Retrospective study of cytologic features of well-differentiatedhepatocellular carcinoma in dogs by Francisco Alvarez

Veterinary Clinical Pathology ISSN 0275-6382

ORIGINAL RESEARCH

Retrospective study of cytologic features of well-differentiated hepatocellular carcinoma in dogs Carlo Masserdotti1, Michele Drigo2 1

San Marco Private Veterinary Laboratory, Padua, Italy; and 2Department of Animal Medicine, Production and Health, Padua University, Padua, Italy

Key Words Acinar, cytology, liver, naked nuclei, palisading Correspondence Carlo Masserdotti, Laboratorio Veterinario San Marco, Via Sorio 114/C, 35141 Padua, Italy E-mail: carlo.masserdotti@gmail.com DOI:10.1111/j.1939-165X.2012.00436.x

Background: Cytologic diagnosis of hepatocellular carcinoma is possible when hepatocytes have prominent criteria of malignancy, but is problematic when hepatocytes have a relatively normal morphologic appearance. Objectives: The aim of this study was to determine the cytologic features of histologically confirmed well-differentiated hepatocellular carcinomas (WD-HCC) in dogs to develop reasonable criteria for an accurate diagnosis when obvious criteria of malignancy in hepatocytes are lacking. Methods: Slides of ultrasound-guided fine-needle aspirates of 15 histologically confirmed WD-HCC in dogs were evaluated retrospectively using a scoring system (0–3) established for 33 cytologic features. For comparison, 15 cytologic samples of non-nodular non-neoplastic liver were evaluated using the same criteria. Statistical analysis included multiple hypothesis testing using the Benjamini–Hochberg method or false discovery rate control to correct for multiple comparisons. Results: The most significant and useful cytologic features in the diagnosis of canine WD-HCC were dissociation of hepatocytes, acinar or palisading arrangements of neoplastic cells, and the presence of naked nuclei and capillaries, together with mild anisocytosis, anisokaryosis, multinuclearity, and increased N:C ratios. Conclusions: Cytologic features of canine WD-HCC were determined. Further studies are warranted to compare these cytologic criteria among WDHCC, hepatomas, and hyperplastic nodules and to examine co-variance of the 33 cytologic features.

Introduction Cytologic diagnosis of hepatocellular carcinoma is possible when hepatocytes have prominent criteria of malignancy,1 but is problematic when hepatocytes have a relatively normal morphologic appearance. In a study of 110 primary hepatic neoplasms in dogs, 55 hepatocellular carcinomas were classified into 11 groups based on architectural patterns in histologic sections in which the neoplastic hepatocytes varied from nearly normal-appearing hepatocytes to highly anaplastic spindle-shaped cells.2 Cytomorphologic descriptions of well-differentiated hepatocellular carcinoma have been reported in human medicine.3 However, in veterinary medicine cytologic criteria for diagnosis of this tumor are available only in textbooks,1,4 and the difficulty in cytologic differentiation of well-differentiated hepatocellular carcinoma

from adenoma and hyperplastic nodules is well recognized.4 Thus, it would be useful to develop cytologic criteria for diagnosis of hepatocellular malignancies. The aim of this study was to determine the cytologic features of histologically confirmed well-differentiated hepatocellular carcinomas in dogs to develop reasonable criteria for an accurate diagnosis when obvious criteria of malignancy in hepatocytes are lacking.

Materials and Methods All cytologic and histologic specimens evaluated in this study were obtained with consent of the owners. Fifteen cytologic samples of well-differentiated hepatocellular carcinoma (WD-HCC) confirmed by histologic examination of excised masses, collected from dogs from September 2002 to October 2009 at the

