In the KEYNOTE-564 trial, treatment with pembrolizumab showed a disease-free survival (DFS) benefit compared to placebo in patients with clear-cell renal-cell carcinoma (ccRCC) at a high risk of recurrence. Yet, overall survival (OS) data were still immature. In this scenario, both the EAU and ESMO weakly recommend to offer adjuvant pembrolizumab following surgery with curative intent in patients meeting the KEYNOTE-564 study criteria. In this study we sought to evaluate the proportion of contemporary patients who might be eligible for adjuvant pembrolizumab in a large multi-institutional European cohort. Moreover, we explore the theoretical prognosis of all patient categories meeting the KEYNOTE-564 study criteria using the Leibovich 2003 and ECOG-ACRIN 2805 (ASSURE) prognostic models.
After Ethical Committee approval, data from consecutive patients who underwent surgery for non-metastatic ccRCC at three high-volume referral Academic Centres between 2015 and 2021 were retrospectively collected. Patients were classified as eligible vs non-eligible for adjuvant pembrolizumab according to the KEYNOTE-564 study criteria. The theoretical estimated 5-year metastasis-free survival (MFS) according to the Leibovich 2003 model and estimated 5-year DFS according to the ASSURE model were calculated for each category of “eligible” patients.
Overall, 695 patients with ccRCC were included. Of these, 177/695 (26%) could have met the KEYNOTE-564 eligibility criteria, most of which (86% with a pT3a N0 stage). Patients with pTany pN+ disease represented 6% of eligible patients. Theoretically, “eligible” patients according to the KEYNOTE-564 study criteria showed a highly heterogenous prognosis according to the Leibovich 2003 and ASSURE models. Patients deemed at “intermediate-high risk” of recurrence may have a 5-y MFS ranging between 12% and 75% according to the Leibovich model, while a 5-y DFS ranging between 29% and 79% according to the ASSURE model. Similarly, patients deemed at “high risk” of recurrence may have a 5-y MFS ranging between 12% and 95% according to the Leibovich model, while a 5-y DFS ranging between 8% and 79% according to the ASSURE model.
Approximately one out of four patients with ccRCC after surgery with curative intent would have been eligible for adjuvant immunotherapy according to the KEYNOTE-564 study criteria. Applying more granular prognostic models, the estimated 5-y DFS of “eligible” candidates might significantly vary, underlying the potential inherent heterogeneity of this disease. While waiting for robust OS data and future results of ongoing adjuvant trials, further steps are needed to optimize patient selection and improve the value proposition of adjuvant pembrolizumab for ccRCC at high risk of recurrence.