Capricor's Duchenne Cell Therapy Voted Down — What It Means for the Field

Capricor's Duchenne cell therapy just got voted down by an FDA advisory committee — and the stock had already lost as much as sixty percent before a single hand went up in that room.

Replimune is off nearly forty percent on its own reviewer documents. Biogen beat the quarter and cut EPS guidance in the same release. Vertex bought into T-cell engagers for twenty-eight million dollars. Busy Wednesday. Let's get into it.

Welcome to The Pharma Closeout for Wednesday, July 29th, 2026. I'm Alex Mercer.

And I'm Maya Patel.

Capricor Therapeutics walked into an advisory committee this morning with deramiocel — its cell therapy for the cardiomyopathy that comes with Duchenne muscular dystrophy. The panel recommended against approval, citing unfavorable benefit-risk and inadequate efficacy evidence. Shares fell somewhere between forty and sixty percent across yesterday and today. That's the number that tells you what happened here.

Because the vote wasn't where this was decided. CBER's briefing document landed first, and by the reporting it was a comprehensively negative assessment of efficacy. That's the moment the application broke. And here's the distinction people outside regulatory affairs consistently underweight — a split panel is survivable. A review division that has already put in writing that the efficacy evidence doesn't support the benefit-risk is a different category of problem. The meeting stops being the argument and becomes the confirmation.

And the setup is what makes it sting. Capricor planted its flag on the cardiac side of Duchenne. Genuinely underserved corner of the disease — that's why the patient community was behind this program going in. STAT's coverage from the meeting has the committee calling the data "fragile." That word will follow this program everywhere it goes now.

Now notice what is not in the panel's reasoning. Nobody raised a safety signal.

Which is unusual.

It's the whole story clinically. You can restrict your way out of a safety problem — narrower population, a monitoring requirement, a warning. You cannot restrict your way out of an efficacy problem. There's no label narrow enough to fix evidence the reviewers have already called fragile.

So my read is the agency's posture in Duchenne is tightening. The flexibility this field has enjoyed on accelerated approval—

I'd push back on that. This isn't a policy story wearing a drug's clothes. If regulatory appetite were the objection, you'd see it in the benefit-risk discussion and nowhere else. It showed up inside the efficacy assessment itself. That's an evidence-architecture failure, and it fails in a permissive year too.

That's a fair correction. I'll take it — and it sharpens the read rather than softening it. If the flexibility is still there but the evidence bar underneath it never moved, then every sponsor modeling regulatory sympathy as a substitute for a comparator arm just got repriced. So here's the competitive read. If you're running strategy on a cell or gene therapy program in rare neuromuscular disease, your risk date is no longer the advisory committee meeting. It's the briefing document release — roughly a full trading day earlier than your model assumes.

And the agency isn't bound by the vote. This is not a therapeutic area where the FDA has always followed its panels. So the open question is narrow: does Capricor come out of this with a confirmatory design CBER will actually accept, or does this become a financing conversation before it gets to be a science conversation again? Whichever one they address first tells you what they think their odds are.

Shifting to the deal side, and something considerably cheaper. Vertex is paying AbCellera twenty-eight million dollars up front for next-generation T-cell engagers in autoimmune disease. Unspecified milestones and sales royalties on top. AbCellera leads discovery and early development, Vertex funds all research and development costs and holds commercialization rights to anything that emerges.

Twenty-eight million buys a seat, not a lead. Look at the comparable spending — UCB paid two-point-two billion to acquire Candid Therapeutics in May, and Merck, Sanofi and Gilead have all cut deals in this space. Vertex is arriving late and cheap.

Which, coming from you, isn't a criticism.

Late and cheap is the correct way to arrive late. And the reason they keep writing these checks hasn't changed. Cystic fibrosis is over ten billion a year and they haven't replicated it anywhere. Alpine Immune Sciences two years ago — that asset has since succeeded in late-stage testing. Then the ten-billion-dollar Crinetics buyout for endocrine disease. One problem, three instruments. There's no platform claim to test yet. This is discovery money.

Staying with the numbers, because Biogen's quarter was a genuine beat. Two-point-seven billion in revenue, up three percent, against consensus of two and a half. Adjusted EPS three sixty versus two eighty-eight expected, per Jefferies. Spinraza four hundred two million, up seven percent sequentially. Ocrevus royalties three hundred eighty-one million — about twenty-one million ahead. MS franchise still delivered nine hundred sixty-three million under copycat pressure. Skyclarys, from the Reata deal, one hundred sixty-eight million against one fifty-seven consensus. Shares up almost five percent, just over two fifteen.

