

0-266; 46

14-78; 29

Omoto et al., 2010 30

Regitnig et al., 2004 31

Santinelli et al., 2008 32

Shen et al., 2015 33

0-264; NR

0-244; 26

12-228; 78

0-277; 35

12-144; 55

St Romain et al., 2012 35

Thomson et al., 2016 36

Vincent-Salomon et al., 2002 37

Yonemori et al., 2008 38

Zidan et al., 2005 39

Simmons et al., 2009

31-58; NR

Nakamura et al., 2013 29

Lower et al., 2009 28

1 S, 40 M

35 M

Lorincz et al., 2006 27

Kulka et al., 2016

15-197; 71

Karagoz Ozen et al., 2014 25

8-323; 59

Jensen et al., 2012 24

no, 1 observer NR yes, 2 observers no, 2 observers no, multiple observers yes yes, 2 observers NR yes, 2 observers NR yes no, 2 observers NR, 1 observer NR yes, 2 observers NR, 2 observers

yes yes yes NR no no NR yes no no no no NR NR no yes

Receptor conversion in breast cancer metastases

203

PART TWO | CHAPTER 6

Supplementary Table S1. The main characteristics of the thirty-nine studies included in this meta-analysis. D. Metastasis characteristics Study

Metastasis characteristics Site of metastasis

HER2 IHC

Amir et al., 2012 1 *

bone, liver, lung, skin

yes

Aurilio et al., 2013 2

bone

yes

Bachmann et al., 2013 3

CNS

yes

Bogina et al., 2011 4

adrenal gland, bone, CNS, GI, gynaecological, liver, lung, skin

yes

Botteri et al., 2012 5

liver

yes

Brogi et al., 2011 6

CNS

yes

Cabioglu et al., 2009 7

bone, CNS, GI, lung

yes

Chan et al., 2012 8

bone, CNS, liver, lung, skin

Chang et al., 2011 9

bone, liver, lung

yes

Cummings et al., 2014 10

adrenal gland, bone, CNS, GI, gynaecological, kidney, liver, lung

yes

Curigliano et al., 2011 11

liver

yes

Curtit et al., 2013 12

CNS, bone, liver, lung, soft tissue

yes

Duchnowska et al., 2012 13

CNS

yes

Edgerton et al., 2003 14

yes

Fabi et al., 2011 15

bone, CNS, liver, lung, gynaecological

yes

Fuchs et al., 2006 16

bone, liver, lung, skin

yes

Gaedcke et al., 2007 17

CNS

yes

Gancberg et al., 2002 18

bone, CNS, GI, gynaecological, liver, lung, soft tissue

yes

Gonzalez-Angulo et al., 2011 19

bone, CNS, liver, lung, GI, gynaecological, soft tissue

yes

Hilton et al., 2010 20

bone

yes

Hoefnagel et al., 2013 21

bone, CNS, GI, gynaecological, liver, lung, skin

yes

Hoefnagel et al., 2012 22 ~

CNS, GI, lung, liver, skin

yes

Idirisinghe et al., 2010 23

adrenal gland, bone, CNS, GI, gynaecological, liver, lung, skin

yes

Jensen et al., 2012 24

bone, CNS, GI, liver, lung, skin

yes

Karagoz Ozen et al., 2014 25

bone, CNS, gynaecological, liver, lung, skin

yes

Kulka et al., 2016 26

bone, CNS, lung

yes

Lorincz et al., 2006 27

bone

yes

Lower et al., 2009 28

yes

Nakamura et al., 2013 29

CNS: 27, lung: 16, liver: 20, bone: 16, GI/gynaecological: 10

yes

Omoto et al., 2010 30

CNS

yes

Regitnig et al., 2004 31

bone, CNS, liver, lung, skin

yes

Santinelli et al., 2008 32

bone, CNS, GI, skin, liver, lung

yes

Shen et al., 2015 33

CNS

yes

Simmons et al., 2009 34

bone, CNS, liver, lung, soft tissue

yes

St Romain et al., 2012 35

bone, CNS, GI, gynaecological, liver, lung, skin

yes

Thomson et al., 2016 36

CNS

yes

Vincent-Salomon et al., 2002 37

liver, lung

yes

Yonemori et al., 2008 38

CNS

yes

Zidan et al., 2005 39

bone, liver, lung, skin

yes

R: retrospective, P: prospective, #: reported for other group than eligible subjects, S: synchrone, M: metachrone, RIS: Remmeleâ&#x20AC;&#x2122;s Immunoreactive Score, NR: not reported. * combination of BRITS study 40. and DESTINY study 41 . ~ same cohort as Hoefnagel et al., 2010 42. numbers do not always add up to 100% due to unknown cases

204

Receptor conversion in breast cancer metastases

APPENDIX 1 Search strategy (((((((((breast cancer[MeSH Terms]) OR ((((((breast*[Title/Abstract] OR mamma[Title/ Abstract] OR mammae[Title/Abstract] OR mammary[Title/Abstract]))) OR breast[MeSH Terms])) AND ((((cancer*[Title/Abstract] OR tumor*[Title/Abstract] OR tumour*[Title/ Abstract] OR malign*[Title/Abstract] OR carcinom*[Title/Abstract] OR neoplas*[Title/ Abstract] OR oncolog*[Title/Abstract] OR adenocarcinoma*[Title/Abstract]))) OR adenocarcinoma[MeSH Terms])))) AND ((neoplasm metastasis[MeSH Terms]) OR ((metasta*[Title/Abstract] OR progressi*[Title/Abstract] OR disseminat*[Title/Abstract] OR relaps*[Title/Abstract] OR recurr*[Title/Abstract]))))) AND ((((receptors estrogen[MeSH Terms]) OR receptors progesterone[MeSH Terms]) OR receptor, erbb 2[MeSH Terms]) OR ((((estrogen*[Title/Abstract] OR oestrogen*[Title/Abstract] OR ERÎą[Title/Abstract] OR ESR1[Title/Abstract] OR estrogen receptor*[Title/Abstract] OR oestrogen receptors*[Title/Abstract])) OR (progesteron*[Title/Abstract] OR progestin[Title/ Abstract] OR PgR[Title/Abstract] OR progesteron receptor*[Title/Abstract] OR progesterone receptor*[Title/Abstract])) OR (her2[Title/Abstract] OR her 2[Title/Abstract] OR her2/neu[Title/Abstract] OR erbb2[Title/Abstract] OR human epidermal growth factor))))) AND (((discordan*[Title/Abstract] OR concordan*[Title/Abstract] OR chang*[Title/Abstract] OR change*[Title/Abstract] OR changing[Title/Abstract] OR conver*[Title/Abstract] OR congruen*[Title/Abstract] OR incongruen*[Title/Abstract] OR dyscongruen*[Title/Abstract] OR discongruen*[Title/Abstract] OR discrepan*[Title/ Abstract] OR diverg*[Title/Abstract])) OR (diversi* OR loss OR variati* OR difference* OR differing OR comparison OR different OR unstable OR stable OR alterat* OR altered OR stable OR stabilit* OR heterogen*)))) AND ((((immunohistochemistry[MeSH Terms]) OR in situ hybridization[MeSH Terms])) OR (((ihc[Title/Abstract] OR immunohistochemi*[Title/Abstract] OR staining[Title/Abstract] OR stained[Title/ Abstract] OR labeled[Title/Abstract] OR labeling[Title/Abstract])) OR (insitu[Title/ Abstract] OR in situ[Title/Abstract] OR hybridis*[Title/Abstract] OR hybridiz*[Title/ Abstract] OR ish[Title/Abstract])))

205

PART TWO | CHAPTER 6

SUPPLEMENTAL REFERENCES 1. Amir E, Clemons M, Purdie CA, et al. Tissue confirmation of disease recurrence in breast cancer patients: Pooled analysis of multi-centre, multi-disciplinary prospective studies. Cancer Treat Rev. 2012;38(6):708-714. 2. Aurilio G, Monfardini L, Rizzo S, et al. Discordant hormone receptor and human epidermal growth factor receptor 2 status in bone metastases compared to primary breast cancer. Acta Oncol. 2013;52(8):1649-1656. doi: 10.3109/0284186X.2012.754990. 3. Bachmann C, Grischke EM, Staebler A, Schittenhelm J, Wallwiener D. Receptor change-clinicopathologic analysis of matched pairs of primary and cerebral metastatic breast cancer. J Cancer Res Clin Oncol. 2013;139(11):1909-1916. 4. Bogina G, Bortesi L, Marconi M, et al. Comparison of hormonal receptor and HER-2 status between breast primary tumours and relapsing tumours: Clinical implications of progesterone receptor loss. Virchows Archiv. 2011;459(1):1-10. doi: 10.1007/s00428-011-10977. 5. Botteri E, Disalvatore D, Curigliano G, et al. Biopsy of liver metastasis for women with breast cancer: Impact on survival. Breast. 2012;21(3):284-288. doi: 10.1016/j. breast.2011.12.014. 6. Brogi E, Murphy CG, Johnson ML, et al. Breast carcinoma with brain metastases: Clinical analysis and immunoprofile on tissue microarrays. Ann Oncol. 2011;22(12):2597-2603. doi: 10.1093/annonc/mdr022 [doi]. 7. Cabioglu N, Sahin AA, Morandi P, et al. Chemokine receptors in advanced breast cancer: Differential expression in metastatic disease sites with diagnostic and therapeutic implications. Ann Oncol. 2009;20(6):10131019. 8. Chan A, Morey A, Brown B, Hastrich D, Willsher P, Ingram D. A retrospective study investigating the rate of HER2 discordance between primary breast carcinoma and locoregional or metastatic disease. BMC Cancer. 2012;12:555. doi: 10.1186/1471-2407-12-555. 9. Chang HJ, Han S, Oh D, et al. Discordant human epidermal growth factor receptor 2 and hormone receptor status in primary and metastatic breast cancer and response to trastuzumab. Jpn J Clin Oncol. 2011;41(5):593599. doi: 10.1093/jjco/hyr020. 10. Cummings MC, Simpson PT, Reid LE, et al. Metastatic progression of breast cancer: Insights from 50 years of autopsies. J Pathol. 2014;232(1):23-31. doi: 10.1002/ path.4288. 11. Curigliano G, Bagnardi V, Viale G, et al. Should liver metastases of breast cancer be biopsied to improve treatment choice? Ann Oncol. 2011;22(10):2227-2233. doi: 10.1093/annonc/mdq751 [doi]. 12. Curtit E, Nerich V, Mansi L, et al. Discordances in estrogen receptor status, progesterone receptor status, and HER2 status between primary breast cancer and metastasis. Oncologist. 2013;18(6):667-674. doi: 10.1634/ theoncologist.2012-0350. 13. Duchnowska R, Dziadziuszko R, Trojanowski T, et al. Conversion of epidermal growth factor receptor 2 and

206

hormone receptor expression in breast cancer metastases to the brain. Breast Cancer Research. 2012;14(4):R119. doi: 10.1186/bcr3244. 14. Edgerton SM, Moore II D, Merkel D, Thor AD. erbB-2 (HER-2) and breast cancer progression. Appl Immunohistochem Mol Morphol. 2003;11(3):214-221. 15. Fabi A, Di Benedetto A, Metro G, et al. HER2 protein and gene variation between primary and metastatic breast cancer: Significance and impact on patient care. Clinical Cancer Research. 2011;17(7):2055-2064. doi: 10.1158/1078-0432.CCR-10-1920. 16. Fuchs IB, Siemer I, Bühler H, et al. Epidermal growth factor receptor changes during breast cancer metastasis. Anticancer Res. 2006;26(6 B):4397-4401. 17. Gaedcke J, Traub F, Milde S, et al. Predominance of the basal type and HER-2/neu type in brain metastasis from breast cancer. Mod Pathol. 2007;20(8):864-870. doi: 3800830 [pii]. 18. Gancberg D, Di Leo A, Cardoso F, et al. Comparison of HER-2 status between primary breast cancer and corresponding distant metastatic sites. Ann Oncol. 2002;13(7):1036-1043. 19. Gonzalez-Angulo AM, Ferrer-Lozano J, Stemke-Hale K, et al. PI3K pathway mutations and PTEN levels in primary and metastatic breast cancer. Molecular Cancer Therapeutics. 2011;10(6):1093-1101. doi: 10.1158/15357163.MCT-10-1089. 20. Hilton JF, Amir E, Hopkins S, et al. Acquisition of metastatic tissue from patients with bone metastases from breast cancer. Breast Cancer Res Treat. 2011;129(3):761765. doi: 10.1007/s10549-010-1264-6 [doi]. 21. Hoefnagel LDC, van der Groep P, van de Vijver MJ, et al. Discordance in ER alpha, PR and HER2 receptor status across different distant breast cancer metastases within the same patient. Annals of Oncology. 2013;24(12):30173023. doi: 10.1093/annonc/mdt390. 22. Hoefnagel LDC, Moelans CB, Meijer SL, et al. Prognostic value of estrogen receptor alpha and progesterone receptor conversion in distant breast cancer metastases. Cancer. 2012;118(20):4929-4935. doi: 10.1002/cncr.27518. 23. Idirisinghe PKA, Thike AA, Cheok PY, et al. Hormone receptor and C-ERBB2 status in distant metastatic and locally recurrent breast cancer pathologic correlations and clinical significance. Am J Clin Pathol. 2010;133(3):416429. 24. Jensen JD, Knoop A, Ewertz M, Laenkholm A. ER, HER2, and TOP2A expression in primary tumor, synchronous axillary nodes, and asynchronous metastases in breast cancer. Breast Cancer Res Treat. 2012;132(2):511-521. doi: 10.1007/s10549-011-1610-3. 25. Karagoz Ozen DS, Ozturk MA, Aydin O, Turna ZH, Ilvan S, Ozguroglu M. Receptor expression discrepancy between primary and metastatic breast cancer lesions. Oncology research and treatment. 2014;37(11):622-6. doi: 10.1159/000368312. 26. Kulka J, Székely B, Lukács LV, et al. Comparison of predictive immunohistochemical marker expression of primary breast cancer and paired distant metastasis using surgical material: A practice-based study. J Histochem Cytochem. 2016;64(4):256-267.

Receptor conversion in breast cancer metastases

27. Lorincz T, Toth J, Badalian G, Timar J, Szendroi M. HER-2/neu genotype of breast cancer may change in bone metastasis. Pathology & Oncology Research. 2006;12(3):149-152. 28. Lower EE, Glass E, Blau R, Harman S. HER-2/neu expression in primary and metastatic breast cancer. Breast Cancer Res Treat. 2009;113(2):301-306. doi: 10.1007/ s10549-008-9931-6. 29. Nakamura R, Yamamoto N, Onai Y, Watanabe Y, Kawana H, Miyazaki M. Importance of confirming HER2 overexpression of recurrence lesion in breast cancer patients. Breast Cancer. 2013;20(4):336-341. doi: 10.1007/ s12282-012-0341-6. 30. Omoto Y, Kurosumi M, Hozumi Y, et al. Immunohistochemical assessment of primary breast tumors and metachronous brain metastases, with particular regard to differences in the expression of biological markers and prognosis. Exp Ther Med. 2010;1(4):561-567. doi: 10.3892/etm_00000088 [doi]. 31. Regitnig P, Schippinger W, Lindbauer M, Samonigg H, Lax SF. Change of HER-2/neu status in a subset of distant metastases from breast carcinomas. J Pathol. 2004;203(4):918-926. doi: 10.1002/path.1592 [doi]. 32. Santinelli A, Pisa E, Stramazzotti D, Fabris G. HER-2 status discrepancy between primary breast cancer and metastatic sites. impact on target therapy. International Journal of Cancer. 2008;122(5):999-1004. doi: 10.1002/ ijc.23051. 33. Shen Q, Sahin AA, Hess KR, et al. Breast cancer with brain metastases: Clinicopathologic features, survival, and paired biomarker analysis. Oncologist. 2015;20(5):466473. doi: 10.1634/theoncologist.2014-0107. 34. Simmons C, Miller N, Geddie W, et al. Does confirmatory tumor biopsy alter the management of breast cancer patients with distant metastases? Ann Oncol. 2009;20(9):1499-1504. doi: 10.1093/annonc/mdp028 [doi].

35. St Romain P, Madan R, Tawfik OW, Damjanov I, Fan F. Organotropism and prognostic marker discordance in distant metastases of breast carcinoma: Fact or fiction? A clinicopathologic analysis. Hum Pathol. 2012;43(3):398404. doi: 10.1016/j.humpath.2011.05.009. 36. Thomson AH, Mcgrane J, Mathew J, et al. Changing molecular profile of brain metastases compared with matched breast primary cancers and impact on clinical outcomes. Br J Cancer. 2016;114(7):793-800. 37. Vincent-Salomon A, Pierga J-, Couturier J, et al. HER2 status of bone marrow micrometastasis and their corresponding primary tumours in a pilot study of 27 cases: A possible tool for anti-HER2 therapy management? Br J Cancer. 2007;96(4):654-659. 38. Yonemori K, Tsuta K, Shimizu C, et al. Immunohistochemical profiles of brain metastases from breast cancer. J Neurooncol. 2008;90(2):223-228. doi: 10.1007/s11060-008-9654-x. 39. Zidan J, Dashkovsky I, Stayerman C, Basher W, Cozacov C, Hadary A. Comparison of HER-2 overexpression in primary breast cancer and metastatic sites and its effect on biological targeting therapy of metastatic disease. Br J Cancer. 2005;93(5):552-556. doi: 6602738 [pii]. 40. Thompson AM, Jordan LB, Quinlan P, et al. Prospective comparison of switches in biomarker status between primary and recurrent breast cancer: The breast recurrence in tissues study (BRITS). Breast Cancer Research. 2010;12(6):R92. doi: 10.1186/bcr2771. 41. Amir E, Miller N, Geddie W, et al. Prospective study evaluating the impact of tissue confirmation of metastatic disease in patients with breast cancer. J Clin Oncol. 2012;30(6):587-592. 42. Hoefnagel LDC, van de Vijver MJ, van Slooten H, et al. Receptor conversion in distant breast cancer metastases. Breast Cancer Research. 2010;12(5):R75. doi: 10.1186/ bcr2645.

207

Chapter 7 Willemijne AME Schrijver, Petra van der Groep*, Laurien DC Hoefnagel*, Natalie D ter Hoeve1, Ton Peeters, Cathy B Moelans, Paul J van Diest * Both authors contributed equally to this study

Influence of decalcification procedures on immunohistochemistry and molecular pathology in breast cancer Mod Pathol. 2016

PART TWO | CHAPTER 7

ABSTRACT Distant breast cancer metastases are nowadays routinely biopsied to re-assess receptor status and to isolate DNA for sequencing of druggable targets. Bone metastases are the most frequent subgroup. Decalcification procedures may negatively affect antigenicity and DNA quality. We therefore evaluated the effect of several decalcification procedures on receptor status and DNA/RNA quality. In 23 prospectively collected breast tumors we compared ERα, PR and HER2 status by immunohistochemistry in (non-decalcified) tissue routinely processed for diagnostic purposes and in parallel tissue decalcified in Christensen’s buffer with and without microwave, EDTA and Formical-4. Furthermore, HER2 fluorescence in situ hybridization and DNA/RNA quantity and quality were assessed. We found that the percentage of ERα positive cells were on average lower in EDTA (p=0.049) and Formical-4 (p=0.047) treated cases, compared to controls, and PR expression showed decreased antigenicity after Christensen’s buffer treatment (p=0.041). Overall, a good concordance (weighted kappa) was seen for ERα, PR and HER2 immunohistochemistry when comparing the non-decalcified control tissues with the decalcified tissues. For two patients (9%), there was a potential influence on therapeutic decision making with regard to hormonal therapy or HER2-targeted therapy. HER2 fluorescence in situ hybridization interpretation was seriously hampered by Christensen’s buffer and Formical-4 and DNA/ RNA quantity and quality were decreased after all four decalcification procedures. Validation on paired primary breast tumor specimens and EDTA treated bone metastases showed that immunohistochemistry and fluorescence in situ hybridization were well assessable, and DNA and RNA yield and quality were sufficient. With this, we conclude that common decalcification procedures have only a modest negative influence on hormone and HER2 receptor immunohistochemistry in breast cancer. However, they may seriously affect DNA/RNA based diagnostic procedures. Overall, EDTA based decalcification is therefore to be preferred since it best allows fluorescence in situ hybridization and DNA/RNA isolation.

210

Decalcification in breast cancer pathology

INTRODUCTION Multiple studies have shown that the expression of predictive tissue markers, such as estrogen receptor alpha (ERÎą), progesterone receptor (PR) and human epidermal growth factor receptor-2 (HER2) may differ between the primary tumor and distant metastases (â&#x20AC;&#x153;receptor conversionâ&#x20AC;?) in a significant proportion of breast cancer patients 1-3. Therefore, several guidelines nowadays advice to biopsy distant metastases to reassess hormone and HER2 receptor status by immunohistochemistry whenever possible 4,5. However, in bone metastases this could potentially lead to inappropriate systemic treatment, since antigenicity may be altered by decalcifying agents that enable sectioning of bone 6,7. On the other hand, some studies report that decalcifying methods can be applied without significant loss of immunoreactivity 8,9. The same contradictory results about influence of decalcifying buffers have been seen when RNA or DNA integrity and interpretation of in situ hybridization were taken into account 10-14. These inconsistencies may be explained by usage of different decalcifying agents. Strong acids such as hydrochloric and nitric acid are traditionally widely used for their rapid decalcifying properties, but they are known to have a detrimental influence on immunoreactivity and DNA integrity 15. Therefore, weak(er) acidic buffers, containing formic or trichloracetic acid, are now more popular. EDTA, a chelating agent with neutral pH, requires longer time periods for the complete removal of calcium salts, but produces the best morphological results 8. Singh et al. recently tested ten commercially available decalcification agents and concluded that the best preservation of nucleic acids is achieved with decalcifying agents that contain either EDTA or formic acid or a combination of both11. Since bone is a frequent metastatic site among breast cancer patients 16,17, we set out to evaluate the influence of three routinely used decalcifying agents (containing formic acid and/or EDTA) on assessment of hormone and HER2 receptor status and DNA/RNA quality in breast cancer.

MATERIALS AND METHODS Material Prospectively, tissue from 23 breast tumors was collected and processed according to routine procedures at the department of Pathology of the University Medical Center Utrecht, The Netherlands. Original diagnoses were made between August 2012 and July 2015. Clinicopathological characteristics are shown in Table 1 (test cohort). Four tumor biopsies (0.4-0.8 cm3 in size) were removed from each breast tumor (Figure 1). All tissue samples were fixed in 4% buffered formaldehyde for 24-48 hours. After fixation, breast cancer tissue collected for diagnostic purposes was processed according to a standard 211

PART TWO | CHAPTER 7

protocol and embedded in paraffin. The additional biopsies of the tumors were processed in Christensen’s buffer (containing formic acid and sodium formate), Ethylene Diamine Tetra acetic Acid disodium salt dehydrate (EDTA) or Formical-4 (containing formic acid, formaldehyde and methanol). Tissue was placed in Christensen’s buffer or Formical-4 overnight in a microwave (Christensen’s buffer with microwave) at 37°C and in EDTA overnight at room temperature on a shaker. To assess the influence of the microwave, one of the biopsies was treated with Christensen’s buffer but withheld from the microwave. Hereafter, all the tissue fragments were washed thoroughly in running tap water, placed in 70% EtOH and routinely processed to paraffin. For each formalin fixed paraffin embedded fragment, a 4-µm hematoxylin and eosin stained slide was reviewed by a single pathologist (PvD) to confirm the presence of malignancy. Because biopsies were taken macroscopically, tumor was not always present (indicated by empty spaces in Tables 2 and 4 and Supplementary Table S1). Tumor negative biopsies were excluded. The use of left-over material requires no ethical approval according to (“opt-out”) Dutch legislation. Use of anonymous or coded left over material for scientific purposes is part of the standard treatment contract with patients and therefore informed consent was not required 18. Table 1. Clinicopathological characteristics of the 23 included primary breast cancer samples (test cohort) and the 8 paired primary tumors and EDTA treated bone metastases (validation cohort) Test cohort N=23

Validation cohort N=8

Feature

Grouping

N or value

Age at diagnosis (in years)

Mean Range

58.7 40-89

53.8 43-63

Tumor size (in cm)

Mean Range

3.5 1.4-14

3.8 1.8-10

Histologic type

Invasive ductal Invasive lobular Ductolobular Other

14 3 3 3

61 13 13 13

4 0 3 1

50 0 38 12

Histologic grade (Bloom & Richardson)

I II III Not available

6 10 7 0

26 44 30

0 2 2 4

0 25 25 50

MAI (per 2mm²)

Mean Range

14.7 0-69

Lymph node status

Positive Negative Not available

10 9 4

44 39 17

5 2 1

63 25 12

Molecular subtype

Luminal A Luminal B Triple negative HER2-driven

20 1 2 0

87 4 9 0

6 2 0 0

75 25 0 0

212

N or value

4.0 0-8

Decalcification in breast cancer pathology

control routinely processed EDTA breast tumor resection specimen

IHC ERα, PR, HER2 per tissue block

ISH HER2 (when IHC 2+/ 3+) DNA & RNA isolation

4 biopsies EDTA Christensen’s buffer with microwave Christensen’s buffer Formical-4 with microwave

CBM

F4 FFPE

Figure 1. Overview of material and methods. Four tumor containing biopsies were taken from breast tumor resection specimens (n=23). Each biopsy was placed in a different decalcifying agent, including EDTA, Christensen’s buffer with microwave (CBM), Christensen’s buffer without microwave (CB) and Formical-4 (F4). The original non-decalcified tumor was used as a control. Specimens were formalin-fixed and paraffinembedded after which sections were cut for immunohistochemistry (IHC), fluorescence in situ hybridization (ISH) and DNA/RNA isolation.

Immunohistochemistry Immunohistochemistry for ERα, PR and HER2 was carried out on full 4-m sections with the Ventana autostainer (Roche, Tucson, USA) according to the manufacturer’s instructions. Rabbit monoclonal antibodies used were against ERα (RTU, SP1; Roche, Tucson, USA), PR (RTU, 1E2; Roche, Tucson, USA), and HER2 (1:50, SP3; ThermoFisher Scientific, Bleiswijk, the Netherlands). Appropriate controls were used throughout. Scoring Scoring of immunohistochemically stained slides was performed by mutual agreement of 2 observers (PvD&WS) in random order, blinded to other data in the paired samples. For ERα and PR, the percentage of positively stained nuclei was assessed. The adequacy of staining was checked by also evaluating the normal breast parenchyma. Samples with 10% or more immune-positive malignant cells were classified as ERα or PR positive. In order to also comply with the ASCO guidelines, we also used the 1% threshold that is now widely used in the USA. HER2 expression was scored using the DAKO scoring system as 0, 1+, 2+ and 3+. HER2 expression was considered positive when 3+. Immunohistochemically assessed surrogate molecular subtypes of breast tumors were assigned as follows: Luminal A (ER+/PR+, HER2−, Ki-67<15), luminal B (ER+/PR+, HER2−, Ki-67>15 or ER+/PR+, HER2+), triple negative or basal type (ER-/PR-, HER2-) and HER2 enriched (ER-/PR-, HER2+). Morphology was judged visually based on the hematoxylin and eosin stained slide, with special emphasis on visibility of mitoses, nuclear morphology, and staining intensity of stroma and epithelium. 213

PART TWO | CHAPTER 7

In situ hybridization According to routine diagnostic procedures, cases with HER2 expression of 2+ or 3+ (patients #4, #5, #6 and #16) were subjected to fluorescence in situ hybridization using a HER2/CEP17 dual FISH probe (Cytocell) on 4-µm slides. Analysis was performed on a Leica DM5500 B microscope system with Application Suite Advanced Fluorescence Software (Leica Microsystems, Rijswijk, The Netherlands). In short, formalin-fixed paraffin-embedded slides were deparaffinized and pretreated with citrate and protease buffers. Next, they were dehydrated and hybridized with 10-μL fluorescence in situ hybridization probe in a ThermoBrite (Abbott Laboratories, Chicago, IL, USA) at 37°C overnight. The next day, slides were washed in saline-sodium citrate buffers, counterstained with DAPI, dehydrated and mounted with Vectashield Mounting Medium (Vector Laboratories, Burlingame, CA, USA). One hundred tumor cell nuclei per tumor were assessed for HER2 gene and CEP17 probe signals at 100x magnification. The HER2/CEP17 ratio was calculated as well. A ratio below 1.8 was defined as a normal copy number, a ratio of 1.8–2.2 as an equivocal copy number and a ratio above 2.2 as gene amplification, as described in ASCO & CAP guidelines 19. DNA and RNA extraction The hematoxylin and eosin stained section from each formalin-fixed paraffin-embedded tissue block was used to guide macro-dissection for DNA and RNA extraction and to estimate tumor percentage. On average, ten and five 10-µm-thick slides were cut for DNA and RNA isolation, respectively (number depending on tumor percentage and size). Because the different biopsies and tumor samples differ in cellularity and size, normalization of input for DNA/RNA isolation was performed by correcting for tumor area and percentage. After deparaffinization in xylene, tumor areas were macro-dissected using a scalpel and areas with necrosis, dense lymphocytic infiltrates, and pre-invasive lesions were intentionally avoided. For DNA isolation, the dissected tissue was placed in 1M NaSCN overnight, after which proteinase K-based extraction was performed according to the QIAamp DNA FFPE tissue kit (Qiagen). RNA extraction was performed according to the miRNeasy FFPE kit (Qiagen). Total DNA and RNA concentration were measured by a spectrophotometer (Nanodrop ND-1000, Thermo Scientific Wilmington, DE, USA) and a fluorometer (Qubit 2.0 Fluorometer, Life Technologies, Bleiswijk, the Netherlands), the latter using the Qubit dsDNA HS assay kit and RNA assay kit. Absorbance at 230, 260 and 280 nm was evaluated (Nanodrop) and the RNA Integrity Number was determined using the Agilent RNA 6000 Nano kit on the Bioanalyzer (Agilent). DNA quality To assess DNA quality and the presence of inhibitors, the QuantideX® qPCR DNA QC Assay (Asuragen, Austin, USA) was used. This is a multiplex qPCR assay with one channel 214

Decalcification in breast cancer pathology

(FAM) that detects an 82bp amplicon from the TBP gene, which assesses DNA quality & quantity, and with another channel (VIC/HEX) that detects a non-human amplicon spiked into each sample, to evaluate the presence of inhibitors. To determine the target copy number in a tested sample, a standard curve was established in every run using DNA standards at 50, 10, 2 and 0.4 ng/μL in duplicate. The QuantideX® assay determines the functional quality of sample DNA using the Quantitative Functional Index (QFI™) Score, which is the fraction of total genomic DNA copies that can be PCR amplified 20 and the quality using the amplifiable copy number. Size ladder PCR To determine DNA fragmentation, a size ladder PCR was performed using the specimen control size ladder kit (In Vivo Scribe, Huissen, the Netherlands) on the Veriti Thermal Cycler (Applied Biosystems, Bleiswijk, the Netherlands) with a 35 cycle PCR reaction. DNA input was corrected for the amplifiable copy number measured with the QuantideX® assay. The PCR product was mixed with Hi-Di Formamide and ROX500 and analyzed on the 3730 DNA Analyzer (ThermoFisher Scientific, Bleiswijk, the Netherlands). Results were processed with Genescan software (Thermo Fisher Scientific, Bleiswijk, the Netherlands). Validation in paired breast tumors and bone metastases To validate these findings on real metastases, we retrospectively selected formalin-fixed paraffin-embedded material of bone metastases from eight patients from our diagnostic pathology archives. Original diagnoses of the primary tumors were made between January 2002 and January 2015 and of the bone metastases between September 2013 and May 2015. All obtained bone metastases were biopsies taken from vertebrae or pelvis and pathological records disclosed decalcification in EDTA for 5 hours to overnight according to routine procedures. Matching primary tumor material was available for four of these patients to allow comparison of receptor staining between primary tumors and bone metastases. Patient characteristics are listed in Table 1 (validation cohort). These samples were subjected to immunohistochemistry, fluorescence in situ hybridization and DNA/RNA quality and quantity assessment as mentioned before. Statistics Results obtained by immunohistochemistry were compared by cross tables, and the concordance percentages and linear weighted kappa-scores were calculated. Concordance was categorized as follows: (almost) perfect (kappa: 0.8-1), substantial (kappa: 0.6-0.8), moderate (kappa: 0.4-0.6) and poor concordance (kappa 0-0.4). In addition, median percentages, 260/280 ratios and RNA Integrity Number were compared using the Wilcoxon signed-rank test. P-values below 0.05 were considered significant. All statistical calculations were done with IBM SPSS Statistics 21 and visualized with GraphPad Prism 6. 215

PART TWO | CHAPTER 7

RESULTS Immunohistochemistry Table 2 shows immunohistochemical expression of ERα, PR and HER2 in the 23 breast tumors per pre-analytic condition (control/no treatment, Christensen’s buffer (with and without microwave), Formical-4 and EDTA). ERα positivity was seen in 87% or 95.7% of the controls and PR positivity in 69.6% or 82.6%, using a 10% or 1% threshold for positivity, respectively. HER2 positivity was seen in 13% of controls. For ERα, exact staining percentages tended to be lower in EDTA (p=0.049) and Formical-4 (p=0.047) treated cases, compared to the control. For PR expression, only Christensen’s

Table 2. Comparison of ERα, PR (percentage of positively stained nuclei) and HER2 (DAKO scores) expression in paired samples of 23 breast cancers routinely processed (control) or undergoing decalcification in EDTA, Christensen’s buffer (CBM with or CB without microwave) or Formical-4 (F4).

100

50 90

EDTA

CBM

100

EDTA

100

CBM

100

EDTA

100

CBM

Control

HER2

Control

216

Control

Sample

ERα

1+ 0

0 0

Decalcification in breast cancer pathology

buffer treated tissue showed decreased antigenicity (p=0.041). However, since conversion from positive in the control to negative in the decalcified tissue is clinically most important, concordance of relative expression was checked with linear weighted kappa. Overall, a substantial to perfect concordance was seen for ERα, PR and HER2 when comparing the controls with all decalcified tissues, with some variation between 1% and 10% thresholds for positivity (2/23 differing cases for ERα and 5/23 for PR) (Tables 2, 3 and Supplementary Table S1 and S2; Figure 2). Especially, some drop in antigenicity was seen in the Christensen’s buffer treated tissue for PR (kappa 0.571; 1% threshold) and HER2 (0.493; exact DAKO scores), in the EDTA treated tissue for ERα (kappa 0.462; 1% threshold) and in the Formical-4 treated group for PR (no concordance at all). Clinically relevant changes (from positive to negative) occurred for ERα in one patient (4%; with Christensen’s buffer with microwave and EDTA) when the 1% threshold for positivity was used, but none was seen for the 10% threshold. For PR, a change from positive to negative appeared in two patients with the 10% threshold (in patient #1 for Christensen’s buffer with and without microwave and EDTA and in patient #18 for Christensen’s buffer only) and in two patients with the 1% threshold (in patient #2 and #3 for Christensen’s buffer with and without microwave and EDTA). Furthermore, the Formical-4 treated tissue showed a serious discordance in PR antigenicity. However, two cases out of this group of nine (patient #16 and #17) showed a change from negative in the control to positive with Formical-4. Nuclear, stromal and epithelial morphology and staining intensity did not seem to differ between samples treated with different buffers (Figure 2).

Figure 2. Immunohistochemical stainings for ERα, PR and HER2 on tumor tissue from patient #5, after decalcification in EDTA and Christensen’s buffer with and without microwave. 20x magnification.

217

PART TWO | CHAPTER 7

In situ hybridization Four control tissues showed a 2+ or 3+ HER2 expression by immunohistochemistry and were subjected to fluorescence in situ hybridization. Comparison of HER2 status by immunohistochemistry and fluorescence in situ hybridization in control tissue compared to decalcified tissues is shown in Supplementary Table 3. Three HER2 immunohisto chemistry positive cases appeared fluorescence in situ hybridization negative in the control and the decalcified samples. In all cases, cell morphology and signal interpretation were severely affected by Christensen’s buffer with microwave and Formical-4 buffers, as depicted in Figure 3. Interpretation of the fluorescence in situ hybridization slides was least hampered by EDTA treatment. RNA and DNA integrity Because the biopsies and original tumor samples differ in cellularity and size we tried to approximate equal tumor input. Paired differences between DNA and RNA yield obtained from tissue treated with different decalcifying agents resembled in pattern between Qubit, Nanodrop and Bioanalyzer (Supplementary Figure S1 and Supplementary Table S4). For both DNA and RNA, highest yields were seen for control tissue, followed by EDTA treated tissue.

Figure 3. HER2 fluorescence in situ hybridization on tumor tissue from patient #16 after decalcification in EDTA, Christensen’s buffer with (CBM) and without microwave (CB) and Formical-4 (F4). 100x magnification is used.

