In KEYNOTE-024 (NCT02142738), pembrolizumab was superior to chemotherapy as 1L therapy for advanced NSCLC with PD-L1 TPS ≥50% and no sensitizing EGFR mutations/ALK translocations. After median follow-up of 11.2 mo, HR was 0.50 for PFS by independent central radiologic review (P<0.001) and 0.60 for OS (P=0.005). We present PFS2 and updated OS.
305 patients were randomized to pembrolizumab 200 mg Q3W (n=154) or investigator-choice platinum-doublet chemotherapy with optional pemetrexed maintenance for nonsquamous histology (n=151). Patients in the chemotherapy arm could cross over to pembrolizumab upon PD (2L+). Kaplan-Meier PFS2 and OS were calculated in all allocated patients. PFS2 was defined as time from randomization to investigator-determined PD after start of 2L therapy, discontinuation of 2L therapy, or death, whichever occurred first; patients alive and without 2L PD were censored at last known survival. There was no adjustment for multiplicity.
As of Jan 5, 2017, 48 patients in the pembrolizumab arm and 97 in the chemotherapy arm had received 2L+ therapy, including 79 who crossed over from chemotherapy to pembrolizumab per protocol. 46 (30%) patients continued on 1L pembrolizumab therapy or in follow-up. Median (95% CI) PFS2 was 18.3 mo (12.7-NE) for patients initially assigned to 1L pembrolizumab and 8.4 mo (6.8-9.8) for those assigned to 1L chemotherapy (HR, 0.54; 95% CI, 0.40-0.72; nominal P<0.001). There were 63 deaths in the pembrolizumab arm and 84 in the chemotherapy arm. Median (95% CI) OS was longer for 1L pembrolizumab (not reached, [19.4 mo-NE]) vs 1L chemotherapy (14.5 mo [9.8-19.6]; HR, 0.63; 95%CI, 0.46-0.88; nominal P=0.003).
Fewer pembrolizumab patients received 2L+ therapy because of the significant improvement in PFS observed for pembrolizumab in the 1L setting. Median PFS2 was substantially improved for pembrolizumab vs chemotherapy. Updated OS maintained consistent superiority of 1L pembrolizumab, despite increased crossover from 1L chemotherapy.