Comparison of iRECIST versus RECIST V.1.1 in patients treated with an anti-PD-1 or PD-L1 antibody: pooled FDA analysis

BACKGROUND: Response criteria developed when cytotoxic chemotherapy was the predominant therapeutic modality to treat patients with cancer, do not capture the full spectrum of tumor response patterns observed with anti-PD-1/PD-L1 antibody treatment. iRECIST was developed to capture both typical and atypical response patterns. METHODS: Target, non-target, and new lesion measurements for 7920 patients receiving anti-PD-1/PD-L1 antibody (n=4751) or anti-CTLA-4 antibody (n=613) or undergoing chemotherapy (n=2556) from 14 randomized controlled trials submitted to the U.S. Food and Drug Administration were used to calculate the best overall response, objective response rate and progression-free survival (PFS) per iRECIST (iPFS) and Response Evaluation Criteria in Solid Tumours (RECIST). Associations between either PFS or iPFS and overall survival (OS) were evaluated using the method adopted by Oba et al.1 RESULTS: Among 4751 anti-PD-1/PD-L1-antibody treated patients, 31.5% (95% CI 30.2% to 32.9%) and 30.5% (95% CI 29.2% to 31.8%) achieved an objective response per iRECIST or RECIST V.1.1, respectively. OS among the 48 patients with objective response by iRECIST only resembled that in patients with responses per RECIST V.1.1. The association between iPFS and OS was R(2)=0.277 and that between PFS and OS was R(2)=0.260. CONCLUSIONS: Patients treated with anti-PD-1/PD-L1 antibodies with initial progressive disease per RECIST V.1.1 can experience prolonged stability or substantial reductions in tumor burden per iRECIST, atypical response patterns associated with prolonged OS. In the subgroup of patients with atypical responses, the application of iRECIST retrospectively in the evaluation of the objective response durations and the magnitude of PFS results in large differences compared with RECIST V.1.1. For the overall pooled population, the magnitude of these differences was modest, although a large proportion of patients had no further tumor assessments following RECIST V.1.1-defined progressive disease. Prospective studies employing iRECIST will be required to assess whether this response criteria more fully captures the benefit of immune checkpoint inhibitors.