218

Decalcification in breast cancer pathology

Table 3. Linear weighted kappa of concordance between expression of ERα, PR and HER2 in split samples of 23 breast cancers routinely processed or undergoing decalcification in EDTA, Christensen’s buffer (with (CBM) or without microwave(CB)) or Formical-4. Shown are the relative expression for ERα and PR with positivity according to 10% and 1% thresholds, the relative expression for HER2 with 3+ cases considered positive and the absolute expression for HER2 with exact DAKO-scores (0, 1+, 2+ and 3+). CBM

EDTA

ERα (1% cut-off)

0.646 (SE 0.324; 95% CI 0.012-1)

0.462 (SE 0.305; 95% CI 0-1)

ERα (10% cut-off)

0.625 (SE 0.333; 95% CI 0-1)

PR (1% cut-off)

0.701 (SE 0.193; 95% CI 0.324-1)

0.571 (SE 0.250; 95% CI 0.082-1)

0.696 (SE 0.195; 95% CI 0.314-1)

PR (10% cut-off)

0.679 (SE 0.170; 95% CI 0.346-1)

0.667 (SE 0.203; 95% CI 0.269-1)

0.667 (SE 0.176; 95% CI 0.323-1)

HER2 (neg vs pos)

0.777 (SE 0.213; 95% CI 0.360-1)

0.625 (SE 0.333; 95% CI 0-1)

0.782 (SE 0.126; 95% 0.536-1)

0.493 (SE 0.223; 95% CI 0.056-0.930)

0.800 (SE 0.106; 95% CI 0.593-1)

HER2 (0, 1+, 2+, 3+)

(Almost) perfect concordance (0.8-1) Substantial concordance (0.6-0.8)

Moderate concordance (0.4-0.6) Poor concordance (0-0.4)

For RNA, 260/280 ratios were significantly higher in the control and the EDTA treated samples compared to the other decalcifying agents (Figure 4A). The RNA Integrity Number did not vary much, although EDTA treated material showed significantly higher values compared to control samples (p=0.019; Figure 4B). Also for DNA, 260/280 ratios were significantly higher in the control and EDTA treated samples, compared to the other decalcified tissues (Figure 4C). With the QuantideX® assay, inhibition was only seen in Formical-4 treated samples (5/9 samples). DNA quality based on amplifiable copies was severely affected by all three decalcifying agents, with the largest drop in copy number for Formical-4, Christensen’s buffer with and without microwave, respectively (Supplementary Figure S2A). When this number was corrected for expected copy number (based on Nanodrop measurements) the same pattern was observed (QFI™; Figure 4D). The size ladder PCR (with input corrected for amplifiable copies) showed a decreased DNA fragment length in Christensen’s buffer with and without microwave treated samples compared to the control (Figure 4E). This average decrease seemed to be caused mainly by loss of 90, 200 and 300 bp fragments (Supplementary Figure S2B). Samples decalcified with Formical-4 showed hardly any amplifiable copies, leading to uninterpretable size ladder PCR data.

219

PART TWO | CHAPTER 7

Validation in paired breast tumors and bone metastases All primary tumors and bone metastases were ERα positive and no significant differences were perceived between absolute ERα staining percentages. Also, no significant differences were seen in PR staining percentages, although patient #1 had a PR-negative primary tumor and a PR positive metastasis. No discordance was observed between HER2 immunohisto chemistry and fluorescence in situ hybridization (Table 4; Supplementary Figure S3). Furthermore, nuclear, stromal and epithelial morphology and staining intensity did not differ between primary tumors and EDTA treated bone metastases. Paired analyses of DNA and RNA quality and quantity as measured with Nanodrop, Qubit and Bioanalyzer (corrected for tumor percentage) between primary tumors and metastases did not point to significant differences (Figure 5). When DNA fragmentation was compared, two primary tumor samples showed a smaller fragment length. Both samples originated from 2002, while all others were diagnosed between 2011 and 2015. A 4.5

RNA Integrity Number (RIN)

3.5

4.0

260/280 ratio

3.0

3.0 RIN

p < 0.01 p < 0.01

3.5

2.5 2.0 1.5 1.0 0.5

RNA 260/280 ratio

p < 0.05

p < 0.001 p < 0.001 p < 0.05

2.5

p < 0.01 p < 0.001

2.0 1.5

Control

EDTA

CBM

Control

EDTA

CBM

Decalcification buffer

DNA 260/280 ratio

DNA Quantidex Functional Index

DNA average fragment length

score (QFI) p < 0.05

p < 0.05 p < 0.05

3.5

p < 0.01 p < 0.05

3.0 2.5 2.0 1.5 Control

EDTA

CBM

Decalcification buffer

p < 0.05

6 p < 0.05 p < 0.001

Control

EDTA

p < 0.05

150

p < 0.001

Fragment length in bp

% actual copies/ expected copies

p < 0.05

4.0 260/280 ratio

p < 0.001

p < 0.05

4.5

CBM

Decalcification buffer

p < 0.05 100

Control

EDTA

CBM

Decalcification buffer

Figure 4. Quality and quantity of RNA and DNA, isolated from tissue pretreated with EDTA, Christensen’s buffer with (CBM) and without microwave (CB) and Formical-4 (F4; n=23). A) 260/280 ratios for RNA, measured by Nanodrop. B) Bioanalyzer RNA Integrity Number (RIN) values. C) 260/280 ratios for DNA, measured by Nanodrop. D) QFI™ (Quantitative Functional Index) in percentages (actual copy number/expected copy number, based on quantity of DNA measured by Nanodrop). E) Average DNA fragment length in base pairs, measured with size ladder PCR. DNA input is corrected for amplifiable copy number, measured with the QuantideX® qPCR assay.

220

Decalcification in breast cancer pathology

Table 4. Comparison of ERÎą, PR (percentage of positively stained nuclei) and HER2 (DAKO scores and FISH) expression in paired samples of 8 routinely processed primary breast cancers and their EDTA treated bone metastases Immunohistochemistry Sample

ERÎą

FISH

HER2

100

>10*

<10*

100

>10*

<10*

>10*

100

<10*

100

>10*

neg*

neg

pos*

pos

* Material not available, information obtained from pathological records. P: primary tumor; M: metastasis

RNA quantity paired samples

260/280 ratio

1000 500 0

Qubit

3.0

2.5

2.0

1.5

Bioanalyzer

DNA quality paired samples

DNA quantity paired samples

DNA fragmentation paired samples 110

260/280 ratio

2.1

1000

500

Nanodrop

Qubit

Primary tumors Metastases

1.0

Nanodrop

Bioanalyzer

1500

ng/uL

3.5

3.0

1.0

Nanodrop

3.5

RIN

ng/uL

10000

1500

RNA quality paired samples

Fragment lenght in bp

2.0

1.9

100 90 80 70 60 50

1.8

Nanodrop

Size ladder PCR

Figure 5. Yield of RNA and DNA, isolated from tissue from paired primary tumors and paired EDTA treated bone metastases (n=8), measured with Nanodrop, Qubit and Bioanalyzer. Tumor areas where normalized and RNA and DNA yield corrected for tumor percentage. Paired differences were calculated with Wilcoxon signed-rank test. A) RNA yield, measured by Nanodrop, Qubit and Bioanalyzer; B) DNA yield, measured by Nanodrop, Qubit and size ladder PCR.

221

PART TWO | CHAPTER 7

DISCUSSION The occurrence of receptor conversion between the primary breast carcinoma and corresponding distant metastases has been widely accepted 21. Therefore, most guidelines now advise to biopsy a distant metastasis at presentation of metastatic disease 4,5. This is challenging when the metastasis is located in bone, since the decalcification process could potentially compromise antigenicity and may, as such, hamper interpretation of (molecular) diagnostics. With an incidence of 6-22% in 5.4 to 8.4 years of follow-up 16,17, bone is one of the most common metastatic sites among breast cancer patients, which emphasizes the relevance of this subject. In the present study we demonstrate that immunohistochemistry of ERα, PR and HER2 is not much affected by tissue decalcification with agents containing formic acid or EDTA. However, quantity and quality of isolated DNA and RNA is affected by all three decalcification buffers tested, although EDTA can be used when results are interpreted with caution. In line with this, HER2 fluorescence in situ hybridization could only be interpreted in EDTA treated tissue. We validated these findings in eight patients with paired primary breast tumors and EDTA treated bone metastases. Immunohistochemistry, fluorescence in situ hybridization and DNA/RNA quality and quantity were comparable in paired cases, validating the EDTA protocol in real life. To our knowledge, this is the first study to describe the influence of multiple decalcifying agents on both immunohistochemistry, fluorescence in situ hybridization and DNA/RNA integrity in the same breast cancer patients in a relative large cohort. A previous study based on ten decalcified breast cancer samples reported an overall negative impact on ERα and PR staining intensity which affected the staining detectability and therefore proportion of tumor cell staining 22. However, time of decalcification was merely one hour in contrast to our protocol of overnight decalcification which approximates the normal diagnostic situation in our clinic. Additionally, Gertych et al. performed decalcification of breast cancer tissue (n=9) during several periods of time and saw the largest decrease in antigenicity in the first six hours 23. Likewise, we observed an absolute decrease in staining percentages for ERα (EDTA and Formical-4 treated tissue) and PR (Christensen’s buffer treated tissue). However, this decrease is only clinically relevant when it affects potential treatment decision making; in other words, when a change is achieved from positive to negative. In our study, this was the case for ERα in one patient (4% of cases) when the 1% threshold for positivity was used. Therefore, we advise to use the 10% threshold, since this leads to fewer falsely stratified patients for hormonal therapy, especially in decalcified tissue. For PR, a change from positive to negative appeared in two patients with the 10% threshold and in two patients with the 1% threshold. However, in three of these cases ERα was still positive, so no influence on treatment decision making was imposed. Furthermore, the Formical-4 treated tissue showed a serious discordance in PR antigenicity. However, two cases showed 222

Decalcification in breast cancer pathology

a change from negative in the control to positive with Formical-4, which can hardly be due to the influence of the decalcifying agent. This was also seen in patient 1 of the paired primary tumor and bone metastases samples. Tumor heterogeneity is the most likely option here. Indeed, PR discordance between biopsies and full resection specimens, also largely to be explained by tumor heterogeneity, is relatively common (15%) 24. Moreover, immunohistochemistry showed three HER2 3+ scored cases (out of 23), although only one fluorescence in situ hybridization amplified case was observed. Discordance between immunohistochemistry and fluorescence in situ hybridization has been described before, but is relatively rare in 3+ samples (9%) 25. However, this discordance was seen both in the decalcified samples and the control sample, so no influence of the decalcification process could explain this discrepancy. And no discordance was perceived in the validation cohort, though this group was small. Furthermore, the interpretation of the Christensen’s buffer with and without microwave and Formical-4 treated fluorescence in situ hybridization slides was seriously impeded, which was confirmed by the absence of signal in the one HER2 amplified case. This is in line with the findings of Brown et al. who reported a failure of their fluorescence in situ hybridization protocol when formic acid was used 12. Future research should address if adjustments in the standard ISH protocol could overcome these problems. DNA quality was affected by decalcification as well; all decalcified tissues showed significantly less amplifiable copies than the controls. Christensen’s buffer with microwave and Formical-4 treated tissue showed the largest drop and the highest fragmentation of DNA, compatible with common knowledge that acidic agents fragment the DNA. Singh et al. compared multiple acidic decalcifying agents on bone biopsies from six patients, but found no differences in DNA and RNA integrity (measured by qPCR) between tissue treated with formic acid based buffers and EDTA 11. However, time of decalcification was again very short (two hours), which may explain these differences. As expected, Christensen’s buffer only showed higher DNA quality than Christensen’s buffer with microwave treated tissue, which can be explained by the acceleration of decalcification by the microwave 26. In immunohistochemistry however, hardly any differences were seen, although one HER2 2+ case converted to 1+. HER2 fluorescence in situ hybridization results for Christensen’s buffer with and without microwave treated tissue were equally affected. In contrast to DNA, no major differences in RNA quality were found between buffers when we compared the RNA integrity Number. In line with our findings, Singh et al. also did not find any dissimilarities in RNA quality by qPCR 11. However, we observed large differences in RNA purity measured by 260/280 ratios. The ratio of the absorbance at 260 and 280 nm (A260/A230 ratios) is commonly used to assess nucleic acid contamination with proteins, organic and phenol 27,28 although studies of Wilfinger et al. have revealed that changes in both the pH and ionic strength can have an influence on these ratios 29. The latter may thus be an alternative explanation for the observed poor outcome of the acidic buffers compared 223

PART TWO | CHAPTER 7

to EDTA and the control. A first limitation of this study is that we do not have samples of all four conditions (EDTA, Christensen’s buffer with and without microwave and Formical-4) of all patients, because biopsies were taken macroscopically and afterwards tumor was not always present. However, we included sufficient numbers of patients to make subgroups still comparable. Second, we did not subject the biopsies to different decalcifying time periods but overnight decalcification similar to the situation in the daily clinical practice. Nevertheless, differences in outcome between studies are probably caused by different decalcifying periods 11,22,23, so it would be worthwhile to further elucidate this aspect. Furthermore, by taking biopsies of the tumor to enable different decalcification conditions, we probably introduced some heterogeneity in samples. We are aware of this side issue of our study design and tried to mark cases suspected of heterogeneity in Supplementary Table S1. In conclusion, decalcification procedures based on Christensen’s buffer, EDTA and Formical-4 in general seem to have relatively little influence on ERα, PR and HER2 analysis by immunohistochemistry, while EDTA performs best. When regarding exact percentages we advise to use a 10% threshold for positivity of ERα and PR to prevent patients to be falsely stratified for hormonal therapy. For molecular diagnostics, EDTA seems to be the best choice.

Acknowledgements This study is supported by Dutch Cancer Society grant UU 2011-5195 and Philips Consumer Lifestyle. We would like to thank Annette Gijsbers-Bruggink for collecting and processing the tissues, and our colleagues from the immunopathology lab of the University Medical Center Utrecht for performing the routine ERα, PR and HER2 immunohistochemical stainings.

224

Decalcification in breast cancer pathology

REFERENCES 1. Hoefnagel LDC, van de Vijver MJ, van Slooten H, et al. Receptor conversion in distant breast cancer metastases. Breast Cancer Research. 2010;12(5):R75. doi: 10.1186/ bcr2645. 2. van Diest PJ, Hoefnagel LD, van der Wall E. Testing for discordance at metastatic relapse of breast cancer matters. J Clin Oncol. 2012;30(24):3031; author reply 3031-2, 30323. doi: 10.1200/JCO.2012.42.6734 [doi]. 3. Vignot S, Besse B, Andre F, Spano JP, Soria JC. Discrepancies between primary tumor and metastasis: A literature review on clinically established biomarkers. Crit Rev Oncol Hematol. 2012;84(3):301-313. doi: 10.1016/j. critrevonc.2012.05.002 [doi]. 4. Carlson RW, Allred DC, Anderson BO, et al. Metastatic breast cancer, version 1.2012: Featured updates to the NCCN guidelines. J Natl Compr Canc Netw. 2012;10(7):821-829. doi: 10/7/821 [pii]. 5. Hammond ME, Hayes DF, Wolff AC, Mangu PB, Temin S. American society of clinical oncology/college of american pathologists guideline recommendations for immunohistochemical testing of estrogen and progesterone receptors in breast cancer. J Oncol Pract. 2010;6(4):195-197. doi: 10.1200/JOP.777003 [doi]. 6. Bussolati G, Leonardo E. Technical pitfalls potentially affecting diagnoses in immunohistochemistry. J Clin Pathol. 2008;61(11):1184-1192. doi: 10.1136/ jcp.2007.047720 [doi]. 7. Gruchy JR, Barnes PJ, Dakin Hache KA. CytoLyt(R) fixation and decalcification pretreatments alter antigenicity in normal tissues compared with standard formalin fixation. Appl Immunohistochem Mol Morphol. 2015;23(4):297-302. doi: 10.1097/PAI.0000000000000082 [doi]. 8. Neves Jdos S, Omar NF, Narvaes EA, Gomes JR, Novaes PD. Influence of different decalcifying agents on EGF and EGFR immunostaining. Acta Histochem. 2011;113(4):484488. doi: 10.1016/j.acthis.2010.04.006 [doi]. 9. Adegboyega PA, Gokhale S. Effect of decalcification on the immunohistochemical expression of ABH blood group isoantigens. Appl Immunohistochem Mol Morphol. 2003;11(2):194-197. 10. Reineke T, Jenni B, Abdou MT, et al. Ultrasonic decalcification offers new perspectives for rapid FISH, DNA, and RT-PCR analysis in bone marrow trephines. Am J Surg Pathol. 2006;30(7):892-896. doi: 10.1097/01. pas.0000213282.20166.13 [doi]. 11. Singh VM, Salunga RC, Huang VJ, et al. Analysis of the effect of various decalcification agents on the quantity and quality of nucleic acid (DNA and RNA) recovered from bone biopsies. Ann Diagn Pathol. 2013;17(4):322-326. doi: 10.1016/j.anndiagpath.2013.02.001 [doi]. 12. Brown RS, Edwards J, Bartlett JW, Jones C, Dogan A. Routine acid decalcification of bone marrow samples can preserve DNA for FISH and CGH studies in metastatic prostate cancer. J Histochem Cytochem. 2002;50(1):113115. 13. Wickham CL, Sarsfield P, Joyner MV, Jones DB, Ellard S, Wilkins B. Formic acid decalcification of bone marrow

trephines degrades DNA: Alternative use of EDTA allows the amplification and sequencing of relatively long PCR products. Mol Pathol. 2000;53(6):336. 14. Alers JC, Krijtenburg PJ, Vissers KJ, van Dekken H. Effect of bone decalcification procedures on DNA in situ hybridization and comparative genomic hybridization. EDTA is highly preferable to a routinely used acid decalcifier. J Histochem Cytochem. 1999;47(5):703-710. 15. Matthews JB, Mason GI. Influence of decalcifying agents on immunoreactivity of formalin-fixed, paraffinembedded tissue. Histochem J. 1984;16(7):771-787. 16. Hagberg KW, Taylor A, Hernandez RK, Jick S. Incidence of bone metastases in breast cancer patients in the united kingdom: Results of a multi-database linkage study using the general practice research database. Cancer Epidemiol. 2013;37(3):240-246. doi: 10.1016/j.canep.2013.01.006 [doi]. 17. Harries M, Taylor A, Holmberg L, et al. Incidence of bone metastases and survival after a diagnosis of bone metastases in breast cancer patients. Cancer Epidemiol. 2014;38(4):427-434. doi: 10.1016/j.canep.2014.05.005 [doi]. 18. van Diest PJ. No consent should be needed for using leftover body material for scientific purposes. for. BMJ. 2002;325(7365):648-651. 19. Wolff AC, Hammond ME, Hicks DG, et al. Recommendations for human epidermal growth factor receptor 2 testing in breast cancer: American society of clinical oncology/college of american pathologists clinical practice guideline update. J Clin Oncol. 2013;31(31):39974013. doi: 10.1200/JCO.2013.50.9984 [doi]. 20. Sah S, Chen L, Houghton J, et al. Functional DNA quantification guides accurate next-generation sequencing mutation detection in formalin-fixed, paraffin-embedded tumor biopsies. Genome Med. 2013;5(8):77. doi: 10.1186/ gm481 [doi]. 21. Van Poznak C, Somerfield MR, Bast RC, et al. Use of biomarkers to guide decisions on systemic therapy for women with metastatic breast cancer: American society of clinical oncology clinical practice guideline. J Clin Oncol. 2015;33(24):2695-2704. doi: 10.1200/ JCO.2015.61.1459 [doi]. 22. Darvishian F, Singh B, Krauter S, Chiriboga L, Gangi MD, Melamed J. Impact of decalcification on receptor status in breast cancer. Breast J. 2011;17(6):689-691. doi: 10.1111/j.1524-4741.2011.01168.x [doi]. 23. Gertych A, Mohan S, Maclary S, et al. Effects of tissue decalcification on the quantification of breast cancer biomarkers by digital image analysis. Diagn Pathol. 2014;9:213-014-0213-9. doi: 10.1186/s13000-014-0213-9 [doi]. 24. Arnedos M, Nerurkar A, Osin P, Aâ&#x20AC;&#x2122;Hern R, Smith IE, Dowsett M. Discordance between core needle biopsy (CNB) and excisional biopsy (EB) for estrogen receptor (ER), progesterone receptor (PgR) and HER2 status in early breast cancer (EBC). Ann Oncol. 2009;20(12):19481952. doi: 10.1093/annonc/mdp234 [doi]. 25. Bahreini F, Soltanian AR, Mehdipour P. A meta-analysis on concordance between immunohistochemistry (IHC) and fluorescence in situ hybridization (FISH) to detect

225

PART TWO | CHAPTER 7

HER2 gene overexpression in breast cancer. Breast Cancer. 2015;22(6):615-625. doi: 10.1007/s12282-014-0528-0 [doi]. 26. Pitol DL, Caetano FH, Lunardi LO. Microwave-induced fast decalcification of rat bone for electron microscopic analysis: An ultrastructural and cytochemical study. Braz Dent J. 2007;18(2):153-157. doi: S010364402007000200013 [pii]. 27. Di Bernardo G, Del Gaudio S, Galderisi U, Cascino A, Cipollaro M. Comparative evaluation of different DNA extraction procedures from food samples. Biotechnol Prog. 2007;23(2):297-301. doi: 10.1021/bp060182m [doi].

226

28. Johnson MT, Carpenter EJ, Tian Z, et al. Evaluating methods for isolating total RNA and predicting the success of sequencing phylogenetically diverse plant transcriptomes. PLoS One. 2012;7(11):e50226. doi: 10.1371/journal.pone.0050226 [doi]. 29. Wilfinger WW, Mackey K, Chomczynski P. Effect of pH and ionic strength on the spectrophotometric assessment of nucleic acid purity. BioTechniques. 1997;22(3):474-6, 478-81.

Decalcification in breast cancer pathology

SUPPLEMENTAL Qubit DNA

Nanodrop DNA

p < 0.01

100 80

Control

EDTA

CBM

p < 0.05 p < 0.001

Control

CBM

Decalcification buffer

Nanodrop RNA

Qubit RNA

300

p < 0.05

ng/uL

200

EDTA

CBM

Decalcification buffer

Bioanalyzer RNA

300 250

p < 0.05 p < 0.01

200

400

p < 0.01 p < 0.01 p < 0.001 p < 0.01

250

600

Control

EDTA

Decalcification buffer

p < 0.05 p < 0.05 p < 0.05

800

ng/uL

p < 0.01 p < 0.001 p < 0.01

150

100

Control

EDTA

CBM

200

ng/uL

4500 4300 1800 1600 1400 1200 1000 800 600 400 200 0

p < 0.001 p < 0.05

p < 0.01

ng/uL

Decalcification buffer

Control

EDTA

CBM

Decalcification buffer

Supplementary Figure S1. Yield of RNA and DNA, isolated from tissue pretreated with EDTA, Christensenâ&#x20AC;&#x2122;s buffer with (CBM) and without microwave (CB) and Formical-4 (F4), measured with Nanodrop, Qubit and Bioanalyzer. Tumor areas where normalized and RNA and DNA yield corrected for tumor percentage. Paired differences were calculated with Wilcoxon signed-rank test. A) DNA yield, measured by Nanodrop and Qubit; B) RNA yield, measured by Nanodrop, Qubit and Bioanalyzer.

227

PART TWO | CHAPTER 7

p < 0.05

9000 3000

p < 0.001

p < 0.05

2000 1000 0

Control

EDTA

CBM

Decalcification buffer

proportion of fragments

p < 0.01 p < 0.001

14000

copies per uL

Amplifiable copies

Proportion of fragments in groups 0.8

600 bp 400 bp

0.6

300 bp 200 bp

0.4

100 bp 90 bp

0.2 0.0

Control

EDTA

CBM

Decalcification buffer

Supplementary Figure S2. Fragmentation and amplifiable copies of DNA isolated from tissue treated with EDTA, Christensen’s buffer with (CBM) and without microwave (CB) and Formical-4 (F4). A) Amplifiable copies and fragmentation of DNA, measured with QuantideX® qPCR assay. B) Proportion of most seen DNA fragment lengths in groups (90, 100, 200, 300, 400, 600 bp). DNA input is corrected for amplifiable copy number, measured with the QuantideX® qPCR assay.

Supplementary Figure S3. HER2 fluorescence in situ hybridization on tissue from an EDTA treated bone metastasis from patient #8. 100x magnification.

228

Decalcification in breast cancer pathology

Supplementary Table S1. Comparison of relative expression of ERα, PR (positivity with 10% and 1% thresholds) in split samples of 23 breast cancers routinely processed (control) or undergoing decalcification in EDTA, Christensen’s buffer (with (CBM) or without microwave (CB)) or Formical-4 (F4).

1% threshold

10% threshold

1% threshold

10% threshold

1% threshold

10% threshold

1% threshold

10% threshold

1% threshold

10% threshold

1% threshold

EDTA

10% threshold

CBM

Control

1% threshold

10% threshold

1% threshold

10% threshold

CBM

EDTA

Control

Sample

ERα

^: discordance between 10% and 1% thresholds for positivity; *: possible heterogeneity.

229

230

control

HER2

control

ERα

18/23

kappa

0.701 2/12

2/12

4/12

1/23

20/23 2/23

0 0.777

1/12

10/12

2/23

0.679

3/23

15/23

2/23

1/23

kappa

1/12

CBM

5/23

21/23

0.646

1/23

1/12

1/23

20/23

CBM

3/23

0, 1+, 2+

1% threshold

10% threshold

1% threshold

10% threshold

kappa

CBM

1/12

8/12

6/12

11/12

10/12

1/12

0.625

kappa

0.571

0.667

kappa

0.625

kappa

0/21

18/21

EDTA

2/21

3/21

1/21

5/21

EDTA

2/21

1/21

3/21

EDTA

3/21

16/21

13/21

2/21

18/21

kappa

0.696

0.667

kappa

0.462

kappa

8/9

1/9

8/9

1/9

7/9

2/9

9/9

kappa

Supplementary Table S2. Comparison of ERα, PR and HER2 status in routinely processed breast cancer tissue and parallel tissue undergoing decalcification in EDTA, Christensen’s buffer (with (CBM) or without microwave(CB)) or Formical-4.

PART TWO | CHAPTER 7

Decalcification in breast cancer pathology

Supplementary Table S3. Comparison of expression of HER2 by immunohistochemistry and fluorescence in situ hybridization in routinely processed breast cancer tissue and parallel tissue undergoing decalcification in EDTA, Christensenâ&#x20AC;&#x2122;s buffer (with (CBM) or without microwave(CB)) or Formical-4 (F4). HER2 Control

CBM

EDTA

Cases

IHC

ISH

IHC

ISH

IHC

ISH

IHC

ISH

Neg

#16

Pos

F4 IHC

ISH

231

PART TWO | CHAPTER 7

Supplementary Table S4. Characteristics of DNA (a) and RNA (b), isolated from tissue pretreated in EDTA, Christensenâ&#x20AC;&#x2122;s buffer with and without microwave and Formical-4 (Test cohort) and paired primary tumors and EDTA treated bone metastases (c) (Validation cohort), measured by Nanodrop, Qubit, Bioanalyzer and Quantidex qPCR assay. A: DNA test cohort Control Nanodrop

Qubit

Quantidex

QFI Score

Minimum Volume to Detect 5%* variant (uL)

0,9

Pass

45,0

1,1

4,4

218,40

2,1

1,52

21,8

Pass

194,0

0,3

253,70

2,1

1,16

30,8

Pass

1068,0

1,4

200,20

2,1

1,11

25,4

Pass

2371,0

3,9

565,10

2,1

1,63

51,4

Pass

2949,0

1,7

220,50

2,0

1,78

17,0

Pass

79,0

0,1

2,5

69,30

2,1

0,63

6,5

Pass

1054,0

5,0

72,40

2,1

0,57

6,1

Pass

519,0

2,4

321,10

2,0

1,2

32,4

Pass

2762,0

2,8

168,70

2,1

1,1

16,3

Pass

1508,0

2,9

314,00

2,1

1,95

31,2

Pass

136,0

0,1

1,5

60,80

2,1

0,63

5,1

Pass

189,0

1,0

1,1

66,80

2,1

1,27

6,9

Pass

220,0

1,1

146,00

2,1

1,64

19,7

Pass

1575,0

3,6

4944,90

2,0

2,19

98,8

At Risk

216269,0

14,4

558,00

2,1

2,09

50,2

Pass

2341,0

1,4

278,60

2,1

1,17

33,4

Pass

2085,0

2,5

73,30

2,0

0,74

5,8

Pass

263,0

1,2

275,60

2,1

1,63

30,4

Pass

1618,0

1,9

428,70

2,06

Pass

9113

1184,30

2,07

2,13

Pass

1560

0,4

1516,70

2,08

2,09

100

Pass

13183

2,9

2434,10

2,09

2,21

Pass

11922

1,6

232

Inhibition

0,24

ng/uL

3,1

260/230

13,80

260/280

ng/uL

tumor %

Amplifiable Copy Number (per uL)

Sample ID

Decalcification in breast cancer pathology

Supplementary Table S4. Continued

A: DNA test cohort EDTA Nanodrop

Qubit

Quantidex Minimum Volume to Detect 5%* variant (uL)

Pass

6,0

0,2

Fail

23,60

2,1

0,47

1,2

Pass

31,0

0,4

6,5

46,60

2,0

0,63

1,1

Pass

1,0

0,0

Fail

25,00

2,2

0,25

2,0

Pass

72,0

0,9

2,8

347,30

2,0

1,88

22,0

Pass

296,0

0,3

1,0

342,90

2,0

1,85

23,2

Pass

363,0

0,3

1,0

15,90

2,5

0,21

1,5

Pass

101,0

2,1

2,0

28,10

2,2

0,52

2,7

Pass

143,0

1,7

1,4

101,10

2,0

1,74

8,6

Pass

100,0

0,3

2,0

10E

349,30

2,1

1,46

26,6

Pass

748,0

0,7

1,0

11E

46,30

2,0

0,96

2,4

Pass

13,0

0,1

Fail

12E

197,20

2,0

1,76

17,1

Pass

276,0

0,5

1,0

13E

313,90

2,1

2,12

19,5

Pass

282,0

0,3

1,0

14E

193,40

2,0

1,81

18,6

Pass

1157,0

2,0

1,0

15E

67,40

2,1

0,72

4,4

Pass

81,0

0,4

2,5

16E

16,10

2,8

0,31

1,8

Pass

49,0

1,0

4,1

17E

70,50

2,1

1,11

7,2

Pass

257,0

1,2

1,0

18E

10,90

1,9

0,8

0,4

Pass

18,0

0,5

11,3

19E

247,60

2,0

1,78

20,8

Pass

712,0

0,9

1,0

20E

231,10

2,06

2,32

Pass

2517

3,6

411,70

2,02

2,25

55,2

Pass

1181

0,9

Amplifiable Copy Number (per uL)

0,6

Inhibition

0,28

ng/uL

2,9

260/230

10,10

260/280

ng/uL

tumor %

QFI Score

Sample ID

21E 22E 23E

233

PART TWO | CHAPTER 7

Supplementary Table S4. Continued A: DNA test cohort Christensen’s buffer Nanodrop

Qubit

Quantidex Minimum Volume to Detect 5%* variant (uL)

Pass

0,0

Fail

2,6

Pass

6,0

0,0

Fail

21,90

2,2

0,38

0,5

Pass

1,0

0,0

Fail

103,50

1,9

0,83

1,5

Pass

0,0

Fail

409,10

1,9

1,66

16,9

Pass

4,0

0,0

Fail

74,70

1,9

1,25

1,8

Pass

0,0

Fail

53,80

1,9

1,45

0,8

Pass

0,0

Fail

66,30

1,9

1,14

1,6

Pass

0,0

Fail

98,10

1,9

1,33

1,0

Pass

2,0

0,0

Fail

10C

113,10

2,0

0,72

9,7

Pass

154,0

0,4

1,3

11C

47,10

1,9

1,37

0,3

Pass

0,0

Fail

12C

37,10

2,1

0,81

0,8

Pass

1,0

0,0

Fail

13C

284,40

1,9

1,73

9,3

Pass

9,0

0,0

Fail

14C

64,80

1,9

0,8

1,6

Pass

9,0

0,0

Fail

15C

17,70

2,5

0,5

0,6

Pass

23,0

0,4

8,8

16C

32,30

2,03

1,81

0,586

Pass

Fail

17C

28,60

2,2

0,68

0,4

Pass

3,0

0,0

Fail

18C

20,00

2,1

0,33

1,6

Pass

30,0

0,5

6,6

19C

304,50

1,9

2,02

13,6

Pass

6,0

0,0

Fail

20C

102,00

1,94

2,86

1,9

Pass

Fail

21C

446,30

1,91

1,96

38,2

Pass

234

0,2

22C

222,40

1,94

1,93

15,2

Pass

0,1

4,1

23C

139,70

1,89

3,22

12,6

Pass

11,3

234

Amplifiable Copy Number (per uL)

0,2

1,48

Inhibition

0,49

2,0

ng/uL

2,2

64,90

260/230

17,20

260/280

ng/uL

tumor %

QFI Score

Sample ID

Decalcification in breast cancer pathology

Supplementary Table S4. Continued A: DNA test cohort Christensenâ&#x20AC;&#x2122;s buffer without microwave Nanodrop

Qubit

Quantidex

QFI Score

0,7

Pass

145,0

1,2

1,4

2,15

10,8

Pass

377,0

0,1

1,0

123,55

1,91

2,1

8,7

Pass

2448,0

3,0

1,0

32,05

1,97

1,25

0,7

Pass

70,0

0,3

2,9

352,35

1,87

1,97

7,5

Pass

6,0

0,0

Fail

449,65

1,93

2,1

10,4

Pass

53,0

0,0

3,8

121,55

1,92

1,29

3,9

Pass

190,0

0,3

1,1

226,80

1,88

2,07

4,7

Pass

24,0

0,0

8,4

175,30

1,84

1,95

3,9

Pass

7,0

0,0

Fail

10M

20,45

1,91

1,12

0,2

Fail

18M

1,4

3,52

0,17

0,0

19M

289,15

1,99

1,98

6,3

Pass

1020,0

1,3

1,0

Inhibition

1,72

1,9

ng/uL

1,85

507,35

260/230

24,70

260/280

ng/uL

tumor %

Amplifiable Copy Number (per uL)

Minimum Volume to Detect 5%* variant (uL)

Sample ID

11M 12M 13M 14M 15M 16M 17M

20M 21M 22M 23M

235

PART TWO | CHAPTER 7

Supplementary Table S4. Continued A: DNA test cohort Formical-4 Qubit

Quantidex Minimum Volume to Detect 5%* variant (uL)

QFI Score

Amplifiable Copy Number (per uL)

Inhibition

ng/uL

260/230

260/280

ng/uL

Nanodrop

tumor %

Sample ID

1F 2F 3F 4F 5F 6F 7F 8F 9F 10F 11F 12F 13F

40,20

2,0

0,32

0,8

Fail

14F

48,10

1,9

1,74

1,2

Fail

15F

67,80

1,92

1,56

1,95

Pass

Fail

16F

31,60

1,96

1,44

1,12

Pass

Fail

17F

21,60

2,2

0,42

0,1

Fail

20F

33,90

1,87

0,26

0,592

Fail

21F

127,90

1,85

2,78

4,86

Fail

22F

53,10

1,83

1,702

1,75

Pass

Fail

23F

87,10

1,87

2,81

4,54

Pass

Fail

18F 19F

236

tumor %

ng/uL

555,9

340,3

248,7

97,8

178,6

208,4

383

109,7

498,8

306,7

303,5

98,8

55,8

130,5

204,4

749,1

469,7

259,4

473,5

86,5

198,3

232,6

134,2

260/280 1,99

1,86

1,9

1,96

2,00

1,78

1,98

1,87

1,96

2,01

2,05

1,95

1,81

1,94

1,91

2,00

2,02

1,94

2,00

2,06

1,92

1,97

2,03

260/230 1,84

1,57

1,92

2,12

1,99

0,83

1,85

1,5

1,95

1,87

1,98

1,84

1,28

1,74

1,53

2,06

1,99

1,88

1,93

2,09

1,64

1,94

2,05

ng/uL 268

366

332

200

186

156

95,8

190

138

128

87,6

46,8

124

166

188

196

184

166

86,4

174

126

178

395

135

122

314

379

346

216

116

ng/uL

Bioanalyzer

N/A

2,5

2,4

3,5

2,3

2,4

2,5

3,1

2,1

2,4

1,1

1,5

2,1

2,2

2,4

2,3

RIN

Qubit

23E

22E

21E

20E

19E

18E

17E

16E

15E

14E

13E

12E

11E

10E

Sample ID tumor % 20

1,79 1,95 1,69 1,92 1,92 1,9 1,9 1,93 1,82 1,94 1,78 1,86 1,88 1,9 1,68 1,84 1,98 1,67 1,92 1,99

1,89

38,8 61,7 118,8 113,6 196,2 282,3 96,1 113,3 89,8 481 121,4 186,8 133,3 254,1 89,1 102 349,3 39,7 319 166,1

342,5

Nanodrop

ng/uL

Nanodrop

260/280

Sample ID

1,79

2,07

1,87

0,63

1,89

1,4

1,34

1,72

1,55

1,58

1,37

1,76

1,42

1,89

1,58

1,71

1,95

1,75

1,85

1,72

1,18

260/230

EDTA Qubit

236

98,2

160

28,4

198

61,2

41,4

148

100

102

70,6

166

55,6

108

65,8

130

174

45,8

62,8

22,4

ng/uL

Control Bioanalyzer

154

348

173

100

189

249

112

292

189

191

235

ng/uL

B: RNA test cohort

2,4

4,3

2,4

2,5

4,1

2,1

2,7

2,4

2,5

2,6

2,5

2,2

2,4

2,1

2,2

2,4

2,3

2,4

2,2

RIN

Supplementary Table S4.

