Feasibility and safety of PD-1/L1 inhibitors for non-small cell lung cancer in front-line treatment: a Bayesian network meta-analysis

Background This Bayesian network meta-analysis (NMA) was conducted to compare efficacy and safety of programmed death 1/ligand 1 (PD-1/L1) inhibitors in previous untreated advanced non-small cell lung cancer (NSCLC) patients. Methods Eligible studies evaluating first-line anti-PD-1/L1 based regimens in advanced NSCLC patients were included. Overall survival (OS), progression free survival (PFS), objective response rate (ORR), as well as treatment-related severe adverse events (tr-SAE) were synthesized within the Bayesian framework. Subgroup analysis was conducted according to PD-L1 expression. Results Twelve studies including 7,490 patients and 9 treatment strategies were enrolled in this study. For the PD-L1 expression non-selective patients, all chemo-immunotherapies were significantly better than chemotherapy for prolonging OS and PFS, except for caremlizumab plus chemotherapy (HR =0.72) failed to show advantages for OS. In addition, pembrolizumab plus chemotherapy showed better PFS than nivolumab plus ipilimumab (HR =0.66). In PD-L1 ≥50% patients, all immunotherapy was better than chemotherapy for OS, except for nivolumab (HR =0.83) and nivolumab plus ipilimumab (HR =0.70). For PFS, pembrolizumab plus chemotherapy (HR =0.39), atezolizumab plus chemotherapy (HR =0.47) and pembrolizumab (HR =0.67) were significantly better than chemotherapy. In PD-L1 1-49% patients, pembrolizumab plus chemotherapy (HR =0.52) and atezolizumab plus chemotherapy (HR =0.70) were better than chemotherapy for PFS. In the PD-L1 positive or negative group, all included corresponding regimens were equivalence according to OS and PFS. Conclusions We conducted a systematic comparison of first line immunotherapy for advanced NSCLC. Chemo-immunotherapies were better than chemotherapy and mono-immunotherapies in most patients. Pembrolizumab might have better efficacy than other PD-1/L1 inhibitors.