Background: Pembro is an anti–PD-1 antibody approved for treating advanced MEL that progressed following IPI and, if BRAF^V600 mutant, a BRAF inhibitor. In KEYNOTE-002 (NCT01704287), pembro doses of 2 mg/kg and 10 mg/kg every 3 wk (Q3W) significantly improved PFS compared with investigator-choice chemotherapy in IPI-R MEL (P < .0001), with no difference between pembro doses (P = .44). Data from KEYNOTE-001 showed that PD-L1 positivity was correlated with a higher ORR and longer PFS in MEL pts. We evaluated efficacy in PD-L1⁺ and PD-L1^– MEL pts enrolled in KEYNOTE-002. Methods: 540 pts with advanced IPI-R MEL were randomized 1:1:1 to pembro 2 or 10 mg/kg every 3 wk (Q3W) or chemotherapy. PD-L1 expression was assessed centrally by IHC using the 22C3 antibody. The cut point for positivity was staining in ≥1% of tumor cells. Response was assessed at wk 12 and every 6 wk thereafter (RECIST v1.1, central review). Primary end points were PFS and OS (final analysis planned after 370 deaths occur). ORR and the relationship between PD-L1 expression and efficacy were secondary end points. Data for the pembro arms were pooled. Results: 421/540 pts enrolled (78%) had tumors evaluable for PD-L1 expression; 291 (69%) were PD-L1⁺, 130 (31%) were PD-L1^–. Pembro prolonged PFS and increased ORR compared with control in PD-L1⁺ and PD-L1^– pts (Table). Among pembro-treated pts, ORR was higher in PD-L1⁺ pts, but the CIs were overlapping, and there was no difference in duration of response (DOR) based on PD-L1 expression. There was no prognostic effect in the control arm. Conclusions: Pembro improved efficacy over chemotherapy in both PD-L1⁺ and PD-L1^– IPI-R advanced MEL. These data indicate that in IPI-R MEL pts, pembro therapy should not be limited to pts with PD-L1⁺ tumors. Clinical trial information: NCT01704287

^aPembro vs control.