Decalcification in breast cancer pathology

237

10C

11C

12C

13C

14C

15C

16C

17C

18C

19C

20C

21C

22C

23C

tumor %

ng/uL

93,7

156

300,7

275,6

148,6

63,4

199,9

134,6

374,8

188,3

189,4

235,9

157,2

146,8

74,3

107,6

144,9

162,2

1132,8

163,5

104,2

84,5

64,8

260/280 1,73

1,74

1,78

1,8

1,61

1,72

1,49

1,85

1,77

1,78

1,76

1,58

1,89

1,72

1,67

1,68

1,82

1,63

1,62

1,83

1,56

260/230 1,34

1,53

1,86

1,94

1,22

0,74

1,41

0,88

1,88

2,06

1,91

1,86

1,3

1,62

1,71

1,59

1,41

1,95

1,54

1,01

1,99

0,98

79,6

135

592

56,6

46,8

34,4

63,8

21,2

156

43,2

62,8

43,6

114

25,8

39,8

59,6

138

41,6

39,4

46,4

18,1

ng/uL

Bioanalyzer

102

188

144

179

111

105

212

150

113

100

ng/uL

Qubit

1,4

1,2

1,3

2,5

2,4

1,1

2,4

N/A

2,3

2,4

2,6

1,4

2,4

2,2

2,3

2,4

2,5

2,2

RIN

Nanodrop

23M

22M

21M

20M

19M

18M

17M

16M

15M

14M

13M

12M

11M

10M

Sample ID tumor % 80

Nanodrop

291

14,2

57,7

62,45

133,6

90,6

218,4

160,4

15,95

93,75

241,75

28,9

ng/uL

Christensen’s buffer without microwave

1,91

1,41

1,68

1,57

1,69

1,73

1,74

1,66

1,61

1,78

1,75

1,65

260/280

Sample ID

1,94

0,51

1,39

1,58

1,77

1,4

1,65

1,42

1,22

1,49

1,68

1,1

260/230

Christensen’s buffer Qubit

53,2

5,62

30,2

35,4

122

57,2

250

118

12,8

70,6

50,6

15,2

ng/uL

B: RNA test cohort

Bioanalyzer

104

110

153

165

252

155

ng/uL

238 2,1

1,1

2,4

2,3

2,2

2,8

2,4

2,3

RIN

Supplementary Table S4. Continued

PART TWO | CHAPTER 7

15F

16F

17F

20F

21F

22F

23F

19F

18F

14F

tumor %

13F

12F

11F

10F

ng/uL

96,5

49,9

112,5

255

126,5

74,9

181,5

263,3

1,68

1,61

1,53

1,67

1,66

1,63

1,66

1,64

1,68

260/280

Nanodrop

1,34

1,06

1,94

1,64

1,22

1,44

1,36

1,71

260/230

Sample ID

Qubit

135

62,2

144

62,6

62,8

116

ng/uL

Formical-4 Bioanalyzer

188

163

249

ng/uL

B: RNA test cohort

N/A

2,3

2,4

2,2

2,3

RIN

Supplementary Table S4. Continued

Decalcification in breast cancer pathology

239

Primary 8

335,3

1,81

45,4

Metastasis 8

Metastasis 7

Primary 7

Metastasis 6

1,95

Primary 6

62,9

Metastasis 5

45,8

Metastasis 1

298,3

273,6

307,1

254,2

112,1

565,8

414,1

71,1

1,23

1,95

1,94

2,05

1,87

1,76

1,96

2,11

Nanodrop

1,78

149,4

1,93

129,4

1,92

302,4

Metastasis 8

1,93

284,4

1,95

556,0 1,92

1,99

91,5

1,93

465,8

Sample ID

Nanodrop

81,4

Metastasis 7

Primary 5

1,92

2,2

7878

Metastasis 5 Metastasis 6

Metastasis 4

2,02

33,2

1,5

306

Metastasis 3

Metastasis 2

Primary 4

440,3

1,26

2,5

1774

Metastasis 1

Metastasis 3

1,95

Qubit

188

2,5

2002

Sample ID

Primary 3

92,1

Nanodrop

1,7

14,9

27,2

-2,49

1,12

79,2

1,56

Bioanalyzer

Metastasis 2

Qubit

Primary 2

Primary 1

Sample ID

DNA

Primary 8

1,93

442,6

Primary 7

3,5

321,7

1,97

1,96

112,8

Nanodrop

44,6

Primary 6

Primary 5

Primary 4

Primary 3

Primary 2

Primary 1

Sample ID ng/uL

RNA

RIN

C: Validation cohort

tumor %

ng/uL

260/280 260/280

tumor % tumor %

260/230 260/230

260/280 1,96

1,97

1,95

1,83

1,97

1,98

1,96

1,84

260/280

ng/uL ng/uL

260/230 1,47

1,87

2,1

1,58

1,66

2,04

1,5

260/230

Qubit

42,4

10,7

15,3

18,4

186

146

10,4

ng/uL 56

60,6

13,5

89,4

ng/uL

2,1 2,4 2,3 2,5 1,1 2,6 1,1 2,5

378 6193 7120 1239 125 1546 165 1865

Bioanalyzer

ng/uL

240 RIN

Supplementary Table S4.

PART TWO | CHAPTER 7

Decalcification in breast cancer pathology

241

Chapter 8 Willemijne AME Schrijver, Karianne Schuurman, Annelot van Rossum, Dutch Distant Breast Cancer Metastases Consortiumâ&#x20AC;˘, Ton Peeters, Natalie ter Hoeve, Wilbert Zwart*, Paul J van Diest*, Cathy B Moelans* â&#x20AC;˘ Members are listed in the acknowledgement * These authors contributed equally to this study

Endocrine therapy imposes an evolutionary selection pressure on breast cancer metastases in malignant peritoneal and pleural effusions to loose steroid hormone receptors Submitted

PART TWO | CHAPTER 8

ABSTRACT Discordance in estrogen receptor alpha (ERα), progesterone receptor (PR), androgen receptor (AR) and human epidermal growth factor receptor 2 (HER2) status between primary breast cancers and solid distant metastases (“conversion”) has been reported previously. Even though metastatic spread to the peritoneal and pleural cavities occurs frequently and is associated with high mortality, the rate of receptor conversion and the prognostic implications thereof remain elusive. We therefore determined receptor conversion in 91 effusion metastases (78 pleural, 13 peritoneal effusions) of 69 patients by immunohistochemistry (IHC) and in situ hybridization. Data were coupled to clinical variables and treatment history. ERα, PR and AR receptor status converted from positive in the primary tumor to negative in the effusion metastases or vice versa in 25-30%, 30-35% and 46-51% of cases, respectively. 19-25% of patients converted clinically relevant from “ERα+ or PR+” to ERα-/PR- and 3-4% from ERα-/PR- to “ERα+ or PR+”. For HER2, conversion was observed in 6% of cases. Importantly, receptor conversion for ERα (p=0.058) and AR (p<0.001) was more often seen in patients adjuvantly treated with endocrine therapy. Analogous to this observation, HER2loss was more frequent in patients adjuvantly treated with trastuzumab (p<0.001). Alike solid distant metastases, receptor conversion for ERα, PR, AR and HER2 is a frequent phenomenon in peritoneal and pleural effusion metastases. Adjuvant endocrine and trastuzumab therapy imposes an evolutionary selection pressure on the tumor, leading to receptor loss in effusion metastases. Determination of receptor status in malignant effusion specimens will facilitate endocrine treatment decision-making at this lethal state of the disease, and is hence recommended whenever possible.

244

Receptor conversion in breast cancer metastases in effusions

INTRODUCTION Each year, around 550.000 women die from the consequences of breast cancer 1, largely due to metastatic relapse. In around 30% of patients with metastatic breast cancer, the pleural cavity is involved 2,3 and less frequently the pericardial and peritoneal cavities 4,5. The presence of metastatic breast carcinoma cells in effusions is associated with poor prognosis and a median survival of 5 months 2,3,6-8. Immunohistochemistry (IHC) plays a valuable role in effusion cytology for the identification of metastatic malignancy. Inclusion of hormone receptor status assessment could direct treatment decision-making. This is underlined by the finding that tamoxifen treatment showed a therapeutic benefit in patients with ERα-positive malignant pleural effusions 9-11. In the clinical management of metastatic breast cancer, the choice of systemic treatment is traditionally based on the tissue characteristics of the primary tumor. Several previous studies have however shown that the expression of predictive tissue markers including ERα, PR and HER2 may differ between the primary breast tumor and solid distant metastases (“receptor conversion”) in a significant proportion of patients 12-14. Prolonged evolutionary pressure invoked by systemic endocrine therapies may effect hormone receptor expression, and with that, alter drug response. Consequently, alterations of hormone receptor expression in metastatic lesions in relation to the primary tumor may directly result in inappropriate endocrine treatment selection. Several guidelines therefore now recommend to biopsy distant metastases, and to reassess hormone and HER2 receptor status by IHC whenever possible 15,16. Androgen receptor (AR) is expressed in 60% of breast cancers and is more frequently expressed in ERα-positive than in ERα-negative tumors. AR signaling pathways show a distinct pattern, depending on the breast cancer subtypes. In ERα-positive breast cancer, AR is thought to antagonize the proliferative effect of ERα and in ERα-negative tumors, AR signaling has a proliferative role 17. In a comparison of ERα- and AR-positive breast cancer and paired local recurrences or solid distant metastases, AR expression is often maintained even when ERα-expression is lost 18,19. This suggests that anti-androgens may be a useful therapeutic strategy for patients with anti-estrogen resistant metastatic disease. Therefore, clinical trials addressing AR-targeted therapies in metastatic breast cancer are currently performed (http://www.cancer.gov/about-cancer/treatment/clinical-trials/, trial IDs NCI-2015-02043 and NCT02605486). However, the role of ERα-inhibitor induced selective pressure on AR receptor status in distant metastases remains to be elucidated. To our knowledge, there have been no studies on the influence of adjuvant endocrine therapies on receptor conversion between primary breast tumors and their corresponding malignant effusions so far, while this is a frequent metastatic site 20,21. Furthermore, also information about differences between receptor expression in solid and effusion metastases is lacking, due to small sample sizes of the reported studies 22,23. 245

PART TWO | CHAPTER 8

Here we report IHC staining of ERÎą, PR, AR and HER2 complemented with HER2 in situ hybridization in 69 patients with primary breast carcinomas and their matched malignant peritoneal and/or pleural effusions and solid distant metastases. We furthermore investigated the influence of adjuvant therapies on receptor conversion. Extensive knowledge of possible receptor conversion in malignant effusions could facilitate optimizing patient tailored therapy strategies for metastatic breast cancer patients.

MATERIALS AND METHODS Material In total, 91 malignant effusion specimens derived from 69 female breast cancer patients were used for this study. Retrospectively, 71 formalin fixed paraffin embedded (FFPE) tissue blocks of pleural and peritoneal effusions from 56 patients were obtained from the departments of Pathology of the University Medical Center Utrecht, Rijnstate Hospital Arnhem, Radboud University Medical Center Nijmegen, Bronovo Hospital The Hague, Meander Medical Center Amersfoort, OLVG Amsterdam, Slotervaart Hospital Amsterdam, Amphia Hospital Breda, Diakonessenhuis Utrecht, Pathology Laboratory Friesland, Groene Hart Hospital Gouda, Clinical Pathology Laboratory Heerlen, Leiden University Medical Center, Canisius Wilhelmina Hospital Nijmegen, St. Franciscus Hospital Rotterdam, Erasmus Medical Center Rotterdam, Clinical Pathology Laboratory Sittard, VieCuri Hospital Venlo, Isala Clinics Zwolle, all in The Netherlands. Fourteen effusion samples from nine patients were collected prospectively in the Netherlands Cancer Institute in Amsterdam and six samples from four patients from the University Medical Center Utrecht. Original diagnoses had been made between December 1989 and February 2016. Cytology samples were initially fixed in isopropanolol and embedded in paraffin by Cellient, (Hologic). For each case, hematoxylin-eosin stained slides of the paraffin blocks were reviewed by a single experienced pathologist (PvD) to confirm the presence of malignancy in all cytology samples. Only samples containing at least 20 tumor cells were selected. BerEP4 monoclonal antibody staining (Monosan), labelling epithelial tissues without reacting with mesothelial cells, was used to confirm presence of tumor cells. All samples were compared with the corresponding primary tumor and, when present, with one or more paired solid distant metastases (fifteen patients). All samples were recut and restained, using current standardized techniques (see below). This study was performed in accordance with the medical ethical guidelines of the University Medical Center Utrecht. The use of anonymous or coded left over material for scientific purposes is part of the standard treatment agreement with patients and therefore ethical approval was not required 24.

246

Receptor conversion in breast cancer metastases in effusions

Immunohistochemistry IHC for ERα, PR, HER2, AR and Ber-EP4 was carried out on full 4-mm sections with the Ventana (Ventana Medical Systems) according to the manufacturer’s instructions with mouse monoclonal antibodies against Ber-EP4/Ep-CAM (1:800, BS14, Monosan) and AR (1:20, AR27, Novocastra) and rabbit monoclonal antibodies against ERα (ready-to-use, SP1, Roche), PR (ready-to-use, 1E2, Roche) and HER2 (1:50, SP3, ThermoFisher). Appropriate controls were used throughout. Scoring Scoring of IHC slides was performed by consensus of two observers (PvD & WS) in random order, blinded to other data. The adequacy of staining in the primary carcinoma was checked by also evaluating the normal breast parenchyma when present. For ERα, PR and AR, the percentage of positively stained nuclei was estimated side by side with the BER-Ep4 stained slide as a reference. Samples with 10% or more immunopositive malignant cells, regardless of staining intensity, were classified as ERα or PR positive (European standard). The same was done for the 1% USA threshold. HER2 expression was scored using the DAKO scoring system as 0, 1+, 2+ and 3+ 25. HER2 expression was considered negative when 0 or 1+, equivocal when 2+ and positive when 3+. We regarded HER2 conversion as a shift from 0/1+/2+ without amplification by FISH to 2+ with amplification/3+ or vice versa. In situ hybridization All cases with 2+/3+ and discordant results in primary tumors compared to paired metastases were subjected to fluorescence in situ hybridization using a HER2/CEP17 dual FISH probe (Cytocell) on 4-µm slides. Analysis was performed on a Leica DM5500 B microscope system with Application Suite Advanced Fluorescence Software (Leica Microsystems). In short, formalin-fixed paraffin-embedded slides were deparaffinized and pretreated with citrate and protease buffers. Next, they were dehydrated and hybridized with 10μl probe in a ThermoBrite (Abbott Laboratories) at 37°C overnight. The next day, slides were washed in saline-sodium citrate buffers, counterstained with DAPI, dehydrated and mounted with Vectashield Mounting Medium (Vector Laboratories). One hundred tumor cell nuclei per tumor were assessed for HER2 gene and CEP17 probe signals at 100x magnification. The HER2/CEP17 ratio was calculated as well. A ratio below 1.8 was defined as a normal copy number, a ratio of 1.8–2.2 as an equivocal copy number and a ratio above 2.2 as gene amplification, according to the ASCO & CAP guidelines 26. Statistics Expression frequency of ERα, PR, AR and HER2 was compared in the primary tumors versus paired effusion and solid metastases using Wilcoxon signed-rank test. Comparison 247

PART TWO | CHAPTER 8

of IHC expression in peritoneal and pleural effusions was performed using Mann-Whitney U test. Dichotomized conversion data (from positive to negative and vice versa) were calculated for 1% and 10% thresholds for positivity and compared by Mc Nemar’s test. As steroid receptor conversion is clinically important if a patient converts from “ERα+ or PR+” to ERα-/PR-, or from ERα-/PR- to “ERα+ or PR+”, we calculated the frequency for these conversions as well. Statistical analysis was performed using IBM SPSS Statistics version 21 and visualized using GraphPad Prism 6.

RESULTS Decreased hormone receptor levels in malignant pleural and peritoneal effusions In total, 69 female breast cancer patients were included in this study with a mean age at diagnosis of the primary tumor of 57 years (Table 1). The primary lesions were predominantly of the ductal type and 38% of patients who underwent sentinel node biopsy had positive lymph nodes. Ninety-one malignant effusions were investigated; 78 of pleural and 13 of peritoneal origin. For sixteen patients, two or more consecutive samples were available. ERα positivity was observed in the vast majority of primary tumors (65% or 71% for the 10% or 1% thresholds for positivity, respectively). Solid metastases were more often ERα negative (p=0.022 for the 10% threshold and p=0.079 for the 1% threshold; ERα positivity of 38% for both thresholds), as was the case for effusion metastases (p=0.024 and p=0.097; ERα positivity of 46% or 57% for the 10% or 1% thresholds for positivity, respectively). PR positivity was generally lower than ERα in the primary tumors (p=0.004 and p=0.154; PR positivity of 39% or 58%, respectively). AR was expressed in 60% or 71% of primary tumors and 10% or 26% of effusion metastases (Figure 1; p<0.001 for the 10% and 1% thresholds for positivity, respectively). Frequent hormone receptor discordance in paired breast cancer and effusion metastases ERα, PR and AR showed a significantly lower expression in effusion samples compared to the paired primary tumors (p<0.001 for all three receptors; n=69). For the 10% threshold for positivity, 30% (21/69) of patients showed ERα conversion, with 26% (18/69) from positive in the primary tumor to negative in the effusions and 4% (3/69) from negative to positive. For PR, conversion rates were similar (total 30%; 21/69), with 25% (17/69) of samples converting from positive to negative and 5% (4/69) from negative to positive. For AR, discordance was even higher with 50% (35/69) of samples converting from positive to negative, and 1% (1/69) conversion from negative to positive. When comparing solid metastases to paired effusion metastases, ERα, PR and AR protein expression diverged in 20% (3/15; Table 2). 248

Receptor conversion in breast cancer metastases in effusions

Table 1. Clinicopathological characteristics of the patients analyzed in this study. Feature

Grouping

N or value

Age at primary diagnosis (in years)

Mean Range

57 32-85

Tumor size (in cm)

Mean Range

3.2 0.6-10.0

Histologic type

Invasive ductal Invasive lobular Invasive ductolobular Not available

52 6 5 6

75 9 7 9

Histologic grade (Bloom & Richardson)

I II III Not available

4 27 30 8

6 39 43 12

Mitotic activity index (per 2mm²)

Mean Range

14 0-60

Lymph node status

Negative Positive Not available

17 26 26

24 38 38

Site of distant solid metastases (n=15)

GI-tract/gynaecological Skin Lung Bone Liver

6 9 2 2 2

29 44 9 9 9

Site of metastases in body effusions (n=69)

Pleural effusion Ascites

78 13

86 14

Time between diagnosis of primary and first effusion metastasis (in months)

Mean Range

55 0-241

Survival time between diagnosis of first effusion metastasis and end of follow-up (in days)

Mean Range

340 0-4477

Adjuvant endocrine therapy

Yes No Unknown

33 18 18

48 26 26

Adjuvant chemotherapy

Yes No Unknown

31 18 20

45 26 29

Adjuvant targeted therapy

Yes No Unknown

8 40 21

12 58 30

For the 1% threshold for positivity, less conversion was seen. 25% (17/69) of patients showed ERα conversion, with 21% (14/69) from positive in the primary tumor to negative in the effusions and 4% (3/69) from negative to positive. For PR, 35% (24/69) converted in total, with 30% (21/69) of samples converting from positive to negative and 5% (3/69) from negative to positive. For AR, 43% (30/69) of samples converted from positive to negative, and 3% (2/69) from negative to positive. When comparing solid metastases to paired effusion metastases, ERα diverged in 20% (3/15), PR in 13% (2/15) and AR in 20% (3/15) of cases (Table 2). 249

PART TWO | CHAPTER 8

Figure 1. ERÎą, PR, AR and HER2 immunohistochemistry on paired primary breast tumors and pleural or peritoneal metastases. 20x magnification is used.

250

Effusion N=64

Solid metastasis N=12

Effusion N=69

Solid metastasis N=15

Effusion N=69

<0.001

0.007

0.50

0.001

0.50

Solid metastasis N=15

Primary N=69

ERα

Effusion N=64

Solid metastasis N=12

Effusion N=69

Solid metastasis N=15

Effusion N=69

Solid metastasis N=15

1% 6 0 17 3 6 1 26 3 6 1 17 2

+ + + + + +

<0.001

0.013

0.25

0.50

Solid metastasis N=15

Primary N=69

HER2

Effusion N=69

Solid metastasis N=15

IHC

0.73

Solid metastasis N=15

Primary N=69

HER2

Effusion N=66

Solid metastasis N=15

FISH

13 0 56 2

+ +

Solid metastasis N=15

Primary N=66

Table 3. HER2 receptor status of primary tumors and solid and effusion metastases of patients analyzed in this study. Data are shown for immunohistochemistry and FISH.

ERα

Solid metastasis N=15

10%

Table 2. Immunohistochemical hormone receptor status (ERα, PR and AR) of primary tumors and solid and effusion metastases of patients analyzed in this study. Data are shown for 10% and 1% thresholds for positivity.

Receptor conversion in breast cancer metastases in effusions

251

PART TWO | CHAPTER 8

Clinically relevant conversion (from “ERα+ or PR+” to ERα-/PR-, or from ERα-/PR- to “ERα+ or PR+”) was perceived in 28% (19/69) of patients for the 10% threshold and in 23% (16/69) of patients for the 1% threshold. 25% (17/69) and 19% (13/69) of patients, respectively, converted from “ERα+ or PR+” to ERα-/PR- and 3% (2/69) and 4% (3/69) of patients from ERα-/PR- to “ERα+ or PR+”. HER2 discordance was seen in 6% (4/69) of cases, were 3% (2/69) shifted from positive to negative and 3% (2/69) from negative to positive (Table 3). Concordance between IHC and FISH for 0, 1+ (being non-amplified) and 3+ cases (being amplified) was high (88%, 23/26, Table 4). Out of 69 patients in our cohort, 27 patients had material available of more than one metastasis (multiple solid or effusion metastases; Supplementary table S1). Figure 2 depicts ERα, PR and AR staining percentages and HER2 DAKO-scores for these patients. Large variation occurred during tumor progression from primary tumor to solid and effusion metastases, with no clear trend over time. Adjuvant endocrine therapy is associated with receptor conversion in pleural metastases Adjuvant endocrine therapy was given to 48% (33/69) of the patients, while 45% (31/69) received adjuvant chemotherapy. For 12% (8/69) of patients (45% or 5/11 of HER2 amplified cases), trastuzumab was prescribed.

Table 4. Differences in HER2 immunohistochemistry and FISH for 2+/3+ or discordant cases. Primary tumor

Solid metastasis

IHC

FISH

3+ 3+ 3+

amp

IHC

FISH

amp

no amp

amp

no amp

amp

IHC

Effusion metastasis

FISH

amp

no amp

amp

no amp

amp amp

252

no amp

Receptor conversion in breast cancer metastases in effusions

ER over time

Nuclear staining percentage

100

Adjuvant endocrine therapy No adjuvant endocrine therapy

2000

Time in days

4000

6000

PR over time

Nuclear staining percentage

100

Adjuvant endocrine therapy No adjuvant endocrine therapy

2000

Time in days

4000

6000

AR over time

Nuclear staining percentage

100

Adjuvant endocrine therapy No adjuvant endocrine therapy

2000

Time in days

4000

6000

HER2 over time

DAKO-score

Adjuvant trastuzumab therapy No adjuvant trastuzumab therapy 0

2000

Time in days

4000

6000

Figure 2. Expression of ERÎą, PR, AR and HER2 in paired primary breast tumors and multiple pleural or peritoneal metastases per patient: progression over time. Grey: adjuvantly treated with endocrine/trastuzumab therapy. Black: not adjuvantly treated with endocrine/trastuzumab therapy.

253

PART TWO | CHAPTER 8

Patients adjuvantly treated with endocrine therapy showed more often conversion of ERα (p=0.006 or p=0.058 for the 10% or 1% thresholds for positivity, respectively) and AR (p=0.001 or p<0.001), but not of PR (p=0.060 or p=0.130). Adjuvant chemotherapy did not show such association (ERα: p=0.835 or p=0.271, PR: p=0.383 or p=0.156 and AR: p=0.557 or p=0.927 for the 10% or 1% thresholds for positivity, respectively). For HER2, adjuvant trastuzumab treatment also influenced the change of receptor status (p<0.001). Again, this effect was not seen for chemotherapy (p=0.117).

DISCUSSION Luminal breast cancer is hallmarked by expression and growth dependency on ERα, which represents one of the cornerstones of adjuvant therapy in the treatment of breast cancer. Now that guidelines allow for the use of at least 5 years of endocrine therapeutics 27, and even 10 years for a subset of patients 28, it is not unlikely that such continuous and longitudinal ERα-inhibition would directly invoke a strong evolutionary pressure on the tumor. For approximately 30% of patients, metastatic relapse of the tumor is observed 29, which also implies that the tumor cells managed to survive and proliferate despite multiple years of ERα inhibition. Since the efficacy of adjuvant endocrine therapy was only established during the end of the last century, leading to an increase in prescription 30, studies including samples before that time could not confirm the evolutionary pressure of these therapies. Next to ERα, also PR, AR and HER2 are treatment targets in the battle against breast cancer. For second or higher lines of therapy, megestrol acetate 31, bicalutamide 32 and trastuzumab 33 are only a few examples of drugs that have demonstrated their clinical utility in breast cancer treatment and continuous research is being performed to develop and optimize new therapies. Furthermore, PR and AR also have the potential to predict response to ERαtargeted therapy; high PR expression in the presence 34 or even absence of ERα 35 is thought to predict an increased probability of benefit from anti-estrogen, while AR protein expression can induce tamoxifen resistance 36. Receptor conversion in solid distant metastases is now a well-known phenomenon and may lead to suboptimal treatment, and most guidelines now recommend to biopsy distant metastases at presentation of metastatic disease 16,37,38. Receptor status in malignant effusion specimens used to be determined only rarely, as in most cases characteristics of the primary tumor were deemed sufficient. However, in this study we demonstrated that receptor conversion in effusions is also a frequent phenomenon. Especially the high AR discordance we found is new and very relevant, since AR-targeted therapies are recently gaining interest for the treatment of ERα-negative and endocrine therapy resistant breast cancer 39. Even more interesting, in contrast to the high AR discordance in effusion metastases, relatively stable AR expression was described in solid metastases 18,19, while ERα, PR and HER2 254

Receptor conversion in breast cancer metastases in effusions

conversion showed roughly the same pattern in effusion and solid metastases 12-14. We show for the first time that receptor conversion for ERα and AR in malignant effusions was more often seen in patients adjuvantly treated with endocrine therapy and for HER2 in patients treated with trastuzumab. For ERα, PR and HER2 this was previously shown in primary breast cancer versus solid metastases 40-46. Conversion occurred most often from positive to negative and could be explained by outgrow of metastatic negative clones from the primary tumor under the selection pressure of prior therapies 47,48. However, ERαinhibitor induced AR conversion was not shown before and the mechanism of this finding remains elusive. Videlicet, tamoxifen and aromatase-inhibitors are not known to affect AR activity and are therefore not thought to inflict evident evolutionary selection pressure. Another explanation for conversion could be clonal dedifferentiation and selection of or evolution to more aggressive phenotypes 47,49-51. Also inadequate sampling of a heterogeneous tumor potentially leads to differences in receptor expression, which may explain some but clearly not all of the differences between primary breast cancer and metastases. Only one previous study reported ERα and PR receptor expression between primary breast tumors and 31 pleural effusion metastases, without mentioning treatment history. With expression rates of 59% and 51% for ERα and PR respectively and receptor conversion of 35% and 42%, these findings corresponded to our results mostly in relation to ERα 52. In patients with multiple effusion samples and solid metastases available, large variation in hormone and HER2 receptor expression occurred during tumor progression from primary tumor to solid and pleural metastases, with no clear trend over time. This could be explained by the different locations of metastases, since it is shown before that tumors with specific hormone expression patterns show a distinct dissemination pattern 47. Furthermore, most patients received multiple lines of therapy, potentially all imposing divergent evolutionary selection pressures on the metastatic cells. Future studies addressing types and number of therapies would yield priceless information about the influence of systemic drugs on tumor progression. Since generally less steroid receptor positivity was seen in peritoneal and pleural effusion metastases compared to their matched primary breast carcinomas, the question arises whether IHC staining on histologically processed cytology specimens is reliable. However, several studies have compared ERα, PR and HER2 status in cell blocks to tissue blocks and found high concordance rates between cytology and histology specimens 53,54. To prevent potential differences caused by such technical issues, we performed IHC staining on fresh FFPE cell block sections, used internal and external controls, assessed only the invasive (not the in situ) component and performed ISH on samples that scored 2+ or 3+ 55. Since heterogeneous expression is not uncommon 56, we included only samples containing at least 20 tumor cells. In summary, we have shown for the first time that ERα, PR, AR and HER2 expression in primary breast cancers is frequently lost in peritoneal and pleural effusion metastases. 255

PART TWO | CHAPTER 8

For ERα, PR and HER2 this is in line with previous findings in solid distant metastases, but AR conversion in late stage of tumor progression is a new observation. We demonstrate that this loss is most likely inflicted by the evolutionary selection pressure of adjuvant endocrine or targeted therapies, as such accounting for acquired therapy resistance. Since more than 35% of hormone receptor positive primary tumors convert to ERα and/or PR negative metastases or vice versa, determination of receptor status in malignant effusion specimens may help to optimize patient tailored hormonal treatment and is therefore recommended whenever possible. Especially the new finding of treatment-induced loss of AR protein expression as shown here, might have ramifications for clinical studies addressing AR-targeted therapies in metastatic breast cancer.

Acknowledgement This study is supported by Dutch Cancer Society grant UU 2011-5195 and Philips Consumer Lifestyle. The Dutch Distant Breast Cancer Metastases Consortium included the Departments of Pathology from the University Medical Center Utrecht, Meander Medical Center Amersfoort, Hospital Gelderse Vallei Ede, Academic Medical Center Amsterdam, Medical Center Alkmaar, Radboud University Nijmegen Medical Center, Canisius Wilhelmina Hospital Nijmegen, VU University Medical Center Amsterdam, The Netherlands Cancer Institute Amsterdam, Groene Hart Hospital Gouda, University Medical Center Groningen, St Antonius Hospital Nieuwegein, Diakonessenhuis Utrecht, Isala klinieken Zwolle, Erasmus Medical Center Rotterdam, Gelre Hospital Apeldoorn and the Laboratories for Pathology Dordrecht, ‘s Hertogenbosch, Terneuzen, Symbiant Zaandam, Sazinon, Hoogeveen, Oost Nederland Enschede (LabPON), all in The Netherlands. We thank Stichting PALGA for the national query for cases.

256

Receptor conversion in breast cancer metastases in effusions

REFERENCES 1. Ferlay J, Soerjomataram I, Dikshit R, et al. Cancer incidence and mortality worldwide: Sources, methods and major patterns in GLOBOCAN 2012. Int J Cancer. 2015;136(5):E359-86. doi: 10.1002/ijc.29210 [doi]. 2. Fentiman IS, Millis R, Sexton S, Hayward JL. Pleural effusion in breast cancer: A review of 105 cases. Cancer. 1981;47(8):2087-2092. 3. Raju RN, Kardinal CG. Pleural effusion in breast carcinoma: Analysis of 122 cases. Cancer. 1981;48(11):2524-2527. 4. Wilkes JD, Fidias P, Vaickus L, Perez RP. Malignancyrelated pericardial effusion. 127 cases from the roswell park cancer institute. Cancer. 1995;76(8):1377-1387. 5. Buck M, Ingle JN, Giuliani ER, Gordon JR, Therneau TM. Pericardial effusion in women with breast cancer. Cancer. 1987;60(2):263-269. 6. Zamboni MM, da Silva CT,Jr, Baretta R, Cunha ET, Cardoso GP. Important prognostic factors for survival in patients with malignant pleural effusion. BMC Pulm Med. 2015;15:29-015-0025-z. doi: 10.1186/s12890-015-0025-z [doi]. 7. Singer TS, Sulkes A, Biran S. Pleural effusion in breast cancer: Influence upon clinical course and survival. Chemioterapia. 1986;5(1):66-69. 8. Bielsa S, Salud A, Martinez M, et al. Prognostic significance of pleural fluid data in patients with malignant effusion. Eur J Intern Med. 2008;19(5):334-339. doi: 10.1016/j.ejim.2007.09.014 [doi]. 9. Di Lorenzo D, Zaniboni A, Simoncini E, et al. Estrogen and progesterone receptors in neoplastic cells of metastatic pleural effusion of breast carcinoma before and after tamoxifen therapy. correlation with the clinical response. Chemioterapia. 1986;5(4):232-236. 10. Shinohara T, Yamada H, Fujimori Y, Yamagishi K. Malignant pleural effusion in breast cancer 12 years after mastectomy that was successfully treated with endocrine therapy. Am J Case Rep. 2013;14:184-187. doi: 10.12659/ AJCR.889249 [doi]. 11. Watanabe T, Wu JZ, Morikawa K, et al. In vitro sensitivity test of breast cancer cells to hormonal agents in a radionucleotide-incorporation assay. Jpn J Cancer Res. 1990;81(5):536-543. 12. Hoefnagel LD, van de Vijver MJ, van Slooten HJ, et al. Receptor conversion in distant breast cancer metastases. Breast Cancer Res. 2010;12(5):R75. doi: 10.1186/bcr2645 [doi]. 13. Hoefnagel LD, Moelans CB, Meijer SL, et al. Prognostic value of estrogen receptor alpha and progesterone receptor conversion in distant breast cancer metastases. Cancer. 2012;118(20):4929-4935. doi: 10.1002/cncr.27518 [doi]. 14. Hoefnagel LD, van der Groep P, van de Vijver MJ, et al. Discordance in ERalpha, PR and HER2 receptor status across different distant breast cancer metastases within the same patient. Ann Oncol. 2013;24(12):3017-3023. doi: 10.1093/annonc/mdt390 [doi]. 15. Hammond ME, Hayes DF, Wolff AC, Mangu PB, Temin S. American society of clinical oncology/college of american pathologists guideline recommendations for

immunohistochemical testing of estrogen and progesterone receptors in breast cancer. J Oncol Pract. 2010;6(4):195-197. doi: 10.1200/JOP.777003 [doi]. 16. Carlson RW, Allred DC, Anderson BO, et al. Metastatic breast cancer, version 1.2012: Featured updates to the NCCN guidelines. J Natl Compr Canc Netw. 2012;10(7):821-829. doi: 10/7/821 [pii]. 17. Pietri E, Conteduca V, Andreis D, et al. Androgen receptor signaling pathways as a target for breast cancer treatment. Endocr Relat Cancer. 2016. doi: ERC-16-0190 [pii]. 18. Dâ&#x20AC;&#x2122;Amato NC, Gordon MA, Babbs B, et al. Cooperative dynamics of AR and ER activity in breast cancer. Mol Cancer Res. 2016. doi: molcanres.0167.2016 [pii]. 19. Grogg A, Trippel M, Pfaltz K, et al. Androgen receptor status is highly conserved during tumor progression of breast cancer. BMC Cancer. 2015;15:872-015-1897-2. doi: 10.1186/s12885-015-1897-2 [doi]. 20. Bertozzi S, Londero AP, Cedolini C, et al. Prevalence, risk factors, and prognosis of peritoneal metastasis from breast cancer. Springerplus. 2015;4:688-015-1449-x. eCollection 2015. doi: 10.1186/s40064-015-1449-x [doi]. 21. Apffelstaedt JP, Van Zyl JA, Muller AG. Breast cancer complicated by pleural effusion: Patient characteristics and results of surgical management. J Surg Oncol. 1995;58(3):173-175. 22. Nakayama Y, Nakagomi H, Omori M, et al. Benefits of using the cell block method to determine the discordance of the HR/HER2 expression in patients with metastatic breast cancer. Breast Cancer. 2016;23(4):633-639. doi: 10.1007/s12282-015-0615-x [doi]. 23. Arihiro K, Oda M, Ogawa K, et al. Discordant HER2 status between primary breast carcinoma and recurrent/ metastatic tumors using fluorescence in situ hybridization on cytological samples. Jpn J Clin Oncol. 2013;43(1):55-62. doi: 10.1093/jjco/hys187 [doi]. 24. van Diest PJ. No consent should be needed for using leftover body material for scientific purposes. for. BMJ. 2002;325(7365):648-651. 25. Hammond ME. ASCO-CAP guidelines for breast predictive factor testing: An update. Appl Immunohistochem Mol Morphol. 2011;19(6):499-500. doi: 10.1097/PAI.0b013e31822a8eac [doi]. 26. Wolff AC, Hammond ME, Hicks DG, et al. Recommendations for human epidermal growth factor receptor 2 testing in breast cancer: American society of clinical oncology/college of american pathologists clinical practice guideline update. J Clin Oncol. 2013;31(31):39974013. doi: 10.1200/JCO.2013.50.9984 [doi]. 27. Fisher B, Dignam J, Bryant J, Wolmark N. Five versus more than five years of tamoxifen for lymph nodenegative breast cancer: Updated findings from the national surgical adjuvant breast and bowel project B-14 randomized trial. J Natl Cancer Inst. 2001;93(9):684-690. 28. Goss PE, Ingle JN, Pritchard KI, et al. Extending aromataseinhibitor adjuvant therapy to 10 years. N Engl J Med. 2016;375(3):209-219. doi: 10.1056/NEJMoa1604700 [doi]. 29. Oâ&#x20AC;&#x2122;Shaughnessy J. Extending survival with chemotherapy in metastatic breast cancer. Oncologist. 2005;10 Suppl 3:20-29. doi: 10/suppl_3/20 [pii]. 30. Johnson TP, Ford L, Warnecke RB, et al. Effect of a

257

PART TWO | CHAPTER 8

national cancer institute clinical alert on breast cancer practice patterns. J Clin Oncol. 1994;12(9):1783-1788. 31. Bines J, Dienstmann R, Obadia RM, et al. Activity of megestrol acetate in postmenopausal women with advanced breast cancer after nonsteroidal aromatase inhibitor failure: A phase II trial. Ann Oncol. 2014;25(4):831-836. doi: 10.1093/annonc/mdu015 [doi]. 32. Arce-Salinas C, Riesco-Martinez MC, Hanna W, Bedard P, Warner E. Complete response of metastatic androgen receptor-positive breast cancer to bicalutamide: Case report and review of the literature. J Clin Oncol. 2016;34(4):e21-4. doi: 10.1200/JCO.2013.49.8899 [doi]. 33. Hortobagyi GN. Trastuzumab in the treatment of breast cancer. N Engl J Med. 2005;353(16):1734-1736. doi: 353/16/1734 [pii]. 34. Luoh SW, Ramsey B, Park B, Keenan E. Quantitative progesterone receptor expression and efficacy of antiestrogen therapy in breast cancer. Breast J. 2014;20(1):4652. doi: 10.1111/tbj.12200 [doi]. 35. Shen T, Brandwein-Gensler M, Hameed O, Siegal GP, Wei S. Characterization of estrogen receptor-negative/ progesterone receptor-positive breast cancer. Hum Pathol. 2015;46(11):1776-1784. doi: 10.1016/j. humpath.2015.07.019 [doi]. 36. De Amicis F, Thirugnansampanthan J, Cui Y, et al. Androgen receptor overexpression induces tamoxifen resistance in human breast cancer cells. Breast Cancer Res Treat. 2010;121(1):1-11. doi: 10.1007/s10549-009-0436-8 [doi]. 37. Van Poznak C, Somerfield MR, Bast RC, et al. Use of biomarkers to guide decisions on systemic therapy for women with metastatic breast cancer: American society of clinical oncology clinical practice guideline. J Clin Oncol. 2015;33(24):2695-2704. doi: 10.1200/ JCO.2015.61.1459 [doi]. 38. Cardoso F, Costa A, Norton L, et al. 1st international consensus guidelines for advanced breast cancer (ABC 1). Breast. 2012;21(3):242-252. doi: 10.1016/j. breast.2012.03.003 [doi]. 39. Garay JP, Park BH. Androgen receptor as a targeted therapy for breast cancer. Am J Cancer Res. 2012;2(4):434445. 40. Curtit E, Nerich V, Mansi L, et al. Discordances in estrogen receptor status, progesterone receptor status, and HER2 status between primary breast cancer and metastasis. Oncologist. 2013;18(6):667-674. doi: 10.1634/ theoncologist.2012-0350 [doi]. 41. Duchnowska R, Dziadziuszko R, Trojanowski T, et al. Conversion of epidermal growth factor receptor 2 and hormone receptor expression in breast cancer metastases to the brain. Breast Cancer Res. 2012;14(4):R119. doi: 10.1186/bcr3244 [doi]. 42. Idirisinghe PK, Thike AA, Cheok PY, et al. Hormone receptor and c-ERBB2 status in distant metastatic and locally recurrent breast cancer. pathologic correlations and clinical significance. Am J Clin Pathol. 2010;133(3):416-429. doi: 10.1309/AJCPJ57FLLJRXKPV [doi]. 43. Jensen JD, Knoop A, Ewertz M, Laenkholm AV. ER, HER2, and TOP2A expression in primary tumor, synchronous axillary nodes, and asynchronous metastases

258

in breast cancer. Breast Cancer Res Treat. 2012;132(2):511521. doi: 10.1007/s10549-011-1610-3 [doi]. 44. Nakamura R, Yamamoto N, Onai Y, Watanabe Y, Kawana H, Miyazaki M. Importance of confirming HER2 overexpression of recurrence lesion in breast cancer patients. Breast Cancer. 2013;20(4):336-341. doi: 10.1007/ s12282-012-0341-6 [doi]. 45. Yonemori K, Tsuta K, Shimizu C, et al. Immunohistochemical profiles of brain metastases from breast cancer. J Neurooncol. 2008;90(2):223-228. doi: 10.1007/s11060-008-9654-x [doi]. 46. Bogina G, Bortesi L, Marconi M, et al. Comparison of hormonal receptor and HER-2 status between breast primary tumours and relapsing tumours: Clinical implications of progesterone receptor loss. Virchows Arch. 2011;459(1):1-10. doi: 10.1007/s00428-011-1097-7 [doi]. 47. Kuukasjarvi T, Karhu R, Tanner M, et al. Genetic heterogeneity and clonal evolution underlying development of asynchronous metastasis in human breast cancer. Cancer Res. 1997;57(8):1597-1604. 48. Rasbridge SA, Gillett CE, Seymour AM, et al. The effects of chemotherapy on morphology, cellular proliferation, apoptosis and oncoprotein expression in primary breast carcinoma. Br J Cancer. 1994;70(2):335-341. 49. Sari E, Guler G, Hayran M, Gullu I, Altundag K, Ozisik Y. Comparative study of the immunohistochemical detection of hormone receptor status and HER-2 expression in primary and paired recurrent/metastatic lesions of patients with breast cancer. Med Oncol. 2011;28(1):57-63. doi: 10.1007/s12032-010-9418-2 [doi]. 50. Amir E, Miller N, Geddie W, et al. Prospective study evaluating the impact of tissue confirmation of metastatic disease in patients with breast cancer. J Clin Oncol. 2012;30(6):587-592. doi: 10.1200/JCO.2010.33.5232 [doi]. 51. Edgerton SM, Moore D,2nd, Merkel D, Thor AD. erbB-2 (HER-2) and breast cancer progression. Appl Immunohistochem Mol Morphol. 2003;11(3):214-221. 52. Schwarz C, Lubbert H, Rahn W, Schonfeld N, Serke M, Loddenkemper R. Medical thoracoscopy: Hormone receptor content in pleural metastases due to breast cancer. Eur Respir J. 2004;24(5):728-730. doi: 24/5/728 [pii]. 53. Moriki T, Takahashi T, Ueta S, Mitani M, Ichien M. Hormone receptor status and HER2/neu overexpression determined by automated immunostainer on routinely fixed cytologic specimens from breast carcinoma: Correlation with histologic sections determinations and diagnostic pitfalls. Diagn Cytopathol. 2004;30(4):251-256. doi: 10.1002/dc.20007 [doi]. 54. Shabaik A, Lin G, Peterson M, et al. Reliability of Her2/ neu, estrogen receptor, and progesterone receptor testing by immunohistochemistry on cell block of FNA and serous effusions from patients with primary and metastatic breast carcinoma. Diagn Cytopathol. 2011;39(5):328-332. doi: 10.1002/dc.21389 [doi]. 55. Fetsch PA, Abati A. The effects of antibody clone and pretreatment method on the results of HER2 immunostaining in cytologic samples of metastatic breast cancer: A query and a review of the literature. Diagn Cytopathol. 2007;35(6):319-328. doi: 10.1002/dc.20638 [doi].

