Japanese Subgroup Analysis of KEYNOTE-048: Ph3 Study of 1L Pembrolizumab for R/M Head and Neck Squamous Cell Carcinoma

Abstract

Background: The global ph 3 KEYNOTE-048 studied pembrolizumab (P) or P+chemotherapy (C) vs EXTREME (E) as 1L treatment (tx) for recurrent/metastatic (R/M) HNSCC (NCT02358031). Methods: Patients (pts) with R/M HNSCC and no prior systemic tx (for R/M) were randomized to P, P+C, or E arms (Rischin et al. JCO. 2019;37:15-suppl, 6000). Primary end points: PFS by blinded review and OS. Data cutoff for this ad hoc interim analysis of Japanese (JPN) pts: June 13, 2018. Results: 67/882 were JPN (23 [P], 25 [P+C], 19 [E]). Median (med) follow-up (FU) (P vs E): 19.4 vs 13.6 mo. OS was longer (P vs E) in CPS≥20 (n=22; HR, 0.22; 95% CI, 0.06-0.77), CPS≥1 (n=37; HR, 0.67; 95% CI, 0.31-1.48), and total population (pop) (n=42; HR, 0.52; 95% CI, 0.25-1.11). PFS was longer (P vs E) in CPS≥20 (HR, 0.57; 95% CI, 0.22-1.43) and similar between arms in CPS≥1 (HR, 1.04; 95% CI, 0.53-2.04) and total pop (HR, 1.19; 95% CI, 0.64-2.23). ORR (P vs E): 29% vs 13% in CPS≥20, 19% vs 25% in CPS≥1, and 17% vs 37% in total pop and med DOR was 8.4 vs 2.6 mo, 8.4 vs 5.5 mo, and 8.4 vs 4.1 mo, respectively. Gr 3-5 tx-related AE (TRAE) rates (P vs E): 22% vs 90%. Med FU (P+C vs E): 12.6 vs 13.3mo. OS was longer (P+C vs E) in CPS≥20 (n=17; HR, 0.57; 95% CI, 0.17-1.90) and similar between arms in CPS≥1 (n=33; HR, 1.13; 95% CI, 0.50-2.53) and total pop (n=41; HR, 1.03; 95% CI, 0.50- 2.13). PFS was longer (P+C vs E) in CPS≥20 (HR, 0.34; 95% CI, 0.09-1.19), CPS≥1 (HR, 0.67; 95% CI, 0.32-1.41), and total pop (HR, 0.74; 95% CI, 0.38-1.42). ORR (P+C vs E): 50% vs 14% in CPS≥20, 32% vs 21% in CPS≥1, and 32% vs 31% in total pop; med DOR was 6.9 vs 2.6 mo, 7.5 vs 4.1 mo, and 7.5 vs 4.1 mo, respectively. Gr 3-5 TRAE rates (P+C vs E): 76% vs 90%. Conclusions: Responses were durable and safety was favorable in JPN pts, supporting P or P+C as 1L tx for R/M HNSCC. Tx with P or P+C improved PFS, OS, and ORR in JPN pts with CPS≥20. Although JPN pt numbers were small, data are consistent with those of the total study pop.