Cohort G of the KEYNOTE-021 study enrolled patients with stage IIIB/IV nonsquamous non–small-cell lung cancer (NSCLC), no activating EGFR mutation or ALK translocation, no prior systemic therapy, measurable disease, ECOG 0 to 1, and adequate tumor sample for PD-L1 status assessment, regardless of PD-L1 expression. Patients were randomized to receive pembrolizumab (200 mg) with carboplatin (AUC 5) and pemetrexed (500 mg/m2) every 3 weeks or carboplatin (AUC 5) and pemetrexed (500 mg/m2 every 3 weeks) alone (chemo). Treatment was given for 4 cycles, followed by maintenance pemetrexed ± pembrolizumab. Pembrolizumab was given for up to 35 cycles.
Randomization was stratified by PD-L1 expression (positive [tumor proportion score (TPS), ≥1%] vs negative [TPS <1%]). Crossover to monotherapy was allowed for eligible patients who experienced disease progression on chemo. The primary end point was objective response rate (ORR). Secondary end points were progression-free survival (PFS), duration of response, and overall survival (OS).
The study enrolled 123 patients with nonsquamous NSCLC (n = 60 pembrolizumab + chemo; n = 63 chemo). After median follow-up of 10.6 months, ORR was nearly doubled with pembrolizumab + chemo versus chemo alone (55% vs 29%, respectively). An ORR of 80% was noted in 20 patients with PD-L1 ≥50%. Risk of disease progression or death was nearly halved, with a hazard ratio for PFS of 0.53 (confidence interval, 0.31-0.91). Median PFS was 13.0 versus 6.8 months for pembrolizumab + chemo versus chemo alone, respectively. OS was similar between the 2 arms: 92% of patients in both arms were alive at 6 months.
Adverse events (AEs) of pembrolizumab + chemo were generally manageable. There was a higher rate of AEs in the pembrolizumab + chemo arm versus chemo (39% vs 26%), but this did not affect discontinuation rates secondary to AEs (10% vs 13%).
Researchers concluded that pembrolizumab + chemo could be an effective treatment option for patients with chemo-naïve advanced nonsquamous NSCLC.
Langer C, et al. WCLC 2016. Abstract MA 09.02. ID 5787.