Receptor conversion in breast cancer metastases in effusions

56. Cormier JN, Hijazi YM, Abati A, et al. Heterogeneous expression of melanoma-associated antigens and HLA-A2 in metastatic melanoma in vivo. Int J Cancer. 1998;75(4):517-524. doi: 10.1002/(SICI)10970215(19980209)75:4<517::AID-IJC5>3.0.CO;2-W [pii].

259

100

HER2

FISH

amp

no amp

100

PR 5

100

AR 90

HER2 0

FISH amp

amp

no amp

Location gynaecological

skin

gynaecological

skin

liver

skin

bone

skin

bone

lung

ER 100

100

Solid metastasis 2

100

Solid metastasis 1 HER2 0

FISH amp

amp

Location GI

skin

lung

liver

skin

Effusion 1

100

Primary tumor

Patient

AR 5

HER2 0

amp

no amp

amp

no amp

FISH

260 peritoneal

pleural

peritoneal

pleural

peritoneal

pleural

peritoneal

pleural

Location

Supplementary Table S1. Staining percentages of ERα, PR, AR, HER2 IHC and HER2 FISH on paired primary breast tumors and multiple pleural or peritoneal metastases per patient.

PART TWO | CHAPTER 8

no amp

pleural

peritoneal

pleural

peritoneal

pleural

peritoneal

pleural

no amp

HER2

FISH

100

Location

ER 75

AR 0

HER2 0

FISH no amp

Location

pleural

Effusion 4

Effusion 3

Effusion 2 HER2 0

AR 1

FISH no amp

Location

pleural

Effusion 5

Patient

Supplementary Table S1. Continued

AR 0

HER2 1+

FISH no amp

Location pleural

pleural

Receptor conversion in breast cancer metastases in effusions

261

Chapter 9 Willemijne AME Schrijver*, Karianne Schuurman*, Suzan Stelloo, Annelot van Rossum, Ekaterina Nevedomskaya, Marjolein Droog, Thomas Kuilman, Oscar Krijgsman, Onno B Bleijerveld, Liesbeth Hoekman, AF Maarten Altelaar, Jelle Wesseling, Lodewyk FA Wessels, Cathy B Moelans, Daniel Peeper, Sabine C Linn, Michel M van de Heuvel, Paul J van Diest, Wilbert Zwart * Both authors contributed equally to this study

Loss of FOXA1 expression in metastatic breast cancer drives acquired endocrine therapy resistance Submitted

PART TWO | CHAPTER 9

ABSTRACT Breast cancer is treated with hormonal therapies to block Estrogen Receptor alpha (ERα) action, but resistance is common. In metastatic disease, a rationale-based drug selection is lacking. Here, we report a novel mechanism of acquired endocrine therapy resistance in metastatic breast cancer, hallmarked by progressive loss of ERα-pioneer factors: GATA3 and FOXA1. Using mass spectrometry, we identified the ERα interactome in breast cancer metastases to the pleural cavity. In disease progression, the ERα protein complex disintegrated while the receptor remained expressed, which was accompanied with a loss of ERα chromatin interactions, as determined by ChIP-seq. These findings could be explained by a loss of FOXA1 expression, exclusively for patients who had received endocrine therapy in the adjuvant setting. FOXA1 loss associated with hormone irresponsiveness in metastatic tumor explants and treatment resistance in metastatic disease. With this, we position FOXA1-loss as a novel biomarker and causally involved in acquired endocrine treatment resistance in metastatic ERα-positive breast cancer.

264

Pioneer factor loss in endocrine resistance

INTRODUCTION Breast cancer is the most common malignancy in women with 1.7 million newly diagnosed cases annually worldwide and over 550,000 patients dying as a consequence of the disease each year 1. Around 70% of all breast tumors are of the luminal subtype, in which Estrogen Receptor alpha (ERα) is considered the main driver of cell proliferation, and consequently the prime drug target in treatment with curative intend 2,3. These patients are generally treated with endocrine therapies in the adjuvant setting; tamoxifen or aromatase inhibitors. Both types of drugs prevent ERα-driven gene transcription, subsequently blocking tumor cell proliferation and tumor progression. Despite an impressive reduction in recurrence risk, relapses do occur after adjuvant endocrine therapy 4. These relapses may either be due to intrinsic or acquired drug resistance 5. Several intrinsic resistance mechanisms to hormonal intervention have been described, including activation of kinase pathways 6-8 and overexpression of co-regulators that are involved in ERα function 9,10. All these resistance mechanisms enable tumor cell proliferation despite hormonal therapy, ultimately giving rise to outgrowth of metastatic breast cancer 11. Over the last years, acquired resistance to therapy in the metastatic setting has gained scientific attention. Since metastatic cancer cells of ERα-positive breast cancer are likely to have evolved to proliferate and survive in presence of endocrine therapeutics, it may not seem surprising that these resistance mechanisms can strongly differ from those responsible for intrinsic treatment resistance 5. Recently, several acquired resistance mechanisms have been identified, including activating mutations within the ESR1 gene (around 20% of cases) 12,13 and loss of ERα expression in metastases (termed receptor conversion; 6-25% of cases) 14 . Cumulatively, the currently known (genomic) aberrations only explain about 40% of endocrine resistant cases in the metastatic setting, and other mechanisms of acquired resistance to hormonal therapies are likely to exist. FOXA1 and GATA3, both well-recognized as luminal breast cancer-defining genes, play crucial roles in genomic functions of ERα. FOXA1 is required for ERα chromatin interactions by rendering the chromatin accessible at designated ERα binding sites 15. GATA3 facilitates FOXA1 chromatin interactions and directly affects chromatin loops that involve ERα 16. Jointly, FOXA1 and GATA3 are essential and sufficient to enable ERα chromatin interactions, responsive gene activation and cell proliferation 17. Here, we uncover a novel mechanism of acquired resistance to endocrine therapy for metastatic breast cancer. Using RIME (Rapid Immunoprecipitation Mass spectrometry of Endogenous proteins) 18 on pleural effusions from breast cancer patients, we characterized endogenous ERα-associated protein complexes in human tumor samples. During metastatic disease progression, a progressive disintegration of the ERα-protein complex was observed that coincided with a loss of ERα-chromatin interactions, decreased RNA Polymerase II recruitment and a loss of chromatin accessibility at these sites. Even though ERα expression 265

PART TWO | CHAPTER 9

persisted, expression of essential ERα-pioneer factors FOXA1 and GATA3 was lost, providing a mechanistic explanation for the loss of ERα functionality. No ERα activity was observed in tumors lacking FOXA1 expression, which coincided with endocrine therapy resistance in these patients. With this, we present a novel mechanism of acquired anti-estrogen resistance in metastatic breast cancer and highlight loss of FOXA1 expression as novel biomarker for endocrine treatment resistance in ERα-positive metastatic breast cancer.

METHODS Pleural tumor cell collection, isolation and processing Pleural fluid effusion specimens of female breast cancer patients were collected prospectively (n=17) and retrospectively (n=96). When available, the paired FFPE material of the primary breast tumor was collected (n=52) (Supplementary Table S1). From 24 patients, pleural effusion samples of multiple time points were obtained, but only the ERα-positive samples were included in this study. Prospectively, fluid isolated from the pleural cavity was collected directly after drainage from patients at the University Medical Center Utrecht and the Netherlands Cancer Institute in Amsterdam. The cells in the pleural effusions were isolated by centrifugation, the erythrocytes were lysed (erythrocyte lysis buffer [pH 7.4]: 75 mM NH4Cl, 5 mM KHCO3, 400 uL 500mM EDTA and 500 mL ddH2O) and remaining cells were either formalin-fixed and paraffin embedded (Cellient, Hologic, 40180I10D0) or cryo-stored at -80 (DMSO). Original diagnoses were made in 2014 and 2015. The retrospectively obtained specimens (FFPEs) were collected in hospitals across the Netherlands: the Groene Hart Hospital Gouda, the Onze Lieve Vrouwe Gasthuis (Amsterdam), the Orbis Medical Center (Sittard), the Medical Center Leeuwarden, the University Medical Center St. Radboud (Nijmegen), the Meander Medical Center (Amersfoort), the Atrium Medical Center (Heerlen), the VieCuri Medical Center (Venlo), the Leiden University Medical Center, Bronovo Hospital (The Hague), Canisius Wilhelmina Hospital, Rijnstate Hospital (Arnhem), Diakonessenhuis (Utrecht), Isala Clinics (Zwolle), St. Franciscus Gasthuis (Rotterdam), Amphia Hospital (Breda), the Netherlands Cancer Institute (Amsterdam) and the University Medical Center Utrecht. Original diagnoses were made between 1988 and 2013. Immunohistochemistry staining was performed on both pleural effusion samples and tissue microarrays (TMAs) 19 from primary tumor and solid metastases. Three core biopsies (0.6 mm in diameter) of histologically representative regions of each tumor were used to construct the TMAs. These TMAs consisted of samples of 97 tumors. For each FFPE sample, haematoxylin-eosin stained slides of the paraffin blocks were reviewed by a single 266

Pioneer factor loss in endocrine resistance

experienced breast pathologist (PvD) to confirm the presence of malignancy. Only samples containing at least 20 tumor cells were selected. Ber-EP4 monoclonal antibody staining (1:40, CCI24, DAKO) that labels epithelial tissues, but does not react with mesothelial cells, was used to confirm presence and quantity of tumor cells in the effusions. This study was performed in accordance with the institutional medical ethical guidelines. The use of anonymous or coded left over material for scientific purposes is part of the standard treatment agreement with patients and therefore informed consent was not required according to Dutch law 20 for patients not actively opting out. Cell processing for mass spectrometry, ChIP and FAIRE Cell pellets were defrosted in solution A (50 mM Hepes, 100 mM NaCl, 1mM EDTA, 0.5 mM EGTA) with 1% formaldehyde, and fixed for 20 minutes. Subsequently, fixation was quenched with 0.2M Glycine. After washing with cold PBS, the nuclear fraction was extracted using LB1 buffer (50 mM HEPES-KOH [pH 7.5], 140 mM NaCl, 1 mM EDTA, 10% glycerol, 0.5% NP-40 or Igepal CA-630, and 0.25% Triton X-100) for 10 min at 4°C. Cells were pelleted and resuspended in 10 ml of LB2 buffer (10 mM Tris-HCL [pH 8.0], 200 mM NaCl, 1 mM EDTA, and 0.5 mM EGTA), and mixed at 4°C for 5 min. Subsequently, cells were pelleted and resuspended in 300 μl of LB3 buffer (10 mM Tris-HCl [pH 8], 100 mM NaCl, 1 mM EDTA, 0.5 mM EGTA, 0.1% Na-deoxycholate, and 0.5% N-lauroylsarcosine) and sonicated in a water bath sonicator (Diagenode Bioruptor Pico). A final volume of 1% Triton X-100 was added, and lysate was centrifuged to purify the debris. Formaldehyde-Assisted Isolation of Regulatory Elements (FAIRE) was performed as previously described 21. The lysate was subjected to three consecutive phenol-chloroformisoamyl alcohol (25:24:1) extractions using 2 ml Phase Lock Gel Light tubes (FPR5101, Flowgen Bioscience). For RIME and ChIP, the supernatant was incubated with 100 μl of magnetic beads (Dynabeads) prebound with 10 μg ER antibody, and immunoprecipitation (IP) was conducted overnight at 4°C. The beads were washed ten times in 1 ml of RIPA buffer. For ChIP and FAIRE, samples were reverse cross-linked overnight at 65°C, treated with RNase A and proteinase K and purified by ethanol precipitation. For mass spectrometry, RIPA-washed beads were washed twice in 100 mM ammonium hydrogen carbonate solution, and beads were stored at -20°C. Mass spectrometry Tryptic digestion (overnight, 370C) of bead-bound proteins was performed by adding 10 µL trypsin solution (mass spec grade, Promega; 10 µg/mL in 100 mM ammonium bicarbonate) to the washed beads. The magnetic beads were collected, and peptidecontaining supernatants were removed and transferred to formic acid aliquots at a final acid concentration of 5%. Tryptic digests were desalted using C18 StageTips (Thermo 267

PART TWO | CHAPTER 9

Scientific) and peptides were dried down in a speed vacuum centrifuge. Prior to mass spectrometry analysis, the peptides were reconstituted in 10% formic acid. Peptides were separated using the Proxeon nLC 1000 system (Thermo Scientific, Bremen) fitted with a trapping (ReproSil-Pur 120 C18-AQ 3µm (Dr. Maisch GmbH, Ammerbuch, Germany); 100 µm x 200 mm) and an analytical column (Agilent Poroshell EC-C18 120 2.7 µm (Agilent Technologies); 50 µm x 500 mm), both packed in-house. The outlet of the analytical column was coupled directly to a Thermo Orbitrap Fusion hybrid mass spectrometer (Q-OT-qIT, Thermo Scientific) using the Proxeon nanoflex source. Nanospray was achieved using a distally coated fused silica tip emitter (generated in-house, o.d. 375 µm, i.d. 20 µm) operated at 2.1 kV. Solvent A was 0.1% formic acid/water and solvent B was 0.1% formic acid/acetonitrile. Samples (25% of total digest) were eluted from the analytical column at a constant flow of 100 nl/min in a 55-min gradient, containing a 30min linear increase from 7% to 60% solvent B, followed by a 23-min wash at 80% solvent B. Survey scans of peptide precursors from m/z 375-1500 were performed at 120K resolution with a 4 x 105 ion count target. Tandem MS was performed by quadrupole isolation at 1.6 Th, followed by HCD fragmentation with normalized collision energy of 33 and ion trap MS2 fragment detection. The MS2 ion count target was set to 104 and the max injection time was set to 50 ms. Only precursors with charge state 2-6 were sampled for MS2. Mono-isotopic precursor selection was turned on; the dynamic exclusion duration was set to 30s with a 10 ppm tolerance around the selected precursor and its isotopes. The instrument was run in top speed mode with 3s cycles. Mass spectrometry data analysis Raw data files were processed using Proteome Discoverer (version 1.4.1.14, Thermo Fisher Scientific). MS2 spectra were searched against the Swissprot database (release 2014_08, 546238 entries) using Mascot (version 2.4.1, Matrix Science, UK) and Homo sapiens as taxonomy filter. Carbamidomethylation of cysteines was set as a fixed modification and oxidation of methionine was set as a variable modification. Trypsin was specified as enzyme and up to two miscleavages were allowed. Data filtering was performed using percolator, resulting in 1% peptide false discovery rate (FDR). Additional filters were search engine rank 1 peptides and ion score >20. MS-ARC visualizations were generated as described previously 18. Proteins were classified based on Gene Ontology (GO) terms, and colored accordingly. The length of the line depicted for each protein is proportional to the square root of the Mascot score of the protein. Labels were drawn for proteins that had high Mascot scores or that were of particular biological relevance.

268

Pioneer factor loss in endocrine resistance

ChIP-sequencing analyses Immunoprecipitated DNA was processed for library preparation as described 22. Samples were sequenced using an Illumina Hiseq 2000 genome analyzer (50bp reads), and subsequently aligned to the Human Reference Genome (assembly hg19, February 2009). Reads were filtered based on MAPQ quality: only reads with MAPQ ≥ 20 were considered to eliminate reads from repetitive elements. Peak calling over input control was performed using DFilter 23 and MACS peak caller version 1.4 24. Only peaks called by both algorithms were used for the analysis. MACS was run with the default parameters, except p=10-7 for ChIP-seq data. DFilter was run with bs=100, ks=50 for FAIRE data and bs=50, ks=30, refine, nonzero for ChIP data. Read counts and number of aligned reads are shown in Supplementary Table S2. For heat map visualizations, SeqMINER was used 25. Other data visualization was performed using IGV 26. Motif analyses were performed using Galaxy Cistrome (cistrome.org), applying the SeqPos motif tool. DNA copy number profiles were generated with CopywriteR 27 using sequence files obtained from input samples of ChIP-seq experiments 27. A bin size of 20kb was used for analysis and DNA copy number profiles were segmented using Circular Binary Segmentation (CBS) 28. Immunohistochemical analyses Sections of 4um were cut from the FFPE blocks of the pleural effusions, primary breast tumors and the TMAs. IHC was performed with the Ventana autostainer (Roche, Tucson, USA) according to the manufacturer’s instructions. Mouse monoclonal antibodies used were against ERα (RTU, SP1; Roche, Tucson, USA), FOXA1 (1:100.000, WMAB-2F83, Seven Hills Bioreagents), GATA3 (1:300, 5852 Cell Signaling Technology, Bioke) and BerEP4 (Epcam, 1:40, CCI24, DAKO). Appropriate controls were used throughout. Scoring of IHC slides was performed by two observers (PvD & WS) blinded to clinicopathologic and molecular data. The percentage of positively stained nuclei was estimated side by side with the Ber-EP4 stained slide as a reference. For the TMAs, the mean score of all three cores was used as the final score per tumor. RNA and DNA isolation 10-µm slides were cut from the FFPE blocks of the pleural effusions and primary breast tumors. The number of slides depended on tumor cellularity, but on average 10 slides for DNA and 4 slides for RNA isolation were used. The slides were deparaffinized using Xylene and the H&E-section was used to guide macro-dissection for DNA and RNA extraction and to estimate tumor percentage. Tumor areas were macro-dissected using a scalpel and areas with necrosis, dense lymphocytic infiltrates, and pre-invasive lesions were intentionally avoided. For DNA isolation, the dissected tissue was put in 1M NaSCN overnight, after which proteinase K-based DNA extraction was performed with the QIAamp DNA FFPE tissue 269

PART TWO | CHAPTER 9

kit (Qiagen, Duesseldorf, Germany), according to the manufacturer’s instructions. RNA isolation was performed using the miRNeasy FFPE Kit (Qiagen, Duesseldorf, Germany). Total DNA and RNA concentrations were measured spectrophotometrically (Nanodrop ND-1000, Thermo Scientific Wilmington, USA) and fluorometrically (Qubit 2.0 Fluorometer, Life Technologies), the latter with the Qubit dsDNA HS and RNA assay kits (Life Technologies). The RNA Integrity Number (RIN) was determined with the Agilent RNA 6000 Nano kit on the Bioanalyzer (Agilent). Methylation-specific PCR (MSP) Sodium bisulfite conversion was performed using the Epitect bisulfite kit (Qiagen) according to the manufacturer’s instructions (Sodium Bisulfite Conversion of Unmethylated Cytosines in DNA Isolated from FFPE Tissue Samples; protocol version 09/2009) with an input of 500 ng DNA. Methylated and unmethylated Epitect control DNA were used as a positive and negative control. TCGA data portal was used to identify the CpG sites of GATA3 and FOXA1 with the best correlation to mRNA expression and with a position near the promoter region (http:// gattaca.imppc.org:3838/wanderer/index.html). For FOXA1, cg03772350 (CpG-island 99, Spearman r: -0.701) and cg23664186 (CpG-island 143, Spearman r: -0.554) met these criteria and for GATA3, cg10089865 (CpG-island 33, Spearman r: -0.64) and cg04213746 (exon 4, Spearman r: -0.79) were chosen. Primers were designed using MethPrimer. For the primer sets and PCR conditions see Supplementary Table S3. 10 µL of each PCR product was loaded onto a 2% agarose gel containing Midori Green, and visualized by UV illumination. Analyses were performed using Image Lab 5.1 (Bio-Rad Laboratories). Statistical analyses IHC for ERα, FOXA1 and GATA3 positivity was assessed with a 1% thresholds for positivity (ASCO guidelines) 29, regardless of staining intensity. Percentages of nuclei expressing ERα, FOXA1 and GATA3 in primary tumors and their metastases in pleural effusions and solid tissues were compared by Wilcoxon signed rank test. Differences between hormonal treatment regimens were determined by Mann-Whitney U test and correlations between FOXA1 and GATA3 loss with Spearman’s Rho. Dichotomization for FOXA1 and GATA3 expression in treated and untreated patients was performed with ROC curves and compared with Fisher’s exact test. Progression during treatment after first pleural effusion was visualized using Kaplan Meier survival plots. All statistical analyses were performed using IBM SPSS Statistics version 21 and GraphPad Prism 6 (GraphPad Software, USA).

270

Pioneer factor loss in endocrine resistance

RESULTS Proteomic isolation of ERα-associated proteins from pleural effusions The ERα interactome represents a multi-megadalton protein complex encompassing a large number of co-regulators, histone modifiers and interacting transcription factors 30. Recently, a novel technology was developed that enables high-throughput identification of ERαinteracting proteins in breast cancer cell lines, termed: RIME (Rapid Immunoprecipitation Mass spectrometry of Endogenous proteins) 18. The RIME protocol makes use of formaldehyde crosslinking to stabilize protein-protein interactions in intact cells, followed by chromatin shearing and immunoprecipitation, after which trypsin-digested proteins are analyzed through mass spectrometry. We successfully translated this approach for applicability on human tumor material, in this case on metastatic breast cancer cells isolated from pleural effusions (Figure 1A). RIME was applied for ERα on tumor cells from pleural fluid of three patients, with IgG as a negative control. ERα peptide count varied per patient, but was consistently enriched over the IgG control that lacks peptides for ERα. Multiple known ERα interactors were identified, including HSP90 31, HSPA8 32 and GREB1 18 (Figure 1B, peptide coverage in Figure 1C), which could be validated using standard co-immunoprecipitations (Figure 1D). Note that GREB1 was not detected as ERα-interacting protein for tumor 2 in the RIME analysis (Figure 1B), which was explained by a lack of GREB1 expression in this tumor (Figure 1D). To minimize noise, ERα-interacting proteins in which any peptide was found in the IgG control were excluded for further analyses. Data reproducibility was high, based on two technical replicates from the same patient sample, (peptide score: Spearman’s rho= 0.895 (Supplementary Figure S1A, B)). Figure 1E illustrates the total interactome of ERα in pleural metastatic breast cancer related to protein function by grouping proteins according to GO annotations. Between two out of three pleural effusions, 81 ERα-interacting proteins were shared (Figure 1F), showing enrichment for the ERα action pathway (Ingenuity analysis: p=8.72*10-4, Figure 1G). Eighty-five out of the 235 proteins that were found interacting with ERα in the clinical specimens were previously identified as ERαinteractors in cell lines (MCF-7, T47D and tamoxifen-resistant MCF-7) using the same approach 18 (Supplementary Figure S2). Progressive loss of Estrogen Receptor complex formation upon tumor progression Three sequential pleural effusion samples were harvested from a patient with progressive metastatic breast cancer, within a period of 36 days (Figure 2A). Tumor cell percentage was initially 70% and remained between 70-80% during all sequential effusions. Immunohistochemical analyses confirmed ERα expression as consistently high and unaltered over time (Figure 2B). Throughout the course of sample collection, no significant changes in DNA copy number could be observed (Figure 2C, Supplementary Figure S3)27. 271

PART TWO | CHAPTER 9

Figure 1. Endogenous protein mass spectrometry using breast cancer pleural effusion samples A. Overview of RIME analyses on pleural effusions. Fluid from pleural cavity is collected, and cells isolated through centrifugation. Cells are formaldehyde-fixed, chromatin sheared, immunoprecipitated for ERα or IgG control and further processed for mass spectrometry. B. Top: Heatmap visualization of Peptide Spectrum Match (PSM) values for each protein identified by mass spectrometry, and ranked on PSM values for proteins found in ERα experiment. Average IgG values are shown. Positions for ESR1, HSP90 and GREB1 are indicated. Bottom: PSM values for ESR1, HSP90 and GREB1. IgG and ERα IP are shown separately. C. Peptide coverage from the mass spectrometry analyses of ERα, HSP90 and GREB1 proteins. The locations of peptides identified are shown in blue. uuu

272

Pioneer factor loss in endocrine resistance

Even though ERα remained equally expressed (Figure 2B), RIME analyses revealed a strong qualitative and quantitative decrease of ERα interactors between the first and second time point in disease progression (Figure 2D, E, F). Loss of Estrogen Receptor chromatin interactions due to loss of FOXA1 and GATA3 ERα recruits many of its interacting proteins while bound to the DNA 33. Since the ERα complex disintegrated during disease progression (Figure 2), we characterized ERα chromatin interactions for all available time points (Figure 3). Using chromatin immunoprecipitation followed by massive parallel sequencing (ChIP-seq) for ERα and its pioneer factor FOXA1 34, RNA Polymerase II and promoter-selective histone modification H3K4me3 (which stimulates ERα-mediated transcription 30), we assessed functionality of the ERα complex, using the first pleural effusion sample (Figure 3A). ERα binding sites were enriched for all these factors at time point zero (Figure 3B), indicating functionality of the complex. However, while ERα/chromatin interactions were readily observed in the first pleural effusion, binding was progressively lost over time, as exemplified at the RARA locus (Figure 3C) and shown on a genome-wide scale (Figure 3D, Figure 3E). Interestingly, active enhancer mark H3K27Ac 35 remained unchanged at these sites (Figure 3E), while chromatin binding of FOXA1 and RNA Polymerase II were lost (Figure 3F) along with diminished chromatin accessibility at these regions, as shown by FAIRE-qPCR for GREB1 and RARA enhancers (Figure 3G).

ttt

Figure 1. Continued

D. Co-immunoprecipitation for ERα and IgG, using sample from tumor 3 and 4. For tumor 3, ERα, HSP90, TRIM28, HSPA8 and GREB1 were tested for immunoprecipitation. For tumor 4, ERα and GREB1 were analyzed. E. Mass Spectrometry-Accurate Radioisotope Counting (MS-ARC) visualization of mass spectrometry results. Data are grouped according to Gene Ontology annotations. Line length indicates PSM score. F. Venn diagram depicting partial overlap of interacting proteins in pleural effusion specimens of 3 patients. G. Pathway analysis of the 81 shared ERα-interacting proteins shown in F. Enrichment of ERα pathway is seen (Ingenuity pathway analysis: p=8.72*10-4).

273

PART TWO | CHAPTER 9

Figure 2. Loss of ERα-interacting protein repertoire during tumor progression A. Treatment history of ERα-positive breast cancer patient, from whom three sequential pleural effusions were drawn. First effusion date was defined as ‘day 0’. B. Immunohistochemical analyses of ERα protein levels in pleural effusions. Tumor cell percentages as scored by the pathologist are indicated. C. DNA copy number profile of the sequential pleural effusions. Y-axis depicts copy number values (2log). D. Heat map depicting Peptide Spectrum Match (PSM) scores for ERα RIME experiments for day 0 and day 21. E. Line plot, illustrating decrease in PSM score for ERα RIME at day 0 versus day 21. F. MS-ARC visualization of peptide count for sequential effusions. Data are grouped according to Gene Ontology annotations. Line length indicates PSM score.

274

Pioneer factor loss in endocrine resistance

Figure 3. ERα/chromatin interactions in tumor progression A. Genome browser snapshot of ERα, FOXA1, H3K4me3 and RNA Polymerase II from the first pleural effusion sample of this patient. B. Graph depicting average read count for ERα, FOXA1, H3K4me3 and RNA Polymerase II ChIP-seq, focused on the ERα binding sites. Data are centered on the top of the peak (arrowhead), with a 2.5 kb window. C. Genome browser snapshot of ERα ChIP-seq from sequential pleural effusions. Y-axis shows tag count. Genomic coordinates are indicated. D. Heat map illustrating progressive loss of ERα binding sites in tumor progression. Genomic regions are selected based on ERα sites as ‘day 0’, and raw data of all time points is shown. All peaks are vertically aligned and depicted within a 5 kb window around the peak. E. Normalized read count data (CPM) at ERα binding sites for all three time points, separately analyzing ERα (top) and H3K27ac (bottom) ChIP-seq data, normalized on the first time point. Gray lines depict all genomic locations separately, purple line shows mean signal. F. Heat map depicting ChIP-seq for FOXA1 (blue) and RNA Polymerase II (green) at ERα binding sites. Time points ‘0’ and ‘21’ days were analyzed. G. FAIRE-qPCR analysis for ERα binding sites at enhancer regions for RARA and GREB1. Error bars illustrate SD values from triplicate measurements.

275

PART TWO | CHAPTER 9

Loss of FOXA1 and GATA3 as novel biomarkers for acquired endocrine therapy resistance During metastatic breast cancer progression, we observed progressive loss of ERα function while the receptor remained expressed (Figure 2). Since ERα action requires FOXA1 15 and GATA3 16, we determined their expression levels in paired primary and pleural metastatic tumor samples from 52 patients by immunohistochemistry (Figure 4; Supplementary Table S1). ERα expression was lost in 21% (11/52) of the patients, consistent with other reports 14. For tumors in which ERα remained expressed, no ESR1 mutations were observed at the previously reported hotspots in the receptor’s ligand-binding domain 36 (Supplementary Figure S4). We did however observe differences in GATA3 and FOXA1 levels: Although levels were retained in some cases (Figure 4A, top), expression was lost in others (Figure 4A, bottom). Taken into consideration all patients, paired analyses of FOXA1 expression in primary and metastatic samples from the same patients showed no significant loss in pleural effusion metastases (p=0.082). This was consistent with the retention of FOXA1 expression seen in solid metastases that arose from an ERα-positive breast cancer, regardless of the sites of metastasis 37. GATA3 however was significantly lost (p=0.004). Changes in FOXA1 levels typically co-occurred with changes in GATA3 (ρ=0.609, p<0.001; Figure 4B). Decreased expression of FOXA1 and GATA3 expression in metastatic samples could not be explained by promoter methylation (Supplementary Figure S5), or by a copy number loss of the GATA3 and FOXA1 loci (data not shown). Importantly, only when patients had received adjuvant endocrine therapy, FOXA1 expression was lost as compared to the primary tumor (Figure 4C), while this was not the case for GATA3. In unpaired analyses, FOXA1 expression in pleural effusions was significantly lower in patients who received adjuvant endocrine therapy as compared to the pleural samples of patients who did not receive any adjuvant endocrine therapeutics (p = 0.001; n=72; Figure 4D). This effect was not seen for GATA3 (p = 0.145; n=72). To determine whether loss of FOXA1 and GATA3 was indicative for non-functionality of ERα in these samples, pleural effusion-derived tumor cells were cultured for 3 days in the presence of estradiol, tamoxifen, estrogen receptor antagonist fulvestrant or DMSO control, after which qPCR was performed for GREB1, IGFBP4 and TFF1; All well-described target genes of ERα (Figure 4E) 18. In parallel to ex vivo hormonal exposure, IHC analyses were performed for ERα, FOXA1 and GATA3 after which the specimens were grouped as ERα+/ GATA3+/FOXA1+ and ERα+/GATA3-/FOXA1-. While gene expression was clearly induced by estradiol and suppressed by both anti-estrogens in the ERα+/FOXA1+/GATA3+ tumor cells, expression of all three ERα target genes was not affected by hormonal stimuli on ERα+/FOXA1-/GATA3- cells. However, since both FOXA1 and GATA3 were lost in the same samples, no functional distinction on the level of ERα action could be made between them. Loss of FOXA1 and GATA3 in progressive metastatic breast cancer renders ERα nonfunctional and would imply resistance to endocrine therapeutics in this setting. This was confirmed in 276

Pioneer factor loss in endocrine resistance

Figure 4. Loss of FOXA1 and GATA3 expression in ERα-positive metastatic breast cancer A. Immunohistochemical analyses of ERα, FOXA1 and GATA3 in primary breast tumors and matched pleural effusion-derived metastatic breast tumor cells. Top: FOXA1 and GATA3 remain expressed in pleural metastases; bottom: FOXA1 and GATA3 are completely lost. B. Quantification of IHC for FOXA1 and GATA3 in ERα-positive pleural effusion metastases compared to matched primary breast tumors. The difference in nuclear staining in primary tumors compared to their matched metastases in pleural effusion specimens is calculated for FOXA1 (ΔFOXA1) and GATA3 (ΔGATA3). Values are colored according to treatment history: green circle for patients that obtained adjuvant hormonal therapy (tamoxifen and/or aromatase-inhibitors), red square for patients that were not adjuvantly treated and blue triangle for unknown treatment history. N=29. C. Waterfall plot with quantification of IHC for FOXA1 (top) and GATA3 (bottom) in ERα-positive primary breast tumors compared to matched pleural effusion metastases. The difference in nuclear staining in primary tumors compared to their matched metastases in pleural effusion specimens is calculated. Values are colored according to treatment history: green for patients that obtained adjuvant hormonal therapy (tamoxifen and/or aromatase-inhibitors), red for patients that were not adjuvantly treated and blue for unknown treatment history. N=29. D. Nuclear staining percentages of FOXA1 (left) or GATA3 (right) in ERα-positive pleural effusions of breast cancer patients. Values are colored according to treatment history: green for patients that obtained adjuvant hormonal therapy (tamoxifen and/or aromatase-inhibitors) and red for patients that were not adjuvantly treated. N=72. E. Heatmap depicting qPCR results from primary cultures of pleural cells derived from ERα-positive metastatic breast cancer. After isolation, pleural metastatic cells were stained and scored for ERα, GATA3 and FOXA1. The cells were cultured for three days in presence of estradiol, tamoxifen, fulvestrant or DMSO control, after which a qPCR was performed for GREB1, IGFBP4 and TFF1. Color scale indicates induction (red) or repression (green) of ERα-responsive gene expression, relative to DMSO control. F. Kaplan Meier survival curves for time on first hormonal therapy after pleural drainage. Patients were divided in two groups according to nuclear staining percentage of the pleural effusion metastases. N=48

277

PART TWO | CHAPTER 9

patients receiving endocrine treatment for metastatic breast cancer. Loss of FOXA1 was significantly associated with a shorter time on endocrine therapy before switching to another type of treatment (Figure 4F, p=0.036; HR=0.567; 95% CI: 0.267-0.964; cut-off for positivity determined with ROC-curves), mostly due to disease progression. These findings could not be explained by clinicopathological differences between FOXA1- and FOXA1 + groups (Supplementary Table S4). For GATA3, this effect was not observed (p=0.566). In contrast to pleural effusions but in agreement with previous reports 37,38, FOXA1 levels were increased in solid metastases (Supplementary Figure S6A, S6E). This effect was most prominently observed in solid metastases to the skin (p=0.001; Supplementary Figure S6E). No difference in GATA3 level was found between primary breast cancer and solid metastases (Supplementary Figure S6B, 6F). Furthermore, FOXA1 and GATA3 levels in solid metastases were not associated with adjuvant hormonal therapy, both in unpaired (Supplementary Figure S6C, S6D) and paired (Supplementary Figure S6G, 6H) analyses. Cumulatively, these data illustrate that GATA3 and FOXA1 are both lost in progressive metastatic breast cancer to the pleural cavity. Loss of FOXA1 expression was particularly apparent for patients who received endocrine therapy in the adjuvant setting, suggesting an evolutionary pressure to lose ERα function in this setting. FOXA1 loss renders the tumor cells irresponsive to hormonal stimuli despite expressing ERα, providing an explanation for the observed endocrine therapy resistance in patients with ERα+/FOXA1-/GATA3- tumors. With this, we present a novel mechanism for acquired endocrine therapy resistance and therapy irresponsiveness for metastatic ERα-positive breast cancer.

