Jazz’s HER2 Bispecific Wins First-Line Gastric & McKesson Pays $2.25B for Precision Medicine

The FDA put two Ziihera regimens on the same label today. One of them has never shown that it helps anyone live longer.

And the analysis that settles that lands within weeks. Also today — McKesson pays two point two five billion dollars for a cancer CRO, and the investors behind Metsera resurface with a one and a half billion dollar Haisco deal. Busy Tuesday. Let's get into it.

Welcome to The Pharma Closeout for Tuesday, August twenty-fifth. I'm Alex Mercer.

And I'm Maya Patel. Three stories today, and two of them are really about who owns what happens after the molecule works.

Jazz and Zymeworks got Ziihera approved in first-line HER2-positive gastroesophageal adenocarcinoma today — on the August twenty-fifth action date we flagged when we covered this earlier in the month. Median overall survival on the triplet: twenty-six point four months, against nineteen point two on Herceptin plus chemo. That's a standard of care getting displaced in a single afternoon. But there are two regimens on that label, and only one of them carries a proven survival benefit.

The second one is the whole story.

It decides how big this drug gets. Ziihera has been a niche product — fifteen million dollars in biliary tract cancer last quarter. Stifel puts the newly eligible US population at a certain number of patients. Jefferies has peak sales at one and a half billion dollars in this indication alone, and Jazz raised full-year guidance before this label even existed.

Read the label instead of the analyst note. The agency cleared the triplet for IHC three-plus and two-plus ISH-positive — that's the lower-expression group — even though the trial skewed heavily toward high expressers. It covers esophageal and junction tumors when most enrolled patients had gastric primaries. The FDA gave Jazz more market than the trial population strictly supports.

That's a generous read of the evidence, then.

It's a broad one. And here's the part I'd hold onto — in PD-L1-negative patients, tumor area positivity under one, median survival was reported. The checkpoint-containing arm performed best in the patients you'd expect it to disappoint.

Wait — best in the negatives?

Biggest benefit there, per the subgroup Jazz presented at ASCO. Which is why I'd push on your franchise framing. Eighty-five percent of patients on the triplet had diarrhea, and it's a boxed warning with fatal cases in the trial. Oncologists will use this drug — but loperamide prophylaxis from cycle one is a real adherence tax in a community practice, and the regimen you'd reach for to avoid a PD-1 is exactly the one without proven survival.

I'll give you the tolerability. Where I hold the line is the economics. Zymeworks banked a two hundred fifty million dollar milestone on this approval and is in line for up to one point three billion more, plus royalties — for a company that spent last year restructuring around royalties instead of selling its own medicines. That's a cash-flow inflection, and it doesn't need the doublet. It needs the triplet to hold.

Then say what it actually is. This is a triplet approval with a doublet attached, and the end of September decides whether that attachment is a business or a line of label text nobody prescribes from.

The competitive read: if you run commercial strategy in upper GI, Herceptin plus chemo stopped being the default today — and the PD-L1 gate that protected Keytruda's position in HER2-positive stomach cancer stopped being a gate at all. Jazz's R&D head told Fierce they're now looking at early-stage gastric, where there's no approved HER2 treatment.

And the thing to put in the calendar is that second interim analysis. Jazz expects to accrue enough events by the end of September. Clean readout, the label loses its asterisk and the drug travels to every patient who can't take a checkpoint inhibitor. Miss, and Jazz spends a year selling one regimen while quietly defending the other — and every field conversation gets organized around a distinction the press release doesn't make.

Which is a deal question as much as a clinical one — and there was a stranger one today. Sentivera, the startup ARCH Venture Partners and Population Health Partners built after Pfizer took Metsera, came out of stealth with a Haisco Pharmaceutical deal worth more than one and a half billion dollars. Forty million in cash upfront. The asset is preclinical.

Preclinical. Forty million dollars, no human data.

The cash isn't the story. The structure is — and it's unusual enough that I want a second read on it. Let's bring in Marcus Webb on this one.

So Haisco is a shareholder, not a seller.

Which is the easy part to underwrite. The target is harder. Type 2 inflammation is Dupixent's house, and a small molecule walking in has to beat an injectable that already controls disease well. Up to one point four six billion dollars in milestones against a molecule with no human data tells you exactly where the risk got parked — Haisco gets paid if the clinic cooperates, and not one day before.

And that's the read-through for anyone tracking these deals. Haisco now ranks among the most prolific China-based partners since the start of 2025, per BioPharma Dive's tracker — and it got there taking ownership in its partners rather than selling assets outright. If you're a US biotech preparing for a China licensing negotiation, the counterparty across the table wants a piece of you now, not a royalty check later.

On the regulatory side, the sharper question is what Jazz still owes the agency. Ziihera's original label — previously treated biliary tract cancer — came through accelerated approval, and the prescribing information still says continued approval may be contingent on verifying benefit in a confirmatory trial. Jazz is running a front-line biliary study built to do exactly that.

So today's win doesn't retire yesterday's obligation.

It doesn't touch it. And the next ask is already visible — breakthrough therapy designation in previously treated HER2-positive colorectal cancer, with Jazz saying it'll keep talking to the FDA about a path there. Breakthrough buys you meetings. It doesn't buy you an endpoint.

The basket trial is the real pipeline, then.

That's where the argument actually gets made. And it reframes what today was worth — Ziihera's value isn't the stomach cancer label, it's whether a HER2 bispecific becomes something close to a tumor-agnostic class. Every confirmatory obligation Jazz clears makes that cheaper to argue. Every one it stumbles on makes the accelerated-approval conversation louder — for Jazz, and for everyone else holding a contingent label right now.

Before we look ahead — McKesson paid two point two five billion dollars today for Precision Medicine Group, the Maryland CRO, with Blackstone selling. Precision folds into the oncology and multispecialty segment McKesson built in last September's restructuring, when it said it was pivoting toward higher-margin biopharma and cancer work.

Look at what's inside it. Biomarker labs, complex trial design, market access consulting, commercialization support. Precision claims on its own site it helped launch fifty-six percent of the drugs the FDA approved in 2024. That's not a services bolt-on — that's a distributor buying a seat at the front of the pipeline, years before the product it eventually ships exists.

The company that moves your drug now designs the trial that creates it.

And nobody has told us what that looks like after closing. It's still subject to regulatory clearances, with no timeline given. The question a sponsor starts asking in diligence is whether bundling distribution, trial execution and commercialization under one roof reads as efficiency or as concentration — and that gets answered by the antitrust review, not the press release.

Looking ahead, the date to hold is the end of September — that second interim survival analysis on the Ziihera doublet.

Two more on my list. Zymeworks' acquisition of Theravance Biopharma is still expected to close in the second half of this year, landing on top of today's milestone. And the first Ziihera breast cancer data could emerge as early as late 2027 — that's the readout that tells you whether this is a GI drug or a platform.

So if you're modeling Ziihera as a stomach cancer product, you're modeling the floor. The label breadth the FDA granted today is the option — and the next five weeks price it.

I'm reading the full prescribing information tonight instead of the press release. The diarrhea management language is where this launch actually gets decided, and label text like that tells you more about a real-world rollout than any analyst note will. Have a good evening.

And that is your Pharma Closeout for Tuesday, August twenty-fifth — Ziihera going from niche biliary drug to first-line stomach cancer standard, McKesson buying its way into clinical research, and ARCH's newest startup writing a one and a half billion dollar check to Haisco. Follow the show on Spotify, Apple Podcasts, or wherever you listen — tomorrow's briefing lands on its own.