Drugmakers are committing billions to next-generation oncology platforms before clinical evidence is mature. The clearest example: Bristol Myers Squibb agreed to pay BioNTech $1.5 billion upfront, $2 billion in noncontingent anniversary payments, and up to $7.6 billion in milestones for pumitamig, an investigational bispecific targeting PD-L1 and VEGF-A. The deal could be worth $11.1 billion, even though pumitamig's randomized proof is still incomplete.
Evidence for the broader PD-(L)1/VEGF strategy is mixed. Akeso and Summit Therapeutics' ivonescimab beat Merck's Keytruda on progression-free and overall survival in a China-only phase 3 trial in first-line non-small cell lung cancer. But in the global HARMONi trial, in EGFR-mutated NSCLC, ivonescimab met its progression-free survival endpoint while the prespecified primary overall survival analysis narrowly missed statistical significance; later follow-up showed a survival trend. Pumitamig has shown encouraging response rates in an interim phase 2 cohort, but without a comparator.
Other fields, including B7-H3 antibody-drug conjugates and in vivo CAR-T, sit at different points on the same proof curve. GSK's Eric Richards says it is important to separate the biological rationale for a target from the clinical evidence for an individual ADC. Mike Rea, CEO of Then, If..., describes the situation as paying 'promise prices on premise evidence': rational only if the buyer can specify what would make the premise fail.