DISCUSSION Metastatic disease is responsible for practically all breast cancer-related deaths. Years of continuous drug exposure in the adjuvant setting provides a massive evolutionary selection pressure on the tumor cells for evading the blocking effects of endocrine treatment 36,39. Recently, several reports identified a distinct set of activating point mutations within the ERα trans-activating domain, only found in ~20% of metastases with acquired endocrine therapy resistance 12,13. In another 6-25% of patients, ERα expression is lost in metastases while the primary tumor was ERα-positive 14. Still, these known mechanisms of endocrine therapy resistance recapitulate only about 40% of all resistance cases, leaving the majority of endocrine therapy resistant cases unexplained. In the current study, we present a thus far unknown mechanism of acquired endocrine therapy resistance, with evident biomarker potential. GATA3 and FOXA1 are both essential for ERα action, enabling its chromatin interactions and consequently responsive gene activity. As such, both GATA3 as well as FOXA1 are termed key-luminal breast cancer defining genes and are expressed in every ERα-positive 278

Pioneer factor loss in endocrine resistance

primary breast cancer 17. In pleural effusion metastases however, FOXA1 and GATA3 expression is lost under evolutionary selection pressure of continuous ERα blockade, resulting in a non-functional receptor with abrogated response to endocrine drugs. Since both FOXA1 and GATA3 are frequently lost in the same ERα-positive tumor, these findings imply either a codependence for their expression, or functional redundancy between them in facilitating ERα action. In clinical response to endocrine therapy, FOXA1 appeared as the best predictor for ERα-receptor functionality. Persistent FOXA1 expression was seen in solid metastases that arose from an ERα-positive breast cancer, regardless of the sites of metastasis 37. This is consistent with our findings in pleural effusion metastases of patients who did not receive adjuvant endocrine therapy. However, adjuvant endocrine therapy gives rise to ERα+/FOXA1-/GATA3- tumors that are irresponsive to hormonal signals. Such associations with treatment history were unexplored in previous studies. Several guidelines advise to reassess ERα status in metastases by immunohistochemistry29,40. This may identify loss of ERα expression, but does not provide any information on ERα functionality in case ERα expression is maintained. Our current study emphasizes that ERα assessment by immunohistochemistry is insufficient to predict hormonal treatment responsiveness in the metastatic setting. Our study positions loss of FOXA1 as novel biomarker for endocrine treatment resistance of metastatic breast cancer in pleural effusions. Since around 11% of breast cancer patients eventually present with symptomatic pleural effusions and at autopsy 36-65% of patients retrospectively suffered from this condition 41,42, assessment of FOXA1 levels in pleural metastases could facilitate tailored treatment for a large group of patients with metastatic disease.

Acknowledgements Supported by Dutch Cancer Society grant UU 2011-5195 and Philips Consumer Lifestyle. Wilbert Zwart is supported by a Dutch Cancer Society KWF/ Alpe d’HuZes Bas Mulder Award and a Netherlands Organization for Scientific Research NWO VIDI grant.

279

PART TWO | CHAPTER 9

REFERENCES 1. Ferlay J, Soerjomataram I, Dikshit R, et al. Cancer incidence and mortality worldwide: Sources, methods and major patterns in GLOBOCAN 2012. Int J Cancer. 2015;136(5):E359-86. doi: 10.1002/ijc.29210 [doi]. 2. Beelen K, Zwart W, Linn SC. Can predictive biomarkers in breast cancer guide adjuvant endocrine therapy? Nat Rev Clin Oncol. 2012;9(9):529-541. doi: 10.1038/ nrclinonc.2012.121 [doi]. 3. Droog M, Beelen K, Linn S, Zwart W. Tamoxifen resistance: From bench to bedside. Eur J Pharmacol. 2013;717(1-3):4757. doi: 10.1016/j.ejphar.2012.11.071 [doi]. 4. Early Breast Cancer Trialistsâ&#x20AC;&#x2122; Collaborative Group (EBCTCG), Davies C, Godwin J, et al. Relevance of breast cancer hormone receptors and other factors to the efficacy of adjuvant tamoxifen: Patient-level meta-analysis of randomised trials. Lancet. 2011;378(9793):771-784. doi: 10.1016/S0140-6736(11)60993-8 [doi]. 5. Dalmau E, Armengol-Alonso A, Munoz M, Segui-Palmer MA. Current status of hormone therapy in patients with hormone receptor positive (HR+) advanced breast cancer. Breast. 2014;23(6):710-720. doi: 10.1016/j. breast.2014.09.006 [doi]. 6. de Leeuw R, Flach K, Bentin Toaldo C, et al. PKA phosphorylation redirects ERalpha to promoters of a unique gene set to induce tamoxifen resistance. Oncogene. 2013;32(30):3543-3551. doi: 10.1038/onc.2012.361 [doi]. 7. Kok M, Zwart W, Holm C, et al. PKA-induced phosphorylation of ERalpha at serine 305 and high PAK1 levels is associated with sensitivity to tamoxifen in ER-positive breast cancer. Breast Cancer Res Treat. 2011;125(1):1-12. doi: 10.1007/s10549-010-0798-y [doi]. 8. Zwart W, Griekspoor A, Berno V, et al. PKA-induced resistance to tamoxifen is associated with an altered orientation of ERalpha towards co-activator SRC-1. EMBO J. 2007;26(15):3534-3544. doi: 7601791 [pii]. 9. Zwart W, Rondaij M, Jalink K, et al. Resistance to antiestrogen arzoxifene is mediated by overexpression of cyclin D1. Mol Endocrinol. 2009;23(9):1335-1345. doi: 10.1210/me.2008-0268 [doi]. 10. Zwart W, Griekspoor A, Rondaij M, Verwoerd D, Neefjes J, Michalides R. Classification of anti-estrogens according to intramolecular FRET effects on phospho-mutants of estrogen receptor alpha. Mol Cancer Ther. 2007;6(5):15261533. doi: 6/5/1526 [pii]. 11. Kennecke H, Yerushalmi R, Woods R, et al. Metastatic behavior of breast cancer subtypes. J Clin Oncol. 2010;28(20):3271-3277. doi: 10.1200/JCO.2009.25.9820 [doi]. 12. Toy W, Shen Y, Won H, et al. ESR1 ligand-binding domain mutations in hormone-resistant breast cancer. Nat Genet. 2013;45(12):1439-1445. doi: 10.1038/ng.2822 [doi]. 13. Robinson DR, Wu YM, Vats P, et al. Activating ESR1 mutations in hormone-resistant metastatic breast cancer. Nat Genet. 2013;45(12):1446-1451. doi: 10.1038/ng.2823 [doi]. 14. Van Poznak C, Somerfield MR, Bast RC, et al. Use of biomarkers to guide decisions on systemic therapy for women with metastatic breast cancer: American society

280

of clinical oncology clinical practice guideline. Journal of clinical oncology : official journal of the American Society of Clinical Oncology. 2015;33(24):2695-704. doi: 10.1200/ JCO.2015.61.1459. 15. Hurtado A, Holmes KA, Ross-Innes CS, Schmidt D, Carroll JS. FOXA1 is a key determinant of estrogen receptor function and endocrine response. Nat Genet. 2011;43(1):27-33. doi: 10.1038/ng.730 [doi]. 16. Theodorou V, Stark R, Menon S, Carroll JS. GATA3 acts upstream of FOXA1 in mediating ESR1 binding by shaping enhancer accessibility. Genome Res. 2013;23(1):12-22. doi: 10.1101/gr.139469.112 [doi]. 17. Kong SL, Li G, Loh SL, Sung WK, Liu ET. Cellular reprogramming by the conjoint action of ERalpha, FOXA1, and GATA3 to a ligand-inducible growth state. Mol Syst Biol. 2011;7:526. doi: 10.1038/msb.2011.59 [doi]. 18. Mohammed H, Dâ&#x20AC;&#x2122;Santos C, Serandour AA, et al. Endogenous purification reveals GREB1 as a key estrogen receptor regulatory factor. Cell Rep. 2013;3(2):342-349. doi: 10.1016/j.celrep.2013.01.010 [doi]. 19. Jiwa LS, van Diest PJ, Hoefnagel LD, et al. Upregulation of claudin-4, CAIX and GLUT-1 in distant breast cancer metastases. BMC Cancer. 2014;14:864. doi: 10.1186/14712407-14-864. 20. van Diest PJ. No consent should be needed for using leftover body material for scientific purposes. for. BMJ. 2002;325(7365):648-651. 21. Giresi PG, Lieb JD. Isolation of active regulatory elements from eukaryotic chromatin using FAIRE (formaldehyde assisted isolation of regulatory elements). Methods. 2009;48(3):233-239. doi: 10.1016/j.ymeth.2009.03.003 [doi]. 22. Zwart W, Theodorou V, Kok M, Canisius S, Linn S, Carroll JS. Oestrogen receptor-co-factor-chromatin specificity in the transcriptional regulation of breast cancer. EMBO J. 2011;30(23):4764-4776. doi: 10.1038/emboj.2011.368 [doi]. 23. Kumar V, Muratani M, Rayan NA, et al. Uniform, optimal signal processing of mapped deep-sequencing data. Nat Biotechnol. 2013;31(7):615-622. doi: 10.1038/nbt.2596 [doi]. 24. Zhang Y, Liu T, Meyer CA, et al. Model-based analysis of ChIP-seq (MACS). Genome Biol. 2008;9(9):R137-2008-99-r137. Epub 2008 Sep 17. doi: 10.1186/gb-2008-9-9-r137 [doi]. 25. Ye T, Krebs AR, Choukrallah MA, et al. seqMINER: An integrated ChIP-seq data interpretation platform. Nucleic Acids Res. 2011;39(6):e35. doi: 10.1093/nar/gkq1287 [doi]. 26. Thorvaldsdottir H, Robinson JT, Mesirov JP. Integrative genomics viewer (IGV): High-performance genomics data visualization and exploration. Brief Bioinform. 2013;14(2):178-192. doi: 10.1093/bib/bbs017 [doi]. 27. Kuilman T, Velds A, Kemper K, et al. CopywriteR: DNA copy number detection from off-target sequence data. Genome Biol. 2015;16:49-015-0617-1. doi: 10.1186/ s13059-015-0617-1 [doi]. 28. Venkatraman ES, Olshen AB. A faster circular binary segmentation algorithm for the analysis of array CGH data. Bioinformatics. 2007;23(6):657-663. doi: btl646 [pii]. 29. Hammond ME, Hayes DF, Wolff AC, Mangu PB, Temin S. American society of clinical oncology/college of

Pioneer factor loss in endocrine resistance

30.

31.

32.

33.

34.

35.

american pathologists guideline recommendations for immunohistochemical testing of estrogen and progesterone receptors in breast cancer. J Oncol Pract. 2010;6(4):195-197. doi: 10.1200/JOP.777003 [doi]. Foulds CE, Feng Q, Ding C, et al. Proteomic analysis of coregulators bound to ERalpha on DNA and nucleosomes reveals coregulator dynamics. Mol Cell. 2013;51(2):185199. doi: 10.1016/j.molcel.2013.06.007 [doi]. Schlatter LK, Howard KJ, Parker MG, Distelhorst CW. Comparison of the 90-kilodalton heat shock protein interaction with in vitro translated glucocorticoid and estrogen receptors. Mol Endocrinol. 1992;6(1):132-140. doi: 10.1210/mend.6.1.1738366 [doi]. Fan M, Park A, Nephew KP. CHIP (carboxyl terminus of Hsc70-interacting protein) promotes basal and geldanamycin-induced degradation of estrogen receptoralpha. Mol Endocrinol. 2005;19(12):2901-2914. doi: me.2005-0111 [pii]. Shang Y, Hu X, DiRenzo J, Lazar MA, Brown M. Cofactor dynamics and sufficiency in estrogen receptor-regulated transcription. Cell. 2000;103(6):843-852. doi: S00928674(00)00188-4 [pii]. Carroll JS, Liu XS, Brodsky AS, et al. Chromosome-wide mapping of estrogen receptor binding reveals long-range regulation requiring the forkhead protein FoxA1. Cell. 2005;122(1):33-43. doi: S0092-8674(05)00453-8 [pii]. Creyghton MP, Cheng AW, Welstead GG, et al. Histone H3K27ac separates active from poised enhancers and predicts developmental state. Proc Natl Acad Sci U S A.

2010;107(50):21931-21936. doi: 10.1073/pnas.1016071107 [doi]. 36. Toy W, Shen Y, Won H, et al. ESR1 ligand-binding domain mutations in hormone-resistant breast cancer. Nat Genet. 2013;45(12):1439-1445. doi: 10.1038/ng.2822 [doi]. 37. Ross-Innes CS, Stark R, Teschendorff AE, et al. Differential oestrogen receptor binding is associated with clinical outcome in breast cancer. Nature. 2012;481(7381):389393. doi: 10.1038/nature10730 [doi]. 38. Tangen IL, Krakstad C, Halle MK, et al. Switch in FOXA1 status associates with endometrial cancer progression. PLoS One. 2014;9(5):e98069. doi: 10.1371/journal. pone.0098069 [doi]. 39. Robinson DR, Wu YM, Vats P, et al. Activating ESR1 mutations in hormone-resistant metastatic breast cancer. Nat Genet. 2013;45(12):1446-1451. doi: 10.1038/ng.2823 [doi]. 40. Carlson RW, Allred DC, Anderson BO, et al. Metastatic breast cancer, version 1.2012 featured updates to the NCCN guidelines. Journal of the National Comprehensive Cancer Network. 2012;10(7):821-829. 41. American Thoracic Society. Management of malignant pleural effusions. Am J Respir Crit Care Med. 2000;162(5):1987-2001. doi: 10.1164/ajrccm.162.5.ats8-00 [doi]. 42. Apffelstaedt JP, Van Zyl JA, Muller AG. Breast cancer complicated by pleural effusion: Patient characteristics and results of surgical management. J Surg Oncol. 1995;58(3):173-175.

281

PART TWO | CHAPTER 9

SUPPLEMENTAL

Supplementary Figure S1. Endogenous protein mass spectrometry of patient #1 and #2 A. MS-ARC visualization of mass spectrometry results from two technical replicates. Data are grouped according to Gene Ontology annotations. Line length corresponds to PSM score. B. Correlation of PSM scores between the replicates. Correlation coefficient is indicated. Dots corresponding with ERÎą and GREB1 are highlighted in red.

282

Pioneer factor loss in endocrine resistance

Supplementary Figure S2. Venn diagram, showing shared and unique proteins identified between our pleural effusion RIME analyses and those found in the MCF-7 and ZR-75-1 breast cancer cell lines.

9 Supplementary Figure S3. Scatterplot visualization of copy number variation data between sequential pleural effusions.

283

PART TWO | CHAPTER 9

Supplementary Figure S4. Sanger sequencing analysis on ERÎą ligand-binding domain hotspot mutations in metastatic breast cancer specimens.

284

Pioneer factor loss in endocrine resistance

Supplementary Figure S5. No change in methylation status observed with the loss of FOXA1 and GATA3 expression in ERα-positive metastatic breast cancer A. Methylation specific PCR of FOXA1 (CpG-island 99) for three sequential pleural effusions in one patient. Effusion dates were defined as ‘day 0, 21 and 36’. M: methylated primer set; U: unmethylated primer set. B. Methylation specific PCR of GATA3 (CpG-island 33) for three sequential pleural effusions in one patient. Effusion dates were defined as ‘day 0, 21 and 36’. M: methylated primer set; U: unmethylated primer set.

285

PART TWO | CHAPTER 9

Supplementary Figure S6. Loss of FOXA1 and GATA3 expression in ERÎą-positive solid distant breast cancer metastases A. Nuclear staining percentages of FOXA1 in paired primary breast tumors and solid distant metastases. B. Nuclear staining percentages of GATA3 in paired primary breast tumors and solid distant metastases. C. Expression of FOXA1 in solid distant breast cancer metastases. Values are colored according to treatment history: green for patients that obtained adjuvant hormonal therapy and red for patients that were not adjuvantly treated. D. Expression of GATA3 in solid distant breast cancer metastases. Values are colored according to treatment history: green for patients that obtained adjuvant hormonal therapy and red for patients that were not adjuvantly treated. E. Nuclear staining percentages of FOXA1 in paired primary breast tumors and solid distant metastases depicted per location of the metastases. Ovary/uterus: n=4, bone: n=4, skin: n=19, brain: n=18, lung: n=5, liver: n=2. F. Nuclear staining percentages of GATA3 in paired primary breast tumors and solid distant metastases depicted per location of the metastases. Ovary/uterus: n=4, bone: n=4, skin: n=19, brain: n=18, lung: n=5, liver: n=2. G. Quantification of IHC for FOXA1 in the two largest groups of solid distant metastases (brain: n=18 and skin: n=19) compared to matched primary breast tumors. The difference in nuclear staining in primary tumors compared to their matched solid distant metastases is calculated for FOXA1 (Î&#x201D;FOXA1). Values are colored according to treatment history: green for patients that obtained adjuvant hormonal therapy (tamoxifen and/or aromatase-inhibitors), red for patients that were not adjuvantly treated. H. Quantification of IHC for GATA3 in the two largest groups of solid distant metastases (brain: n=18 and skin: n=19) compared to matched primary breast tumors. The difference in nuclear staining in primary tumors compared to their matched solid distant metastases is calculated for GATA3 (Î&#x201D;GATA3). Values are colored according to treatment history: green for patients that obtained adjuvant hormonal therapy (tamoxifen and/or aromatase-inhibitors), red for patients that were not adjuvantly treated.

286

Pioneer factor loss in endocrine resistance

GATA3

FOXA1

GATA3

FOXA1

GATA3

FOXA1

GATA3

100 90

100 75

100 90

100 75

100 95

100 100 90

100 90

100 75

100

100 75

100 90

100 75

100 35

100 75

100 20

100 2

100 75

100 90

FOXA1

Effusion 5

Effusion 4

GATA3

Effusion 3

FOXA1

Effusion 2

Effusion 1

GATA3

Primary

FOXA1

Patient #

Supplementary Table S1. Characteristics of FFPEs of paired primary breast tumors and pleural effusion metastases. Nuclear staining percentages, difference between staining in primary breast tumor and pleural effusion metastases (Δ) and treatment data are shown.

287

PART TWO | CHAPTER 9

GATA3

FOXA1

GATA3

FOXA1

GATA3

FOXA1

GATA3

FOXA1

GATA3

Effusion 5

FOXA1

Effusion 4

Effusion 3

GATA3

Effusion 2

FOXA1

Effusion 1

Primary

Patient #

Supplementary Table S1. Continued

100 100 90

100 50

100 10

100 0

100 90

100 50

100 90

100 100 90

100 75

288

Pioneer factor loss in endocrine resistance

GATA3

FOXA1

GATA3

FOXA1

GATA3

FOXA1

GATA3

FOXA1

GATA3

Effusion 5

FOXA1

Effusion 4

Effusion 3

GATA3

Effusion 2

FOXA1

Effusion 1

Primary

Patient #

Supplementary Table S1. Continued

100 50

100 90

100 75

100 0

100 90

100 20

289

PART TWO | CHAPTER 9

Supplementary Table S1. Continued Hormone therapy

-40

-45

tamoxifen, aminoglutethimide

megestrol acetate

tamoxifen, vorozole, anastrozole

-25

-75

megestrol acetate

-10

-70

tamoxifen

-70

-25

tamoxifen

anastrozole

-70

-5

tamoxifen

-15

tamoxifen

anastrozole

goserilin, anastrozole, tamoxifen, megestrol acetate, faslodex

anastrozole, tamoxifen

-30

-35

letrozole

-90

-20

tamoxifen, anastrozole

exemestane, megestrol acetate, tamoxifen

-95

-90

tamoxifen

letrozole

-100 -90

tamoxifen

tamoxifen, goserelin

tamoxifen, exemestane

anastrozole, tamoxifen, exemestane

-90

-75

tamoxifen

anastrozole, tamoxifen

tamoxifen, anastrozole

tamoxifen, megestrol acetate, faslodex, exemestane, letrozole

tamoxifen

letrozole

tamoxifen

exemestane, anastrozole

-15

-10

-25

anastrozole, exemestane

-40

tamoxifen, lucrin, anastrozole, exemestane

-95

-15

-80

letrozole

anastrozole

-10

-15

-40

-15

-60

-25

tamoxifen, aminoglutethimide

tamoxifen

aminoglutethimide, megestrol acetate

ovariectomy, tamoxifen, aminoglutethimide, megestrol acetate

tamoxifen

Patient #

Δ GATA3

Difference Δ FOXA1

290

Adjuvant

Palliative

Pioneer factor loss in endocrine resistance

Supplementary Table S1. Continued

Δ GATA3

Hormone therapy

Δ FOXA1

Difference

Patient #

Adjuvant

tamoxifen

tamoxifen, aminoglutethimide

tamoxifen

megestrol acetate, aminoglutethimide

tamoxifen

anastrozole

tamoxifen

letrozole, megestrol acetate

anastrozole, tamoxifen

tamoxifen, anastrozole

tamoxifen

anastrozole

tamoxifen

letrozole, tamoxifen, exemestane

anastrozole

tamoxifen, anastrozole, fulvestrant, exemestane

tamoxifen, goserelin, letrozole, exemestane

tamoxifen

exemestane, anastrozole

tamoxifen, exemestane

anastrozole, exemestane, tamoxifen

letrozole, tamoxifen

tamoxifen, goserelin

anastrozole, goserelin, tamoxifen, exemestane

tamoxifen, letrozole

tamoxifen, exemestane

tamoxifen

exemestane

letrozole

letrozole, tamoxifen, exemestane

tamoxifen

tamoxifen, anastrozole, letrozole

tamoxifen

tamoxifen, arimidex

tamoxifen, anastrozole, letrozole

tamoxifen

tamoxifen, anastrozole

anastrozole, exemestane, faslodex

tamoxifen

tamoxifen, exemestane, letrozole

tamoxifen, anastrozole

exemestane, anastrozole

Palliative

291

PART TWO | CHAPTER 9

Supplementary Table S1. Continued

Δ GATA3

Hormone therapy

Δ FOXA1

Difference

Patient #

tamoxifen, exemestane

tamoxifen, anastrozole

tamoxifen

tamoxifen, letrozole, anastrozole, exemestane, faslodex

tamoxifen

tamoxifen, letrozole

tamoxifen

tamoxifen, anastrozole, faslodex, megestrol acetate

aromatase-inhibitor

letrozole

tamoxifen

tamoxifen, anastrozole, exemestane

tamoxifen

anastrozole, megestrol acetate

tamoxifen

tamoxifen, anastrozole

tamoxifen

letrozole

faslodex, anastrozole

exemstane, tamoxifen, letrozole

letrozole, tamoxifen

tamoxifen

anastrozole, exemestane, tamoxifen, megestrol acetate

tamoxifen

tamoxifen, letrozole, exemestane

letrozole, tamoxifen

anastrozole

292

Adjuvant

Palliative

Pioneer factor loss in endocrine resistance

Supplementary Table S2. ChIP-sequencing variables. Read counts and number of aligned reads are shown. factor

timepoint

reads

alligned reads

13009345

12130042

93,24098946

10346142

9786427

94,59010905

8271806

7706412

93,16480585

H3K27Ac

23128245

21896834

94,67572658

H3K27Ac

20136132

18759448

93,16311594

H3K27Ac

22288060

20603766

92,44306593

FOXA1

11213486

10815050

96,44681413

FOXA1

7550444

7096153

93,98325449

RNA Pol II

9320854

8923526

95,73721464

RNA Pol II

11066296

9831186

88,83899364

H3K4me3

5866904

5563463

94,82791946

input

10267200

9800461

95,45407706

Supplementary Table S3. Primer sets and PCR conditions for methylation specific PCR for FOXA1 and GATA3. Gene

Exon CpG-site

Correlation to expression (Spearman R)

Primers

PCR conditions

FOXA1

99 (cg03772350)

-0.701

M-Fw: TTATGAATGGTTTGGGTTTTTAC M-Rv: GACTTAACGTACGAATAACTACGCT U-Fw: TTTATGAATGGTTTGGGTTTTTAT U-Rv: AAACAACTTAACATACAAATAACTACACT

1x 95°C - 7 min 35x 98°C - 15 sec 57°C - 1 min

FOXA1

143 (cg23664186)

-0.554

M-Fw: TTTGAGATTTTAGTTCGGATTTTTC M-Rv: AAAAAAACATCTCCCATAACACG U-Fw: TTGAGATTTTAGTTTGGATTTTTTG U-Rv: AAAAAAAACATCTCCCATAACACAC

1x 95°C - 7 min 35x 98°C - 15 sec 62°C - 1 min

GATA3

33 (cg11679455)

-0.640

M-Fw: TGTTTTATAGGGAGTTAGGTGTGTC M-Rv: CGAAAAACCGTAATAAATAAACGTC U-Fw: TTGTTTTATAGGGAGTTAGGTGTGTT U-Rv: CCAAAAAACCATAATAAATAAACATC

1x 95°C - 7 min 35x 98°C - 15 sec 58°C - 1 min

GATA3

(cg04213746)

-0.790

M-Fw: GTTATTTCGTTGTTATAGTGGGGTC M-Rv: TCCAAATACTATCAACTTTCCTACGTA U-Fw: TTATTTTGTTGTTATAGTGGGGTTGA U-Rv: TTCCAAATACTATCAACTTTCCTACATA

1x 95°C - 7 min 35x 98°C - 15 sec 60°C - 1 min

M: methylated. U: unmethylated. Fw: forward. Rv: reversed

293

PART TWO | CHAPTER 9

Supplementary Table S4. Influence of clinicopathological characteristics of the primary tumors on survival curve differences for FOXA1 and GATA3

metastases to other organs (yes/no)

disease-free interval in days (primary to first metastasis)

GATA3

632

18-7-1990

4066

182

2-3-1990

liver, bone

583

6-2-1992

1096

skin, bone

120

12-4-1992

396

bone, brain

477

21-1-1994

3885

1206

8-8-1996

1987

thoracic wall

113

22-7-1994

3408

thoracic wall, bone, pericarditis carcinomatosa

548

1-9-1995

1726

thoracic wall, bone, liver

228

12-11-1994

432

bone, lymfangitis carcinomatosa

1-3-1995

1023

120

19-3-1998

3466

bone

21-12-1999

851

supraclavicular lymph node

215

20-10-2008

712

bone

1285

ongoing

5493

bone

655

ongoing

4157

bone, lung

236

1-12-2014

331

skin

649

ongoing

1044

552

ongoing

5132

171

2-10-2014

3864

axillary and mediastinal lymph nodes, bone, liver, skin

164

14-3-2014

2270

bone

25-2-2015

6292

retina, retrosternal lymph nodes, ovaria, bone

21-5-2015

1007

bone, liver, peritoneum

176

ongoing

903

1-9-2015

bone, liver

100

1216

15-3-2014

179

1-5-2012

243

15-9-2010

452

31-12-2011

560

10-1-2012

1832

ongoing

294

Code

FOXA1

Stop therapy

Patient #

Days elapsed

characteristics

Pioneer factor loss in endocrine resistance

Supplementary Table S4. Continued

metastases to other organs (yes/no)

disease-free interval in days (primary to first metastasis)

GATA3

132

20-12-2007

4883

bone

205

27-12-2006

3351

1498

12-10-2012

4031

infraclavicular lymph nodes

101

5-12-2007

3119

bone, lymfangitis carcinomatosa, liver, lung

1226

29-5-2009

3150

186

24-7-2006

2172

lung, liver

100

1-3-2008

2978

bone, liver

268

29-10-2007

1922

lung

150

15-6-2007

1230

skin, bone, omentum, peritoneum

3-10-2007

358

bone, lymfangitis carcinomatosa, liver

1-10-2007

bone, liver

384

1-12-2009

907

liver

18-12-2007

1544

lung, liver

100

460

1-12-2015

5470

bone

562

18-9-2015

3126

4-2-2013

1033

lung, skin, bone

22-1-2014

204

mediastinal and retroperitoneal lymph nodes

100

25-8-2014

bone

Code

FOXA1

Stop therapy

Patient #

Days elapsed

characteristicsÂ

Code 1: deceased, 0: alive

295

PART TWO | CHAPTER 9

Supplementary Table S4. Continued

size (in mm)

ER-status

PR-status

HER2 status

ER-status

PR-status

HER2 status

ductal

46-47

44-45

positief

ductal

negatief

41-42

ductal

100

ductal

III

ductal

positief

65-66

ductal

1-2

100

ductal

positief

ductal

lobular II

100

ductal

100

ductal

III

100

ductal

III

100

100 10

ductal

100

ductal

III

ductal

296

age at diagnosis pleural effusion

MAI

type

stage

pleural effusion

age at diagnosis

primary tumor

Patient #

Pioneer factor loss in endocrine resistance

Supplementary Table S4. Continued

pos

100

pos

ductal

pos

neg

ductal

neg

pos

neg

100

ductal

pos

neg

lobular II

pos

100

ductal

pos

neg

lobular -

pos

ductal

pos

neg

lobular -

150 pos

neg

100

ductal

pos

neg

ductal

III

pos

neg

ductal

pos

100

ductal

pos

neg

ductal

pos

neg

ductal

pos

neg

ductal

III

pos

neg

HER2 status

PR-status

ER-status

age at diagnosis pleural effusion

size (in mm)

III

HER2 status

ductal

PR-status

lobular -

ER-status

stage

type

MAI

pleural effusion

age at diagnosis

primary tumor

Patient #

297

Chapter 10 Simone U Dalm, Willemijne AME Schrijver, Anieta M Sieuwerts, Maxime P Look, Angelique CJ Ziel - van der Made, Vanja de Weerd, John W Martens, Paul J van Diest, Marion de Jong and Carolien HM van Deurzen

Prospects of Targeting GRPR, CXCR4 and SSTR2 for nuclear imaging and therapy in metastatic breast cancer Submitted

PART TWO | CHAPTER 10

ABSTRACT The gastrin releasing peptide receptor (GRPR), chemokine c-x-c motif receptor type 4 (CXCR4) and somatostatin receptor 2 (SSTR2) are overexpressed on primary breast cancer, making them ideal candidates for receptor-mediated nuclear imaging and therapy. The aim of this study was to determine whether these receptors are also suitable targets for metastatic breast cancer. mRNA expression of human breast cancer samples were studied by in vitro autoradiography and associated with radioligand binding. Next, GRPR, CXCR4 and SSTR2 mRNA levels of 60 paired primary breast cancers and metastases from different sites were measured by quantitative reverse transcriptase polymerase chain reaction. Receptor mRNA expression levels were associated with clinic-pathological factors and expression levels of primary tumors and corresponding metastases were compared. Binding of GRPR and SSTR radioligands to tumor tissue correlated significantly with receptor mRNA expression. High GRPR and SSTR2 mRNA levels were associated with estrogen receptor (ESR1)-positive tumors (p<0.001 for both receptors). High GRPR and SSTR2 mRNA levels were associated with estrogen receptor (ESR1)-positive tumors (p<0.001 for both receptors). There was no significant difference in GRPR and CXCR4 mRNA expression of primary tumors versus paired metastases. Regarding SSTR2 mRNA expression, there was also no significant difference in the majority of cases, apart from liver and ovarian metastases which showed a significantly lower expression compared to the corresponding primary tumors (p=0.02 and p=0.03, respectively). Targeting the GRPR, CXCR4 and SSTR2 for nuclear imaging and/or treatment has the potential to improve breast cancer care in primary as well as metastatic disease.

300

GRPR, CXCR4 and SSTR2 in breast cancer metastases

INTRODUCTION Breast cancer is the second most common cancer found in women and the fifth cause of cancer related death 1. The disease is very heterogeneous. Different subtypes with distinctive morphological and molecular characteristics exist. The four major intrinsic breast cancer subtypes are luminal A, luminal B, human epidermal growth factor 2 (HER2)-driven and basal-like breast cancer 2,3. Treatment and prognosis of the disease are highly dependent on these subtypes; luminal A and luminal B tumors have a better prognosis than basal-like breast cancer 2,3. Although multiple therapy options for breast cancer exist, 20-30% of breast cancer patients experience relapse with metastatic disease 4. Peptide receptor scintigraphy and peptide receptor radionuclide therapy are methods successfully used in the clinic for imaging and treatment of neuroendocrine tumors 5. These methods are based on targeting receptors that are overexpressed on cancer cells using radiolabeled peptide analogues. Regarding breast cancer, multiple studies have demonstrated overexpression of the gastrin releasing peptide receptor (GRPR), chemokine c-x-c motif receptor type 4 (CXCR4) and somatostatin receptor 2 (SSTR2). In line with this, several pre-clinical as well as clinical studies demonstrated feasibility of imaging and/or treatment of breast cancer with GRPR and SSTR2 radioligands with promising results, and indicated specific breast cancer patients groups that can benefit from the application of these radioligands 6-11. However, previous studies were solely based on primary breast cancer while breast-cancer related death is largely caused by metastatic disease. Targeting the GRPR, CXCR4 and SSTR2 could thus especially be advantageous for treatment of metastatic breast cancer. In this study, we examined the GRPR, CXCR4 and SSTR2 mRNA expression levels of primary tumors and paired metastases, in order to evaluate whether nuclear imaging and therapy might also be beneficial for metastatic breast cancer.

