Background: Platinum doublets provide 6-mo PFS and 12-mo OS for treatment-naive, metastatic NSCLC without driver mutations. Preliminary data from KEYNOTE-001 revealed robust antitumor activity and manageable toxicity for the anti–PD-1 antibody pembro in treatment-naive NSCLC. In these patients (pts), PD-L1 positivity correlated with improved ORR, PFS, and OS. We explored the relationship between PD-L1 expression levels and efficacy in treatment-naive pts in KEYNOTE-001. Methods: 101 treatment-naive, PD-L1⁺ metastatic NSCLC pts were randomized 1:1 to pembro 10 mg/kg Q2W or Q3W (11 pts randomized to 2 or 10 mg/kg Q3W). Pembro was given until unacceptable toxicity, PD, or pt/investigator decision. PD-L1 expression was assessed by IHC (22C3 antibody). Staining of tumor cells or stroma by a prototype assay was used for enrollment. In a prospectively defined validation set (n = 90), the percentage of PD-L1–stained tumor cells was determined with a clinical trial assay. Response was assessed centrally every 9 wk by RECIST 1.1. Results: There were no significant between-dose differences in ORR, PFS, or OS; 11% of pts experienced gr 3-5 drug-related AEs. In the validation set (38-wk median follow-up), ORR and DCR were highest and PFS was longest in pts with ≥50% PD-L1 tumor cell staining (Table); there was a trend toward longer OS (immature data). Conclusions: Pembro demonstrateddurable antitumor activity and manageable toxicity profile as first-line therapy for PD-L1⁺ metastatic NSCLC, with the greatest efficacy observed in pts with PD-L1 staining in ≥50% of tumor cells. These data demonstrate that PD-L1 expression in tumor cells identifies those NSCLC pts most likely to respond to pembro. The efficacy and safety of pembro as first-line therapy for PD-L1⁺NSCLC is being examined in ongoing, controlled clinical trials. Clinical trial information: NCT01295827

NR, not reached.