ctDNA in advanced bladder cancer: What does the future hold?

Adanma Ayanambakkam, MD

In this interview, Adanma Ayanambakkam, MD, discusses the current state of treatment of advanced bladder cancer, including the need for predictive and prognostic markers for immune therapy as well as the potential showed by circulating tumor DNA (ctDNA). Ayanambakkam is an assistant professor of medicine in the section of hematology/oncology at the University of Oklahoma in Oklahoma City.

Please discuss the unmet need for predictive and prognostic markers for immune therapy.

This topic has been a persistent focus of interest for numerous prominent figures in the scientific community. Over nearly a decade of employing immune checkpoint inhibitors, we have come to recognize that these treatments exhibit remarkable efficacy in specific patients but prove ineffective in others. Notably, the effectiveness of pembrolizumab [Keytruda] in the context of metastatic urothelial cancer varies significantly among patients, with some showing substantial responses.

Efforts have been made to enhance the effectiveness of immune checkpoint inhibition, such as by combining it with chemotherapy. However, these trials have shown limited success in intensifying the impact of pembrolizumab. The introduction of the EV-103 trial shed light on a new approach. It revealed that enfortumab vedotin-ejfv [Padcev], an immune-mediated modulator, can influence the tumor microenvironment, potentially enhancing the efficacy of pembrolizumab. Preclinical and clinical phase 2 evidence indicates that combining pembrolizumab with enfortumab vedotin-ejfv primes the immune system more effectively, resulting in improved responses. Median overall survival surpassed 2 years in patients receiving the combination, and EV-103 trial data demonstrated that enfortumab vedotin-ejfv plus pembrolizumab was markedly more effective than enfortumab vedotin-ejfv alone. These 2 treatments, pembrolizumab and enfortumab vedotin-ejfv, appear to complement each other, yielding significant benefits when used together. However, an in-depth examination of these exciting data reveals substantial toxicity, with 64% of patients experiencing grade 3 or higher treatment-related adverse events, including peripheral neuropathy. A considerable proportion of patients required dose reductions (30%), and some had to discontinue treatment (24%). The regimen's tolerability is a significant challenge when attempting to intensify these therapies.

Upon reviewing frontline immunotherapy-based approaches in metastatic urothelial cancer, an exploratory subgroup analysis has identified a cohort of patients who respond exceptionally well to pembrolizumab monotherapy. This group may include patients with a combined proportion score exceeding 10, high PD-L1 expression, or extensive tumor immune infiltrate. These individuals, often referred to as exceptional responders, may not require treatment intensification. Identifying such patients early on could help avoid treatment intensification and associated adverse events, while also pinpointing those who may benefit from adding chemotherapy or enfortumab vedotin-ejfv combinations. This is crucial because, when compared with immunotherapy monotherapy, the intensified approach of pembrolizumab plus enfortumab vedotin-ejfv results in a significantly higher rate of treatment-related adverse events (more than 64%). Therefore, there is a pressing need to identify a biomarker to predict patient responses accurately.

However, developing a reliable biomarker has proven challenging, particularly in the context of immune checkpoint inhibition. Various biomarkers, including tumor mutational burden, PD-L1 expression on tumor and tumor-infiltrating immune cells, combined proportion scores, and tumor proportion scores, have been explored. Yet, these biomarkers are hampered by intratumoral and temporal heterogeneity. PD-L1 expression varies across different tumor sites and time points, making it an unreliable predictor. Although PD-L1-positive patients are more likely to respond to immune checkpoint inhibition, reports indicate that even PD-L1-negative patients can respond effectively. Pembrolizumab plus enfortumab vedotin-ejfv is approved for all individuals regardless of their PD-L1 levels, underscoring the complexity of this landscape. Whether PD-L1 is low or high, patients respond well to this combination, and it remains a valuable option in the realm of immune checkpoint inhibition.

Let’s move on to discussion of cell-free DNA and ctDNA. Could you start by explaining the rationale for cell-free DNA in healthy adults and in cancer patients?