MATERIALS AND METHODS Human breast cancer cases Retrospectively, we selected 74 formalin-fixed paraffin-embedded (FFPE) primary breast cancers and 77 corresponding metastases from an existing database of the University Medical Center Utrecht and from the pathology archive of the Erasmus Medical Center 12,13 . Fresh frozen (FF) tissue of 6 paired primary tumors and regional lymph node metastases were also included. Each specimen was reviewed by a pathologist (CvD) to confirm the presence of malignancy and to determine the percentage of tumour cells (cut-off point of >50% tumor cells). Inclusion criteria were: availability of clinico-pathological data, the presence of enough tumor tissue and good RNA quality to reliably determine RT-qPCR 301

PART TWO | CHAPTER 10

levels (see below). After applying these inclusion criteria, 68 primary tumors and 60 metastases remained, resulting in 60 paired primary breast cancers and metastases from different sites, including brain (n=12), regional lymph nodes (n=20), liver (n=10), ovary (n=5), lung (n=5) and other sites (n=8, consisting of bone (n=2), uterus (n=1), gastrointestinal tract (n=2) and distant lymph node metastases (n=3)). Clinico-pathological characteristics included age, primary tumor size, histological subtype, histological grade according to Bloom & Richardson 14, estrogen receptor (ER) status, HER2 status, and regional lymph node status. The use of anonymous or coded left over material for scientific purposes is part of the standard treatment agreement with patients and therefore informed consent was not required according to Dutch law 15,16. Stored tissue samples were collected by WS and CvD. RNA isolation and quantitative reverse transcriptase polymerase chain reaction (RT-qPCR) Ten 10 µm slides were cut from the FFPE and 10×20 μm from the FF primary breast cancers and paired metastases. The first and last sections (5 μm) were stained with hematoxylin and eosin to guide macro-dissection of the tumor cells for RNA extraction. Total RNA was isolated from the macro-dissected FFPE sections with the AllPrep DNA/RNA FFPE Kit (Qiagen) and from the FF sections with RNA-B (Campro Scientific) according the manufacturer’s instructions. Nucleic acid concentrations were measured with a Nanodrop 1000 system. cDNA was generated for 30 min at 48°C with RevertAid H minus (ThermoFisher Scientific) and gene-specific pre-amplified with Taqman PreAmp Master mix (ThermoFisher Scientific) for 15 cycles, followed by Taqman probe–based real time PCRs according the manufacturer’s instructions in a MX3000P Real-Time PCR System (Agilent). The following gene expression assays were evaluated (all from ThermoFisher Scientific): GRPR, Hs01055872 m1; CXCR4, Hs00237052 m1; SSTR2, Hs0099356 m1; ESR1, Hs00174860_m1; ERBB2, Hs01001580_m1, and quantified relative to the average expression of GUSB, Hs9999908_m1; HMBS, Hs00609297_m1 and TBP, Hs00427620_m1 using the delta Cq method (dCq=2ˆ(average Cq reference genes – Cq target gene)). Samples that resulted in amplifiable products within 25 cycles for this reference gene set at an input of 50 ng total RNA (91.2 % of the samples) were considered to be of good quality to reliably determine RT-qPCR levels. Additional quality and quantity control measurements that were taken to ensure reliable RT-qPCR data analysis are described in the Supplementary methods. In this study, we used ESR1 and ERBB2 mRNA expression levels to determine ESR1 and ERBB2 status (using a cut-off dCq for ESR1>1 and ERBB2>3.5 by optimal binning for n=92 and n=87 overlapping samples, respectively (Supplementary Figure S1). Radioligands and in vitro autoradiography The radiolabeled GRPR antagonist, JMV4168 17, and the radiolabeled SSTR2 agonist, DOTA-Tyr3-octreotate (Mallinckrodt) were radiolabeled with 111In (Covidien) using 302

GRPR, CXCR4 and SSTR2 in breast cancer metastases

quenchers to prevent radiolysis as previously described 18,19. Specific activity was 80 MBq/ nmol for both radiotracers. Radiochemical purity and radiometal incorporation, measured by instant thin-layer chromatography on silica gel and high-pressure liquid chromatography as previously described, were >90% 18. Slides (10 μM) of FF primary breast cancer and paired metastases (n=6 each) were used for autoradiography experiments. Tissue sections were incubated with 100 μL incubation buffer containing 10-9 M of the radiolabeled peptide for 1 h, with and without 10-6 M unlabeled tracer to determine specificity of binding. Results of the autoradiography experiments were quantified using Optiquant (Perkin Elmer) and the percentage added dose (%AD) of the radioligand bound to the tumor tissue was used as an indirect measurement for the level of protein expression. Radioligand binding to primary tumors and paired metastasis was compared and correlated with the measured GRPR and SSTR2 mRNA expression levels in corresponding FF tumor material. Furthermore, mRNA receptor expression measured in FF tumor material was correlated with mRNA receptor expression measured from FFPE tumor material of the same tumor. These correlation analyses were performed to demonstrate that mRNA expression of FFPE material could be used as a surrogate for radiotracer binding. The autoradiography experiments and quantification of the results were performed as described in the supplementary methods. In vitro autoradiography experiments were not performed for CXCR4, since the CXCR4 radioligand available to us, pentaxifor, showed reduced receptor affinity when radiolabeled with 111In. Statistics For the analysis, the STATA statistical package v14.1 and SPSS version 23 were used. Variables were checked for normality prior to analysis. To compare mean values between two or more groups, the Student t-test or analysis of variance ANOVA were used. To compare values for primary and metastatic disease the paired t-test was applied. Pearson and Spearman correlations were calculated when appropriate. P≤0.05 were considered statistically significant.

RESULTS In vitro autoradiography Six pairs of primary breast cancers and regional lymph node metastases (n=12 samples) with varying mRNA receptor expression were analyzed for their ability to bind the GRPR radioligand, 111In-JMV4168, and the SSTR2 radioligand, 111In-DOTA-Tyr3-octreotate, using in vitro autoradiography. Figure 1A shows the in vitro autoradiography results for four of the paired samples. From the six paired samples analyzed, two cases showed specific binding 303

PART TWO | CHAPTER 10

304

GRPR, CXCR4 and SSTR2 in breast cancer metastases

of the GRPR and SSTR2 radioligands in both the primary tumor and the lymph node metastases. In three cases there was no binding of GRPR and SSTR2 radioligands in both the primary tumors and the lymph node metastases. In one case binding of the GRPR radioligand was observed in the primary tumor but not in the lymph node metastasis, while binding of the SSTR2 radioligand was observed in the lymph node metastasis, but not in the primary tumor. When the %AD of the radiotracer bound to the FF tumor tissue was correlated with the mRNA receptor expression of the FF tumor material, a significant positive correlation was found for both GRPR (Spearman rs=0.83, p=0.0008) and SSTR2 (Spearman rs=0.87, p=0.0003) (Figure 1B and D). Furthermore, correlation analysis of mRNA receptor expression levels quantified in FF and FFPE material of the same tumor, resulted in a significant positive correlation for both GRPR (Spearman rs=0.77, p=0.0034) and SSTR2 (Spearman rs=0.72, p=0.0082) (Figure 1C and E). Association of GRPR, CXCR4 and SSTR2 mRNA expression with clinico-pathological factors Table 1 and Supplementary Table S1 show the patient characteristics, including the association of GRPR, CXCR4 and SSTR2 mRNA expression of primary breast cancers with clinico-pathological factors. High GRPR mRNA expression levels were significantly associated with low histologic grade, lobular subtype, ESR1-positive and ERBB2-negative tumors. High SSTR2 mRNA expression levels were also significantly associated with lobular subtype and ESR1-positive tumors. CXCR4 mRNA expression of the primary breast cancer showed no association with the studied clinico-pathological factors. GRPR, CXCR4 and SSTR2 mRNA expression levels of the metastases were correlated with ESR1 and ERBB2 expression of the metastasis itself (Table 2). Similar to the primary tumors, high GRPR and SSTR2 mRNA levels were significantly associated with ESR1-positive metastases. Furthermore, high GRPR mRNA status was significantly associated with ERBB2-negative metastases. Unlike the primary tumors, high SSTR mRNA levels were significantly associated with ERBB2-negative metastatic lesions. In line with the primary breast cancers studied, CXCR4 mRNA expression levels of the metastases showed no significant association with ESR1 and ERBB2 status.

Figure 1. In vitro autoradiography of primary breast cancer and corresponding regional lymph node metastases. A. hematoxylin and eosin (H&E) staining and autoradiography results after incubating cells with the GRPR radioligand, 111In-JMV4168, and the SSTR2 radioligand, 111In-DOTA-Tyr3-Octreotate. B+D. Correlation of quantified autoradiography results (% AD) with mRNA expression of fresh frozen (FF) tissue. C+E. Correlation of mRNA expression of fresh frozen (FF) and formalin fixed paraffin embedded (FFPE) tissue of the same tumor. ttt

305

PART TWO | CHAPTER 10

Table 1. Association of GRPR, CXCR4 and SSTR2 mRNA expression with clinico-pathological factors of primary breast cancer a SSTR2 mRNA log2

0.88

-2.51

2.15

≤ 40

-1.63

4.02

0.34

1.10

-2.23

2.13

41-55

-1.96

3.75

0.53

0.89

-2.08

2.38

56-70

-4.96

2.85

0.29

0.80

-3.07

1.94

> 70

-2.72

4.32

1.09

0.67

-2.90

2.04

Age at surgery (years)

0.48

Mean

3.8

-2.90

Mean

100

All patients in this cohort

Characteristic

Mean

Percentage of patients

CXCR4 mRNA log2

No of patients

GRPR mRNA log2

0.08

0.32

0.22

Tumour size c ≤ 2 cm

-2.62

3.97

0.34

0.88

-2.61

2.24

2 ≤ 5 cm

-3.47

3.44

0.53

0.91

-2.32

2.16

> 5 cm

-2.00

4.22

0.53

0.89

-2.95

2.09

0.51

0.70

0.72

Histopathological subtypes d Ductal

-3.34

3.76

0.50

0.90

-2.82

2.15

Lobular

-0..80

3.44

0.49

0.84

-1.04

1.74

Other

-2.38

3.72

-0.32

0.14

-2.04

1.26

0.04

0.96

0.01

Bloom & Richardson grade e I + II

-1.23

3.57

0.18

0.86

-1.80

2.20

III

-3.66

3.91

0.62

0.89

-2.88

2.18

0.04

0.10

0.12

ESR1 status e Negative

-6.52

1.68

0.54

0.90

-3.83

2.09

Positive

-0.79

3.04

0.41

0.87

-1.79

1.80

<0.001

0.58

<0.001

ERBB2 status e Negative

-2.12

3.60

0.50

0.89

-2.44

2.26

Positive

-5.04

2.62

0.81

0.89

-2.7

2.23

P Regional lymph node status

0.006

0.32

0.73

Negative

-4.40

3.06

0.66

1.01

-2.96

2.33

Positive

-2.67

4.02

0.40

0.82

-2.48

2.01

0.13

0.33

0.44

Due to missing values numbers do not always add up to 68. b Receptor expression of ductal breast cancer and lobular breast cancer was compared using the student t-test. c P for Pearson correlation. d P for variance of ANOVA. e P for student t-test.

306

GRPR, CXCR4 and SSTR2 in breast cancer metastases

Table 2. Associations of receptor mRNA expression with ESR1 and ERBB2 status in breast cancer metastases ER status a

ERBB2 status a

Negative

Positive

Mean

Negative

Positive

Mean

p SD

All metastases No of patients

GRPR

-6.26

2.76

35 -1.87

3.90

<0.001 -2.67

48 3.77

-7.98

2.15

<0.001

CXCR4

0.27

1.28

0.29

1.07

0.95

0.17

1.14

0.74

1.13

0.15

SSTR2

-3.95

1.64

-2.89

2.23

0.04

-2.91

1.96

-5.11

1.49

<0.001

Regional lymph node metastases

No of patients

GRPR

-7.0

1.96

-1.70

3.53

0.003

-1.70

3.54

-7.02

1.84

0.002

CXCR4

0.04

0.91

0.14

0.90

0.85

0.10

0.88

0.21

1.00

0.86

SSTR2

-4.03

0.97

-2.92

2.67

0.20

-2.77

2.47

-4.62

1.92

0.16

All distant metastases b

No of patients

GRPR

-6.12

2.92

-2.01

4.27

0.001

-3.15

3.84

-8.52

2.26

<0.001

CXCR4

0.31

1.36

0.41

1.20

0.82

0.21

1.26

1.05

1.14

0.12

SSTR2

-3.94

1.76

-2.86

1.85

0.07

-2.98

1.69

-5.40

1.26

0.001

P for student t test. Numbers do not add up to 60 because for 1 patient ESR1 and ERBB2 were unknown. a

GRPR, CXCR4 and SSTR2 mRNA expression of primary breast cancer vs. corresponding metastases Figure 2 shows the box plots of GRPR, CXCR4 and SSTR2 mRNA expression in primary tumors and corresponding metastases. Comparison of receptor mRNA expression levels of primary tumors and corresponding metastases showed no significant difference in GRPR and CXCR4 mRNA levels between primary tumors and corresponding regional lymph node and distant metastases in the brain, lung, liver and ovaries. However, in the group of metastases from other sites, GRPR mRNA expression levels were significantly lower in the metastases compared to the corresponding primary breast cancer (p=0.02). Regarding SSTR2 mRNA levels, there were no significant differences in SSTR2 mRNA expression of the primary tumor and the paired metastasis in regional lymph nodes, brain, lung and other locations. However, SSTR2 mRNA levels of liver and ovarian metastases were significantly lower compared to the expression in the corresponding primary breast cancer (p=0.02 and p=0.03, respectively). 307

PART TWO | CHAPTER 10

Next, we compared the receptor mRNA expression levels between distant metastases from various metastatic sites amongst each other. GRPR mRNA levels were significantly higher in the ovarian metastases (p=0.03) and CXCR4 mRNA expression levels were significantly higher in liver metastases (p=0.05). There were no significant differences in SSTR2 mRNA expression levels in distant metastases from different sites.

Figure 2. GRPR, CXCR4 and SSTR2 mRNA levels in primary breast cancer (PBC) and corresponding metastases (BCM). Significant differences are indicated by *

308

GRPR, CXCR4 and SSTR2 in breast cancer metastases

In some cases studied (n=11), there was a discordance regarding ESR1 status of primary breast cancers and corresponding metastases. When studying the effect of change in ESR1 status on receptor mRNA expression in primary tumors and paired metastasis, GRPR and SSTR2 mRNA expression changed accordingly (higher GRPR/SSTR2 mRNA expression in ESR1-positive lesions compared to ESR1-negative lesions) in the majority of the tumors. However, this difference was only significant (p<0.05) for ESR1-positive primary breast cancers with corresponding ESR1-negative metastases (n=6). Discordance regarding ERBB2 status was seen in 6 paired samples. In these samples a change in ERBB2 status of primary breast cancers and corresponding metastases did not have a consistent effect on GRPR mRNA expression levels.

DISCUSSION Targeting of GRPR, CXCR4 and SSTR2 overexpressed on breast cancer cells with radioligands can offer novel imaging and therapy options for breast cancer. Previous clinical and preclinical studies reported promising results. However, these studies were restricted to primary breast cancer, while metastases are the main cause of breast cancer -related death. In this study, we compared GRPR, CXCR4 and SSTR2 mRNA expression levels in a unique dataset of primary breast cancer and corresponding metastases to determine whether receptor-based imaging and/or therapy could also be useful for metastatic breast cancer. For this purpose, we selected FFPE material of primary breast cancers and corresponding metastases from different sites, and compared mRNA receptor expression levels of the paired samples. Prior to this, we confirmed that mRNA expression levels of tumor tissue properly represent radioligand binding, by correlating in vitro autoradiography results with mRNA expression levels of selected primary tumors and corresponding metastases with varying mRNA receptor expression. This was only done for GRPR and SSTR2 because our CXCR4 radioligand could not be labelled with a radionuclide suited for in vitro autoradiography. However, based on a previous study by Philipp-Abbrederis et al. 20, who reported on CXCR4 mRNA expression in breast cancer cell lines and CXCR4 targeted imaging of corresponding xenograft models, we assumed that CXCR4 mRNA expression can be used as a surrogate for CXCR4 radioligand binding. Next, we determined the GRPR, CXCR4 and SSTR2 mRNA expression of the paired primary breast cancers and metastases. When we associated receptor mRNA expression levels of primary breast cancers and metastases with clinico-pathological factors, we observed a significantly higher GRPR and SSTR2 expression in both ESR1-positive primary breast cancer and metastases. These findings are in agreement with our previous findings 11 and findings by Kumar et al. 21 and Stoykow et al. 7. The latter publication describes a clinical study in which the GRPR radioligand, 68Ga-RM2 was successfully used for imaging of breast 309

PART TWO | CHAPTER 10

cancer lesions and imaging success-rate associated positively with ER and PR status. Furthermore, Prignon et al. 22 demonstrated that 68Ga-AMBA, a GRPR agonist, was better suited for monitoring response to hormonal treatment than 18F-FDG PET in an ER-positive breast cancer model. In another study, van den Bossche et al. 23 published data indicating an estrogen-dependent regulation of SSTR expression in breast cancer cell lines. Since ERpositive breast cancer accounts for approximately 75% of the breast cancer population, applying receptor targeted nuclear imaging and/or therapy using GRPR or SSTR2 radioligands could be beneficial for the majority of the breast cancer population 2. In contrast to our previous study we did not find a significant association of high CXCR4 mRNA expression with ESR1-negative tumors. A potential explanation for this is the smaller sample size analyzed in this study. Nevertheless, CXCR4 might be a good candidate for targeting of ER-negative breast cancers, that usually have a poor prognosis 2. In paired primary tumors and metastases a change in ESR1 expression from positive to negative resulted in a significant decrease in GRPR and SSTR2 mRNA levels. This may indicate an ESR1 dependent expression of GRPR and SSTR2, which is consistent with literature 23,24. Difference in ERBB2 status in primary tumors and paired metastasis did not show a clear effect on GRPR mRNA expression, although these numbers were too small for reliable conclusions. Comparison of GRPR, CXCR4 and SSTR2 mRNA levels of primary tumors and corresponding metastases resulted in similar GRPR and CXCR4 mRNA expression in primary tumors and paired regional lymph nodes and distant metastases of the brain, lung, liver and ovaries. However, GRPR mRNA expression was significantly higher in primary tumors compared to corresponding metastases from other sites. Since this group is very diverse, containing metastases from distant lymph nodes, bone, uterus and metastases from the gastrointestinal tract, it is not possible to draw solid conclusions. Regarding SSTR2, mRNA expression levels were significantly lower in liver and ovarian metastases compared to the paired primary breast cancer. Combining our findings, both GRPR and SSTR2 are promising targets for nuclear imaging and/or therapy in primary and metastatic ER-positive breast cancer, but GRPR seems more suitable due to its retained expression in the metastases. This finding is also supported by a previous study by our group, in which we demonstrated GRPR expression in 48/50 breast cancers 6, while SSTR2 was only expressed in 26/53 breast cancers (SU Dalm, CHM van Deurzen, M. Melis, J. W. Martens and M. de Jong, unpublished data, 2014). Since a substantial portion of breast cancer patients experience relapse with metastatic disease, it is important to develop new treatment options for this late stage of disease. We showed that receptor mRNA expression levels were similar in primary tumors and corresponding metastases in the majority of the cases, implying that targeting these receptors for disease monitoring or therapy might improve breast cancer patient care.

310

GRPR, CXCR4 and SSTR2 in breast cancer metastases

Biopsy material or excised tumors can be used to determine receptor expression by immunohistochemistry, RNA in situ, in vitro autoradiography or RT-qPCR 25. Disease monitoring of receptor-positive tumors can then be performed by single photon emission computed tomography/computer tomography (SPECT/CT), positron emission tomography (PET)/CT or PET/magnetic resonance imaging using radioligands targeting these receptors. Also, dedicated breast PET cameras can be used. These dedicated cameras have improved sensitivity and specificity compared to whole body PET, because of a restricted field of view, resulting in higher cancer detection 26. Furthermore, tumors can be treated with therapeutic radioligands. Another option is to use GRPR, CXCR4 or SSTR2 radioligands for visualization of sentinel node metastases or as a guide for breast cancer surgery (e.g. preoperative imaging, radioguided surgery) in patients with receptor positive primary tumors 27,28. The next step would be to perform clinical studies to investigate the feasibility of imaging primary tumors and metastases with radioligands targeting these receptors. One important aspect is to study physiological uptake of the radioligands in other organs, since this is of great importance for successful nuclear imaging and treatment. However, previous studies using radioligands targeting these receptors on other tumor types did not report on any alarming physiological uptake 5,7,29 The presented data indicates that nuclear based imaging and therapy has the potential to improve breast cancer patient care in primary as well as in metastatic disease, by targeting GRPR, CXCR4 and SSTR2. Both GRPR and SSTR2 radioligands, but especially GRPR radioligands, are promising for imaging and treatment of ER-positive primary and metastatic breast cancer. Furthermore, imaging and treatment of metastases derived from CXCR4-positive primary tumors using CXCR4 radioligands seems to be feasible.

10 Acknowledgements This study is supported by Dutch Cancer Society grant UU 2011-5195 and Philips Consumer Lifestyle. The authors thank Natalie D. ter Hoeve and Erik de Blois for their technical assistance.

311

PART TWO | CHAPTER 10

REFERENCES 1. IACR. GLOBOCAN 2012: Estimated cancer incidence, mortality and prevalence worldwide in 2012: International agency for research on cancer. 2. Yersal O, Barutca S. Biological subtypes of breast cancer: Prognostic and therapeutic implications. World J Clin Oncol. 2014;5(3):412-424. doi: 10.5306/wjco.v5.i3.412 [doi]. 3. Perou CM, Sorlie T, Eisen MB, et al. Molecular portraits of human breast tumours. Nature. 2000;406(6797):747752. doi: 10.1038/35021093 [doi]. 4. Early Breast Cancer Trialists’ Collaborative Group (EBCTCG). Effects of chemotherapy and hormonal therapy for early breast cancer on recurrence and 15-year survival: An overview of the randomised trials. Lancet. 2005;365(9472):1687-1717. doi: S0140-6736(05)66544-0 [pii]. 5. Bison SM, Konijnenberg MW, Melis M, et al. Peptide receptor radionuclide therapy using radiolabeled somatostatin analogs: Focus on future developments. Clin Transl Imaging. 2014;2:55-66. doi: 10.1007/s40336-0140054-2 [doi]. 6. Dalm SU, Martens JW, Sieuwerts AM, et al. In vitro and in vivo application of radiolabeled gastrin-releasing peptide receptor ligands in breast cancer. J Nucl Med. 2015;56(5):752-757. doi: 10.2967/jnumed.114.153023 [doi]. 7. Stoykow C, Erbes T, Bulla S, Mäcke H, Drendel V, Bronsert P, et al. Gastrin-releasing peptide receptor expression is associated with estrogen receptor status in breast cancer: Findings of a PET/CT pilot study (PW082). annual congress of the european association of nuclear medicine; hamburg, Germany2015, october 10-14. . 8. Skanberg J, Ahlman H, Benjegard SA, et al. Indium-111octreotide scintigraphy, intraoperative gamma-detector localisation and somatostatin receptor expression in primary human breast cancer. Breast Cancer Res Treat. 2002;74(2):101-111. 9. Van Den Bossche B, Van Belle S, De Winter F, Signore A, Van de Wiele C. Early prediction of endocrine therapy effect in advanced breast cancer patients using 99mTcdepreotide scintigraphy. J Nucl Med. 2006;47(1):6-13. doi: 47/1/6 [pii]. 10. Azad BB, Chatterjee S, Lesniak WG, et al. A fully human CXCR4 antibody demonstrates diagnostic utility and therapeutic efficacy in solid tumor xenografts. Oncotarget. 2016;7(11):12344-12358. doi: 10.18632/oncotarget.7111 [doi]. 11. Dalm SU, Sieuwerts AM, Look MP, et al. Clinical relevance of targeting the gastrin-releasing peptide receptor, somatostatin receptor 2, or chemokine C-X-C motif receptor 4 in breast cancer for imaging and therapy. J Nucl Med. 2015;56(10):1487-1493. doi: 10.2967/ jnumed.115.160739 [doi]. 12. Hoefnagel LD, van der Groep P, van de Vijver MJ, et al. Discordance in ERalpha, PR and HER2 receptor status across different distant breast cancer metastases within the same patient. Ann Oncol. 2013;24(12):3017-3023. doi: 10.1093/annonc/mdt390 [doi].

312

13. Schrijver WA, Jiwa LS, van Diest PJ, Moelans CB. Promoter hypermethylation profiling of distant breast cancer metastases. Breast Cancer Res Treat. 2015;151(1):41-55. doi: 10.1007/s10549-015-3362-y [doi]. 14. BLOOM HJ, RICHARDSON WW. Histological grading and prognosis in breast cancer; a study of 1409 cases of which 359 have been followed for 15 years. Br J Cancer. 1957;11(3):359-377. 15. FEDERA. human tissue and medical research: Code of conduct for responsible use (2011). 16. van Diest PJ. No consent should be needed for using leftover body material for scientific purposes. for. BMJ. 2002;325(7365):648-651. 17. Marsouvanidis PJ, Nock BA, Hajjaj B, et al. Gastrin releasing peptide receptor-directed radioligands based on a bombesin antagonist: Synthesis, (111)in-labeling, and preclinical profile. J Med Chem. 2013;56(6):2374-2384. doi: 10.1021/jm301692p [doi]. 18. de Blois E, Chan HS, Konijnenberg M, de Zanger R, Breeman WA. Effectiveness of quenchers to reduce radiolysis of (111)in- or (177)lu-labelled methioninecontaining regulatory peptides. maintaining radiochemical purity as measured by HPLC. Curr Top Med Chem. 2012;12(23):2677-2685. doi: CTMCEPUB-20130104-19 [pii]. 19. De Blois E, Schroeder RP J, De Ridder CM A, Van Weerden WM, Breeman WA P, De Jong M. Improving radiopeptide pharmacokinetics by adjusting experimental conditions for bombesin receptor-mediated imaging of prostate cancer. Q J Nucl Med Mol Imaging. 2013. doi: R39Y9999N00A0058 [pii]. 20. Philipp-Abbrederis K, Herrmann K, Knop S, et al. In vivo molecular imaging of chemokine receptor CXCR4 expression in patients with advanced multiple myeloma. EMBO Mol Med. 2015;7(4):477-487. doi: 10.15252/ emmm.201404698 [doi]. 21. Kumar U, Grigorakis SI, Watt HL, et al. Somatostatin receptors in primary human breast cancer: Quantitative analysis of mRNA for subtypes 1--5 and correlation with receptor protein expression and tumor pathology. Breast Cancer Res Treat. 2005;92(2):175-186. doi: 10.1007/ s10549-005-2414-0 [doi]. 22. Prignon A, Nataf V, Provost C, et al. (68)ga-AMBA and (18)F-FDG for preclinical PET imaging of breast cancer: Effect of tamoxifen treatment on tracer uptake by tumor. Nucl Med Biol. 2015;42(2):92-98. doi: 10.1016/j. nucmedbio.2014.10.003 [doi]. 23. Van Den Bossche B, D’haeninck E, De Vos F, et al. Oestrogen-mediated regulation of somatostatin receptor expression in human breast cancer cell lines assessed with 99mTc-depreotide. Eur J Nucl Med Mol Imaging. 2004;31(7):1022-1030. doi: 10.1007/s00259-004-1500-6 [doi]. 24. Nagasaki S, Nakamura Y, Maekawa T, et al. Immunohistochemical analysis of gastrin-releasing peptide receptor (GRPR) and possible regulation by estrogen receptor betacx in human prostate carcinoma. Neoplasma. 2012;59(2):224-232. 25. Korner M, Waser B, Schonbrunn A, Perren A, Reubi JC. Somatostatin receptor subtype 2A immunohistochemistry

GRPR, CXCR4 and SSTR2 in breast cancer metastases

using a new monoclonal antibody selects tumors suitable for in vivo somatostatin receptor targeting. Am J Surg Pathol. 2012;36(2):242-252. doi: 10.1097/ PAS.0b013e31823d07f3 [doi]. 26. Vercher-Conejero JL, Pelegri-Martinez L, Lopez-Aznar D, Cozar-Santiago Mdel P. Positron emission tomography in breast cancer. Diagnostics (Basel). 2015;5(1):61-83. doi: 10.3390/diagnostics5010061 [doi]. 27. Mariani G, Erba P, Villa G, et al. Lymphoscintigraphic and intraoperative detection of the sentinel lymph node in breast cancer patients: The nuclear medicine perspective. J Surg Oncol. 2004;85(3):112-122. doi: 10.1002/jso.20023 [doi].

28. Hindie E, Groheux D, Brenot-Rossi I, Rubello D, Moretti JL, Espie M. The sentinel node procedure in breast cancer: Nuclear medicine as the starting point. J Nucl Med. 2011;52(3):405-414. doi: 10.2967/jnumed.110.081711 [doi]. 29. Lapa C, Luckerath K, Kleinlein I, et al. (68)ga-pentixaforPET/CT for imaging of chemokine receptor 4 expression in glioblastoma. Theranostics. 2016;6(3):428-434. doi: 10.7150/thno.13986 [doi].

313

PART TWO | CHAPTER 10

SUPPLEMENTAL Supplementary methods Quality and quantity control measurements for reliable quantitative reverse transcriptase polymerase chain reaction (RT-qPCR) For reliable RT-qPCR measurements, only samples that resulted in amplifiable products within 25 cycles for the used reference gene set at an input of 50 ng total RNA (91.2% of the samples) were considered to be of good quality for reliable determination of RT-qPCR levels. Furthermore, a serially diluted fresh frozen (FF) and formalin fixed paraffin embedded (FFPE) breast tumor sample was included in each experiment to evaluate the linear amplification and efficiencies for all genes included in the panel, and absence of amplification in the absence of reverse transcriptase. All gene transcripts were 100% efficient amplified (range 89%-113%) and were negative in the absence of reverse transcriptase. To ensure unbiased results from FF and FFPE samples, these 2 data sets were normalized based on the expression levels measured in a set of n=13 matched FF-FFPE samples. Estrogen receptor (ER/ESR1) and receptor tyrosine-protein kinase erbB-2 status/human epidermal growth factor 2 (ERBB2/HER2) status of the investigated samples Because data regarding ER and HER2 protein expression of our data set was incomplete, ESR1 and ERBB2 mRNA expression was used to determine ESR1 and ERBB2 mRNA status (using a dCq cut-off for ESR1>1 and ERBB2>3.5 by optimal binning for n=92 and n=87 overlapping samples, respectively (Supplementary Figure S1)). Because ER and HER2 are determined on protein level in daily clinical practice (using a scoring system according to national and international guidelines 1,2, we investigated whether the ESR1 and ERBB2 mRNA status accurately reflected the ER and HER2 protein status as reported in the pathology reports in samples with known receptor protein status. These cut-offs resulted for ESR1 in a sensitivity of 0.88 and specificity of 0.85 and for ERBB2 in a sensitivity of 0.89 and specificity of 0.97. In vitro autoradiography Fresh frozen tumor sections (10 Îźm) were incubated with 100 ÎźL 10-9 M of the radioligands for 1 h, without and with 10-6 M unlabeled tracer. Octreotide (Covidien) and Tyr4-bombesin (Sigma-Aldrich) were used to block the somatostatin receptor and the gastrin releasing peptide receptor, respectively. Subsequently unbound radioligand was removed and slides were exposed to super-resolution phosphor screens (Perkin Elmer) for at least 24 h. Next, screens were read using the cyclone (Perkin Elmer) and the results were quantified using OptiQuant Software (Perkin Elmer). For this tumor containing regions, identified with the help of hematoxylin and eosin staining of adjacent tumor sections, were encircled and the 314

GRPR, CXCR4 and SSTR2 in breast cancer metastases

digital light units/mm2 (DLU/mm2) were measured. Specific binding was determined by subtracting DLU/mm2 of blocked tissue sections from the DLU/mm2 of the unblocked sections (DLU/mm2unblocked – DLU/mm2blocked = DLU/mm2specific). Standards containing 1 μL drops of the radiotracer solution were also quantified and used to determine the percentage of added dose that was bound to the tumors (%AD = (DLU/mm2specific / (DLU/mm2standard × 100)) * 100%).

Supplementary Figure S1. Correlation of ER and HER2 protein status with ESR1 and ERBB2 mRNA levels. Arrows indicate used cut-off value.

315

PART TWO | CHAPTER 10

Other metastases n=8

Ovarian metastases n=5

Liver metastases n=10

Lung metastases n=5

Brain metastases n=12

â&#x2030;¤ 55

â&#x2030;Ľ 56

< 2 - 5 cm

100

> 5 cm

100 10

100 2

Characteristic of primary tumour

No of patients

Percentage of patients

Regional lymph node metastases n=20

Supplemental Table S1. Overview of clinico-pathological characteristics of the primary breast cancers associated with the site of the paired metastases a

Age at surgery (years)

Tumour size

Histopathological subtypes Ductal

Lobular

Other

Bloom & Richardson grade I + II

III

Negative

Positive

100 7

Negative

Positive

100 7

ESR1 mRNA status

ERBB2 mRNA status

Due to missing values and rounding off numbers do not add up to 100%.

316

GRPR, CXCR4 and SSTR2 in breast cancer metastases

SUPPLEMENTAL REFERENCES 1. NABON. Breast cancer dutch guideline, version 2.0. 2012. 2. Wolff AC, Hammond ME, Hicks DG, et al. Recommendations for human epidermal growth factor receptor 2 testing in breast cancer: American society of clinical oncology/college of american pathologists clinical practice guideline update. Arch Pathol Lab Med. 2014;138(2):241-256. doi: 10.5858/arpa.2013-0953-SA [doi].

317

Chapter 11

Summarizing discussion

PART TWO | CHAPTER 11

In this thesis, a subset of genotypic and phenotypic differences between primary breast tumors and paired distant metastases are described, with the aim to gain more insight into the metastatic process and to find biomarkers that could aid in prognostication and treatment of metastatic breast cancer. Here, the main conclusions of the different parts will be reviewed and discussed.

PART ONE (Epi)genotyping of distant breast cancer metastases Cancer is thought to arise from a series of somatic genetic and epigenetic modifications that allow the cell to evade normal cell growth and organization 1,2. To drive cancer initiation and progression, a cancer cell needs to maintain these genetic and epigenetic alterations for as long as they confer a selective advantage 3. Exploration of the (epi)genetic differences between primary tumors and paired metastases could therefore yield valuable information about tumor progression and could lead to markers for metastatic breast cancer detection, prognosis and therapy. Multiple genetic alterations have been implicated in carcinogenesis and disease progression4. In Chapter 2, we compared the mutational profiles of actionable cancer targets between HER2-enriched and triple negative breast tumors and their matched distant metastases to brain and skin. We found extensive inter-patient differences and intra-patient similarities in somatic mutations and copy number alterations. Metastases often demonstrated an increase in variant frequency, copy number and novel mutations compared to the primary tumors, suggesting clonal evolution. Multiple of these newly harbored genetic aberrations are potentially actionable, implying that mutational profiling of metastatic samples would be of added value to tailor treatment. In Chapter 3 we investigated mRNA overexpression of APOBEC3B (apolipoprotein B editing catalytic subunit 3B), which provokes higher than normal mutation rates through C-to-T transition, which accelerates tumor evolution and could potentially be abrogated by therapeutic intervention 5. We demonstrated that a distinct APOBEC3B mRNA expression pattern of the primary tumor is largely retained during progression, with a significant increase in the paired metastases. Furthermore, distant metastases showed higher expression than loco-regional lymph node metastases. This implies a role for APOBEC3B not only at the stage of the primary tumor but also, and according to our data even more dominantly, during the tumor evolution in metastatic breast cancer. DNA methylation has a similar potential in serving as a selectable driver during clonal expansion or metastatic dissemination 3. In Chapter 4, we therefore performed promoter hypermethylation profiling of tumor suppressor genes in patients with brain, liver, lung or skin metastases. Our results showed that MS-MLPA based hypermethylation of tumor suppressor genes is generally lower in distant metastases compared to the primary tumors. 320

Summarizing discussion

This loss or rearrangement of hypermethylation pattern may be explained by the fact that, in contrast to the previously mentioned stable DNA mutations, epigenetic processes are sometimes plastic or reversible over time 6,7. The question remains if this is a random process or a response to specific signals. Due to its reversibility, methylation is probably not an epigenetic factor that could be used for therapy against metastatic tumor spread. However, since different metastasizing localizations showed varying methylation patterns, screening for a specific pattern that predicts most likely site of metastasis could be a useful clinical tool. Another promising epigenetic point of engagement for cancer treatment are microRNAs (miRs). Dysregulation of miRs occurs in various types of cancer and is associated with tumor initiation, drug resistance, and metastasis 8. In Chapter 5 we performed expression profiling of miRs in primary breast tumors and matched multiple metastases. We demonstrated miR expression to be largely retained in metastases, but the expression of known â&#x20AC;&#x2DC;metastamiRsâ&#x20AC;&#x2122; was generally higher in the metastases compared to the primary tumor. Ferracin et al. already showed that primary tumors of different origin display a distinct miR expression profile and that metastases retain a large part of these miRs 9. Furthermore, we found the abundance of specific miRs in the primary tumor to be metastasis location-specific. For example, hsa-miR-106b-5p could be used as a predictive factor for lung metastasis. This knowledge could potentially be exploited to gain more insight into the metastatic behavior of breast cancer. Overall, some interesting similarities were seen between the distribution of studied (epi) genetic markers in part one of this thesis. The detected pattern of somatic mutations, copy number alterations, APOBEC3B mRNA and miR expression was largely retained in the metastasis relative to the primary breast tumor. Furthermore, a significant amount of novel alterations and/or an increase of alteration frequency was demonstrated. Since cancer cells are believed to evolve according to Darwinian positive selection, further genetic evolution throughout progression might be expected 10. Also, during progression a cancer cell is exposed to a wide range of influencing factors, like the tumor microenvironment or systemic treatments. Thus, the additional aberrations that we found may be a mechanism provoked by these influences that confers an added survival advantage 11. In Chapter 2, there was a trend towards important driver mutations remaining concordant, while probable passenger mutations diverged between primary tumors and metastases. Moreover, in Chapter 5 expression profiling demonstrated that known metastasis-promoting miRs were generally higher in the metastases compared to the primary tumors. Altogether, this leads to the assumption that certain (epi)genetic markers are positively or negatively selected for during tumor progression and that beneficial aberrations for the primary tumor might be equally advantageous for the metastasis at a distant location. Only hypermethylation of tumor suppressor genes was less often seen in paired metastases, but the mechanism for this finding remains unclear. One possible explanation could be 321

PART TWO | CHAPTER 11

that the spread of tumor cells may take place even prior to methylation, since it has been demonstrated before that DTCs (disseminated tumor cells) display significantly fewer genetic aberrations than primary tumor cells 12-15. This finding would favor the parallel progression model as stated in the introduction 16. Another contributing finding to this hypothesis is, that some primary tumors in Chapter 2 showed mutations that were not encountered in the metastasis. This could be explained by clonal divergence within the primary tumor, either with other clones being selected for metastatic dissemination or with the metastasis branching off at an earlier time point. On the other hand, we may have introduced selection bias by choosing one of many available tissue blocks per (heterogeneous) tumor. Intra- and inter block differences should not be overlooked, since a single random sample may not be representative of the whole tumor 17-21. However, considering that the primary tumors and metastases showed to be (epi)genetically closely related in Chapters 2, 3 and 5, the influence of heterogeneity and the presence of early parallel progression are probably negligible. Another recurring theme is that breast cancer is a very heterogeneous disease. This large genetic diversity was seen before in a targeted gene panel, where no two tumors shared an identical genetic profile 22. Even tumors classified into the same molecular subgroups based on a combination of (epi)genomic, transcriptomic and proteomic arrays, showed extensive inter-patient discordance 23. Interestingly, in Chapters 3, 4 and 5, metastasis locationspecific similarities were found, implying some orderliness and predictability in what sometimes seems to be a random combination of driver mutations 24. It is well known that mutations can follow a location specific pattern in primary tumors 25 and it is thought that molecular characteristics of the cancer cell and of the target tissue cooperate to determine the location of metastasis 26,27. Paired assessment of normal and tumorous tissue of breast and metastasis-location would be contributory to comprehend the interplay of a metastasis with the microenvironment of a target organ. Since a large part of studied (epi)genetic aberrations are retained in the paired metastasis, this could help predict metastasis location or identify the origin of cancers with unknown primary tumor 28.

