Abstract
Background: Immune checkpoint inhibitors (ICIs) have dramatically expanded the therapeutic landscape of non-small-cell lung cancer (NSCLC). In a previous study, gender, smoking history, and PD-L1 status were found to influence the efficacy of single-agent ICI in NSCLC. This meta-analysis evaluated the clinical and molecular factors that could predict a benefit from adding ICIs to first-line chemotherapy in metastatic NSCLC. Patients and Methods: The pooled hazard ratio (HR) of progression-free survival (PFS) and overall survival (OS) among the selected subgroups were analyzed using the random effects model. The correlation between PD-L1 expression and outcome was analyzed by meta-regression. Results: Seven phase III randomized controlled trials comparing chemo-immunotherapy (CIT) with chemotherapy in untreated stage 4 NSCLC were included. CIT evenly improved PFS irrespective of age, gender, histology, smoking history, and performance status. Among patients with baseline hepatic metastasis treated with Atezolizumab-containing CIT, PFS improvement was only detected with the addition of Bevacizumab. Whereas patients with EGFR/ALK-driven cancer exhibited greater PFS with the addition of ICI to a Bevacizumab (BEV)-based regimen, the derived benefit was no longer statistically significant among those treated with non-BEV-based regimens. Although the superior PFS conferred by CIT was noticeable across all PD-L1 expression subgroups, this benefit correlated with PD-L1 level and was more pronounced in the "PD-L1 high" cohort. Except for patients harboring EGFR/ALK aberrations or squamous histology, CIT consistently improved OS across the other selected subgroups. Conclusions: The survival advantage associated with first-line CIT in metastatic NSCLC was observed in different patient populations, including those for which single-agent ICI has marginal therapeutic benefit. Our findings support the role of chemotherapy with or without VEGF blockade as enhancers of ICI activity in NSCLC.