Cell-free DNA comprises unencapsulated double-stranded DNA fragments that circulate in the bloodstream. Typically, these fragments are approximately 150 to 200 base pairs in length and are released passively due to processes like apoptosis, necrosis, and phagocytosis. They tend to associate with protein complexes or membrane vesicles, which are subsequently eliminated through nuclear actions or renal excretion. All individuals harbor cell-free DNA in their blood, primarily originating from hematopoietic cells, and can be released under both pathological and physiological conditions. A notable example of a physiological state associated with increased circulating free DNA is pregnancy. Pathological conditions such as ischemia, trauma, inflammation, and infections, including COVID-19, also lead to elevated levels of circulating free DNA. Ordinarily, the blood contains approximately 10 ng/mL of circulating free DNA. The release of circulating free DNA from both normal and cancerous cells occurs during cell death, involving DNA fragmentation by nucleases that cleave nucleosomal fragments. Importantly, this fragmentation is nonrandom. Analyzing the ends of circulating free DNA fragments allows us to discern the nucleosomal profile of their cell of origin, facilitating the differentiation between normal and cancerous cell sources.

ctDNA constitutes a subset of a patient's circulating free DNA. In individuals with cancer, circulating free DNA is approximately 50 times more abundant than in healthy individuals, and ctDNA is a subset of circulating free DNA. ctDNA is characterized by being much more fragmented and shorter compared to cell-free DNA. The extent of ctDNA in circulation varies based on the tumor type and stage. Different cancers release varying amounts of ctDNA, with the quantity of ctDNA released correlating with the number of tumor cells undergoing senescence or apoptosis. In cases where senescence is inhibited or apoptosis is induced through treatments like radiation or chemotherapy, there is an observed increase in the amount of ctDNA released by cancer cells. Some types of cancer are more prone to releasing ctDNA into the bloodstream, with urothelial cancers being among those with a high ctDNA release. ctDNA can be detected not only in blood but also in other bodily fluids like saliva, urine, pleural fluid, ascitic fluid, and cerebrospinal fluid. This provides a readily accessible biomarker, which, thanks to modern sequencing technology, can be examined noninvasively. ctDNA-based approaches enhance the ability to detect certain critical disease states, such as leptomeningeal disease, surpassing the capabilities of conventional methods like random baseline cytology or flow cytometry analysis. In essence, ctDNA allows us to delve deeper into the molecular aspects of blood cells, offering insights into the extent of cancer burden.

ctDNA can be categorized into 2 main groups: ctDNA assays and tumor-informed ctDNA approaches, as well as tumor-uninformed ctDNA approaches. The tumor-informed approach involves sequencing and genomic profiling of the baseline tumor tissue to identify somatic genomic characteristics unique to the tumor, distinguishing them from passenger mutations like CHIP mutations. In this method, a specific set of tumor mutations is identified, typically around 16 mutations in the case of assays like Natera. A personalized panel is then generated based on these mutations, and blood samples are sequenced using this panel. Computational algorithms are employed to determine whether a sufficient number of these mutations are present in the blood, signifying the presence of ctDNA originating from the tumor. This approach provides a snapshot of whether the tumor is releasing ctDNA into the bloodstream and allows for the quantification of the cancer burden in the blood.

On the other hand, the tumor-uninformed approach involves broad sequencing, such as whole-exome sequencing or next-generation sequencing of blood samples. ctDNA is identified based on fragment length and unique characteristics, with subsequent analyses capable of distinguishing CHIP mutations and quantifying ctDNA. Although this method is more comprehensive, it can be more labor-intensive and costly. However, it offers advantages such as the ability to identify a broader range of mutations and even track the clonal evolution of specific tumors.

In the genitourinary field, there are approved methods for detecting mutations like BRCA and ATM in peripheral blood samples, using platforms like Foundation or Guardant360. These mutations can guide the use of targeted agents like olaparib [Lynparza] or rucaparib [Rubraca]. Several companies now offer CLIA-certified ctDNA assays for routine clinical use, including Natera, Inivata, Guardant Health, and Naveris. However, there is a growing interest in developing a commercial platform for assessing ctDNA in clinical contexts.

What are ctDNA’s roles in response to neoadjuvant therapy, detecting minimal residual disease, recurrence monitoring after adjuvant therapy, and treatment response monitoring?

ctDNA holds a unique position as it offers a molecular means to quantify the extent of residual disease burden. It's known that conventional imaging techniques don't always effectively predict or quantify the remaining disease burden, and often, even after robust treatment responses, a residual disease burden persists. The crucial, unanswered question in such cases is whether this residual material comprises viable tumor cells or necrotic granulomatous tissue. ctDNA made its debut in the metastatic setting, and its success prompted consideration for applications in adjuvant settings, including cancer surveillance and screening. In recent years, multiple studies have explored the utility of ctDNA in various contexts, including metastatic, localized advanced cancer, the neoadjuvant setting, and the adjuvant setting.