And in the same release they cut EPS guidance to twelve-to-thirteen dollars, down from fifteen eighty-five to sixteen eighty-five. That's Apellis dilution and research costs — not demand. The two Apellis drugs contributed thirty million and ninety-seven million, both ahead of estimates. Leqembi grew fifteen percent year over year to a hundred eighty-four million globally, and the FDA has now cleared patients to start at home on the under-the-skin form. RBC's read is that the launch comes off a solid base. So the honest framing: the base business bought Biogen time, and the Apellis deal has already spent some of it.

Same earnings roundup, two more names worth pulling out. AstraZeneca — Pascal Soriot spent Monday's call defending eighty billion dollars in sales by 2030, and he was careful to say risk-adjusted. That's after eplontersen fell short in a big study in a deadly heart condition and the shares came off more than ten percent. Then GSK, where oncology expansion is now the centerpiece of the accelerate growth plan — at a company built on HIV, respiratory and vaccines. And the consequence there is uncomfortable. GSK is asking Wall Street to underwrite billion-dollar revenue ambitions on a franchise it's still assembling, which raises the cost of every miss and puts every cost line inside that plan under a much brighter light.

On the regulatory side — the exact mechanism that broke Capricor is already running at Replimune, one day ahead of it. Shares fell nearly forty percent after briefing documents published ahead of Thursday's advisory committee. FDA reviewers wrote that RP1's effect on survival wasn't "interpretable," and that with no control group in the key study you can't isolate the drug's contribution to tumor response. Leerink's Daina Graybosch wrote that a negative vote is the most likely outcome.

And this is attempt number three. RP1 has been rejected by U.S. regulators twice. Replimune went back in late May after what the company called productive talks with the agency and — reportedly — White House intervention.

That's the part worth sitting with. The political route got the meeting on the calendar. It didn't touch the evidentiary problem, because the problem is the trial design. No amount of access re-randomizes a single-arm study.

Two blowups inside forty-eight hours, both triggered by documents rather than votes.

Elsewhere in that same roundup — ProMIS Neuroscience traded up as much as forty-seven percent Tuesday on a six-month interim look at 136 participants in a Phase 1 in early Alzheimer's, with no cases of ARIA-E. Twelve-month data comes in the first quarter of 2027. And Atea reported its two-drug regimen — bemnifosbuvir with ruzasvir — was statistically non-inferior to Epclusa. Ninety-three point nine percent of patients with no detectable virus at six months, versus ninety-four point eight. Per the reporting, that's the first success for any company in a head-to-head hepatitis C trial, and Evercore's Jonathan Miller called the results strong even in challenging modern populations. Atea shares fell about six percent anyway.

A head-to-head win and the stock goes down. The market isn't arguing with the data — it's arguing with the commercial slot.

Which is a separate fight, and a winnable one. The Capricor and Replimune fight isn't. So here's what to carry out of today: if your filing rests on a single-arm dataset, the decision point has moved upstream. These applications are being resolved in the briefing document now. Model your risk to the document release date, not the meeting date — anyone still carrying it on the vote is a full trading day behind the market.

Looking ahead — Thursday is the Replimune panel, and after this week nobody in that room should be surprised. Also worth tracking: Altimmune's pemvidutide at 2.4 milligrams cut heavy drinking days by 4.2 per week at six months, against 2.8 on placebo. William Blair called the clinical benefit unequivocal, and said in the same note that Altimmune will require additional capital to fund late-stage work. Lilly is already running two late-stage studies of brenipatide in the same indication. So the read is that cardiometabolic mechanisms are moving into addiction — and the balance sheet gets there first, regardless of who had the better six-month curve.

What I want out of Thursday isn't the vote count. It's whether the committee's objection is the design or the drug. Those are two completely different answers, and only one of them is survivable for every single-arm oncology filing sitting in the queue behind it.

And that is your Pharma Closeout for Wednesday, July 29th — Capricor voted down on deramiocel, Replimune off nearly forty percent on its own reviewer documents, Biogen beating and cutting in the same breath, GSK betting the growth plan on oncology, and Vertex buying into T-cell engagers for twenty-eight million. Two applications broke this week before anyone voted. That's the story of the day and probably the quarter. Follow wherever you listen and tomorrow's briefing lands on its own.

It is. Have a good evening — I'll be back in the Replimune briefing documents tonight, because on this week's evidence, that's where the decision has already been made.