322

Summarizing discussion

PART TWO Phenotyping of distant breast cancer metastases As mentioned in the introduction, the preferred site of distant metastasis as well as the prognosis and response to treatment, amongst others, are attributed to the molecular subtype of breast cancer 29-31. This subtyping is classically based on ERα (estrogen receptor alpha), PR (progesterone receptor) and HER2 (human epidermal growth factor receptor 2) status. However, a change of hormone or HER2 receptor status in the paired metastasis compared to the primary tumor is often encountered 32,33. In Chapter 6 the frequent occurrence of receptor conversion is confirmed with random effects pooled discordance proportions for ERα, PR and HER2 of 19%, 31% and 10%, respectively. Large heterogeneity was seen between patients, receptors, techniques used and subsites of metastasis leading to dispersed discordance percentages. Therefore, the usage of standardized techniques for receptor assessment and implementation of personalized cancer care seem to be of the utmost importance. A factor that could lead to overestimation of receptor conversion is the decalcification process of bone metastases. Especially acidic buffers may potentially compromise antigenicity and as such hamper interpretation of (molecular) diagnostics 34,35. In Chapter 7 we demonstrated that immunohistochemistry of ERα, PR and HER2 is not significantly affected by tissue decalcification with agents containing formic acid or EDTA. However, quantity and quality of isolated DNA and RNA was diminished, especially on account of acidic buffers. To prevent patients from being falsely stratified for systemic therapy, we advise to use EDTA decalcification for DNA or RNA based techniques and in situ hybridization, and to interpret findings with caution. Receptor conversion in malignant effusions has rarely been reported, although it is a frequent metastatic site 36-39. In Chapter 8 we demonstrated that receptor conversion in effusions follows the same trend as in distant solid metastases with discordance proportions of 25-30%, 30-35%, 47-60% and 6-12% for ERα, PR, AR and HER2, respectively. About 27% of patients showed changes in receptor status likely to have consequences for the response to hormonal treatment. Furthermore, under selective pressure of adjuvant endocrine therapy for ERα, PR and AR and adjuvant chemotherapy for HER2, significantly more conversion was perceived. This may be explained by the idea that clones resistant to the applied treatment preferentially disseminate to distant locations 40,41. Other interpretations could be clonal dedifferentiation and selection of or evolution to more aggressive phenotypes 40,42-44. Unfortunately, the presence of hormone or HER2 receptor expression does not guarantee response to systemic therapies 45,46. A subset of receptor positive breast cancers do not benefit from endocrine therapy at all (termed intrinsic resistance) and most metastatic breast tumors ultimately develop resistance to hormonal or targeted therapies (termed acquired resistance) 47. A profound insight into the mechanisms of resistance will aid 323

PART TWO | CHAPTER 11

optimisation of treatment and eventually survival of metastatic breast cancer patients. In Chapter 9 we presented a thus far unknown mechanism of acquired endocrine therapy resistance. In pleural effusion metastases, we showed that FOXA1 and GATA3 expression is lost under evolutionary selection pressure of adjuvant endocrine therapy. FOXA1 and GATA3 are key-luminal breast cancer defining genes 48 and loss of these markers resulted in a non-functional ERα receptor with diminished response to endocrine drugs. Finally, we focused on less acquainted biomarkers GRPR (gastrin releasing peptide receptor), CXCR4 (chemokine c-x-c motif receptor type 4) and SSTR (somatostatin receptor 2), which have proven themselves successful in receptor-based imaging and treatment 49. These methods are based on radiolabeled peptide analogs that can target previously mentioned receptors overexpressed on cancer cells. Since multiple studies have demonstrated overexpression in primary breast tumors 50,51, we compared the expression in paired primaries and metastases in Chapter 10. Large resemblance was seen in corresponding samples, implying that targeting CXCR4, SSTR and especially GRPR for disease monitoring or therapy potentially improves patient care. Using radioligands targeting these receptors, breast cancer metastases can be visually monitored or they can help to guide surgery. Furthermore, patients can be treated with peptide analogs linked to therapeutic radionuclides 52,53. In contrast to the relatively stable (epi)genetic aberrations mentioned in part one of this thesis, the phenotypic markers tested in part two showed high discordance levels between primary breast tumors and paired distant metastases. It could be that the proteins we tested were more prone to selective pressure of adjuvant therapies than mutations and epigenetic variations. However, this seems counter intuitive, since protein expression is the result from an (epi)genetic readout. Possibly, other influencing markers like histone modifications, mRNA expression or ubiquitination, not addressed in this thesis, could be an important link here 54,55. In recent years, obtaining tissue from metastases to confirm presence of disease and to establish receptor status, has become a key recommendation in multiple clinical guidelines56-59. However, it is as yet unclear whether outcomes are better with treatment regimens based on the receptor status of the metastases in comparison to the primary tumor. When discordance is evident, it is advised to preferentially use the receptor status of the metastasis to guide therapy on 56. In primary breast cancer, approximately 50% of patients with ERα-positive tumors benefit from endocrine drugs 60. Particularly a combination of ERα- and PR-positivity predicts good response to hormonal therapies 61,62. As shown in Chapters 6 and 8, PR expression is often lost in paired metastases, positioning PR as an important marker in predicting response failure to endocrine therapy in metastatic disease. Furthermore, in Chapter 9 we emphasized that ERα functionality is not always guaranteed when ERα expression is 324

Summarizing discussion

maintained in breast cancer metastases. ERα assessment is therefore not always sufficient to predict hormonal treatment responsiveness in the metastatic setting, so analyzing FOXA1 and GATA3 expression should also be considered. Ideally, in cancer care, we need a predictable biomarker pattern for therapeutic targeting that can help many patients at the same time. However, we showed that no patient and no tumor looks alike and the key in targeted treatment must therefore be a tailored approach per tumor and per patient. With current technologies we are more and more able to select optimal drug and drug dosage per patient, thereby improving patient outcome. Unfortunately, significant obstacles withhold widespread implementation of personalized cancer care. Amongst them are the high costs of targeted therapies and the relatively large group of investigated biomarkers, with only few of them being independently validated and/or possessing the potential to reliably identify patients who are likely to respond to treatment 63. Furthermore, as we showed in this thesis, large inter-patient variability is seen in biomarker expression, minimizing the chance of only one biomarker to cover all escape mechanisms of a tumor. The challenge in translational medicine is therefore, to no longer only work from ‘bench to bedside’, but rather to maintain the dialogue with clinical practitioners and get back to the ‘bench’ again when necessary. Eventually, this will lead to an appropriate prognostic and therapeutic combination for each patient and at each stage of the disease. Despite the potentially increased cost of personalized cancer care, avoiding unnecessary treatments will pay for this in the end 64.

325

PART TWO | CHAPTER 11

REFERENCES 1. Jones PA, Baylin SB. The epigenomics of cancer. Cell. 2007;128(4):683-692. doi: S0092-8674(07)00127-4 [pii]. 2. Hanahan D, Weinberg RA. Hallmarks of cancer: The next generation. Cell. 2011;144(5):646-674. doi: 10.1016/j. cell.2011.02.013 [doi]. 3. Aryee MJ, Liu W, Engelmann JC, et al. DNA methylation alterations exhibit intraindividual stability and interindividual heterogeneity in prostate cancer metastases. Sci Transl Med. 2013;5(169):169ra10. doi: 10.1126/scitranslmed.3005211 [doi]. 4. Â Vogelstein, K. W. kinzler, the genetic basis of human cancer (McGraw-hill, medical pub. division, new york, 2002). . 5. Harris RS. Molecular mechanism and clinical impact of APOBEC3B-catalyzed mutagenesis in breast cancer. Breast Cancer Res. 2015;17:8-014-0498-3. doi: 10.1186/ s13058-014-0498-3 [doi]. 6. Hemberger M, Dean W, Reik W. Epigenetic dynamics of stem cells and cell lineage commitment: Digging waddingtonâ&#x20AC;&#x2122;s canal. Nat Rev Mol Cell Biol. 2009;10(8):526537. doi: 10.1038/nrm2727 [doi]. 7. Feinberg AP, Irizarry RA, Fradin D, et al. Personalized epigenomic signatures that are stable over time and covary with body mass index. Sci Transl Med. 2010;2(49):49ra67. doi: 10.1126/scitranslmed.3001262 [doi]. 8. Takahashi RU, Miyazaki H, Ochiya T. The roles of MicroRNAs in breast cancer. Cancers (Basel). 2015;7(2):598-616. doi: 10.3390/cancers7020598 [doi]. 9. Ferracin M, Pedriali M, Veronese A, et al. MicroRNA profiling for the identification of cancers with unknown primary tissue-of-origin. J Pathol. 2011;225(1):43-53. doi: 10.1002/path.2915 [doi]. 10. Stephens PJ, Tarpey PS, Davies H, et al. The landscape of cancer genes and mutational processes in breast cancer. Nature. 2012;486(7403):400-404. doi: 10.1038/ nature11017 [doi]. 11. Desmedt C, Yates L, Kulka J. Catalog of genetic progression of human cancers: Breast cancer. Cancer Metastasis Rev. 2016;35(1):49-62. doi: 10.1007/s10555016-9609-1 [doi]. 12. Schmidt-Kittler O, Ragg T, Daskalakis A, et al. From latent disseminated cells to overt metastasis: Genetic analysis of systemic breast cancer progression. Proc Natl Acad Sci U S A. 2003;100(13):7737-7742. doi: 10.1073/ pnas.1331931100 [doi]. 13. Schardt JA, Meyer M, Hartmann CH, et al. Genomic analysis of single cytokeratin-positive cells from bone marrow reveals early mutational events in breast cancer. Cancer Cell. 2005;8(3):227-239. doi: S15356108(05)00261-8 [pii]. 14. Stoecklein NH, Hosch SB, Bezler M, et al. Direct genetic analysis of single disseminated cancer cells for prediction of outcome and therapy selection in esophageal cancer. Cancer Cell. 2008;13(5):441-453. doi: 10.1016/j. ccr.2008.04.005 [doi]. 15. Weckermann D, Polzer B, Ragg T, et al. Perioperative activation of disseminated tumor cells in bone marrow of patients with prostate cancer. J Clin Oncol.

326

2009;27(10):1549-1556. doi: 10.1200/JCO.2008.17.0563 [doi]. 16. Naxerova K, Jain RK. Using tumour phylogenetics to identify the roots of metastasis in humans. Nat Rev Clin Oncol. 2015;12(5):258-272. doi: 10.1038/ nrclinonc.2014.238 [doi]. 17. Ding L, Ellis MJ, Li S, et al. Genome remodelling in a basal-like breast cancer metastasis and xenograft. Nature. 2010;464(7291):999-1005. doi: 10.1038/nature08989 [doi]. 18. Moelans CB, de Groot JS, Pan X, van der Wall E, van Diest PJ. Clonal intratumor heterogeneity of promoter hypermethylation in breast cancer by MS-MLPA. Mod Pathol. 2014;27(6):869-874. doi: 10.1038/ modpathol.2013.207 [doi]. 19. Blighe K, Kenny L, Patel N, et al. Whole genome sequence analysis suggests intratumoral heterogeneity in dissemination of breast cancer to lymph nodes. PLoS One. 2014;9(12):e115346. doi: 10.1371/journal. pone.0115346 [doi]. 20. Kroigard AB, Larsen MJ, Laenkholm AV, et al. Clonal expansion and linear genome evolution through breast cancer progression from pre-invasive stages to asynchronous metastasis. Oncotarget. 2015;6(8):56345649. doi: 3111 [pii]. 21. Goswami RS, Patel KP, Singh RR, et al. Hotspot mutation panel testing reveals clonal evolution in a study of 265 paired primary and metastatic tumors. Clin Cancer Res. 2015;21(11):2644-2651. doi: 10.1158/1078-0432.CCR-142391 [doi]. 22. Muller KE, Marotti JD, de Abreu FB, et al. Targeted nextgeneration sequencing detects a high frequency of potentially actionable mutations in metastatic breast cancers. Exp Mol Pathol. 2016;100(3):421-425. doi: 10.1016/j.yexmp.2016.04.002 [doi]. 23. Cancer Genome Atlas Network. Comprehensive molecular portraits of human breast tumours. Nature. 2012;490(7418):61-70. doi: 10.1038/nature11412 [doi]. 24. Kroigard AB, Larsen MJ, Thomassen M, Kruse TA. Molecular concordance between primary breast cancer and matched metastases. Breast J. 2016;22(4):420-430. doi: 10.1111/tbj.12596 [doi]. 25. Kandoth C, McLellan MD, Vandin F, et al. Mutational landscape and significance across 12 major cancer types. Nature. 2013;502(7471):333-339. doi: 10.1038/ nature12634 [doi]. 26. Minn AJ, Kang Y, Serganova I, et al. Distinct organspecific metastatic potential of individual breast cancer cells and primary tumors. J Clin Invest. 2005;115(1):44-55. doi: 10.1172/JCI22320 [doi]. 27. Lu X, Kang Y. Organotropism of breast cancer metastasis. J Mammary Gland Biol Neoplasia. 2007;12(2-3):153-162. doi: 10.1007/s10911-007-9047-3 [doi]. 28. Marquard AM, Birkbak NJ, Thomas CE, et al. TumorTracer: A method to identify the tissue of origin from the somatic mutations of a tumor specimen. BMC Med Genomics. 2015;8:58-015-0130-0. doi: 10.1186/ s12920-015-0130-0 [doi]. 29. Kennecke H, Yerushalmi R, Woods R, et al. Metastatic behavior of breast cancer subtypes. J Clin Oncol.

Summarizing discussion

2010;28(20):3271-3277. doi: 10.1200/JCO.2009.25.9820 [doi]. 30. Yersal O, Barutca S. Biological subtypes of breast cancer: Prognostic and therapeutic implications. World J Clin Oncol. 2014;5(3):412-424. doi: 10.5306/wjco.v5.i3.412 [doi]. 31. Savci-Heijink CD, Halfwerk H, Hooijer GK, Horlings HM, Wesseling J, van de Vijver MJ. Retrospective analysis of metastatic behaviour of breast cancer subtypes. Breast Cancer Res Treat. 2015;150(3):547-557. doi: 10.1007/ s10549-015-3352-0 [doi]. 32. Aurilio G, Disalvatore D, Pruneri G, et al. A meta-analysis of oestrogen receptor, progesterone receptor and human epidermal growth factor receptor 2 discordance between primary breast cancer and metastases. Eur J Cancer. 2014;50(2):277-289. doi: 10.1016/j.ejca.2013.10.004. 33. Houssami N, Macaskill P, Balleine RL, Bilous M, Pegram MD. HER2 discordance between primary breast cancer and its paired metastasis: Tumor biology or test artefact? insights through meta-analysis. Breast Cancer Res Treat. 2011;129(3):659-674. doi: 10.1007/s10549-011-1632-x. 34. Bussolati G, Leonardo E. Technical pitfalls potentially affecting diagnoses in immunohistochemistry. J Clin Pathol. 2008;61(11):1184-1192. doi: 10.1136/ jcp.2007.047720 [doi]. 35. Gruchy JR, Barnes PJ, Dakin Hache KA. CytoLyt(R) fixation and decalcification pretreatments alter antigenicity in normal tissues compared with standard formalin fixation. Appl Immunohistochem Mol Morphol. 2015;23(4):297-302. doi: 10.1097/PAI.0000000000000082 [doi]. 36. Fentiman IS, Millis R, Sexton S, Hayward JL. Pleural effusion in breast cancer: A review of 105 cases. Cancer. 1981;47(8):2087-2092. 37. Raju RN, Kardinal CG. Pleural effusion in breast carcinoma: Analysis of 122 cases. Cancer. 1981;48(11):2524-2527. 38. Wilkes JD, Fidias P, Vaickus L, Perez RP. Malignancyrelated pericardial effusion. 127 cases from the roswell park cancer institute. Cancer. 1995;76(8):1377-1387. 39. Buck M, Ingle JN, Giuliani ER, Gordon JR, Therneau TM. Pericardial effusion in women with breast cancer. Cancer. 1987;60(2):263-269. 40. Kuukasjarvi T, Karhu R, Tanner M, et al. Genetic heterogeneity and clonal evolution underlying development of asynchronous metastasis in human breast cancer. Cancer Res. 1997;57(8):1597-1604. 41. Rasbridge SA, Gillett CE, Seymour AM, et al. The effects of chemotherapy on morphology, cellular proliferation, apoptosis and oncoprotein expression in primary breast carcinoma. Br J Cancer. 1994;70(2):335-341. 42. Sari E, Guler G, Hayran M, Gullu I, Altundag K, Ozisik Y. Comparative study of the immunohistochemical detection of hormone receptor status and HER-2 expression in primary and paired recurrent/metastatic lesions of patients with breast cancer. Med Oncol. 2011;28(1):57-63. doi: 10.1007/s12032-010-9418-2 [doi]. 43. Amir E, Miller N, Geddie W, et al. Prospective study evaluating the impact of tissue confirmation of metastatic disease in patients with breast cancer. J Clin Oncol.

2012;30(6):587-592. doi: 10.1200/JCO.2010.33.5232 [doi]. 44. Edgerton SM, Moore D,2nd, Merkel D, Thor AD. erbB-2 (HER-2) and breast cancer progression. Appl Immunohistochem Mol Morphol. 2003;11(3):214-221. 45. Rastelli F, Crispino S. Factors predictive of response to hormone therapy in breast cancer. Tumori. 2008;94(3):370-383. 46. Vogel CL, Cobleigh MA, Tripathy D, et al. Efficacy and safety of trastuzumab as a single agent in first-line treatment of HER2-overexpressing metastatic breast cancer. J Clin Oncol. 2002;20(3):719-726. 47. Dalmau E, Armengol-Alonso A, Munoz M, Segui-Palmer MA. Current status of hormone therapy in patients with hormone receptor positive (HR+) advanced breast cancer. Breast. 2014;23(6):710-720. doi: 10.1016/j. breast.2014.09.006 [doi]. 48. Kong SL, Li G, Loh SL, Sung WK, Liu ET. Cellular reprogramming by the conjoint action of ERalpha, FOXA1, and GATA3 to a ligand-inducible growth state. Mol Syst Biol. 2011;7:526. doi: 10.1038/msb.2011.59 [doi]. 49. Bison SM, Konijnenberg MW, Melis M, et al. Peptide receptor radionuclide therapy using radiolabeled somatostatin analogs: Focus on future developments. Clin Transl Imaging. 2014;2:55-66. doi: 10.1007/s40336-0140054-2 [doi]. 50. Reubi C, Gugger M, Waser B. Co-expressed peptide receptors in breast cancer as a molecular basis for in vivo multireceptor tumour targeting. Eur J Nucl Med Mol Imaging. 2002;29(7):855-862. doi: 10.1007/s00259-0020794-5 [doi]. 51. Salvucci O, Bouchard A, Baccarelli A, et al. The role of CXCR4 receptor expression in breast cancer: A large tissue microarray study. Breast Cancer Res Treat. 2006;97(3):275-283. doi: 10.1007/s10549-005-9121-8 [doi]. 52. Hindie E, Groheux D, Brenot-Rossi I, Rubello D, Moretti JL, Espie M. The sentinel node procedure in breast cancer: Nuclear medicine as the starting point. J Nucl Med. 2011;52(3):405-414. doi: 10.2967/jnumed.110.081711 [doi]. 53. Mariani G, Erba P, Villa G, et al. Lymphoscintigraphic and intraoperative detection of the sentinel lymph node in breast cancer patients: The nuclear medicine perspective. J Surg Oncol. 2004;85(3):112-122. doi: 10.1002/jso.20023 [doi]. 54. Santos GC,Jr, da Silva AP, Feldman L, Ventura GM, Vassetzky Y, de Moura Gallo CV. Epigenetic modifications, chromatin distribution and TP53 transcription in a model of breast cancer progression. J Cell Biochem. 2015;116(4):533-541. doi: 10.1002/jcb.25003 [doi]. 55. Qin J, Zhou Z, Chen W, et al. BAP1 promotes breast cancer cell proliferation and metastasis by deubiquitinating KLF5. Nat Commun. 2015;6:8471. doi: 10.1038/ncomms9471 [doi]. 56. Van Poznak C, Somerfield MR, Bast RC, et al. Use of biomarkers to guide decisions on systemic therapy for women with metastatic breast cancer: American society of clinical oncology clinical practice guideline. J Clin Oncol. 2015;33(24):2695-2704. doi: 10.1200/ JCO.2015.61.1459 [doi].

327

PART TWO | CHAPTER 11

57. Cardoso F, Costa A, Norton L, et al. 1st international consensus guidelines for advanced breast cancer (ABC 1). Breast. 2012;21(3):242-252. doi: 10.1016/j. breast.2012.03.003 [doi]. 58. Hammond ME, Hayes DF, Wolff AC, Mangu PB, Temin S. American society of clinical oncology/college of american pathologists guideline recommendations for immunohistochemical testing of estrogen and progesterone receptors in breast cancer. J Oncol Pract. 2010;6(4):195-197. doi: 10.1200/JOP.777003 [doi]. 59. Carlson RW, Allred DC, Anderson BO, et al. Metastatic breast cancer, version 1.2012: Featured updates to the NCCN guidelines. J Natl Compr Canc Netw. 2012;10(7):821-829. doi: 10/7/821 [pii]. 60. Bardou VJ, Arpino G, Elledge RM, Osborne CK, Clark GM. Progesterone receptor status significantly improves outcome prediction over estrogen receptor status alone for adjuvant endocrine therapy in two large breast cancer databases. J Clin Oncol. 2003;21(10):1973-1979. doi: 10.1200/JCO.2003.09.099 [doi].

328

61. Bartlett JM, Brookes CL, Robson T, et al. Estrogen receptor and progesterone receptor as predictive biomarkers of response to endocrine therapy: A prospectively powered pathology study in the tamoxifen and exemestane adjuvant multinational trial. J Clin Oncol. 2011;29(12):1531-1538. doi: 10.1200/JCO.2010.30.3677 [doi]. 62. Brankovic-Magic M, Jankovic R, Neskovic-Konstantinovic Z, Nikolic-Vukosavljevic D. Progesterone receptor status of breast cancer metastases. J Cancer Res Clin Oncol. 2002;128(1):55-60. doi: 10.1007/s00432-001-0299-9 [doi]. 63. Schilsky RL. Personalized medicine in oncology: The future is now. Nat Rev Drug Discov. 2010;9(5):363-366. doi: 10.1038/nrd3181 [doi]. 64. Fleck R, Bach D. Trends in personalized therapies in oncology: The (venture) capitalistâ&#x20AC;&#x2122;s perspective. J Pers Med. 2012;2(1):15-34. doi: 10.3390/jpm2010015 [doi].

Summarizing discussion

329

Chapter 12

Future perspectives

PART TWO | CHAPTER 12

For this manuscript we used a very valuable and exceptional database of paired primary tumor and metastasis samples from around 400 patients, retrospectively collected at the department of pathology of the University Medical Center Utrecht, and obtained from multiple pathology laboratories throughout The Netherlands 1,2. However, the retrospective acquisition of tissues led to some important limitations. First, introduction of selection bias likely occurred, as the majority of metastases is normally not resected or biopsied. Second, since most metastases were biopsied instead of surgically resected, limited material was present for multiple techniques (DNA/RNA isolation, immunohistochemistry etc.). And, a biopsy is shown not always to be representative for the whole tumor 3. Furthermore, unaffected breast or metastasis-location tissue was not always present as a control, thus valuable comparison of tumor microenvironment could not be performed in a standardized fashion. Also, we were restricted to certain assays for archived tissue, since not all genomic tests are suitable for FFPE material 4,5. Finally, we could not always retrieve patient information to support our laboratory findings, by reason of masked patient information or absence of informed consent, because some patients were already deceased. To overcome these problems, prospectively collected, fresh frozen as well as FFPE patient material is needed with sufficient follow-up data on survival and therapy response. Clinical guidelines increasingly advice to obtain tissue from metastases, paving the way for potential newly pitched databases 6-9. A recently contracted partnership between the Alexander Monro Breast Cancer Hospital (AMBZ) and the University Medical Center Utrecht (UMCU) will enable accessibility of greater numbers of samples for scientific research in the interests of better cancer care 10. It would be most valuable to include patients that present with primary breast cancer, follow them over time until they show progression into metastatic disease, and then obtain material from all metastatic sites and blood. Wagle et al., started an initiative like this in the USA, called the â&#x20AC;&#x2DC;Metastatic Breast Cancer Projectâ&#x20AC;&#x2122;. This initiative entails a website that allows patients with metastatic breast cancer across the country to participate. Enrolled patients complete a questionnaire, send a saliva sample which is used to extract germline DNA and sign informed consent to release medical records and tumor biopsy. Whole exome and transcriptome sequencing is performed on tumor and germline and response and resistance to therapies is distilled from the medical records 11. In this project patients are included that already presented with metastatic disease. Therefore, information about patients that do not form metastases is not included. Prerequisite to get a complete picture about the metastatic cascade, is to include and compare normal and tumorous tissue of patients that did and did not form metastases. For Chapter 8 and 9 we already started collecting a new database of fresh material from pleural and peritoneal effusion metastases which is immensely valuable for future research. In contrast to the secured cells in solid metastases, effusion metastases already show an anchorage independent growth pattern, a signature that identifies cells with metastatic potential 12. We already showed that these cells could be used for culture experiments, 332

Future perspectives

enabling insight into the metastatic process. In the near future, we would like to start a clinical study with cell culture of effusion metastases in the presence of chemo- or endocrine therapies to ascertain a response to treatment. Clinically, this could potentially lead to personalized cancer treatment, since feedback to the clinician about resistance or sensitivity to a drug could help guide treatment decisions. In Chapter 9 we showed that pleural effusion metastases became resistant to endocrine drugs under evolutionary selection pressure of adjuvant therapies. The ERα receptor became non-functional, paralleled by loss of expression of FOXA1 and GATA3. However, the question remained which cell proliferation cascade took over the mitogenic potential of ERα in this setting. A possible explanation might be activation of the PI3K/mTOR pathway 13 . We therefore analyzed paired pMTOR and pERK expression in primary tumors and pleural effusion of treated versus untreated patients. However, in both patient populations, loss of FOXA1 and/or GATA3 and thus ERα responsiveness coincided with a loss of pMTOR and/or pERK. Future studies could be directed on other pathways that may provide an escape route of the cancer cell that lost ERα directed proliferation. Amongst others, markers from the PI3K/AKT- or Ras/p42/44 MAPK-pathway could be explored 14. Next generation sequencing could definitely play a role in the search for resistance mechanisms, with its potential for fast data processing and numerous possibilities in targeted cancer treatments 15. However, guidelines state that there is no routine clinical role for genomic or expression profiling yet, in the selection of treatment for hormone receptor positive metastatic breast cancer 16. In ERα-negative breast cancer on the other hand, we showed that a large part of metastases harbored new, potentially actionable genetic aberrations relative to the primary tumor. Unfortunately, most of these mutations and amplifications have not been shown to predict response to therapy in breast cancer yet. Therefore, future research should address the potential efficacy of these targeted therapies to evaluate if mutational profiling of metastatic samples would be of added value to tailor treatment. Eventually, this could lead to some kind of tool to predict outcome based on the different genetic aberrations between primary tumors and matched metastases, similar to the 70-gene signature test (MammaPrint) which has been shown to improve prediction of clinical outcome in women with early-stage breast cancer 17. In order to thoroughly examine genotypic and phenotypic differences, it would be most informative to assess histological information, ERα, PR and HER2 status, copy number data, whole genome, (mi)RNA and Chip sequencing and DNA methylation altogether in primary tumors and paired metastases of the same patient cohort, as such starting a “Metastasis Genome Atlas” project. In this thesis, we focused on divergent biomarkers and for every single study we used a different set of patients from the database (due to limited material availability). The (epi)genetic and protein expression findings could therefore not be compared to each other. The integration of information across different platforms was done before in primary tumors, which yielded valuable information including the discovery of molecular subtypes 18. 333

PART TWO | CHAPTER 12

Lastly, an important topic that we did not address in this thesis is tumor cell dormancy. In the clinical setting, dormancy describes an unusually long latency period between the treatment of primary tumors and metastatic recurrences. From a cell biology perspective, it is a senescence-like state of disseminated tumor cells in a distant tissue microenvironment 19 . Dormancy is an important factor to consider when interpreting linear or parallel models of progression. A dormant cell does not proliferate and is therefore thought not to be able to gain genotypic and phenotypic alterations 20. However, the fact that some metastases occur relatively late implies that certain disseminated tumor cells are not fully capable to give rise to a secondary tumor at the moment of dissemination. Yet, these cells remain viable and acquire their capacity progressively through possible additional (epi)genetic modifications or signals from the microenvironment 21. All the same, whether a metastasis arose after a long latency period, because it disseminated late during cancer progression (linear progression model) or because it underwent a period of dormancy at the distant site, is difficult to distinguish 20. Future research should dig into this deeper, by comparing genotypic and phenotypic biomarkers of early and late metastases (synchrone and metachrone), preferentially of the same patients. Currently, dormant cells seem to escape systemic therapies, because these drugs are directed on highly proliferating cells 22. When better understood, dormancy could in the future be used as a therapeutic target. Overall, we encourage to biopsy metastatic lesions whenever possible. As we have shown, molecular and phenotypic profiling of metastatic tissue provides invaluable mechanistic insight into the biology underlying metastatic progression and has the potential to identify novel, potentially druggable, drivers of progression. As such, we will be able to make a small step forward towards cancer becoming a chronic disease.

334

Future perspectives

REFERENCS 1. Hoefnagel LDC, Moelans CB, Meijer SL, et al. Prognostic value of estrogen receptor alpha and progesterone receptor conversion in distant breast cancer metastases. Cancer. 2012;118(20):4929-4935. doi: 10.1002/cncr.27518. 2. Hoefnagel LDC, van de Vijver MJ, van Slooten H, et al. Receptor conversion in distant breast cancer metastases. Breast Cancer Research. 2010;12(5):R75. doi: 10.1186/ bcr2645. 3. Tamaki K, Sasano H, Ishida T, et al. Comparison of core needle biopsy (CNB) and surgical specimens for accurate preoperative evaluation of ER, PgR and HER2 status of breast cancer patients. Cancer Sci. 2010;101(9):2074-2079. doi: 10.1111/j.1349-7006.2010.01630.x [doi]. 4. van Beers EH, Joosse SA, Ligtenberg MJ, et al. A multiplex PCR predictor for aCGH success of FFPE samples. Br J Cancer. 2006;94(2):333-337. doi: 6602889 [pii]. 5. Kerick M, Isau M, Timmermann B, et al. Targeted high throughput sequencing in clinical cancer settings: Formaldehyde fixed-paraffin embedded (FFPE) tumor tissues, input amount and tumor heterogeneity. BMC Med Genomics. 2011;4:68-8794-4-68. doi: 10.1186/1755-87944-68 [doi]. 6. Van Poznak C, Somerfield MR, Bast RC, et al. Use of biomarkers to guide decisions on systemic therapy for women with metastatic breast cancer: American society of clinical oncology clinical practice guideline. J Clin Oncol. 2015;33(24):2695-2704. doi: 10.1200/ JCO.2015.61.1459 [doi]. 7. Cardoso F, Costa A, Norton L, et al. 1st international consensus guidelines for advanced breast cancer (ABC 1). Breast. 2012;21(3):242-252. doi: 10.1016/j. breast.2012.03.003 [doi]. 8. Carlson RW, Allred DC, Anderson BO, et al. Metastatic breast cancer, version 1.2012: Featured updates to the NCCN guidelines. J Natl Compr Canc Netw. 2012;10(7):821-829. doi: 10/7/821 [pii]. 9. Hammond ME, Hayes DF, Wolff AC, Mangu PB, Temin S. American society of clinical oncology/college of american pathologists guideline recommendations for immunohistochemical testing of estrogen and progesterone receptors in breast cancer. J Oncol Pract. 2010;6(4):195-197. doi: 10.1200/JOP.777003 [doi]. 10. Persbericht samenwerking AMBZ-UMCU (DEF). bilthoven, 14 maart 2016. available from: http://Www. alexandermonro.nl/over-ons/nieuws/.

11. Wagle et al., J clin oncol 34, 2016 (suppl; abstr LBA1519). 12. Mori S, Chang JT, Andrechek ER, et al. Anchorageindependent cell growth signature identifies tumors with metastatic potential. Oncogene. 2009;28(31):2796-2805. doi: 10.1038/onc.2009.139 [doi]. 13. Beelen K, Hoefnagel LD, Opdam M, et al. PI3K/AKT/ mTOR pathway activation in primary and corresponding metastatic breast tumors after adjuvant endocrine therapy. Int J Cancer. 2014;135(5):1257-1263. doi: 10.1002/ ijc.28769 [doi]. 14. Osborne CK, Schiff R. Mechanisms of endocrine resistance in breast cancer. Annu Rev Med. 2011;62:233247. doi: 10.1146/annurev-med-070909-182917 [doi]. 15. Doherty M, Metcalfe T, Guardino E, Peters E, Ramage L. Precision medicine and oncology: An overview of the opportunities presented by next-generation sequencing and big data and the challenges posed to conventional drug development and regulatory approval pathways. Ann Oncol. 2016;27(8):1644-1646. doi: 10.1093/annonc/ mdw165 [doi]. 16. Rugo HS, Rumble RB, Macrae E, et al. Endocrine therapy for hormone receptor-positive metastatic breast cancer: American society of clinical oncology guideline. J Clin Oncol. 2016;34(25):3069-3103. doi: 10.1200/ JCO.2016.67.1487 [doi]. 17. Cardoso F, vanâ&#x20AC;&#x2122;t Veer LJ, Bogaerts J, et al. 70-gene signature as an aid to treatment decisions in early-stage breast cancer. N Engl J Med. 2016;375(8):717-729. doi: 10.1056/NEJMoa1602253 [doi]. 18. Cancer Genome Atlas Network. Comprehensive molecular portraits of human breast tumours. Nature. 2012;490(7418):61-70. doi: 10.1038/nature11412 [doi]. 19. Klein CA. Framework models of tumor dormancy from patient-derived observations. Curr Opin Genet Dev. 2011;21(1):42-49. doi: 10.1016/j.gde.2010.10.011 [doi]. 20. Naxerova K, Jain RK. Using tumour phylogenetics to identify the roots of metastasis in humans. Nat Rev Clin Oncol. 2015;12(5):258-272. doi: 10.1038/ nrclinonc.2014.238 [doi]. 21. Lorusso G, Ruegg C. New insights into the mechanisms of organ-specific breast cancer metastasis. Semin Cancer Biol. 2012;22(3):226-233. doi: 10.1016/j. semcancer.2012.03.007 [doi]. 22. Goss PE, Chambers AF. Does tumour dormancy offer a therapeutic target? Nat Rev Cancer. 2010;10(12):871-877. doi: 10.1038/nrc2933 [doi].