An influential study by Bratman and colleagues, published in Nature Cancer, stands out as one of the initial prospective phase 2 clinical trials to collect ctDNA in a systematic manner.¹ This study focused on metastatic patients treated with immune checkpoint inhibitors, particularly pembrolizumab, and employed a tumor-informed ctDNA approach. The investigators established that baseline ctDNA served as a prognostic biomarker: patients with low baseline ctDNA experienced better overall survival, whereas those with high baseline ctDNA had inferior outcomes. Notably, patients who achieved ctDNA clearance after three cycles of pembrolizumab demonstrated outstanding responses, with nearly 100% survival at a median follow-up of 25 months. Patients with decreased but persisting ctDNA exhibited mixed responses, not as favorable as those who successfully cleared their ctDNA. The study further revealed that ctDNA clearance often preceded radiographic response by 3 to 6 months, providing a valuable tool for assessing patient outcomes through routine blood draws during immune checkpoint inhibition treatment. In essence, this approach offered both a baseline prognostic biomarker to gauge patient well-being and a predictive biomarker to assess ongoing treatment responses. Physicians could identify which patients were unlikely to respond to immune checkpoint inhibitors and make informed decisions about the course of treatment without waiting for radiographic responses. It's important to note that this trial did not use ctDNA to guide interventions; instead, it collected ctDNA data prospectively and analyzed it retrospectively to evaluate its impact. One noteworthy observation from this analysis was that patients with RECIST-defined disease progression but negative ctDNA experienced longer overall survival. This suggests that when ctDNA levels decrease while radiographic images indicate disease progression, pseudoprogression is likely occurring.

Other studies in the same context have further confirmed that ctDNA can accurately describe pseudoprogression across various disease sites, particularly in lung cancer and melanoma. They have shown that an increase in ctDNA levels correlates with genuine disease progression, whereas a decrease in ctDNA, alongside a simultaneous increase in radiographic imaging, is more indicative of pseudoprogression. In sum, these investigations underscore the dual roles of ctDNA as both a predictive and prognostic biomarker.

After demonstrating the effectiveness of ctDNA in the metastatic setting, there was significant interest in its application in localized settings to detect cancer recurrence following treatment for localized cancer. The IMvigor010 clinical trial [NCT02450331] serves as an illustrative example of this context.² IMvigor010 is a randomized phase 3 study involving patients with muscle-invasive bladder cancer post-cystectomy. They were randomized to receive adjuvant atezolizumab [Tecentriq] for a year or undergo observation. Unfortunately, this trial did not yield a statistical benefit in terms of disease-free survival or overall survival for the patients. Consequently, atezolizumab is not recommended for all individuals undergoing cystectomy upfront.

However, a subgroup analysis of IMvigor010, as published by Powles and colleagues in Nature, offers intriguing insights. They examined patients who had undergone cystectomy, achieved complete cure, and showed no residual disease or radiographic evidence. The patients were categorized based on the presence or absence of ctDNA. The analysis revealed that patients with negative baseline ctDNA, before any treatment initiation, fared better than those with positive ctDNA. This finding underscores the prognostic role of ctDNA, which extends from the metastatic setting to the non-metastatic setting. Additionally, the study demonstrated that although atezolizumab did not enhance survival outcomes in the ctDNA-negative group, it did make a significant difference in the ctDNA-positive group. In essence, not everyone requires adjuvant atezolizumab, but individuals with positive ctDNA after cystectomy are considered higher-risk patients. Among this subgroup of ctDNA-positive patients who received a year of adjuvant atezolizumab, they were more likely to benefit, experiencing improvements in disease-free survival and overall survival. Moreover, the clearance of ctDNA was associated with more favorable clinical outcomes. If a patient was ctDNA-positive and subsequently became ctDNA-negative, their outcomes were markedly improved.

The study also emphasized that whether a patient was ctDNA-negative at baseline or became ctDNA-negative after a few cycles of atezolizumab, their outcomes were similar. What truly mattered was reaching the ctDNA-negative status, which was associated with better survival outcomes. It's noteworthy that only patients in the atezolizumab arm demonstrated ctDNA clearance. The implication is that not everyone necessitates adjuvant immune checkpoint inhibition, and current pathological criteria are used to determine the need for adjuvant nivolumab at this stage. However, evidence now suggests that individuals with positive ctDNA are at a higher risk and might benefit from a year of adjuvant immunotherapy. An ongoing phase 3 randomized trial, IMvigor011 [NCT04660344], is addressing this specific scenario by randomizing ctDNA-positive patients to receive atezolizumab or not, whereas ctDNA-negative patients undergo surveillance.