Cytology of canine hepatocellular carcinoma

Sant’Antonio Veterinary Clinic, Salo`, Italy and at the San Marco Veterinary Laboratory, Padua, Italy, were identified from the archives of both institutions and reviewed. All samples were collected by ultrasoundguided fine-needle aspiration (FNA) of hepatic masses5,6 using a 25-gauge spinal needle (BD Spinal Needle, Becton Dickinson, Madrid, Spain). Aspirated material was spread on glass slides, air-dried, and stained with May–Gru¨nwald Giemsa (Fisher Diagnostics, Middletown, VA, USA). For each sample, cytologic features selected based on personal experience of the authors and on published reports in human medicine7 were evaluated. Necrosis was defined cytologically as clumped or dispersed cellular debris in the background and by cells without recognizable features. Acinar/ pseudoacinar and palisading arrangements of hepatocytes were indentified as small clusters of hepatocytes organized around a small central space and as hepatocytes aligned in rows, respectively.8 Trabeculae were represented by elongated clusters of cells with a width of 3–20 cells. Dissociation of hepatocytes was identified as loss of cohesiveness of peripheral cells, either singly or in small groups, from clusters of hepatocytes. Multinuclearity was defined as cells with 3 or more nuclei. Each feature was evaluated at 4009 in 20 randomly selected fields, described, and given a score of 0–3 (Table 1). Chromatin pattern, irregularity of nuclear shape, nuclear pseudoinclusions, and nucleoli were evaluated at 10009 in 20 random fields. Lipofuscin, vacuolar change characterized by decreased cytoplasmic density (glycogen-type), and lipid vacuoles were evaluated morphologically based on previous descriptions1,4 and scored. For comparison, 15 cytologic samples from nonneoplastic hepatic parenchyma were collected by ultrasound-guided FNA of liver during staging of extrahepatic diseases, such mast cell tumor, or for clinical management of hepatic disease. Hepatic nodules were not aspirated. Surgical biopsy, using a 16-gauge biopsy needle (Bard Magnum, Bard Peripheral Vascular Inc, Tempe, AZ, USA), was also performed. Final diagnoses confirmed the non-neoplastic nature of the pathologic process and ranged from normal liver (4 cases), nonspecific vacuolar hepatopathy (6 cases), suppurative hepatitis (2 cases), and chronic idiopathic hepatitis (3 cases). Statistical analysis was performed using a commercially available software program (SPSS, version 12.0.0, SPSS Inc, Chicago, IL, USA). Multiple hypothesis testing using the Benjamini–Hochberg method or false discovery rate (FDR) control to correct for multiple comparisons was performed.9 “U” was the value of the statistical test. An FDR of .018 was used as a critical

Masserdotti and Drigo

value to assess whether the obtained P values were significant.

Results Ages of the 6 female and 9 male dogs with hepatocellular carcinoma ranged from 9 to 18 years (mean, 12 years; median, 11 years). Ages of dogs with nonneoplastic hepatic disease ranged from 9 to 14 years (mean, 11 years; median, 10 years). Cytologic examination revealed significant differences in criteria between WD-HCC and non-neoplastic hepatic disease (Table 2). Criteria for which significant differences were found in the cytologic samples of WD-HCC were the presence of necrosis (P = .016), architectural features, such as acinar (P = .035) (Figure 1) and palisading (P = .001) (Figure 2) arrangements, dissociation (P < .001) (Figure 3), and the presence of capillaries (P = .016) (Figure 4) among epithelial clusters. The presence of naked nuclei (P < .001) (Figure 5), anisokaryosis (P = .004), which was usually mild, increased N:C ratios (P < .001), and multiple nuclei (P < .001) (Figure 6) also were significant criteria. Criteria of malignancy that were not significantly different included the presence of multiple and prominent nucleoli (P = .024), anisocytosis (P = .038), cytoplasmic basophilia (P = .277), irregularity in chromatin pattern (P = .073), irregularity of nuclear shape (P = .148), and the presence of mitotic figures (P = .15). Fusiform cells were found among neoplastic hepatocytes in 46.6% of WD-HCC and were considered indicators of fibrosis (Figure 7).

Discussion In this study, we identified cytologic criteria for the diagnosis of WD-HCC. HCC is an uncommon tumor in dogs, and the most typical presentation is a large mass in a single lobe or in contiguous lobes of the liver.10,11 Hepatocellular neoplasms may be diagnosed by cytologic examination,12 especially when there is marked cellular atypia, ie, high N:C ratios, macronuclei and macronucleoli, deeply blue cytoplasm, and marked anisocytosis and anisokaryosis. However, if atypia is not marked, nonneoplastic hyperplastic nodule or hepatocellular adenoma (hepatoma) are possible cytologic differential diagnoses,1,4 a conundrum that also occurs in human medicine.7,13–16 Histologically, HCC can be differentiated from hepatocellular adenoma by infiltrative

