ctDNA dynamics in advanced bladder cancer

Prof. Jørgen Bjerggaard Jensen Chairman of the Nordic Urothelial Research Group (NORTH-REG) Department of Urology, Aarhus University Hospital, Aarhus (DK)

At the time of diagnoses, approximately one quarter of the patients with muscle invasive bladder cancer (MIBC) presents with metastasis or fixed tumour and are thereby not candidates for local curative intended treatment with cystectomy or radiotherapy. Of the remaining approximately 75% of MIBC patients with supposedly localized disease, between one third to two thirds will experience a metastatic relapse despite intended radical local treatment.

The main reason for this is the presence of occult metastatic disease at the time of diagnoses. These micro deposits of metastasis is not visible despite modern FDG-PET/CT or other currently available imaging modalities. This will inevitably lead to a recurrence visible on imaging if left untreated which ultimately can lead to a fatal outcome.

Selection of patients for adjuvant treatment In order to reduce the high treatment failure rate in MIBC patients, neoadjuvant chemotherapy has been recommended to high-risk patients prior to local radical treatment. Another strategy is to offer adjuvant systemic treatment after surgery in patients with a high risk of recurrence based on conventional risk parameters (e.g. non-organ confined disease or remnant MIBC following neoadjuvant chemotherapy).

However, both strategies will inevitably lead to both overtreatment in classical high-risk patients as well as undertreatment in a fraction of the classical low-risk patients. This is based on the fact that not all apparently high-risk patients experience recurrence why these patients could do without a potentially harmful and expensive treatment if better selection was possible. Moreover, some supposedly low-risk patients will develop a recurrence. Nevertheless, because the risk is low per se, adjuvant treatment is typically not given to these patients, neither as standard in clinical trials. Instead, low-risk patients, which despite this may develop recurrence, are treated at the time of recurrence.

At this time point, the metastatic burden is higher, the disease is more molecularly heterogeneous, and therefore theoretically more treatment refractory compared to treatment close to the radical local treatment. If all patients with remnant metastatic disease, and only these, instead could be treated at a very early time point this will reduce the number of patients undergoing superfluous and potentially harmful treatment alongside with a reduction in cost as the non-recurrent high risk patients will be omitted from additional treatment. Methods to identify these patients have been lacking but the future looks bright regarding this with the development of new molecular methods for detection of circulating tumour DNA (ctDNA).

“Live” monitoring of treatment response in metastatic disease Another current challenge in advanced urothelial cancer is whether patients undergoing systemic oncological treatment have a response justifying the continuation of a systemic treatment; or whether they should be undergoing another potentially more effective treatment or maybe abandon treatment all together. Thus, patients without tumour reduction effect of the given treatment but suffering from severe side effects could be spared the latter. Current standard regarding oncological response to systemic treatment is response estimated by e.g. RESIST criteria from imaging. This is, however, associated with a certain lead time before true tumour reduction can be seen and a certain interval between imaging is required in order to see any changes.

Circulating tumour DNA One very promising biomarker in modern molecular medicine is circulating tumour DNA (ctDNA). Small fragments of DNA from tumour cells is released into circulation. The tumour DNA contains tumour-specific mutations and other genomic alterations, which can be used as highly specific biomarkers. Typically, ctDNA from plasma samples is used as a marker of metastatic disease whereas urine-based ctDNA is more useful to estimate local tumour presence in the urothelium.

Figure 1: Design of Danish Bladder Cancer Study no. 14 – TOMBOLA. All patients are diagnosed with localized MIBC and undergo standard neoadjuvant chemotherapy followed by radical cystectomy. During a close follow-up regimen, immunotherapy with Atezoluzimab is started if ctDNA is detected. During neoadjuvant chemotherapy and immunotherapy, ctDNA is measured in additional observational plasma samples to estimate the potential use of ctDNA in treatment monitoring.

Figure 2: Patient example with multiple time point measurement of ctDNA following diagnosis of MIBC and standard treatment with neoadjuvant chemotherapy and radical cystectomy. Noticeably, there is a 2.5 year lead time from postoperative positive ctDNA until metastasis is visible on imaging and immunotherapy is initiated. Concordant with an increase in ctDNA during immunotherapy, there was progression on imaging and the patient underwent re-induction with conventional GemcitabinCisplatin resulting in apparently complete response on imaging but still detectable dissemination on ctDNA despite a much lower level. Unfortunately, this indicates that the patients will have an identificable recurrence within the near future. (Courtesy of Associate Prof. Karin Birkenkamp-Demtröder).