As the effectiveness of adjuvant immunotherapy became apparent, the exploration of neoadjuvant approaches gained momentum. Notably, the ABACUS trial, as reported by Powles and colleagues in Nature, delved into the role of neoadjuvant atezolizumab in muscle-invasive urothelial cancer before cystectomy.³ The trial successfully achieved its primary end point of pathological complete response. However, it was observed that patients with a positive ctDNA baseline exhibited poorer clinical outcomes. Interestingly, atezolizumab treatment was associated with reduced ctDNA levels in patients who achieved a pathological response. This not only emphasizes the prognostic role of baseline ctDNA before cystectomy but also underscores its predictive role in identifying patients likely to respond favorably to atezolizumab treatment based on achieving a pathological response. Patients receiving atezolizumab were more likely to experience ctDNA clearance, which was expected as it was the sole treatment offered. Nevertheless, this reinforces the utility of ctDNA as a prognostic biomarker in the metastatic setting, a potentially predictive biomarker in the adjuvant setting, and highlights its significance in the neoadjuvant context.

A notable study by Christensen and colleagues, published in the Journal of Clinical Oncology in 2019, explored early detection of metastatic relapse and therapeutic efficacy monitoring using ctDNA in urothelial cancer.⁴ The findings revealed that post-cystectomy, ctDNA analysis accurately identified all patients with metastatic relapse during disease monitoring, achieving a remarkable sensitivity of 100% and specificity of 98%. Moreover, ctDNA provided a median lead time of approximately 96 days compared with standard radiographic imaging, meaning it could predict metastatic relapse well in advance, even up to 90 days before radiographic evidence. This advanced detection allows for intervention at a molecular level, potentially preventing the development of advanced metastases or complications. Additionally, the study demonstrated that in high-risk patients, particularly those who were ctDNA-positive before or during treatment, the dynamics of ctDNA during chemotherapy correlated with more substantial pathological downstaging and better clinical outcomes. Essentially, ctDNA emerges as a predictive biomarker that can inform how an individual will respond to chemotherapy and whether pathological downstaging is achievable.

In summary, ctDNA has established itself as a valuable prognostic and potentially predictive biomarker, regardless of the disease stage. It serves a pivotal role in the metastatic, neoadjuvant, pre-cystectomy, and adjuvant post-cystectomy settings. A negative ctDNA status consistently aligns with improved outcomes. Importantly, whether a patient has a ctDNA-positive baseline or attains a ctDNA-negative status, either post-cystectomy or after subsequent treatment, significant enhancements in overall outcomes are evident.

What are the next steps with ctDNA?

In summary, ctDNA serves as a significant prognostic and predictive biomarker in the context of metastatic and muscle-invasive bladder cancer. It plays a crucial role in guiding decisions regarding neoadjuvant and adjuvant chemotherapy or immunotherapy. ctDNA can predict recurrence-free survival outcomes and assess the potential for pathological downstaging, making it clinically valuable. Presently, many oncologists across different specialties are integrating commercial ctDNA platforms into their routine practices.

However, it's important to acknowledge that ctDNA, although promising, is still in its early stages, and standardization across various commercial platforms remains a challenge due to differences in sensitivity and modalities. Additionally, insurance coverage for ctDNA analysis may be limited, and cost considerations are relevant. Although existing data are reassuring, treatment decisions based on ctDNA changes lack large-scale randomized phase 3 trial validation. Ongoing trials such as TOMBOLA and IMvigor011 are paving the way for more robust evidence specific to ctDNA-guided treatment decisions.

ctDNA is increasingly becoming a valuable biomarker in clinical trials. It offers a noninvasive means to gain real-time insights into tumor dynamics, assess treatment response, detect resistance mechanisms, and advance our comprehension of cancer biology. Ultimately, the goal is to tailor more effective and personalized cancer therapies based on ctDNA platforms, acknowledging their limitations and economic implications as we integrate them into real-world practice. In the coming years, ctDNA is poised to revolutionize how treatment decisions are made in oncology.

References

1. Bratman SV, Yang SYC, Iafolla MAJ, et al. Personalized circulating tumor DNA analysis as a predictive biomarker in solid tumor patients treated with pembrolizumab. Nat Cancer. 2020;1(9):873-881. doi:10.1038/s43018-020-0096-5

2. Bellmunt J, Hussain M, Gschwend JE, et al. Adjuvant atezolizumab versus observation in muscle-invasive urothelial carcinoma (IMvigor010): a multicentre, open-label, randomised, phase 3 trial. Lancet Oncol. 2021;22(4):525-537. doi:10.1016/S1470-2045(21)00004-8

3. Powles T, Assaf ZJ, Davarpanah N, et al. ctDNA guiding adjuvant immunotherapy in urothelial carcinoma. Nature. 2021;595(7867):432-437. doi:10.1038/s41586-021-03642-9

4. Christensen E, Birkenkamp-Demtröder, Sethi H, et al. Early detection of metastatic relapse and monitoring of therapeutic efficacy by ultra-deep sequencing of plasma cell-free DNA in patients with urothelial bladder carcinoma. J Clin Oncol. 2019;37(18):1547-1557. doi:10.1200/JCO.18.02052