Masserdotti and Drigo

Cytology of canine hepatocellular carcinoma

Table 1. Quantitative and qualitative scoring system established for 33 for cytologic features for evaluation of fine-needle aspirates of canine liver. Score Cytologic Feature Proteinaceous debris Blood in background Bile casts Necrosis Cellularity Trabecular arrangements Rows of cuboidal cells Acinar arrangements Palisading arrangements Dissociation Fibrosis Capillaries Cytoplasmic basophilia Lipofuscin Lipidosis Vacuolar change (glycogentype) Cytoplasmic inclusions Anisocytosis Increased N:C ratio Anisokaryosis Chromatin pattern Irregular nuclear shapes

Absent Absent Absent Absent

Widespread Widespread Widespread Widespread 20–50 cells/HPF 2–3/10 HPF 2–3/10 HPF 2–3/10 HPF 2–3/10 HPF 3–5 single cells/10 hepatocytes in a cluster

Abundant Abundant Abundant Abundant > 50 cells/HPF > 3/10 HPF > 3/10 HPF > 3/10 HPF > 3/10 HPF

Absent

Scattered Scattered Scattered Scattered 10–20 cells/HPF 1/10 HPF 1/10 HPF 1/10 HPF 1/10 HPF 1–3 single cells/10 hepatocytes in a cluster 1 fusiform cell/20 hepatocytes

1–3 fusiform cell/20 hepatocytes

None Homogeneous Absent Absent Absent

1/HPF Patchy < 30% of hepatocytes < 30% of hepatocytes < 30% of hepatocytes

2-5/HPF

> 5 fusiform cells/20 hepatocytes > 5/HPF

30–60% of hepatocytes 30–60% of hepatocytes 30–60% of hepatocytes

> 60% of hepatocytes > 60% of hepatocytes > 60% of hepatocytes

Absent Absent Absent Absent

30–60% of hepatocytes 20–60% of hepatocytes 20–60% of hepatocytes

> 60% of hepatocytes > 60% of hepatocytes > 60% of hepatocytes

Coarse

Clumped

1–3/HPF

> 3/HPF

Absent Normal Absent Absent Absent Absent

< 30% of hepatocytes < 20% of hepatocytes < 20% of hepatocytes Present in about 10% of hepatocytes Granular Present in about 10% of hepatocytes Present in about 10% of hepatocytes Present in about 10% of hepatocytes 1/HPF 0–1/10HPF Prominent 1–2/HPF 2–3 cells/HPF 2–3 cells/HPF 2–3 cells/HPF

Multiple 2–10/HPF 4–10 cells/HPF 4–10 cells/HPF 4–10 cells/HPF

Multiple and prominent > 10/HPF > 10 cells/HPF > 10 cells/HPF > 10 cells/HPF

Absent Absent

2–3 cells/HPF 1 cell/HPF

4–10 cells/HPF 2–3 cells/HPF

> 10 cells/HPF > 3 cells/HPF

Absent Absent Absent Absent No single cells

Compact Absent

Nuclear pseudoinclusions

Absent

Nuclear eccentricity

Absent

Multiple nuclei (3 or more) Mitotic figures Nucleoli Naked nuclei Neutrophilic inflammation Macrophagic inflammation Lymphoplasmacytic inflammation Eosinophilic inflammation Mast cells

HPF indicates high-power field (4009).

growth, regional or diffuse pleomorphism, and the presence of atypical cells or mitotic figures.17 Necrosis was observed in one-third of the cytologic samples of carcinomas, and this finding was not unexpected and has been previously reported for histologic sections of this neoplasm.11 In people, necrosis was observed in 17% of WD-HCC, 37% of moderately differentiated HCC, and 50% of poorly differentiated HCC.13

Although acinar arrangements were observed only occasionally, this feature was absent in samples of non-neoplastic liver. These arrangements, better defined as “pseudoacinar,” are described in cytologic preparations of human WD-HCC and also in moderately or poorly differentiated tumors,15 and, when present, must be distinguished from metastatic glandular tumors, cholangiocarcinoma, and the adenocarcinoma component of combined

