Association of PD-L1 expression with survival benefit from PD-1/PD-L1 inhibitors in advanced cancer: systematic review and meta-analysis of phase III randomized clinical trials

Whether PD-L1 testing is needed to identify patients receiving PD-1/PD-L1 inhibitors is an area of debate. PubMed and Embase were searched for phase III randomized clinical trials. We assessed the heterogeneity of overall survival (OS) between patients with high and low PD-L1 expression using an interaction test. Seventy studies representing 44791 patients were included. Both the CPS and TPS can predict better survival from anti-PD-1/PD-L1 therapy in patients with high PD-L1 expression. However, only CPS 1 has the ability to select patients who are unlikely to respond to anti-PD-1/PD-L1 therapy, while an OS advantage can be obtained from PD-1/PD-L1 inhibitors both in patients with high and low PD-L1 expression defined by CPS 5, CPS 10 and TPS. CPS 1 is recommended to select patients with the likelihood of benefiting from PD-1/PD-L1 inhibitors while excluding patients who may not respond.