COSMIC-313 and the Future of Triplet Therapy in RCC: Efficacy, Toxicity, and Biomarkers

At the 2025 American Society of Clinical Oncology Genitourinary Cancers Symposium, a roundtable of expert panelists convened to discuss the latest research and practice updates related to advanced kidney cancer. Dr. Karine Tawagi of University of Illinois was joined by Drs. Regina Barragan-Carrillo, Benjamin Maughan, Laurence Albiges, and David McDermott in the discussion.

In the second part of the roundtable, the panel considers the current limitations of ctDNA in kidney cancer due to low tumor shedding, but expresses optimism for future advancements in detection technology. They also provide an in-depth analysis of the final results from the COSMIC-313 trial, highlighting challenges in managing toxicity and the potential role of immune biomarkers, particularly M2-like macrophages, in guiding treatment strategies.

View the next segment on The Role of RCC Second-Line Therapy in Frontline Decision-Making: Does the Approach Matter?

—

Dr. Tawagi: Speaking of risk-adapted approaches, we are seeing some of these trials in bladder cancer with ctDNA. What are your thoughts on ctDNA when it comes to kidney cancer?

Dr. Barragan-Carrillo: Right now, it has not translated to the clinic. We know that kidney cancer is a tumor that does not shed much, so there is not much utility in that sense. Personally, I never look at it in the clinic. The same applies to information from next-generation sequencing. While it is informative and provides new insights, it does not change my approach to patients. How are you approaching this?

Dr. Maughan: Similarly for me, I conduct a lot of this testing mainly to identify patients who may benefit from a specific clinical trial in a later-line therapy. However, in terms of FDA-approved or regulatory body-approved patient management, it does not significantly impact how we take care of patients today. This is largely due to the sensitivity issue with current ctDNA technology in clear cell kidney cancer, which is not particularly high.

Dr. Albiges: As of now, that is the case. However, I like to think that with deeper or more advanced technologies, we may be able to change that in the future.

Dr. Maughan: Yes, and there are indications that advancements in technology will improve its utility. For example, Signatera has made modifications by shifting from an off-the-shelf ctDNA approach to a BESPOKE approach. Based on the small amount of data presented, this method is not dramatically more sensitive, but it is an improvement. What this demonstrates to me is that with some modifications, the technology could become more applicable to this patient population.

Dr. Tawagi: Absolutely. Is anyone else checking next-generation sequencing in kidney cancer? Not yet? Okay.

We have a lot of exciting data coming up on Saturday, including COSMIC-313, which you will be presenting. Why do we not start with some updates on COSMIC-313?

Dr. Albiges: The COSMIC-313 trial is the first triplet phase three trial, which randomized patients to receive cabozantinib plus nivolumab plus ipilimumab, the triplet, versus nivolumab plus ipilimumab plus placebo, the doublet. This is an important study because, for the first time, the comparator was a doublet, which is already a standard of care. The initial 15-month follow-up demonstrated a benefit in progression-free survival. Now, we are reporting the final analysis at 45 months of follow-up. The benefit in progression-free survival remains, with a hazard ratio of 0.81, showing an increase from a median PFS of 11 months with nivolumab and ipilimumab to 16.6 months with the triplet. However, regarding overall survival, there is no difference between the two treatment arms.

This study is crucial because it paved the way for triplet therapy. We also learned that managing toxicity in a blinded study is challenging. Dose reductions were required for cabozantinib or placebo pills, and the use of ipilimumab was adjusted. The drug exposure data from this trial showed that the median daily dose of cabozantinib was only 22 milligrams per day, indicating that patients were receiving a lower exposure. Additionally, only 58% of patients in the triplet arm received all four cycles of ipilimumab, compared to 75% in the doublet arm. This highlights the difficulty in managing toxicity with triplet therapy. Half of the patients in the triplet arm had to discontinue at least one drug in the regimen. The toxicity was mostly driven by the combination, with notable elevations in liver enzymes. As we move toward triplet therapy, it is essential to consider how we will manage these toxicities.

Another interesting aspect of the study is the biomarker analysis using RNA sequencing. This included two analyses. The first connected to your earlier biomarker question, examining molecular clusters developed in the IMmotion150 and IMmotion151 studies. The seven molecular clusters were not associated with a differential benefit between the triplet and doublet treatments. The second analysis involved immune deconvolution, assessing the relative abundance of different immune cells and correlating them with outcomes such as overall survival and progression-free survival.

Interestingly, what stood out was the presence of M2-like macrophages, which were associated with outcomes. This cell population is typically linked to immune-suppressive activity and poor prognosis features, such as IMDC poor-risk factors and visceral metastases. In our dataset, adding cabozantinib in patients with high levels of this immune-modulatory cell population appeared to counteract the detrimental effect of M2-like macrophages. While this requires further validation, it aligns with findings from previous myeloid signature research, such as in the bevacizumab-plus-atezolizumab trial.

Dr. McDermott: Those were not my signatures, but—

Dr. Albiges: We are going to say they are yours.