CheckMate-025 and METEOR: The Overall Survival Barrier Has Finally Been Broken in RCC

by Arnold DuBell | Sep 28, 2015
Co-authored by Stephanie Hawthorne, Ph.D. 

Sutent® (sunitinib, Pfizer) and Votrient® (pazopanib, Novartis) are currently considered the standards of care for first-line therapy in renal cell carcinoma (RCC). However, neither drug (nor any other agents currently approved) has been able to show an overall survival (OS) benefit in clinical trials; they were approved by regulatory agencies based on a significant improvement in progression-free survival (PFS). Since 2005, seven agents have been approved, including Sutent and Votrient, on the basis of PFS alone.

Moreover, the presence of two established agents with strong utilization in the first-line setting contributes to the need for agents that can be utilized following relapse. Two agents are currently approved for use in this setting – Afinitor® (everolimus, Novartis), based on significant improvements in PFS compared with placebo, and Inlyta® (axitinib, Pfizer), based on significant improvement in PFS compared with Nexavar® (sorafenib, Bayer/Amgen). Given the lack of proven OS benefits for any of these agents, an opportunity exists for an agent that can establish a significant benefit for this endpoint. It is in this framework that  two trials presented in the Presidential Symposium Saturday at the 2015 European Cancer Congress finally show the possibility of improving OS. The trials were CheckMate-025, which evaluated Opdivo® (nivolumab, Bristol-Myers Squibb/Ono), and METEOR, which evaluated Cometriq® (cabozantinib, Exelixis).

CheckMate-025 was the first of these two studies to be presented.1 This trial compared Opdivo (3 mg/kg every two weeks) to daily oral Afinitor (10 mg) in 821 advanced or metastatic clear-cell RCC patients who had been treated with one or two prior anti-angiogenic therapies. The trial was stopped early by the data-monitoring committee in July 2015 as the statistical boundary for declaring OS superiority of Opdivo had been reached. Opdivo reduced the risk of death by 27% (median OS: 25.0 months versus 19.6 months, HR 0.73, p=0.0018). Most subgroups favored Opdivo, although the arms for elderly patients (75 years or older), patients with favorable MSKCC risk status, or patients who had received two prior therapies (confidence interval overlapped the mark for non-significance)showed no apparent difference ,. Tumor PD-L1 positivity, defined as PD-L1 membrane expression in 1% or more of cells stained using the Dako immunohistochemical kit, was not predictive of activity. The objective response rate was improved (25% versus 5%, OR 5.98, p<0.0001), but the duration of response (12.0 months in both arms) and PFS (4.6 months versus 4.4 months, HR 0.88, p=0.1135) were not improved.

Opdivo was better tolerated than Afinitor, with fewer treatment-related Grade 3/4 adverse events than Afinitor (19% versus 37%). The more common toxicities of any grade for Opdivo included fatigue (33%), nausea (14%), pruritus (14%), diarrhea (12%), and decreased appetite (12%). The improved efficacy and tolerability were reflected in an improved score in quality of life as measured by the FKSI-DRS questionnaire (p<0.05).

METEOR was presented next in the session.2 METEOR compared oral daily Cometriq (60 mg) with oral daily Afinitor (10 mg) in 658 advanced clear-cell RCC patients who have progressed on a prior VEGFR tyrosine kinase inhibitor (TKI) within six months of enrollment. Despite that inclusion criterion, there was no limit to the number of prior therapies as 29% of patients on the Cometriq arm received two or more prior VEGFR TKI treatments. Other therapies were also allowed but had been less frequently offered to the enrolled patients.

METEOR achieved its primary endpoint, which was PFS, with a 3.6-month improvement (HR 0.58, p<0.001). Subgroup analysis suggested that Cometriq improved PFS in most patients, but PFS may not have been significantly improved in those with two or more prior therapies or a poor MSKCC risk status. The response rate was also improved (21% versus 5%, p<0.001). OS was not yet reached but trended to significance with Cometriq (HR 0.67, p=0.005). The medians could not be estimated due to frequent early censoring, and the interim boundary for significance was set at p=0.00019). As the Kaplan-Meier curves were clearly separated from one another and these data are immature, it is expected that a significant OS benefit ultimately will be reached.

Cometriq had some toxicities but was generally well-tolerated. Grade 3 or 4 adverse events were higher in the Cometriq arm (68% versus 58%) and were reflected in a higher percentage of dose reductions (60% versus 25%). Select Grade 3/4 toxicities of note, compared with Afinitor, included hypertension (15% versus 3%), diarrhea (11% versus 2%) and palmar-plantar erythrodysesthesia (8% versus less than 1%); anemia (5% versus 16%) and hyperglycemia (less than 1% versus 5%) were improved in the Cometriq arm.

If METEOR does not reach its threshold for statistical significance, the decision for physicians will be easy: Choose Opdivo for its improvement in OS. If the OS benefit in METEOR becomes statistically significant (and the Kaplan-Meier curves strongly suggest that it will do so), then physicians will have a more difficult decision to make. The high excitement level for immuno-oncology agents in general, coupled with its seemingly more favorable tolerability, may lead physicians to opt first for Opdivo. On the other hand, Cometriq does have its own advantages, perhaps most notably the ability to delay progression (by 3.6 months at the median in the METEOR trial).\


CheckMate 025









mOS, mos.




Median not available


mPFS, mos.







Total Grade 3/4 AEs (%)








Making two assumptions – that the METEOR trial ultimately shows a significant OS benefit and that regulatory agencies will approve the two drugs within a few months of each other – these two agents will have a hefty competition for market share in second-line RCC. The toxicity profile for Opdivo may favor its use in many patients, although Cometriq might end up being the agent of choice for select patients, such as those who are older than 75 years of age, where Opdivo appears less effective. In a way, this is a good problem, at least for oncologists and their patients. It’s always better to question which of several good agents to choose from rather than having few options or none at all. 


  1. Sharma P, et al. CheckMate 025: a randomized, open-label, phase III study of nivolumab (NIVO) versus everolimus (EVE) in advanced renal cell carcinoma (RCC). Vienna, Austria: European Cancer Congress; 2015. Abstract 3LBA. 
  2. Choueiri T, et al. Cabozantinib versus everolimus in patients with advanced renal cell carcinoma: Results of the randomized phase 3 METEOR trial. Vienna, Austria: European Cancer Congress; 2015. Abstract 4LBA. 

Leave a comment

© 2017 Kantar Health, A Kantar Company