335

Appendices

SUMMARY IN DUTCH / NEDERLANDSE SAMENVATTING ACKNOWLEDGEMENTS / DANKWOORD CURRICULUM VITAE LIST OF PUBLICATIONS

APPENDICES | SUMMARY IN DUTCH

SUMMARY IN DUTCH / NEDERLANDSE SAMENVATTING Introductie Met 1,7 miljoen nieuwe gevallen per jaar wereldwijd is borstkanker een van de meest voorkomende vormen van kanker bij vrouwen. Hoewel door screening en de ontwikkeling van nieuwe behandelingen de borstkankergerelateerde sterfte sterk is afgenomen, ontwikkelt ongeveer 30% van de borstkankerpatiĂŤnten uiteindelijk metastasen (uitzaaiingen). Uitgezaaide borstkanker is tot op heden ongeneeslijk en verantwoordelijk voor het grootste deel van deze sterfte. Daarom is het ophelderen van de onderliggende mechanismen die leiden tot het ontstaan van uitzaaiingen een van de belangrijkste uitdagingen binnen het borstkankeronderzoek. Kanker ontstaat door mutaties (veranderingen) in het DNA, het genetische materiaal van een cel. Deze mutaties kunnen aangeboren zijn en worden overgedragen aan de volgende generatie (erfelijke varianten van kanker), of nieuw ontstaan door schadelijke invloeden van buitenaf (bijvoorbeeld zonlicht of sigarettenrook). Bepaalde mutaties kunnen ervoor zorgen dat een cel zich ongeremd gaat delen, waardoor een tumor (gezwel) ontstaat die het gezonde weefsel verdringt. Er is dan sprake van een kwaadaardige tumor oftewel kanker. Indien deze in aanraking komt met bloed- of lymfevaten kan hij zich verspreiden door het hele lichaam. De kanker is dan uitgezaaid. Uitzaaiingen worden grofweg in twee groepen verdeeld: locoregionale of nabijgelegen uitzaaiingen en uitzaaiingen op afstand. Nabijgelegen uitzaaiingen komen het vaakst voor in de lymfeklieren van de oksel en in de borstwand. Afstandsuitzaaiingen uitten zich meestal in de botten, de hersenen en de lever. Dit proefschrift gaat met name over de uitzaaiingen op afstand. Het uitzaaiingsproces wordt vaak weergegeven als meerdere opeenvolgende, afgebakende stappen die cellen afkomstig van de primaire tumor doorlopen. De kankercel die in aanraking komt met bloed- of lymfevaten moet eerst los komen van de primaire tumor. Dan dient hij de eigenschap te bezitten om een vat binnen te dringen. In bloedvaten is het immuunsysteem actief, waarbij een kankercel ongezien moet blijven om te overleven. Vervolgens zal hij ergens anders in het lichaam blijven steken, om daar weer door de vaatwand naar buiten te dringen. Hierna probeert een kankercel te overleven in het orgaan waar hij terecht is gekomen. Dit orgaan heeft vaak hele andere weefselkenmerken dan de borst (de oorspronkelijke locatie), waardoor de kankercel zich moet aanpassen om te kunnen overleven. Uiteindelijk is de uitzaaiingscascade dus eerder een zeer ingewikkeld proces dan een afgebakende reeks van stappen, waarbij een complex scala van interacties tussen de kankercellen en de omgeving het beloop bepaalt. Mammacarcinoom (borstkanker) zaait het vaakst uit naar de longen, de lever, de hersenen en/of de botten. Dit is niet goed te verklaren door de route die voor een kankercel het gemakkelijkst af te leggen is. Goed doorbloede organen zoals de milt en het hart vertonen namelijk bijna nooit uitzaaiingen. Het lijkt daarom wel alsof borstkankercellen een voorkeur 338

Nederlandse samenvatting

hebben voor specifieke locaties, mogelijk door een (on)zichtbare interactie met de weefsels aldaar. Ook andere soorten kankers lijken een dergelijke voorkeur te hebben. Zo zaaien prostaattumoren meestal uit naar de botten, testis- (zaadbal) en niertumoren naar de longen en darmtumoren naar de lever. De precieze interacties tussen uitzaaiingen en doelorganen blijven nog onopgehelderd. Tegenwoordig worden borstkankers ingedeeld in verschillende typen. Er is namelijk gebleken dat de voorkeursplaats van en het risico op uitzaaiingen, alsook de prognose en de respons op behandeling kunnen worden toegeschreven aan het type borstkanker. Deze subtypering is hoofdzakelijk gebaseerd op de gevoeligheid voor hormonen. Hormoongevoelig betekent dat de hormonen oestrogeen en progesteron de kanker via receptoren kunnen stimuleren om te groeien en te delen. Wanneer deze hormonen geen invloed hebben op de kanker spreek je van hormoonongevoelig. Een hormoongevoelige kanker is dus oestrogeenreceptor (ERα) positief en/of progesteronreceptor (PR) positief. Vooral ERα wordt beschouwd als een belangrijke positieve prognostische marker en een voorspeller voor de respons op anti-hormonale therapie (ofwel endocriene therapie). De belangrijkste vormen van anti-hormonale therapie zijn tamoxifen en aromatase-remmers. Hoewel ongeveer 75% van de borstkankers ERα-positiviteit vertonen, is of wordt een groot deel van deze tumoren resistent tegen therapie. Een deel van deze resistentie wordt bepaald door verlies van de ERα-positiviteit. Deze kankers zijn vaak nog wel gevoelig voor androgenen (mannelijke geslachtshormonen). Androgenen komen in kleine hoeveelheden voor bij vrouwen en zijn van betekenis voor het libido, het ovulatieproces en de lichaamsbeharing. De androgeenreceptor komt tot uiting in ongeveer 60% van de borstkankers. Een andere belangrijke marker die wordt gebruikt voor subtypering en die als aangrijpingspunt kan dienen voor therapie, is Humane Epidermale groeifactor Receptor 2 (HER2). Dit is een eiwit dat op de buitenkant van cellen zit, celgroei kan stimuleren en daarmee het ontstaan van kanker kan beïnvloeden. Ongeveer 12% van de borstkankers is HER2-positief en is daarmee waarschijnlijk gevoelig voor anti-HER2-therapie. De meest bekend therapie is trastuzumab. ERα, PR, AR en HER2 zijn fenotypische markers (fenotype: waarneembare kenmerken), aangezien deze receptoren buiten de celkern in het cytoplasma zitten. Naast fenotypische markers bestaan er ook verschillende genotypische factoren (genotype: erfelijke informatie, gelokaliseerd in de celkern), zoals mutaties, waarop borstkanker geclassificeerd kan worden. Het mutatieprofiel kan veel informatie geven over het ontstaan en de agressiviteit van een tumor. Uitgebreid onderzoek laat reeds zien dat er verschillen en overeenkomsten zijn tussen hormoongevoeligheid en mutatieprofiel van primaire borsttumoren en lymfeklieruitzaaiingen van dezelfde patiënt. Er is echter nog steeds een relatieve schaarste aan studies gericht op afstandsuitzaaiingen, veroorzaakt door moeilijkheden bij het 339

APPENDICES | SUMMARY IN DUTCH

verkrijgen van weefsel (denk aan moeilijk te bereiken hersenen diepliggende botuitzaaiingen). Een gedetailleerde genotypische en fenotypische karakterisering van borstkankeruitzaaiingen zorgt ervoor dat therapie beter gericht en/of gepersonaliseerd kan worden, wat uiteindelijk zal leiden tot minder therapieresistentie en lagere sterfte. In dit proefschrift bespreken we de overeenkomsten en verschillen van uiteenlopende genotypische en fenotypische kenmerken in gepaarde primaire tumoren en afstandsuitzaaiingen van borstkanker.

DEEL 1 (Epi)genotypering van afstandsuitzaaiingen van borstkanker In het eerste deel van dit proefschrift richtten we ons op de genetische (DNA) en epigenetische (aanpassingen in of op het DNA) overeenkomsten en verschillen tussen primaire borsttumoren en gepaarde uitzaaiingen op afstand. In hoofdstuk 2 hebben we de mutatieprofielen bekeken van hormoonongevoelige tumoren die naar huid en hersenen zijn uitgezaaid, met de nadruk op mutaties waar inmiddels een gerichte therapie tegen bestaat. We zagen uitgebreide verschillen in mutaties tussen patiënten (inter-patiënt verschillen) en juist grote overeenkomsten in mutatieprofielen tussen de borsttumoren en uitzaaiingen van dezelfde patiënt (intra-patiënt overeenkomsten). In hoofdstuk 3 onderzochten we de aanwezigheid van APOBEC3B, een enzym dat een zeer hoge mutatiefrequentie veroorzaakt, waardoor tumorontwikkeling wordt versneld. We hebben aangetoond dat APOBEC3B-expressie van de primaire tumor grotendeels behouden blijft in uitzaaiingen van dezelfde patiënt, met een aanzienlijke toename in aantal kopieën. DNA methylatie is een epigenetische manier van het lichaam om het aflezen van DNA aan of uit te zetten. Dit kan belangrijk zijn bij kanker wanneer tumor suppressor genen (tumor onderdrukkende genen) worden ‘uitgezet’ of oncogenen (tumor ondersteunende genen) worden ‘aangezet’. In hoofdstuk 4 hebben we gekeken naar methylatie van tumor suppressor genen. Onze resultaten toonden dat methylatie van deze genen doorgaans verloren gaat in afstandsuitzaaiingen ten opzichte van de primaire tumoren. Dit verlies of deze herschikking van het methylatiepatroon kan worden verklaard door het feit dat, in tegenstelling tot de eerdergenoemde stabiele mutaties, epigenetische processen soms omkeerbaar zijn. Methylatie kan daarom niet als therapeutisch doel worden gebruikt tegen uitzaaiingen, maar kan mogelijk wel dienen als voorspeller van uitzaaiingslocatie. Er werd namelijk een specifiek methylatiepatroon gezien in de primaire tumoren die naar de hersenen, de huid, de lever of de longen waren uitgezaaid. Andere veelbelovende epigenetische aangrijpingspunten voor de behandeling van borstkanker zijn microRNAs (miRs). MiRs zijn zeer kleine stukjes RNA (boodschappers van DNA) die het aflezen van RNA kunnen blokkeren. Ontregeling van of door miRs komt in verschillende vormen van kanker voor en wordt geassocieerd met tumorinitiatie, 340

Nederlandse samenvatting

therapieresistentie en het uitzaaiingsproces. In hoofdstuk 5 onderzochten we de aanwezigheid van miRs in primaire borsttumoren en meerdere uitzaaiingen van dezelfde patiënt. We lieten zien dat de aanwezigheid van bepaalde miRs grotendeels blijft gehandhaafd in uitzaaiingen, maar vaak in een toenemende frequentie. Tevens demonstreerden we dat specifieke miRs in de primaire tumor uitzaaiingslocatie-specifiek zijn. MiR-106b-5p kan bijvoorbeeld worden gebruikt als een voorspellende factor voor longuitzaaiingen. Samenvattend werden er een aantal interessante overeenkomsten gezien tussen de bestudeerde (epi)genetische markers in het eerste deel van dit proefschrift. Het gedetecteerde patroon van mutaties, miRs en APOBEC3B-expressie werd grotendeels behouden in de uitzaaiing ten opzichte van de primaire borsttumor. Bovendien werd een aanzienlijke hoeveelheid nieuwe afwijkingen en/of een toename in frequentie van bestaande afwijkingen aangetoond. Zoals genoemd in de inleiding wordt een kankercel tijdens het uitzaaiingsproces blootgesteld aan een groot aantal beïnvloedende factoren, zoals de nieuwe omgeving waarin een cel terecht komt of bepaalde behandelingen met medicijnen (bijvoorbeeld antihormonale of chemotherapie). De extra afwijkingen die we vonden kunnen daarom een overlevingsvoordeel opleveren voor de uitzaaiing om deze invloeden te kunnen doorstaan. Een ander terugkerend thema is dat borstkanker een zeer heterogene ziekte is; geen twee borsttumoren toonden dezelfde (epi)genetische profielen. In deze context is het zeer interessant dat in de hoofdstukken 3, 4 en 5 uitzaaiingslocatie-specifieke overeenkomsten werden gevonden. Dit impliceert dat er toch enigszins orde en voorspelbaarheid bestaat in wat soms een willekeurige combinatie van genetische afwijkingen lijkt. Deze orde zou uiteindelijk kunnen worden gebruikt om de locatie van een onbekende primaire tumor te identificeren.

DEEL 2 Fenotypering van afstandsuitzaaiingen van borstkanker Zoals is vermeld in de inleiding worden de voorkeursplaats van uitzaaiingen en de prognose en respons op behandeling onder meer toegeschreven aan het subtype van borstkanker. Deze subtypering is grotendeels gebaseerd op ERα, PR en HER2. Een verschil in hormoonof HER2-receptor status in de uitzaaiing ten opzichte van de primaire tumor van dezelfde patiënt komt vaak voor. Dit wordt receptor conversie genoemd. In hoofdstuk 6 werd het veelvuldig voorkomen van receptor conversie bevestigd met conversiepercentages voor ERα, PR en HER2 van respectievelijk 19%, 31% en 10%. Tevens werd een grote mate van heterogeniteit gezien tussen patiënten, receptoren, gebruikte technieken en uitzaaiingslocaties. Daarom is het gebruik van gestandaardiseerde technieken voor de beoordeling van receptorstatus van het allergrootste belang voor het bewerkstelligen van gepersonaliseerde kankerzorg. 341

APPENDICES | SUMMARY IN DUTCH

Receptor conversie kan worden overschat door het ontkalkingsproces van botuitzaaiingen. Bot is een veelvoorkomende uitzaaiingslocatie van borstkanker. Om de weefselkenmerken van botuitzaaiingen te kunnen beoordelen wordt het vaak gedecalcificeerd (ontkalkt) met behulp van ontkalkingsbuffers. Vooral zure buffers worden ervan verdacht de beoordeelbaarheid van receptorpositiviteit te kunnen compromitteren. In hoofdstuk 7 hebben we aangetoond dat immuunhistochemie van ERα, PR en HER2 niet significant beïnvloed wordt door ontkalking met middelen die mierenzuur of EthyleenDiamineTetraAzijnzuur (EDTA) bevatten. Echter, kwantiteit en kwaliteit van het geïsoleerd DNA en RNA was wel verminderd. Om te voorkomen dat patiënten onterecht een anti-hormonale therapie wordt onthouden, dienen bevindingen op ontkalkt materiaal met de nodige voorzichtigheid te worden geïnterpreteerd. Receptor conversie van uitzaaiingen in pleuravocht (longvocht) en ascites (buikvocht) is niet eerder beschreven, terwijl dit veel voorkomende uitzaaiingslocaties zijn in de late fase van kanker. In hoofdstuk 8 hebben we aangetoond dat receptor conversie in uitzaaiingen in longvocht en buikvocht qua incidentie dezelfde trend volgt als solide afstandsuitzaaiingen, met conversiepercentages van respectievelijk 25-30%, 30-35%, 47-60% en 6-12% voor ERα, PR, AR en HER2. Ongeveer 27% van de patiënten vertoonden receptorstatusveranderingen die invloed kunnen hebben op de gevoeligheid voor anti-hormonale behandeling. Het is daarom zeer belangrijk om, indien mogelijk, de receptorstatus van uitzaaiingen te bepalen om therapie hier eventueel op aan te passen. Zoals beschreven in de inleiding, geeft de expressie van hormoonof HER2-receptoren geen garantie op gevoeligheid voor therapie. Sommige tumoren zijn vanaf het begin ongevoelig voor bepaalde therapieën (de zogenaamde intrinsieke resistentie) en de meeste uitgezaaide borsttumoren ontwikkelen uiteindelijk resistentie tegen anti-hormonale of mutatie-gerichte medicijnen (aangeduid als verworven resistentie). Een diepgaand inzicht in de mechanismen van resistentie zal leiden tot optimalisatie van behandeling en uiteindelijk tot lagere sterftecijfers. In hoofdstuk 9 presenteerden we een tot nu toe onbekend mechanisme van verworven resistentie tegen anti-hormonale therapie. In uitzaaiingen in longvocht hebben we aangetoond dat FOXA1- en GATA3-expressie verloren gaan onder druk van anti-hormonale therapie. FOXA1 en GATA3 zijn hulpeiwitten van de ERα-receptor en verlies van deze markers resulteerde in een niet-functionele ERαreceptor met verminderde respons op anti-hormonale geneesmiddelen. Tenslotte hebben we ons gericht op minder bekende biomarkers Gastrine-Loslatende Peptide Receptor (GRPR), C-X-C motief Chemokine Receptor type 4 (CXCR4) en SomatoSTatine Receptor 2 (SSTR) in hoofdstuk 10. Deze receptoren komen veel voor op kankercellen en kunnen radioactief gelabeld worden. Specifieke vormen van therapie en beeldvorming kunnen gericht worden op deze labels om de kanker te behandelen of op te sporen. Het was reeds bekend dat deze receptoren tot uiting kwamen op primaire borsttumoren en nu hebben wij de expressie vergeleken met uitzaaiingen van dezelfde 342

Nederlandse samenvatting

patiënten. Grote gelijkenis werd gezien in kankersamples van dezelfde patiënt, wat impliceert dat het richten van therapie of beeldvorming op CXCR4, SSTR en vooral GRPR potentieel zou kunnen bijdragen aan gepersonaliseerde patiëntenzorg. In tegenstelling tot de relatief stabiele (epi)genetische afwijkingen in het eerste deel van dit proefschrift, laten de fenotypische markers die getest zijn in het tweede deel grote verschillen zien tussen primaire borsttumoren en gepaarde uitzaaiingen op afstand. Het kan zijn dat de markers die we getest hebben vatbaarder zijn voor selectiedruk van anti-hormonale of chemotherapie dan (epi)genetische mutaties en variaties. Dit lijkt echter contra-intuïtief, aangezien fenotypische kenmerken het resultaat zijn van de aflezing van (epi)genetische informatie. In de afgelopen jaren is het verkrijgen van weefsel van uitzaaiingen in de richtlijnen doorgedrongen om uitgezaaide borstkanker optimaal te kunnen behandelen. Bij discrepanties in receptorstatus wordt nu geadviseerd om bij voorkeur de kenmerken van de uitzaaiing te volgen om therapie op te richten. Diepgaand onderzoek moet echter uitwijzen of het richten van therapie op de eigenschappen van de uitzaaiing in plaats van op de primaire tumor ook een overlevingswinst oplevert. Idealiter zouden onderzoekers een therapie willen ontwikkelen die werkt tegen alle borstkankertypen. We hebben echter laten zien dat geen enkele patiënt of kanker dezelfde fenotypische en/of genotypische kenmerken vertoont. De sleutel tot gerichte behandeling moet dus een op maat gesneden aanpak zijn per patiënt, ook wel gepersonaliseerde kankerzorg genoemd. Dat klinkt gemakkelijk, maar tot nu toe is er veel onderzoek verricht naar biomarkers die uiteindelijk toch niet effectief genoeg zijn voor toepassing in de kliniek. Daarnaast zijn de kosten voor gepersonaliseerde behandeling erg hoog. Meer onderzoek is daarom een vereiste, waarbij het vermijden van onnodige behandelingen zich uiteindelijk zal uitbetalen.

343

APPENDICES | ACKNOWLEDGEMENT

ACKNOWLEDGEMENTS / DANKWOORD Lieve vrienden, familie en collega’s, Afgelopen drie jaar zijn ontzettend mooi en leerzaam, maar vooral ook hectisch geweest. Zonder de hulp en steun van vele lieve mensen om mij heen was ik waarschijnlijk in zeven sloten tegelijk gelopen, volledig ondervoed geraakt en had dit proefschrift niet bestaan. Ik wil bij deze alle mensen bedanken die in welke vorm dan ook betrokken zijn geweest bij de totstandkoming van dit proefschrift. Daarbij wil ik enkele personen in het bijzonder in het zonnetje zetten*: Geachte prof. dr. van Diest, beste Paul. Dank voor het vertrouwen en de mogelijkheid om dit promotietraject bij jou te volgen! Ondanks je volle agenda wist je altijd tijd te maken voor wetenschap, maar vooral voor een goed gesprek. Je hebt me vanaf het begin gestimuleerd om zelf na te denken en je gevleugelde uitspraak ‘Niets is ooit altijd of nooit’ zal ik niet snel vergeten. Gekscherend zei je dat ik een van de weinige vrouwen ben die je niet aan het huilen hebt gekregen. Helaas heb ik je kort geleden toch teleur moeten stellen, maar dat was natuurlijk alleen omdat ik nog niet weg wilde bij de pathologie. Geachte prof. dr. van der Wall, beste Elsken. Ook al heb je afgelopen jaar weinig tijd gehad voor onderzoek, ik heb je vooral tijdens de Breast Cancer Research meeting ervaren als een zeer betrokken en vertrouwen gevende begeleider. Als arts en als persoon zie ik je als voorbeeld! Dank voor de mentale steun. Beste dr. Moelans, lieve Cathy. Ik had me geen betere begeleider kunnen wensen! Wat heb jij een boel geduld gehad met dit onwetende, veel te directe, Hollandse labgroentje. Ik heb veel van je geleerd op onderzoeksgebied en natuurlijk wat betreft de Vlaamse taal. Ik draai mijn hand nu niet meer om voor het ‘inkloppen in de computer’ en het ‘dingen bij elkaar kappen’. Ik wens je heel veel succes met je nieuwe huisje en het klussen! Geachte leden van de leescommissie: prof.dr. Kranenburg, prof. dr. Linn, prof. dr. Borel Rinkes, prof. dr. de Vries en dr. Zwart. Hartelijk dank voor het zitting nemen in mijn commissie en het beoordelen van mijn proefschrift. Geachte opponenten, hartelijk dank voor het bijwonen van mijn verdediging en het voeren van mijn oppositie. Lieve Willy, wat had ik zonder jou gemoeten de afgelopen jaren? Als een ware leeuwin verdedig jij de agenda van Paul, maar er is altijd wel een gaatje te vinden. Meestal werden 344

Dankwoord

de lunchpauzes opgeofferd, maar de snoepjes tijdens het wachten hebben veel goed gemaakt. Dank voor je flexibiliteit, je interesse en de zorg voor mijn plantje! Lieve Natalie, van analist, naar onderzoeker naar Pathasser/mammameisje. Wat heb jij veel geleerd afgelopen jaar! Het is een groot gemis voor de mammagroep dat jij je nu minder op onderzoek kan richten. Jij zorgde altijd dat kleuringen en blokjes binnen no time klaar waren. Ons experiment met de Vibratome is helaas geen succes geworden, maar het uitje naar Eindhoven met een heerlijke lunch was het helemaal waard. Beste prof. dr. Goldschmeding, prof.dr. Offerhaus, dr. de Weger, dr. Willems, dr. Nguyen, en dr. Bovenschen, dank voor de waardevolle wetenschappelijke bijdrage aan de aio/ postdoc meetings. Beste Roel, Jan en Folkert, steunpilaren van het PRL! Wat zouden we zonder jullie moeten beginnen? Dank voor het geduld, de uitleg en jullie tijd om de vaak domme vragen te beantwoorden. Roel, onderzoek doen met jou is altijd een vrolijke aangelegenheid! Mijn excuses voor de soms rondslingerende pleuravochten in het kweeklab. Folkert, ik ga je aanstekelijke lach op de vroege morgen erg missen! Jan, ik ben benieuwd naar de wijnproeverij. Hopelijk zien we elkaar daar! Beste deelnemers aan de Breast Cancer Research meeting, lieve Robert, Milou, Patrick, Miranda, Gwen, Shoko, Marijn, Mirthe en Jolien. Dank voor de waardevolle opmerkingen en de opbouwende kritiek. Veel succes met de toekomstige onderzoeksplannen en carrières! Lieve Marina, Wendy en Petra, de paden op de lanen in! Jullie hebben het wandelen tot een kunst verheven. De ‘a Sister’s Hope Walk’ was een persoonlijke overwinning voor mij. Dank voor de wandelgezelligheid en de goede gesprekken! Beste medewerkers van de weefselfaciliteit: Annette, Jan, Doménico, Gladys, Frederic en Jeffrey. Dank voor jullie flexibiliteit en bereidheid om te helpen met mijn onderzoek. Zonder jullie snelle service bij het opzoeken en doorvoeren van blokjes en coupes had ik de decalcificatiestudie en het pleuravochtproject niet tot een goed einde kunnen brengen. Beste medewerkers van de diagnostische laboratoria histologie, cytologie en moleculaire pathologie. Dank voor jullie geduld en hulp bij het beantwoorden van vragen en het gebruikmaken van jullie faciliteiten. In het bijzonder wil ik graag Ton, Marja, Yvon, Angelique en Petra bedanken voor het meedenken met de verschillende studies. Beste co-auteurs: Karianne, Annelot, Britta, Carolien, Anieta, Simone en Karijn. Dank voor de prettige samenwerking! 345

APPENDICES | ACKNOWLEDGEMENT

Lieve Marise en Aernoud, paranimftoppers! Jullie zijn mijn maatjes en partners in crime bij de pathologie. Marise, ontzettend knap hoe jij je staande hebt weten te houden tussen twee pittige professoren; heen en weer geslingerd tussen Baltimore en Utrecht. Heel veel succes bij de interne in Amersfoort, ik denk dat ze aan jou een zeer goede, betrokken collega hebben! Aernoud, pappa met roze cellen! Altijd in voor een dropje en synchroon luisteren naar 3FM; ik heb ongelooflijk veel lol met je gehad als buurman op het PRL. Veel succes met de afronding van je project! Lieve PRLâ&#x20AC;&#x2122;ers: Joost, Liling, Hiroshi, Ellen, Emma, Pauline, Rob, Koos, Justin, Huiying, Robert, Jeroen, Manon, Quirine en Laurien. Ik heb een geweldige tijd met jullie gehad op het lab en in de kroeg! Wie durft nog te beweren dat pathologie saai is?! De vele Micaffe, Pitstop, nespresso en ijsjes in de zon momenten hebben me op de been weten te houden de afgelopen jaren. Joost, bijna achterbuurman op werk en thuis. Succes met de toekomstige kaakchirurgieopleiding en op naar de 200+ publications! Liling, good luck with your new house and the new student that you need to supervise. Hiroshi, I hope you will get your PhD next year. Good luck at the department of nephrology in Japan in the future! Ellen, lieve labmamma. Heel knap hoe jij van onderwerp hebt weten te switchen en je staande hebt kunnen houden tussen al die domme artsen. Laten we snel weer een spelletje doen! Emma, bridgekoningin. Al heb je de spelregels al vaak genoeg uitgelegd, ik snap er nog steeds niets van. Succes met de hypoxia working station en je toekomstige medische carriĂ¨re. Ik ben benieuwd welke richting het gaat worden. Lieve Pauline, gek, slim, bierverslaafd praatmaatje met je grote mond! Menig bierglas is in dat gapende gat verdwenen en nooit meer terug gevonden. Ik kijk uit naar onze volgende hardloopsessie en de voorstelling van het Ro Theater. Rob, al ben je alweer een jaar bezig met de TOVA, we missen je hier nog elke dag. Geen koffie of biermomentje sloeg jij over en met hockey ook altijd van de partij. Ik wens je veel succes bij de kaakchirurgie, en met je vorderingen op het datingvlak. Koos, ik hoop dat je aanstekelijke lach en je research kwaliteiten je uiteindelijk een opleidingsplek op zullen leveren. Succes bij de pathologie! Justin, getrouwde man. Ik vond het een eer om je bruiloft bij te wonen. Succes met de afronding van je promotietraject! Huiying, how was your trip through Europe? I miss the sight of your sleepy head in the mornings. I hope you enjoy working at the immunology department! Robert, ik heb nog nooit iemand zo gek een banaan zien eten! Jammer dat je postdoc-positie in het Karolinska je niet heeft gebracht wat je hoopte. Succes met de banenzoektocht! Jeroen, hoewel jij het niet zo nauw neemt met de IHC protocollen kreeg je altijd een goede kleuring. Succes met de volgende sollicitatieronde! Manon, van HTX naar de moleculaire pathologie. Knap hoe jij duizend dingen tegelijk kan. Succes met de drukte! Quirine, gezellige, Leidse deerne. Van de chirurgie naar de pathologie was in het begin een enorme cultuurshock, maar je weet je mannetje te staan. Fijn dat jij de BCRM-planning over hebt kunnen en willen nemen! Laurien, mede door jouw gestructureerde database-indeling heb ik deze studies kunnen 346

Dankwoord

verrichten. Enorm bedankt! Succes bij het KWF. Nog een klein verzoekje: zou je Lucas alsjeblieft een keertje kunnen laten winnen met squash? Beste Kenny, toen ik mijn kruisband scheurde was jij mijn redder in nood. Ik verdenk je er nog steeds van dat je geniet van het pijnigen van je patiënten, maar dat kraken/ dubbelvouwen heeft me wel weer op de been gebracht. Ontzettend bedankt voor een intensief jaar trainen! Lieve collega’s van de chirurgie in het St. Antonius, dank voor een mooie en intensieve tijd! Maar Roks, Benders, Keijsers, van Vugt, Zhu en Ochtman, wanneer gaan we nou eindelijk weer eens een biertje drinken? Lieve collega’s van het Jeroen Bosch, dank voor de gezelligheid en de flexibiliteit om diensten te ruilen de afgelopen tijd! Ik heb het werken met jullie vanaf het begin als een warm bad ervaren (letterlijk wat betreft de bubbels tot aan het plafond in de Ardennen). Ik kijk uit naar vele borrels en drukke diensten samen! Lieve Shoqies, vanaf het begin zijn jullie enorm bedrokken geweest bij mij en mijn onderzoek. Dank voor de interesse, de fijne snaaiavondjes en de mooie clubweekenden! Lieve Rosie, ook al kan ik jouw cum laude niet evenaren, jij bent mijn grote voorbeeld als multitasker en doorzetter. Succes bij de gyn! Lieve Bekkers, op de app ben jij het meest actief en je maakt van ieder evenement een feestje. Zet hem op met de banenjacht! Lieve Siem, als derde clubhuisbewoner bied jij altijd een luisterend oor en ben je ontzettend attent. Hopelijk tot snel in Uutje. Lieve Hilles, als burgervrouw in Amersfoort heb je het toch mooi voor mekaar in jullie huis vol gadgets. Ik kom graag snel een keertje langs! Lieve Linnie, jij werkt jezelf naar de top bij de belastingdienst. Geniet van de voorbereidingen voor jullie huwelijk. Lieve Mies, verliefd, verloofd en bijna getrouwd. Altijd bezig met ‘supervetteshit’ en ambitieus in je werk; zet hem op! Lieve Eef, geen beer op de weg die jij niet aankan. Ik ben zeer trots op je keuze om een andere baan te zoeken! Lieve Bel, fijn om te zien dat je jezelf gevonden hebt in the Big Apple en dat je je scriptie zo goed hebt afgerond. Ik wens je veel succes met de sollicitaties voor je droombaan. Lieve Ronnie, in het verre noorden vordert jouw PhD-traject gestaag, maar dat kan ook niet anders met een bijna-professor als vriend. Ik ben benieuwd wat je carrière je gaat brengen. Lieve Lot, met één been in Marokko weet jij toch de opleiding orthopedie glansrijk te doorlopen. Ik bewonder de manier waarop jij je eigen (toekomst)plan trekt. Lieve Tess, heerlijke yup die je bent. Jij hebt je schaapjes op het droge en weet met een lach elk gaatje te vullen. Lieve Rolo, afgelopen jaar heb je een nieuwe baan gevonden en ben je gaan samenwonen. Veel succes met alle veranderingen!

347

APPENDICES | ACKNOWLEDGEMENT

Lieve lievelingslichtings, Kar en Step (het mobiele duo) en El. Vanaf het begin waren wij heel hecht en nog steeds kijk ik uit naar onze goede gesprekken in de sauna. Dank voor jullie onvoorwaardelijke steun! Lieve Kar, als maatje in Utrecht voel ik me helemaal thuis in jullie nieuwe huis. Dank voor de fijne zwemavondjes toen ik last had van mijn knie, de heerlijke maaltijden die je kookt, de spontane koffietjes op zondagmiddag en je belletjes vanuit de auto. Lieve Step, ik bewonder de manier waarop jij je afgelopen tijd door een moeilijke periode hebt geslagen. Jij weet altijd goed te verwoorden waar het probleem zit en laat je niet snel gek maken. Een goede rode wijn is de beste remedie. Zet hem op bij Microsoft! Lieve El, als carrièretijger ook dit keer weer succes. Ik ga je ontzettend missen als je straks in Singapore/Fontainebleau zit, maar ik beloof bij deze plechtig dat ik je op kom zoeken. Wie weet kunnen we daar ook wel stroekie doeken! Lieve dispuutsgenootjes van Sarasvati, dank voor jullie steun en interesse! Lieve (bedrijfs)hockey teamies, sorry voor de afwezigheid van de laatste tijd. Na de winterstop ben ik weer helemaal van de partij! Ik krijg ontzettend veel energie van jullie en we gaan dit jaar natuurlijk weer voor het kampioenschap! Lieve (oud)huisgenootjes van Huize Hofstee, de Cliviazusters en Club R 1.0, 2.0 en 3.0. Ik zou me niet voor kunnen stellen hoe het is om alleen te wonen. Dank voor de vele huisavonden, pubquizen, biertjes/wijntjes/borrelplanken, Stef Stuntpilootsessies en bergen thee met chocolade. Altijd een luisterend oor en veel gezelligheid. Lieve Sas, wat hebben wij een mooie tijd gehad in Uganda, Tanzania en Barcelona. Al zijn we op sommige vlakken tegenpolen, ik kijk altijd uit naar gezellige avondjes samen eten en wijntjes drinken. Erg knap hoe jij een opleidingsplek klinische genetica binnen gekopt hebt. Veel succes met je onderzoek! Lieve Fré, ook jij bent erg druk met onderzoek. En dan ook nog een opleidingsplek! De laatste tijd leefde jij uit een koffer tijdens de verbouwing van jullie huisje. Ik hoop dat je inmiddels gesetteld bent in een mooi paleisje. Lieve Meik, net als Fré heb jij voor scopieëren als vakgebied gekozen. Tot die tijd nog even zo veel mogelijk buitenlandse congressen meepakken; knap hoor! Laten we snel weer wat gaan drinken met zijn drieën! Lieve Taar, harde werker! Wat goed dat jij Studelta en de Pabo met zo’n sportief en sociaal leven kunt combineren. Ook al vind ik het jammer dat je weg gaat, ik wens je ontzettend veel plezier met Thijs in een nieuw appartementje. Lieve Loes, jij hebt van Club R echt een knus paleisje gemaakt. Altijd in voor bank hangen en lekker uit eten, maar ik hoop binnenkort ook eindelijk eens te winnen met squash! Lieve Siem, ik heb een geweldige tijd met je gehad in Utrecht. Als lief dispuutsgenootje ben je nooit ver weg en je doet het ook heel goed als surrogaatclubgenoot. Succes bij de KNO!

348

Dankwoord

Lieve familie Thijssen, wat bof ik met zo’n grote, betrokken familie! Jullie hebben me gemotiveerd om te gaan studeren en zijn altijd geïnteresseerd in mijn carrièrevorderingen. Lieve Imke en Marloes, ik hoop snel weer een fijne nichtjesdag met jullie te plannen! Lieve opa, ik ben erg blij en trots dat u aanwezig zult zijn bij mijn promotie. Ook al wil het lichaam niet altijd meer even snel, de geest blijft relatief sterk. Dank voor uw interesse in mijn onderzoek! Lieve Hidde, Stijn en Tjomme, wat ben ik ongelooflijk blij met jullie! Al sloegen wij elkaar vroeger bijna de hersens in, nu besef ik me hoezeer ik het heb getroffen met zulke knappe en fijne broers! Lieve Hidde, ik vind het bewonderenswaardig hoe hard jij nu voor je studie werkt en daarnaast al 3 jaar midden in de verbouwing zit. Ik hoop dat de badkamer in 2038 wel klaar zal zijn ;) Succes met je nieuwe baan! Lieve Stijn, met je geweldige gevoel voor humor weet jij altijd iedereens hart te winnen. Ik vind het fijn om te zien dat je het nu zo naar je zin hebt op je werk en hoop dat in het voorjaar je contract zal worden verlengd. Wat je ook voor toekomstpad zult kiezen, weet dat ik onvoorwaardelijk achter je sta! Lieve Tjomme, ik ben trots op de manier waarop jij de veranderingen van afgelopen jaar hebt weten te doorstaan. Een wisseling van baan (en toch weer niet), verhuisd naar Zaandam en je relatie uit (en misschien wel weer iets nieuws); jij laat je niet meer zo snel gek maken. Ik kijk uit naar nog vele biosbezoekjes en een wijncursus binnenkort. Lieve moeders, ik heb dit proefschrift aan je opgedragen omdat ik ontzettend dankbaar ben voor de liefde, het vertrouwen en de wijsheid die je me meegegeven hebt. Zonder jouw steun was ik nooit zo ver gekomen en ik heb veel geleerd van jouw zelfstandigheid en doortastende aanpak. Ik kijk erg op tegen de manier waarop jij als fulltime werkende moeder met vier kinderen, en nu ook een samengesteld gezin, alle balletjes hoog weet te houden. En je weet het hè, ik hou van je en ik ben trots op je! Lieve Lucas, dankjewel dat jij lief en (onderzoeks)leed met me deelt en het beste in me naar boven haalt! Ik vind het fijn dat jij me bijdehante feitjes vertelt, die ik eigenlijk niet wil weten. Ik lach om (bijna) al je flauwe woordgrapjes. Ik geniet als we samen kibbelen over…, tja, waarover niet eigenlijk? Het maakt me niet uit dat jij wel je hoofd-, maar niet je wilde haren verliest. Wat ooit begon als onderzoekscollega’s, zal met de afronding van dit proefschrift niet zomaar eindigen! Veel liefs!

* Disclaimer: mogelijk is deze lijst niet volledig. Mijn oprechte excuses indien ik iemand vergeten ben en bij deze alsnog veel dank voor alles!

349

APPENDICES | CURRICULUM VITAE

CURRICULUM VITAE Willemijne Alberta Maria Elisabeth Schrijver was born on April 14th, 1987 in Nijmegen, the Netherlands. After obtaining her gymnasium diploma at the Merletcollege in Cuijk she started the study Psychology in Groningen in 2005. In 2006, she obtained her propaedeutic diploma in psychology and enrolled in medical school. During her studies she engaged in several extracurricular activities. She continued her fifth year of studies in the Isala Klinieken in Zwolle and went to Uganda and Tanzania in the summer of 2012 for an internship in social medicine. Willemijne completed her final year of medical school at the department of Plastic Surgery and Otorhinolaryngology at the University Medical Center Utrecht. In 2013 she obtained her medical degree and worked as a physician at the surgery department of the St. Antonius Hospital in Nieuwegein for a year. In 2014, she began her PhD project under supervision of prof. dr. van Diest and dr. Moelans on a Dutch Cancer Foundation grant. In three years she laid the foundation for this dissertation and successfully applied for a ‘a Sister’s Hope’ research grant, which resulted in chapter 5. Several projects have been presented at national and international conferences. In July 2016, Willemijne started working as a physician at the surgery department of the Jeroen Bosch hospital in Den Bosch.

350

List of publications

LIST OF PUBLICATIONS Schrijver WA, van Diest PJ, Dutch Distant Breast Cancer Metastases Consortium, Moelans CB. Unravelling site-specific breast cancer metastasis: a microRNA expression profiling study. Oncotarget. 2016. Schrijver WA, van der Groep P, Hoefnagel LD, et al. Influence of decalcification procedures on immunohistochemistry and molecular pathology in breast cancer. Mod Pathol. 2016. doi: 10.1038/modpathol.2016.116 [doi]. Schrijver WA, Jiwa LS, van Diest PJ, Moelans CB. Promoter hypermethylation profiling of distant breast cancer metastases. Breast Cancer Res Treat. 2015;151(1):41-55. doi: 10.1007/ s10549-015-3362-y [doi]. van Kalkeren TA, van der Houwen EB, Schrijver WA, Verkerke GJ, van der Laan BF. In vivo test of a new hands-free tracheostoma inhalation valve, a randomised crossover study. Clin Otolaryngol. 2013;38(5):420-424. doi: 10.1111/coa.12168 [doi].

351

Genotyping and phenotyping of distant breast cancer metastases ÂŠ Willemijne A.M.E. Schrijver, 2016