Cytology of canine hepatocellular carcinoma

Masserdotti and Drigo

Table 2. Scores for 33 cytologic features evaluated in fine-needle aspirates of 15 hepatocellular carcinomas and 15 samples from non-neoplastic, nonnodular liver in dogs. Score Cytologic Feature

Diagnosis

P†

Proteinaceous debris

WD-HCC Control WD-HCC Control WD-HCC Control WD-HCC Control WD-HCC Control WD-HCC Control WD-HCC Control WD-HCC Control WD-HCC Control WD-HCC Control WD-HCC Control WD-HCC Control WD-HCC Control WD-HCC Control WD-HCC Control WD-HCC Control WD-HCC Control WD-HCC Control WD-HCC Control WD-HCC Control WD-HCC Control WD-HCC Control WD-HCC Control WD-HCC Control WD-HCC Control WD-HCC Control

3 (20%) 7 (46.7%) 0 0 12 (80%) 13 (86.7%) 10 (66.7%) 15 (100%) 0 0 0 0 3 (20%) 5 (33.3%) 11 (73.3%) 15 (100%) 7 (46.7%) 15 (100%) 0 12 (80%) 8 (53.3%) 12 (80%) 10 (66.7%) 15 (100%) 10 (66.7%) 7 (46.7%) 7 (46.7%) 2 (13.3%) 8 (53.3%) 12 (80%) 8 (53.3%) 3 (20%) 12 (80%) 15 (100%) 5 (33.3%) 10 (66.7%) 3 (20%) 13 (86.7%) 5 (33.3%) 14 (93.3%) 0 0 11 (73.3%) 14 (93.3%) 12 (80%) 11 (73.3%) 7 (46.7%) 2 (13.3%) 6 (40%) 15 (100%) 13 (86.7%) 15 (100%)

5 (33.3%) 8 (53.3) 4 (26.7%) 4 (26.7%) 1 (6.7%) 2 (13.3%) 4 (26.7%) 0 0 1 (6.7%) 5 (33.3%) 3 (20%) 5 (33.3%) 7 (46.7%) 3 (20%) 0 7 (46.7%) 0 6 (40%) 2 (13.3%) 5 (33.3%) 3 (20%) 3 (20%) 0 5 (33.3%) 8 (53.3%) 8 (53.3%) 5 (33.3%) 5 (33.3%) 2 (13.3%) 5 (33.3%) 7 (46.7%) 1 (6.7%) 0 7 (46.7%) 5 (33.3%) 2 (13.3%) 2 (13.3%) 10 (66.7%) 1 (6.7%) 12 (80%) 15 (100%) 4 (26.7%) 1 (6.7%) 3 (20%) 4 (26.7%) 8 (53.3%) 13 (86.7%) 7 (46.7%) 0 2 (13.3%) 0

6 (40%) 0 11 (73.3%) 10 (66.7%) 2 (13.3%) 0 1 (6.7%) 0 5 (33.3%) 4 (26.7%) 6 (40%) 10 (66.7%) 6 (40%) 2 (13.3%) 1 (6.7%) 0 1 (6.7%) 0 9 (60%) 1 (6.7%) 1 (6.7%) 0 1 (6.7%) 0 0 0 0 7 (46.7%) 2 (13.3%) 0 2 (13.3%) 3 (20%) 2 (13.3%) 0 3 (20%) 0 9 (60%) 0 0 0 3 (20%) 0 0 0 0 0 0 0 2 (13.3%) 0 0 0

1 (6.7%) 0 0 1 (6.7%) 0 0 0 0 10 (66.7%) 10 (66.7%) 4 (26.7%) 2 (13.3%) 1 (6.7%) 1 (6.7%) 0 0 0 0 0 0 1 (6.7%) 0 1 (6.7%) 0 0 0 0 1 (6.7%) 0 1 (6.7) 0 2 (13.3%) 0 0 0 0 1 (6.7%) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

