Despite the survival gain observed with antiPD-1 agents in R/M HNSCC, responses are modest and no predictive biomarkers have been validated. We investigated immune and genomic biomarkers of response in a prospective cohort of R/M HNSCC patients (pts) treated with pembrolizumab 200 mg IV Q3W in the INSPIRE study.
Pts had blood samples (BS) collected at baseline (BL) and on treatment (OT) at cycle 3 (week 9); fresh tumor biopsies (FTB) collected at BL and OT (week 6-9). Analyses included tumor whole exome sequencing and immunophenotyping by flow cytometry of FTB and BS; tumor PD-L1 staining (clone 22C3) using modified proportion score. Response rate (RR) using RECIST 1.1. Median progression-free survival (PFS) and overall survival (OS) estimated by Kaplan-Meier method. Time to progression (TTP) was estimated using cumulative incidence function. Univariable analyses of genomic and immune parameters were conducted to identify response predictors.
Seventeen pts were enrolled: median age 62 years (48-71); smoking history > =10 pack-year (PY)= 11; oral cavity=7, oropharynx=6 (5 HPV+), larynx/hypopharynx=4. Platinum-refractory= 15; > = 2 prior lines of therapy= 4; PD-L1 > = 1%= 9. RR = 3 partial responses (PR), 10 stable disease (SD), 4 progressive disease (PD). Median follow-up was 4.9 months (m)(0.6-19.3); TTP 9.7m (3.5-not-reached (NR)), PFS 4m (2.3-9.5); OS 7.9m (3.4-NR). Pts with PR had >10 PY, PD-L1 >50% and no local recurrence. Pts with PR+SD had less proliferating ki67+, FoxP3+ T-regulatory cells (Tregs) in blood compared to PD pts (p0.05). Seven pts had FTB for analysis (1 PR, 3 SD, 3 PD): patient with PR (HPV+ oropharynx) had less immunosuppressive Tregs via lower expression of CTLA-4 and CD39 in BL tumor and increased proliferating Tregs at paired OT FTB. Tumor somatic coding mutation burden and percent genome copy alteration in FTB at BL (N = 9) did not predict response (PR+SD) but a trend was observed with percent loss of heterozygosity (p0.15).
This preliminary data suggest peripheral and tumor T-regs might play a critical role in R/M HNSCC pts treated with anti-PD1 agents. Dynamic immune-cell changes were more informative than BL genomic markers.
Clinical trial identification
Legal entity responsible for the study
Princess Margaret Cancer Centre.
A. Spreafico: Consultant and advisory boards: Merck, Bristol-Myers Squibb, Novartis. L.L. Siu: Advisory board: Merck; Funding to institution (Princess Margaret Cancer Centre) to conduct clinical trials: Merck. A. Hansen: Research support: Genentech/Roche, Merck, GlaxoSmithKline, Bristol-Myers Squibb, Novartis, Boston Biomedical, Boehringer-Ingelheim. All other authors have declared no conflicts of interest.