Setanaxib + pembrolizumab misses its tumor-shrinkage endpoint but shows exploratory survival signals in CAF-rich head and neck cancer

Setanaxib (GKT137831) is an oral dual NADPH oxidase (NOX)1/NOX4 inhibitor intended to reduce reactive oxygen species (ROS)–driven, fibroblast-associated immune suppression, and it was combined with pembrolizumab (anti–PD-1) in a randomized, double-blind Phase 2 trial in recurrent/metastatic CAF-rich HNSCC with CAF tumor levels ≥5% (NCT05323656; n=55). The primary endpoint (best % change in tumor size) did not improve: LS mean change -7.9% with setanaxib vs -12.9% placebo (LS mean difference 5.1; 80% CI -11.9 to 22.0; p=0.7008), but time-to-event signals favored setanaxib (PFS HR 0.58, 80% CI 0.38–0.89; p=0.1023 with a prespecified significance threshold <0.2; OS HR 0.45, 80% CI 0.24–0.85; p=0.1057). Transcriptomics showed increased CD8+ TILs at week 9 with setanaxib (p=0.037) and not placebo; serious AEs were similar (16 vs 17), and four deaths occurred on placebo; for real-world context, first-line pembrolizumab in R/M HNSCC delivers median OS ~12.3 months in CPS≥1 and ~14.9 months in CPS≥20 (KEYNOTE-048)—so a Phase 2 OS HR signal is interesting, but needs confirmation with mature medians and standard CI/p-value reporting.