Two Pharma Giants, One Drug Class, Totally Different Bets
BioNTech/BMS and Pfizer are building nearly identical lung cancer drugs but just made opposite bets on how to prove they work. Their ASCO 2026 split on trial endpoints could reshape how the next generation of cancer therapies reaches patients.
Imagine two poker players sitting at the same table, holding similar hands, and making completely opposite bets. That's what just happened at ASCO 2026.
BioNTech/Bristol Myers Squibb and Pfizer are both racing to build the next great lung cancer drug. Both are developing bispecific antibodies for first-line non-small cell lung cancer (NSCLC). Both are gunning for the throne currently occupied by Merck's Keytruda. And yet, they just made a very public split on one of the most consequential decisions in drug development: what counts as winning.
The Fork in the Road
To understand why this matters, you need a quick primer on how cancer trials measure success. There are two main scorecards.
Progression-free survival (PFS) measures how long patients live without their tumors growing. Think of it as "time before the cancer fights back." It reads out faster because you don't have to wait for the worst outcome.
Overall survival (OS) is the gold standard: did patients actually live longer? It takes years to mature, but nothing in oncology carries more weight with regulators, doctors, and insurers.
At ASCO 2026, BioNTech and BMS announced that their Phase 3 trial for pumitamig (a bispecific targeting PD-L1 and VEGF-A) would use PFS as the sole primary endpoint, with OS relegated to a "key secondary." Meanwhile, Pfizer confirmed that OS will remain a dual primary endpoint in its Symbiotic-Lung-01 Phase 3 trial for PF-08634404 (targeting PD-1 and VEGF).
Same drug class. Same disease. Same enemy (Keytruda + chemo). Opposite philosophies.
The Need for Speed
BioNTech/BMS aren't being reckless. They're being impatient, and they have reasons.
BioNTech's CMO Özlem Türeci explained the logic at ASCO: PFS is "the earliest and most sensitive acceptable endpoint" for this type of drug. By putting their entire statistical alpha on PFS (translation: concentrating all their firepower on one target instead of splitting it across two), they get a cleaner, faster shot at proving the drug works.
Get tomorrow's biotech intelligence before your competitors.
Join thousands of biotech professionals who start their day with our free, daily briefing.
If PFS hits, they can sit down with the FDA sooner. They don't have to wait for enough patients to die to generate a mature OS readout. It's the pharmaceutical equivalent of taking the early flight: you get there faster, even if the legroom isn't as nice.
The company isn't ignoring survival, though. They expanded enrollment from 982 to 1,260 patients in ROSETTA Lung-02, their pivotal trial testing pumitamig plus chemo versus pembrolizumab plus chemo. That larger population gives them enough statistical power to test OS as a secondary endpoint later. BMS CMO Cristian Massacesi called it a classic hierarchical plan: "simply how you test PFS first, and then OS after."
Türeci also noted the redesign came after discussions with the FDA, suggesting the agency is comfortable with this approach.
Pfizer Plants Its Flag
Pfizer's oncology development chief Johanna Bendell took the opposite stance at ASCO. For the company's Symbiotic-Lung-01 trial (PF-08634404 plus chemo versus pembrolizumab plus chemo in first-line NSCLC), OS stays as a dual primary endpoint. Not secondary. Not exploratory. Primary.
This is a deliberate, slower, more expensive bet. It means longer follow-up, more events, and a later regulatory submission. So why make it?
Because if Pfizer proves its bispecific helps patients live longer than Keytruda does, the conversation is over. Payers open their wallets. Oncologists change their practice. Health-technology assessors in Europe (who are famously skeptical of PFS-only data) take it seriously. A survival win is the difference between "promising new option" and "the next standard of care."
Pfizer is essentially saying: we don't just want to slow down tumors. We want to prove we keep people alive longer. OS or bust.
A Cautionary Tale From China
Analysts watching this split aren't doing so in a vacuum. They have a recent, uncomfortable data point to consider.
Ivonescimab, a PD-1/VEGF bispecific from Akeso, took the PFS-primary route in its Harmoni trials. In Harmoni-6, the strategy looks promising against a PD-1 comparator. But in Harmoni-3, the drug missed its interim PFS analysis in squamous NSCLC. The PFS-first playbook didn't guarantee a quicker win; it exposed the program to an early, very public stumble.
That's the risk BioNTech/BMS are taking. A PFS miss at interim would be devastating, and they wouldn't have an OS co-primary to fall back on.
Early Signs From the Clinic
The Phase 2 portion of ROSETTA Lung-02 offered some encouraging signals for pumitamig. Among 40 evaluable first-line NSCLC patients, confirmed response rates ranged from roughly 57% to 68% depending on histology, with a disease control rate of 100% across the cohort. The selected Phase 3 dose is 1500 mg given every three weeks alongside chemotherapy.
Pfizer shared updated Phase 2 data for PF-08634404 as well, though detailed efficacy numbers are less mature. Investors see more class-level validation than molecule-specific proof at this stage, which partly explains why Pfizer feels the need to set a higher evidentiary bar.
Why This Matters Beyond Two Trials
This isn't just a BioNTech vs. Pfizer story. It's a question that will define how the entire next generation of cancer drugs gets developed.
The FDA has increasingly accepted PFS as a primary endpoint in confirmatory oncology trials. Among 16 novel cancer drugs approved in 2025, only two traditional approvals relied on OS as the primary endpoint. For bispecific antibodies specifically, the agency required PFS-primary with OS-secondary for linvoseltamab's confirmatory trial in multiple myeloma. The regulatory wind is blowing toward PFS.
But oncologists and payers still treat OS as the ultimate currency. In first-line metastatic NSCLC, where Keytruda already delivers a proven survival benefit, showing up with "we slowed progression" may not be enough to change behavior. You might need to prove you actually extended lives.
Two Bets, One Winner?
The strategic calculus boils down to a simple question: is it better to be first, or to be definitive?
BioNTech/BMS are optimizing for speed. In a class where at least half a dozen bispecifics are chasing Keytruda's crown (including AstraZeneca's volrustomig and Johnson & Johnson's already-approved amivantamab in EGFR-mutant disease), getting through the FDA's door first has enormous commercial value.
Pfizer is optimizing for conviction. If PF-08634404 can show OS superiority over the world's best-selling cancer drug, it becomes very hard for PFS-only competitors to claim equivalence. That kind of data doesn't just win market share; it redefines the standard of care.
Of course, there's a third possibility: neither drug beats Keytruda convincingly, and the whole PD-(L)1/VEGF bispecific thesis needs a rethink. The next few years of data will tell.
For now, we're watching two of the world's biggest pharma companies play the same game with different rules. One wants the early lead. The other wants the knockout punch. Grab your popcorn.
Clinical & Regulatory
4 min read
The First Pill That Stops COVID Before It Starts
The FDA just approved the first oral antiviral that prevents COVID after exposure, not just treats it. Shionogi's Xocova cut symptomatic infection by 67% in a pivotal trial, filling a gap that's been wide open since Omicron killed off the antibody options.
