Randomized, phase 2 study of carboplatin and pemetrexed with or without pembrolizumab as first-line therapy for advanced NSCLC: KEYNOTE-021 cohort G

Abstract

Background: Pembrolizumab (pembro) monotherapy exhibits robust antitumor activity in PD-L1─expressing advanced NSCLC. Cohort G of the multicenter, open-label, phase 1/2 KEYNOTE-021 study (NCT02039674) evaluated the efficacy and safety of pembro + carboplatin and pemetrexed (CP) vs CP alone as first-line therapy for advanced nonsquamous NSCLC. Methods: Key eligibility criteria were stage IIIB/IV, chemotherapy-naive, nonsquamous NSCLC, ECOG PS 0-1, and no EGFR mutation or ALK translocation. Patients (pts) were randomized to 4 cycles of carboplatin AUC 5 + pemetrexed 500 mg/m2 Q3W ± 24 mo of pembro 200 mg Q3W; maintenance pemetrexed was allowed in both arms. Randomization was stratified by PD-L1 tumor proportion score ≥1% vs <1%. Eligible pts with radiologic progression on CP could crossover to pembro monotherapy. Primary end point was ORR, with PFS as the key secondary end point; both were assessed per RECIST v1.1 by blinded independent central review. ORR and PFS superiority thresholds were one-sided α = 0.025. Results: 123 pts were enrolled: 60 in the pembro + CP arm, 63 in the CP arm. Demographics were generally balanced between treatment arms. As of Aug 8, 2016, median follow-up was 10.6 mo (range, 0.8-19.3); median exposure was 8.0 mo for pembro + CP and 4.9 mo for CP. In the CP arm, 43 pts discontinued therapy; 32 received subsequent anti–PD-1 therapy as part of crossover (n = 20) or off study (n = 12). Pembro + CP significantly improved ORR (55% vs 29%; P = 0.0016) and PFS (HR 0.53, 95% CI 0.31-0.91, P = 0.0102; median 13.0 vs 8.9 mo). Overall survival was similar; 6-mo survival rates were 92% in each arm. Without adjusting for exposure, for pembro + CP vs CP, treatment-related AEs led to discontinuation in10% vs 13%, were of grade ≥3 severity in 39% vs 26%, and led to death in 2% (sepsis, n = 1) vs 3% (sepsis and pancytopenia, n = 1 each). The most common any-grade treatment-related AEs were fatigue (64% vs 40%), nausea (58% vs 44%), and anemia (32% vs 53%). Conclusions: Pembro 200 mg Q3W + CP demonstrated a statistically significant, clinically relevant ORR and PFS benefit over CP alone and had a manageable, safety profile as first-line therapy in patients with advanced nonsquamous NSCLC.