

Bristol Myers Squibb just ended an eleven-year bispecific antibody partnership with Zymeworks, torching $313 million in potential milestones. But the real story is what BMS's ruthless portfolio purge tells us about where big pharma is placing its bets.
When your partner of eleven years says they need to "prioritize" other things, you know the breakup is coming. You just didn't think it'd show up in an earnings report.
Bristol Myers Squibb quietly ended its bispecific antibody collaboration with Zymeworks in April 2026, pulling the plug on a relationship that traces back to 2014. The program, a Phase 1 cancer drug targeting two proteins at once, got dropped from BMS's pipeline as part of what the company calls routine portfolio prioritization.
The question isn't really why BMS walked away. It's what this tells us about the brutal calculus big pharma is running right now.
To understand this breakup, you need to know the backstory. Zymeworks didn't sign up with BMS. It signed up with Celgene, way back in 2014, to co-develop up to eight bispecific antibodies using Zymeworks' proprietary Azymetric platform. Think of bispecific antibodies like a Swiss Army knife for the immune system: instead of grabbing one target on a cancer cell, they grab two at the same time.
Celgene paid $8 million upfront, with each candidate eligible for up to $164 million in milestone payments. In 2018, the deal expanded to include two more candidates and a two-year extension, costing Celgene an additional $4 million. Then came BMS's $74 billion acquisition of Celgene in 2019, and suddenly this was BMS's problem to manage.
Of the original ten possible candidates, only one survived to reach the clinic: a bispecific antibody targeting CD40 and fibroblast activation protein (FAP) in solid tumors. It made it to Phase 1. And then BMS decided it wasn't worth finishing.
BMS didn't single out Zymeworks for punishment. The company is in the middle of a sweeping, sometimes painful, corporate makeover.
Here's the context: BMS is staring down a patent cliff. Blockbusters like Revlimid, Eliquis, Opdivo, and others are either losing exclusivity or facing generic erosion. The company's strategy is to shift more than half its revenue to a "Growth Portfolio" of newer drugs by 2026. That requires focus, and focus means cutting.

LEO Pharma just paid $50 million for a company that turns herpes simplex virus into a gene therapy delivery system for devastating skin diseases. It's preclinical, it's audacious, and it might be the future of dermatology.


Join thousands of biotech professionals who start their day with our free, daily briefing.
BMS launched a restructuring program in early 2025 targeting $2 billion in annual cost savings by 2027. So far, the results have been blunt. Significant layoffs in New Jersey have been announced through WARN filings since early 2025, with additional cuts scheduled through 2026. Pipeline programs that don't scream "first-in-class" or "best-in-class" are getting the axe.
A Phase 1 bispecific inherited from a pre-acquisition deal? That's the definition of expendable.
Losing potential milestones sounds devastating. But potential milestones are a bit like lottery tickets; they only pay out if everything goes perfectly. In practice, Zymeworks' business stretches well beyond this single BMS program.
The company pulled in $69.6 million in milestone payments in 2025 from a roster of partners that includes GSK, Johnson & Johnson, Daiichi Sankyo, and BeOne Medicines. Its Azymetric platform (the technology that makes these dual-targeting antibodies possible) is validated by a real, approved product: zanidatamab (Ziihera), a HER2-targeting bispecific that's on the market.
Zymeworks also has a deep bench of next-generation programs in the works. Two new bispecifics, ZW209 and ZW1528, have investigational new drug applications on track for 2026. A couple of antibody-drug conjugates (ZW191 and ZW251) are in Phase 1. And the company presented data on trispecific T cell engagers and a novel RAS-inhibitor ADC platform at AACR in April 2026.
In other words, Zymeworks has options. Losing the BMS deal stings, but it doesn't blow a hole in the hull.
The irony of BMS dropping a bispecific program is that BMS is simultaneously betting billions on bispecifics elsewhere. In June 2025, BMS struck an $11.1 billion deal with BioNTech for a PD-L1/VEGF-A bispecific antibody targeting solid tumors. That's the same modality, different molecule, wildly different price tag.
And BMS isn't alone. The bispecific antibody space has seen enormous deal activity. Takeda paid up to $11.4 billion for Innovent's PD-1/IL-2 bispecific. Genmab acquired Merus for $8 billion, largely for a late-stage bispecific with Breakthrough Therapy Designation. Eli Lilly signed a $2.6 billion platform deal with ABL Bio.
The pattern is clear: big pharma is hungry for bispecifics, but only the ones with strong clinical data or late-stage momentum. Early-stage, inherited programs from old deals? Those are getting left behind.
With the BMS collaboration officially over, Zymeworks gets full rights to the CD40/FAP program back. The company could partner it with a new pharma collaborator, develop it internally, or shelve it. Given the competitive landscape in bispecifics and the early stage of the program, finding a new suitor won't be automatic.
But Zymeworks' real value isn't in any single molecule. It's in the Azymetric platform itself, which converts standard antibodies into multispecific formats by tweaking amino acids in their structure. It's like having a universal adapter that plugs into multiple outlet types. That versatility, proven by zanidatamab's approval and a growing clinical pipeline, is what keeps big pharma coming back.
The BMS chapter is closed. For Zymeworks, the story is far from over.
Indivior's experimental opioid addiction drug flopped in Phase 2, and now the company is gutting its research team. Wall Street didn't flinch, but the long-term implications of becoming a one-drug company are harder to ignore.