54.5

.010

107

.775

103

.543

.016

110

.900

111.5

.964

82.5

.189

82.5

.035

52.5

.001

16.5

.000

79.5

.098

.016

.277

.002

.155

67.5

.047

.073

67.5

.038

27.5

.000

.004

.073

.148

105

.671

.126

.000

97.5

.150

Blood in background Bile casts Necrosis Cellularity Trabecular arrangements Rows of cuboidal cells Acinar arrangements Palisading arrangements Dissociation Fibrosis Capillaries Cytoplasmic basophilia Lipofuscin Lipidosis Vacuolar change (glycogen-type) Cytoplasmic inclusions Anisocytosis Increased N:C ratio Anisokaryosis Chromatin pattern Irregular nuclear shapes Nuclear pseudoinclusions Nuclear eccentricity Multiple nuclei (3 or more) Mitotic figures

(continued)

Masserdotti and Drigo

Cytology of canine hepatocellular carcinoma

Table 2 (continued) Score Cytologic Feature

Diagnosis

P†

Nucleoli

WD-HCC Control WD-HCC Control WD-HCC Control WD-HCC Control WD-HCC Control WD-HCC Control WD-HCC Control

7 (46.7%) 12 (80%) 0 14 (93.3%) 7 (46.7%) 2 (13.3%) 3 (20%) 6 (40%) 10 (66.7%) 6 (40%) 13 (86.7%) 14 (93.3%) 11 (73.3%) 15 (100%)

2 (13.3%) 3 (20%) 4 (26.7%) 1 (6.7%) 8 (53.3%) 8 (53.3%) 12 (80%) 8 (53.3%) 4 (26.7%) 8 (53.3%) 2 (13.3%) 1 (6.7%) 4 (26.7%) 0

6 (40%) 0 5 (33.3%) 0 0 5 (33.3%) 0 1 (6.7%) 1 (6.7%) 1 (6.7%) 0 0 0 0

0 0 6 (40%) 0 0 0 0 0 0 0 0 0 0 0

.024

.000

.008

.406

84.5

.190

105

.550

.017

Naked nuclei Neutrophilic inflammation Macrophagic inflammation Lymphoplasmacytic inflammation Eosinophilic inflammation Mast cells

U is the value of the statistical test. P-value is bolded when statistically significant (< .018 according to Benjamini–Hochberg method to correct for multiple comparisons).9

†

Figure 1. Fine-needle aspirate of a well-differentiated hepatocellular carcinoma in a dog. Note the microacinar arrangement of neoplastic cells surrounding a central area of eosinophilic material. May–Gru¨nwald Giemsa, 9100 objective.

hepatocholangiocarcinoma. In one report a pseudoglandular pattern was observed in 38% of WD-HCC but also in 7% of cytologic samples from non-neoplastic liver7; in another report pseudoacini were seen in 53% of WD-HCC.13 Palisading cells, observed in more than half the WD-HCC in our study, but in low numbers, represent an aberrant alignment of neoplastic cells or an abortive attempt to form bile canaliculi,18 sometimes containing bile plugs,13 rather than true glandular structures. These arrangements are probably not correlated with glandular changes, although an adenoid variant of

Figure 2. Fine-needle aspirate of a well-differentiated hepatocellular carcinoma in a dog. Note the palisading cells with nuclei arranged in a row. May–Gru¨nwald Giemsa, 9100 objective.

canine HCC has been described in which hepatocytes are arranged in acinar structures.11 Similar to what is found in human HCC,19 trabecular arrangements were not a distinguishing feature of canine HCC. Dissociation was identified in only a small percentage of samples of non-neoplastic liver, but was a feature of all HCC, similar to what has been described in tumors in people, in which dissociation is increased in moderately differentiated HCC owing to loss of cellular cohesion.13 Capillaries were observed in low numbers in about one-third of the WD-HCC in this study, except for one case in which they were numerous and branching.

Cytology of canine hepatocellular carcinoma

Masserdotti and Drigo

A Figure 3. Fine-needle aspiration of a well-differentiated hepatocellular carcinoma in a dog. Notice the dissociation of the neoplastic hepatocytes among naked nuclei. May–Gru¨nwald Giemsa, 9100 objective.

B Figure 5. Fine-needle aspirate of a well-differentiated hepatocellular carcinoma in a dog. Naked nuclei are dispersed among neoplastic cells (A) or are tightly crowded around clusters of cells (B). May–Gru¨nwald Giemsa, 9100 objective.

