COSMIC-313: Investigating Progression-Free Survival Benefits in Renal Cancer Therapy - Toni K. Choueiri

June 2, 2023

Alicia Morgans and Toni Choueiri discuss the findings of the COSMIC-313 study, as published in the New England Journal of Medicine. The study examines the efficacy of adding cabozantinib to the Nivolumab-Ipilimumab (NIVO-IPI) regimen for treating advanced renal cell carcinoma. It finds that the triple combination (NIVO-IPI-CABO) significantly improves progression-free survival, especially in intermediate-risk patients, but the overall survival data remains immature. As the triple combination exhibits higher toxicity, questions are raised about the optimal application of the three-drug regimen. Further exploration in other studies like the PDIGREE study is suggested to investigate possible sequencing instead of simultaneous application. The counterintuitive absence of benefit in poor-risk patients is also highlighted as a future area of investigation. The researchers plan to update on overall survival and molecular biomarker studies in the following year and are excited about the potential knowledge to be gained from COSMIC-313.


Toni K. Choueiri, MD, Director, Lank Center for Genitourinary Oncology, Dana-Farber Cancer Institute, Boston, MA

Alicia Morgans, MD, MPH, Genitourinary Medical Oncologist, Medical Director of Survivorship Program at Dana-Farber Cancer Institute, Boston, Massachusetts

Read the Full Video Transcript

Alicia Morgans: Hi, I'm so excited to be here with Dr. Toni Choueiri to talk about COSMIC-313 in the recent New England Journal of Medicine publication. Thank you so much for being here with me today.

Toni Choueiri: Dr. Morgans, thank you again.

Alicia Morgans: Well, wonderful. Tell me a little bit about this study, another study in advanced renal cell carcinoma, such an important population. What did the study investigate?

Toni Choueiri: Yeah, no, for the first time, now that we know the doublets are standard, VEGF/IO or IO/IO, in renal cell cancer, we thought to ask the next important question, is three better than two? We know two is better than one, in general. Is three better than two? And so we investigated adding cabozantinib at 40 milligram a day dose to a combination that is standard in intermediate and poor metastatic RC, which is 80% of all patients that have metastatic RCC. Nivolumab-Ipilimumab was the control arm, the standard we use. This was a 1:1 randomization of NIVO-IPI, let's call it NIVO-IPI-CABO versus NIVO-IPI. Primary endpoint was progression-free survival and there was other secondary, your usual endpoints, with a stratification, most important stratification is intermediate versus poor.

The study primary endpoint was met NIVO-IPI-CABO is superior to NIVO-IPI in term of progression-free survival, statistically significant, I would argue clinically also relevant difference in PFS. The other endpoints, such as response rate, were numerically higher. CRs were not higher. One of the reasons could be that this is a population post CARMINA, where a lot of patient had primary in place, and that's very hard to have a response in the primary. Also, overall survival wasn't put in the publication because this was very preliminary, but from a press release earlier, very, very immature overall survival was not met.

So the next step here, if you ask me what are the next steps, is to look at overall survival. Now, interesting enough, and we looked into that, so far, we don't know why the benefit of PFS, even the numerical difference in response rate was only in intermediate-risk patients. You would think it would be mostly in poor-risk patients. That's not the case. Is it maybe because poor-risk patients, there was an imbalance in that population. Not really. There was no imbalance in baseline characteristics. Maybe the poor-risk population didn't have the chance to receive much cabozantinib and went off quickly? That also was not the case as we're trying to look into that. One theory is the biology. It's possible that poor-risk patients do not have much angiogenic profile in their tumor and cabozantinib or any VEGF inhibitor wouldn't be a great idea. So, next step on COSMIC-313 is molecular biomarker and certainly overall survival update. Expect that maybe next year, hopefully, but depends on the events.

Alicia Morgans: Yeah, absolutely. I think that will be so important because if you're thinking of adding a third agent onto a doublet that's already standard of care, you know PFS is going to be improved if OS is not really different, is this something we could sequence instead of doing all at once? What do you think?

Toni Choueiri: That's actually the question that another trial, an Alliance trial that co-led by us and Dr. Zhang, that asked the question in the PDIGREE study, that you start by NIVO-IPI and then if you have a CR, you continue NIVO for a year. If you have PD by RECIST, then you go to cabozantinib. And if not, normally you continue nivolumab if you have a response that is not a CR or stable disease. Well, in that population we can randomize patients to continue nivolumab or add cabozantinib, with the primary endpoint being overall survival. The study is accruing very well. I think this is so important, because going back to COSMIC-313, the toxicities were also higher in the combination arm, and that's not a surprise. So, of course, I think if you have a PFS benefit and you don't have an OS with a higher toxicity, you have to ask yourself where the three drugs work the best. And today we don't have an answer.

Alicia Morgans: That makes sense, too. Is this the kind of thing you might use in a setting where you have an intermediate-risk patient, you really need that higher response rate because of concerns for rapidly progressive disease? Is there a situation where you might use this combination now?

Toni Choueiri: Yeah, PD as best response were very low in the triplet arm, and that's not a surprise because VEGF plus IO have all a very low PD rate. We looked a bit if the sarcomatoid could be better, certainly they can get NIVO-IPI only. At this point, I think without the OS benefit, without looking at the intermediate risk only for OS, it's hard to justify it. I think in the future if there's inconsistent OS benefit, then we can start envisioning this becoming a standard.

On the other hand, it's very possible also that if OS is to happen, OS will happen only in the intermediate risk. Then we have to all sit down and see what to do. Is OS going to track the same as PFS and response rate? And again, why not in poor risk? This is a bit counterintuitive. So I think COSMIC-313, important study, three versus two the first time, met the primary endpoint, but still a lot of question. Hopefully 2024 will alleviate some of these.

Alicia Morgans: Absolutely. And I assume you do have some correlatives built in and some plans for follow-up analyses?

Toni Choueiri: Absolutely. We do have some correlatives, of course, outside just PD-L1, we have collected tissue, we have collected the blood, we have also some patients that underwent biopsies, has fresh tissue, tissue in good shape. Remember, because a front-line trial, a lot of patients do present with kidney in place, so we're trying to do a lot of tests. One of the them was Dr. Brown at Yale. He's an expert in F-Seq, so we're trying to do F-Seq in some of these tumors. So, certainly this is something we're looking at. My passion is biomarker, having came out now with none in practice, but we cannot stop.

Alicia Morgans: Very good. Wonderful. Well, thank you so much for sharing this. I think COSMIC-313 is definitely something that will continue to enlighten us in the future and I really appreciate hearing your thoughts on it today.

Toni Choueiri: Thank you Dr. Morgans, thank you UroToday and PCF for always being at the front.