Predictors of survival to immunotherapy and chemoimmunotherapy in non-small cell lung cancer: A meta-analysis

Background Many patients with non-small cell lung cancer (NSCLC) derive poor benefit from immunotherapy (IO). For some of them, adding chemotherapy (CT) can improve the outcomes, but the reliability of programmed death-ligand 1 (PD-L1) expression as the only biomarker to distinguish these patients is unsatisfactory. We sought to detect clinicopathological and molecular predictive factors of survival that might be added to PD-L1 expression in the selection of patients who should receive IO alone or chemoimmunotherapy (CIT). Methods We conducted a systematic search of randomized controlled clinical trials investigating IO, alone or with CT, vs CT alone in treatment-naive advanced NSCLC patients. Meta-analyses and meta-regression analyses were performed to investigate IO alone vs CT, CIT vs CT, and IO alone vs CIT. Results A total of 14 367 patients with advanced NSCLC across 25 randomized controlled clinical trials were included. Squamous histology, male sex, current and former smoker status, PD-L1 expression of 50% or more, and high tumor mutational burden (TMB) correlated with improved survival with IO alone compared with CT. Conversely, female sex, no smoking history, negative PD-L1 expression, and low TMB correlated with unsatisfactory outcomes with IO alone vs CT but not with CIT vs CT. CIT improved survival vs IO alone in female patients, never smokers, those having a PD-L1 expression of 1% or more (but not with a PD-L1 of >= 50%) or a low TMB and in patients with central nervous system metastasis. Conclusions These findings suggest some clinicopathological and molecular features that, added to PD-L1 expression, could help in the selection of the most appropriate first-line IO-based treatment for advanced NSCLC patients.