Figure 4. Fine-needle aspirate of a well-differentiated hepatocellular carcinoma in a dog. A large cluster of neoplastic hepatocytes is traversed by a long capillary (arrows) with RBCs in the lumen. May–Gru¨nwald Giemsa, 940 objective.

Although present in low numbers in WD-HCC, capillaries were absent in smears from non-neoplastic liver and likely represent increased vascularization of the tumor, rather than sinusoidal dilation, due to neoangiogenesis.19 Similar findings have been reported in people, and the observation of vessels traversing tissue fragments is a useful criterion in the diagnosis of WDHCC.20 The presence of naked nuclei with distinct borders is an important cytogic feature of WD-HCC in dogs and was observed in all cases in our series, with the nuclei surrounding clusters of neoplastic cells or dispersed individually in the background. Presumably, neoplastic cells lose their cytoplasm, leaving behind variably sized nuclei, sometimes with a prominent nucleolus. The presence of naked nuclei has been reported as a prominent feature of human HCC.7,13,21 Free nuclei

were found in only a single sample from non-neoplastic liver and were present in low numbers among clusters of hepatocytes, were associated with severe lipidosis, and had no atypical features. This finding is in agreement with other reports of naked nuclei observed in low percentages of cytologic samples from non-neoplastic liver smears in people.7,13 Multinucleated hepatocytes with 3 or more slightly atypical nuclei were observed frequently and represented a distinctive feature of canine WD-HCC compared with findings in non-neoplastic liver. Binucleated hepatocytes have been described in normal canine liver, particularly in old dogs,22 and multinucleated cells (3 or more nuclei) may be found in human viral hepatitis with marked regenerative activity23 or in hepatic dysplasia.24 Anisokaryosis was observed in two-thirds of the samples from WD-HHC in this study, but involved < 10% of the cells; it was present in only one sample of non-neoplastic liver (again in < 10% of the cells). Other features, such as the presence of multiple prominent nucleoli, irregular chromatin patterns and nuclear outlines, and the presence of mitotic fig-

Masserdotti and Drigo

Cytology of canine hepatocellular carcinoma

Figure 6. Fine-needle aspirate of a well-differentiated hepatocellular carcinoma in a dog. Note the cluster of neoplastic cells, including a trinucleate cell (arrow), surrounded by naked nuclei. Mild anisokaryosis is present. May–Gru¨nwald Giemsa, 9100 objective.

Figure 7. Fine-needle aspirate of a well-differentiated hepatocellular carcinoma in a dog. Note the fusiform cells with eosinophilic cytoplasm and oval nuclei, indicative of fibrosis, among the small clusters of neoplastic hepatocytes. May–Gru¨nwald Giemsa, 9100 objective.

ures were not useful criteria in the diagnosis of canine WD-HCC. Nuclear inclusions, including globular or crystalline rectangular structures, were observed both in neoplastic and non-neoplastic cells; some authors have reported that the crystalline structures have no known role1 and that globular structures may represent invagination of cytoplasm, containing glycogen and mitochondria, into the nucleus of either normal or abnormal hepatocytes.25 Cytoplasmic basophilia and accumulation of materials like lipofuscin, glycogen-type material, or lipid vacuoles were observed in both neoplastic and non-neoplastic cells in our series and were not discriminating features. Cytoplasmic vacuoles containing glycogen or lipid were observed in 27% of human WD-HCC13; in people, abundant lipid or glycogen can cause the cytoplasm to appear clear, as seen in the so-called clear cell carcinoma of the liver.26 Fine eosinophilic globules were identified in the cytoplasm of hepatocytes in low numbers of WD-HCC in our study. Cytoplasmic inclusions described in human HCC include Mallory Bodies,27 accumulation of a1-antitrypsin,28 fibrinogen,29 or hyaline bodies.30 These inclusions are unrelated to hepatocellular dysfunction, but rather seem to be a manifestation of metabolic errors in neoplastic cells.31 In our series, neutrophilic infiltrates were more prominent in samples from non-neoplastic liver, but this finding was expected as the comparison group included dogs with hepatitis. Mast cells were observed in low numbers of samples of WD-HCC. Scattered mast cells may be observed when hepatic fibrosis, found in almost half of the samples from WD-HCC, is present.32 Mast cells also play a critical role in tumor

