A 100% disease control rate across 40 response-evaluable patients is the kind of number that stops a room — and that is exactly what the Phase 2 interim readout from ROSETTA Lung-02 delivered for pumitamig, the PD-L1xVEGF-A bispecific being co-developed by BioNTech and Bristol Myers Squibb. The confirmed objective response rate of 72.7% in squamous NSCLC at the lower 1400 mg dose, paired with a cORR of 100% in patients with TPS ≥ 50%, is not a signal that needs squinting to see. For a disease where first-line pembrolizumab plus chemotherapy historically yields ORRs in the 40–60% range depending on subtype, these Phase 2 numbers represent a meaningful step up — with the caveat that cross-trial comparisons at nine months median follow-up are exactly that.
The trial design makes the data more credible than it might first appear. The Phase 2 portion randomized patients 1:1 across two dose levels rather than running a single arm, and efficacy held at both doses with the response advantage clustering at the lower dose. That dose-response inversion — better outcomes at 1400 mg than 2000 mg — is the kind of empirical finding that justifies the dose-selection work, and the team has landed on a 1500 mg flat dose for Phase 3. The PD-L1 agnostic activity matters enormously for trial design: responses at TPS < 1% (47.6% cORR) mean the pivotal program does not need to gate on biomarker enrichment, which simplifies enrollment and broadens the addressable population.
Safety warrants attention without alarm. Grade ≥ 3 TRAEs in 48.8% of patients sounds high until you note that only 23.3% were attributed specifically to pumitamig, and discontinuation reached just 9.3%. The VEGF-A arm of the mechanism raises the expected bleeding concern; nine bleeding events in 43 patients, with only one grade 3, is a manageable rate compared to what bevacizumab historically produced in NSCLC. The irAE rate of 37.2% is broadly consistent with checkpoint-based combinations.
Three Phase 3 trials are now enrolling simultaneously, including a direct head-to-head against pembrolizumab in PD-L1 ≥ 50% patients (ROSETTA Lung-202) — the clearest competitive confrontation in the program. That trial’s progression-free survival readout will determine whether pumitamig earns label language that displaces pembrolizumab monotherapy, the most consequential regulatory question the entire program faces.
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.