Regarding the concept of ctDNA as a biomarker, it is important to recognize that the techniques for identifying ctDNA in its current use is based on a highly specialized individual design of assays designed individually to each patient or expensive sequencing approaches. This is time consuming and requires advanced laboratory and bioinformatic methods. However, the techniques and facilities are constantly evolving, leading to a more and more easily available technique in studies and hopefully in a near future daily practice.

“Another current challenge in advanced urothelial cancer is whether patients undergoing systemic oncological treatment have a response justifying the continuation of a systemic treatment.”

While identification of the highly tumour-specific and patient-specific mutations in blood samples is associated with an extreme high positive predictive value (and a positive quantity related correlation with tumour burden) the negative predictive value is not perfect. Thus, selection of new tumour clones during ongoing systemic treatment can result in false negative findings or at least a nonlinear correlation between ctDNA dynamics and tumour burden. It is therefore important to continue to use conventional imaging in parallel to the introduction of this promising diagnostic technique, at least until more prospective studies and clinical trials have been conducted.

ctDNA in selection of patients for additional treatment following radical cystectomy ctDNA represents the mutational spectrum of the tumour and is thus highly tumour-specific. Detection of ctDNA following local radical treatment has proved to be associated with certainty of remnant carcinoma cells and thus leading to recurrence following a variable lead time. In a previous study, this lead time was proven to be variable between few months to years, but very important, all positive findings in patients lead to later or simultaneous clinical recurrence. [1]

With this in mind, the Danish Bladder Cancer Group has established an ongoing intervention study – the TOMBOLA trial – , a national, interdisciplinary collaboration between five clinical centres (Aarhus, Aalborg, Odense, Herlev, Rifshosp.) and the Dept. of Molecular Medicine (MOMA), Aarhus. In the TOMBOLA trial, plasma ctDNA positivity following neoadjuvant chemotherapy and cystectomy in MIBC patient as indication to administer Atezolizumab as early postoperative additional treatment. [2] (Fig. 1) In the TOMBOLA setup, multiple time points for ctDNA measurement are used over the whole standard-ofcare follow-up period after cystectomy, but the long-term hope is that we will be able to identify the majority of high-risk patients by positive ctDNA very early, a short time after CX in order to have the longest possible lead time to the time where a recurrence would have been visible on the first or second (4m/8m) conventional imaging. This will in theory lead to prolonged relapse-free survival, better long-term outcome and will reduce the current logistic challenges in the postoperative setting. Furthermore, a “continuous” monitoring surveillance scheme may be associated with a better quality of life.

A somewhat similar study is the ongoing IMvigor011 that is also using positive ctDNA to trigger early immunotherapy. [3] In the IMvigor011, the patients are restricted to the classical high-risk patients that otherwise typically all would undergo adjuvant treatment. Introduction of widespread standard use of adjuvant immunotherapy can actually hamper the field of ctDNA as studies will be forced to show non-inferiority of selected use of immunotherapy compared to a very liberal use. Introduction of new diagnostic methods like ctDNA therefore calls for an intellectual reset of knowledge learned from studies without the use of these techniques. In the IMvigor011 study, randomization is fortunately made against no adjuvant treatment, thus making it a superiority study. a more or less binary outcome, the quantity of ctDNA in blood samples can be used as a surrogate marker of metastatic burden and thus as an indicator of treatment response. In patients treated with immunotherapy, a pseudo-progression on imaging can be seen in otherwise highly responsive patients and irrespectively of this, there is need for a certain interval between imaging to estimate response. Contrary to this, ctDNA has a very short half-time of about two hours. Therefore, more or less “live” monitoring is possible very early in different active treatments and intervals between these where the patient is undergoing follow-up only. (Fig. 2)

For the moment, this is not considered as decisive for shift in treatment but should at least lead to shorter intervals between imaging and assessments if a rising level is identified in a patient, or may be used to spare patients for some CT scans. Hopefully, ongoing research will shed more light into this field in the near future and bring ctDNA into the future standard treatment monitoring.

Due to space constraints, the entire reference list can be made available to interested readers upon request by sending an email to: communications@ uroweb.org.