neoangiogenesis33 and may be important in the transition from sinusoidal to capillary-type hepatic endothelial cells.34 Cirrhosis has been reported in canine HCC, but was not correlated with histologic type.2 In human WD-HCC, inflammatory cells were not noted in one report, whereas in moderately and poorly differentiated forms variable degrees of inflammation were present.13 In conclusion, the most useful cytologic features in the diagnosis of canine WD-HCC were dissociation of hepatocytes, acinar, or palisading arrangements of neoplastic cells, and the presence of naked nuclei and capillaries, together with mild anisocytosis, anisokaryosis, multinuclearity, and increased N:C ratios. Further studies are warranted to compare these cytologic criteria among WD-HCC, hepatomas, and hyperplastic nodules. In addition, the inter-relationship among the cytologic features was not examined in this study owing to evaluation of only 15 cases, and a larger study is required to evaluate co-variance of the 33 cytologic features.

Disclosure: The authors have indicated that they have no affiliations or financial involvement with any organization or entity with a financial interest in, or in financial competition with, the subject matter or materials discussed in this article.

References 1. Meyer DJ. The liver. In: Raskin RE, Meyer DJ, eds. Canine and Feline Cytology. A Color Atlas and Interpretation

Cytology of canine hepatocellular carcinoma

Guide. 2nd ed. St. Louis, MO: Saunders Elsevier; 2010:242–243. 2. Patnaik AK, Hurvitz AI, Lieberman PH, Johnson GF. Canine hepatocellular carcinoma. Vet Pathol. 1981;18: 427–438.

entiated hepatocellular carcinoma and macroregenerative nodule. Definition of standardized criteria from a study of 100 cases. Acta Cytol. 2000;44:515–523.

3. Kulesza P, Torbenson M, Sheth S, Erozan YS, Ali SZ. Cytopathologic grading of hepatocellular carcinoma on fine-needle aspiration. Cancer. 2004, Aug 25;102(4): 247–252.

17. Charles JA, Cullen JM, Van den Ingh TSGAM, Van Winkle T, Desmet VJ. Morphological classification of neoplastic disorders of the canine and feline liver. WSAVA Standards for Clinical and Histological Diagnosis of Canine and Feline Liver Disease. St. Louis, MO: Saunders Elsevier; 2006:117–124.

4. Blue JT, French TW, Meyer DJ. The liver. In: Cowell RJ, Tyler RD, Meinkoth JH, eds. Diagnostic Cytology and Hematology of the Dog and Cat. 2nd ed. St. Louis, MO: Mosby; 1999:183–194.

18. Goodman ZD, Terracciano LM. Tumours and tumourlike lesions of the liver. In: Burt AD, Portmann BC, Ferrell LD, eds. McSween’s Pathology of the Liver. 5th ed. Philadelphia, PA: Elsevier; 2007:771–782.

5. Yue X, Zheng S. Cytologic diagnosis by transthoracic fine needle sampling without aspiration. Acta Cytol. 1989;33:806–808.

19. Yang GCH, Yang GY, Tao LC. Cytologic features and histological correlations of microacinar and microtrabecular types of well-differentiated hepatocellular carcinoma in fine-needle aspiration biopsy. Cancer. 2004;102:27–33.

