Randomized Phase II Study of Concurrent Versus Sequential Pembrolizumab in Combination With Chemoradiation in Locally...

The optimal timing of pembrolizumab with chemoradiation (CRT) in locally advanced (LA) head and neck squamous cell carcinoma (HNSCC) is unknown.

Our phase II trial randomly assigned patients 1:1 to concurrent pembrolizumab (200 mg once every 3 weeks × 8) starting 1 week before CRT (cisplatin 40 mg/m² once weekly + radiation 70 Gy) versus sequential pembrolizumab starting 2 weeks after CRT. Human papillomavirus (HPV)+ (>10 pack-years or T4 or N3) and HPV(-) LA HNSCC were included, stratified by HPV and N stage. In our pick-the-winner design, if both arms met the trivariate primary end point (1-year locoregional failure <60%, progression-free survival [PFS] ≥60%, and dose limiting toxicity rate ≤20%), the arm with numerically superior 1-year PFS would be selected. Survival end points were compared by a univariate Cox model. Pretreatment and on-treatment tumor biopsies (week 2 of CRT) were evaluated by multispectral imaging and compared using two-sided paired t-tests.

Treated patients (41 concurrent and 39 sequential) were 71% oropharynx (53% HPV+), 92.5% stage IV (46% T4, 76% N2), similar by arm. Both arms met the trivariate primary end point, with superior 1-year PFS in the sequential arm (84% v 71%) and favorable 4-year outcomes: locoregional control (96% v 64%; hazard ratio [HR], 0.11 [95% CI, 0.01 to 0.89]; P = .012), PFS (69% v 49%; HR, 0.55 [95% CI, 0.25 to 1.22]; P = .132), and overall survival (83% v 71%; HR, 0.51 [95% CI, 0.19 to 1.37]; P = .17). There was a significant increase in macrophages, PD-L1+ macrophages, and PD-L1+ tumor cells with treatment in the concurrent but not the sequential arm.

CRT with sequential pembrolizumab met criteria for further study. Immunosuppressive changes in the TME differed between arms, reflecting the impact of one dose of pembrolizumab in the concurrent arm.