Programmed Death 1 and Programmed Death Ligand 1 Inhibitors in Advanced and Recurrent Urothelial Carcinoma: Meta-analysis of Single-Agent Studies

We performed a systematic review and meta-analysis on the response rates of patients with treatment-refractory urothelial carcinoma treated with programmed cell death 1 (PD-1) and programmed death ligand 1 (PD-L1) inhibitors. We reviewed the literature for prospective studies evaluating PD-1/PD-L1 inhibitors in refractory urothelial carcinoma patients, which formed the basis for US Food and Drug Administration approval of 5 different antagonistic antibodies targeting PD-1 or PD-L1 (atezolizumab, durvalumab, avelumab, nivolumab, and pembrolizumab). We considered studies examining PD-1/PD-L1–treated patients, which we identified using the following key terms in the Pubmed, Scopus, Web of Science, ClinicalTrial.gov, and Cochrane Library databases. Eligible studies had ≥ 20 patients each and reported response rates, duration of response, and overall survival (OS). We performed fixed and random-effects meta-analyses to model the point estimates for objective response rate and complete response. The median progression-free survival (PFS) and OS for studies reporting these statistics were evaluated. We found 10 eligible studies that met our inclusion criteria, providing extractable numerators and denominators for response rates, PFS, and OS for 1934 patients with metastatic urothelial carcinoma. The objective response rate was 18% (95% confidence interval, 15-22) for second-line or later therapies. The random-effects estimate for complete response was 4% (95% confidence interval, 3-5), including all disease locations and all PD-1 and PD-L1 inhibitors. Median OS and PFS were < 13 months and 3 months, respectively, across all studies, irrespective of PD-L1 expression. We found that the estimated response rates of agents included in this meta-analysis seem to be more favorable than other salvage therapies.