A 34% reduction in the risk of death — HR 0.66 — achieved not against placebo or chemotherapy alone, but against an active PD-1 inhibitor plus chemotherapy backbone, is a result that most oncology trialists would have assigned low prior probability. That is the number that makes HARMONi-6 genuinely disruptive. The phase III trial enrolled 532 patients with advanced squamous NSCLC, a population historically resistant to meaningful immunotherapy gains beyond the first checkpoint inhibitor wave, and randomized them to ivonescimab plus chemotherapy versus tislelizumab plus chemotherapy. At a median follow-up of 21.36 months, median OS was 27.9 months versus 23.7 months — a gap that widened rather than narrowed over time, with 24-month OS rates of 64.7% versus 48.6%.
The design deserves scrutiny because it is what elevates this above prior ivonescimab signals. Using an active PD-1 comparator eliminates the confounding question of whether dual VEGF/PD-1 blockade simply outperforms chemotherapy alone. It does not. It outperforms a regimen that already includes checkpoint inhibition. The PFS result — 11.1 versus 6.9 months, HR 0.60 — was the primary endpoint and already reported; OS was a key secondary, pre-specified, and cleared its alpha-spending threshold of 0.0049 comfortably at P=0.0017. That sequential gating matters for regulatory interpretation. Both endpoints fell in the same direction, with consistent hazard ratios across every prespecified subgroup including PD-L1 TPS below 1% (HR 0.64) and patients with three or more metastatic sites (HR 0.47).
The safety data introduce the one genuine complication. Grade ≥3 treatment-related adverse events hit 69.2% in the ivonescimab arm versus 58.9% with tislelizumab plus chemotherapy — a ten-percentage-point gap that cannot be dismissed as noise, even if discontinuation and treatment-related death rates were similar between arms. Squamous NSCLC patients often carry baseline pulmonary compromise and prior smoking-related comorbidities; that toxicity differential will shape real-world prescribing and payer coverage arguments more than the survival curves alone. Subsequent therapy rates were actually lower in the ivonescimab arm, meaning the OS benefit is not an artifact of better salvage access.
The single marker to track now is whether Summit Therapeutics‘ US development program — running its own ivonescimab trials under a licensing structure — produces concordant OS data in a Western population, because regulatory approval in the United States will hinge entirely on that replication, not on HARMONi-6 alone.
Jon Napitupulu is Director of Media Relations at The Clinical Trial Vanguard. Jon, a computer data scientist, focuses on the latest clinical trial industry news and trends.