6. Mair S, Dunbar F, Becker PJ. Fine needle cytology – is aspiration suction necessary? A study of 100 masses in various sites. Acta Cytol. 1989;33:809–813. 7. Cohen MB, Haber MM, Holly EA, Ahn DK, Bottles K, Stoloff AC. Cytologic criteria to distinguish hepatocellular carcinoma from nonneoplastic liver. Am J Clin Pathol. 1991;95:125–130. 8. Masserdotti C. Architectural pattern in cytology: correlation with histology. Vet Clin Pathol. 2006;35: 388–396. 9. Benjamini Y, Hochberg Y. Controlling the false discovery date: a pratical and powerful approach to multiple testing. J Royal Stat Soc Series B (Methodological). 1955;57: 289–300. 10. Strombeck DR. Clinicopathologic features of primary and metastatic neoplastic disease of the liver in dogs. J Am Vet Med Assoc. 1978;173:267–269. 11. Cullen JM, Popp JA. Tumors of the liver and gall bladder. In: Meuten DJ, eds. Tumors in Domestic Animals. 4th ed. Ames, IA: Iowa State Press; 2002:483–508. 12. Roth L. Comparison of liver cytology and biopsy diagnoses in dogs and cats: 56 cases. Vet Clin Pathol. 2001;30: 35–38. 13. Wee A, Nilsson B, Tan LK, Yap I. Fine needle aspiration biopsy of hepatocellular carcinoma. Diagnostic dilemma at the ends of the spectrum. Acta Cytol. 1994;38: 347–354. 14. Noguchi S, Yamamoto R, Tatsuta M, et al. Cell features and patterns in fine-needle aspirates of hepatocellular carcinoma. Cancer. 1986;58:321–328. 15. Wee A, Nilsson B. Highly well-differentiated hepatocellular carcinoma and benign hepatocellular lesions. Can they be distinguished on fine needle aspiration biopsy. Acta Cytol. 2003;47:16–26. 16. Longchampt E, Patriarche C, Fabre M. Accuracy of cytology vs. microbiopsy for the diagnosis of well-differ-

Masserdotti and Drigo

20. De Boer WB, Segal A, Frost FA, Sterrett GF. Cytodiagnosis of well differentiated hepatocellular carcinoma. Cancer. 1999;87:270–277. 21. Pedio G, Landolt U, Zobeli L, Gut D. Fine needle aspiration of the liver. Significance of hepatocytic naked nuclei in the diagnosis of hepatocellular carcinoma. Acta Cytol. 1988;32:437–442. 22. Stockhaus C, Teske E, Van Den Ingh T, Rothuizen J. The influence of age on the cytology of the liver in healthy dogs. Vet Pathol. 2002;39:154–158. 23. Atterbury CE, Enriquez RE, Desutonagy GI, Conn HO. Comparison of the histologic and cytologic diagnosis of liver biopsies in hepatic cancer. Gastroenterology. 1979;76: 1352–1357. 24. Berman JJ, McNeill RE. Cirrhosis with atypia: potential pitfall in the interpretation of liver aspirates. Acta Cytol. 1986;32:11–14. 25. Kelly WR. The liver and biliary system. In: Jubb KVF, Kennedy PC, Palmer N, eds. Pathology of Domestic Animals. 4th ed, Vol 2. San Diego, CA: Academic Press; 1993:319–406. 26. Murakata LA, Ishak KG, Nzeako UC. Clear cell carcinoma of the liver: a comparative immunoistochemical study with renal clear cell carcinoma. Mod Pathol. 2000; 13:874–881. 27. Nakanuma Y, Ohta G. Expression of Mallory bodies in hepatocellular carcinoma in man and its significance. Cancer. 1986;57:81–86. 28. Kelly JK, Davies JS, Jones AW. Alpha-1-antitrypsin deficiency and hepatocellular carcinoma. J Clin Pathol. 1979;32:373–376. 29. Stromeyer FW, Ishak KG, Gerber MA, Mathew T. Ground-glass cells in hepatocellular carcinoma. Am J Clin Pathol. 1980;74:254–258.

Masserdotti and Drigo

Cytology of canine hepatocellular carcinoma

30. MacDonald K, Bedard YC. Cytologic, ultrastructural and immunologic features of intracytoplasmic hyaline bodies in fine needle aspirates of hepatocellular carcinoma. Acta Cytol. 1990;34:197–200.

33. Peng SH, Deng H, Yang JF, et al. Significance and relationship between infiltrating inflammatory cells and tumor angiogenesis in hepatocellular carcinoma tissues. World J Gastroenterol. 2005;11:6521–6524.

31. Ishak KG, Goodman ZD, Stocker JT. Tumors of the Liver and Intrahepatic Bile Ducts. American Registry of Pathology. Washington, DC: AFIP; 2001:206–208.

34. Grizzi F, Franceschini B, Chiriva-Internati M, Liu Y, Hermonat PL, Dioguardi N. Mast cells and human hepatocellular carcinoma. World J Gastroenterol. 2003;9: 1469–1473.

32. Masserdotti C, Bertazzolo W. Cytologic features of hepatic fibrosis in dogs: description of 11 cases [abstract]. Vet Clin Pathol. 2009;38(S1):E10–E11.