Replimune RP1 on the Brink: FDA Calls Melanoma Data Uninterpretable

FDA reviewers just told an advisory committee that Replimune's pivotal melanoma data cannot be interpreted. The panel votes Thursday.

Second day running that the agency's own scientists have taken an accelerated approval package apart in public. Meanwhile Gossamer Bio came back from a Phase 3 failure with a filing path and its drug rights back, and IQVIA just beat. Let's get into it.

Welcome to The Pharma Closeout for Tuesday, July 28th. I'm Alex Mercer.

And I'm Maya Patel.

Replimune shares fell sharply in early trading this morning. And the trigger wasn't a data readout — it was a document. FDA briefing materials for Thursday's Cellular, Tissue, and Gene Therapies panel posted today, and the reviewers wrote that the overall survival analysis from the single-arm IGNYTE study is, in their words, not interpretable.

We've been tracking this a month, and the escalation is the story. This isn't the agency asking for more data. RP1 has already faced multiple rejections. The resubmission was accepted with a committed decision timeline — and then the FDA scheduled an advisory committee anyway.

That's the tell.

You don't convene a panel on a package you're preparing to approve. You convene one to build a record.

Here's the ask. Accelerated approval for vusolimogene oderparepvec — RP1 — with Bristol Myers Squibb's Opdivo, in unresectable advanced cutaneous melanoma after progression on PD-1 therapy. Reviewers said the response assessment used in IGNYTE confounds interpretation of the reported efficacy and limits their ability to verify results. Then the line that actually does the damage: without a reliable historical control, it cannot be determined whether RP1 contributes to any observed effect at all.

Sit with that one. It's not "the effect is modest." It's "we can't attribute the effect to your drug." In a combination built on an anti-PD-1 backbone, that's existential, not procedural.

Replimune's defense is ethical rather than statistical. The company wrote it isn't feasible or ethical to randomize this population to anti-PD-1 monotherapy that offered no hope of response. So the comparator arm was never run.

I'd push back on treating that as a complete defense. The argument isn't wrong on its face — nobody wants to randomize progressed patients back onto a therapy they've already failed. But that's an objection to one comparator, not to randomization. Nivolumab plus investigator's choice. A delayed-start design. An external control agreed with the agency prospectively. Those all existed. The company is presenting a design constraint as a design impossibility, and the reviewers aren't accepting the substitution.

That's a fair correction. I'll take it. So the sharper version is that this isn't really an argument about whether the drug works — it's an argument about whether a sponsor can retroactively earn interpretability from a trial that was never built to produce it.

That's the frame. And here's the detail I'd flag for anyone reading the briefing package tonight. The criticism lands on the response assessment methodology, not just the magnitude. When reviewers say the assessment confounds interpretation, they're questioning how responses were adjudicated — which contaminates the objective response data and the survival analysis at the same time. You can argue a small effect size at a panel. You cannot re-adjudicate scans that were read under a method the agency has already rejected.

BMO Capital Markets told clients this morning the documents read negatively, and that Replimune likely faces an uphill battle Thursday — the FDA appears aligned with its prior assessment from the previously published complete response letters. Same signal, priced immediately.

Which is the part every accelerated approval sponsor should be uncomfortable about. The agency didn't produce a new objection. It restated the old one, publicly, in front of a panel. That's not a negotiation posture.

So the competitive read. If you're running commercial strategy on any single-arm accelerated approval asset in oncology, the takeaway is that the agency will now litigate your response-assessment methodology in public briefing documents before the vote. Which means pre-submission alignment on adjudication just moved from regulatory housekeeping to launch-timeline risk.

Thursday is the open question. A panel can still vote against staff analysis — we watched the peptide committee do something close to that last week. But the verdict is committed by August 2nd, and the agency has already put its position in writing. What's genuinely unresolved is whether leadership overrides its own reviewers. Committees advise. Leadership decides. That decision hasn't been made.

Shifting to the deal side, and to a company that got the opposite week. Gossamer Bio — we last covered seralutinib in early May. Yesterday, July 27th, they announced they've reacquired worldwide development and commercial rights from Chiesi, and they're filing an NDA in pulmonary arterial hypertension in September. Shares surged in Tuesday's pre-market session, with the stock still trading at a relatively small market capitalization.

The regulatory language in that release is doing all the work. Per the company, FDA characterized the degree of statistical significance and the magnitude of treatment effect in PROSERA as review issues rather than filing issues. That's a narrow distinction and it's worth being precise about — the agency is saying you may file. Not that you'll be approved. The submission rests on PROSERA plus confirmatory evidence from Phase 2 TORREY.

The structure is the interesting part. Gossamer paid no upfront cash to take the rights back. Chiesi makes a one-time five million dollar payment settling prior obligations, and keeps a capped royalty on worldwide net sales plus regulatory and commercial milestones. The fifty-fifty US profit share dissolves. Separately, stockholders approved the convertible note exchange — per the company that cut aggregate principal debt by roughly one hundred fifteen point nine million — and authorized a reverse split. Cash was approximately fifty-seven million as of June 30th.

Two parties looked at the same asset and priced it differently. Gossamer consolidating global economics right before a filing is what you do when you believe the asset survives review. Chiesi accepting a capped royalty instead of half the US market is the counter-signal. Somebody's probability estimate is wrong.

If it's accepted for filing, a decision would still be a considerable way off. So for anyone modeling PAH: a program that read out as a Phase 3 disappointment just became a live 2027 entrant, fully owned, with a cleaner balance sheet behind it.

And on a fifty-seven million dollar cash position going into a filing year — the financing question doesn't disappear because the rights came home.

No, it gets sharper. Owning worldwide economics is only an advantage if you can fund the commercial build that collects them.

On the regulatory side, I want to stay with the sequencing, because Replimune isn't isolated. BioSpace reported that the rebuff came a day after FDA scientists similarly questioned the Phase 3 trial supporting Capricor's Duchenne cell therapy — and BioPharma Dive carried the same item in its roundup. That's the agency's review staff undercutting approval packages on consecutive days, in both cases before the committee ever sits down.

And it lands against something BioPharma Dive is tracking separately — five FDA decisions to watch in the third quarter, framed explicitly around how flexible the agency is willing to be under new leadership. That question is getting answered in real time this week, and not in the direction the market assumed.

Here's the precedent layer, though. The peptide panel last week narrowly recommended broader use of six of the seven peptides under consideration, despite what BioPharma Dive described as a lack of substantive evidence. So we have a committee going soft on a thin evidence base, and staff going hard at two accelerated approval packages inside forty-eight hours. Those can only both be true if flexibility at this agency is being allocated by something other than the evidence.

That's the uncomfortable version.

It's the version sponsors have to plan around. If your regulatory strategy assumes this FDA is uniformly more permissive, this week says the permissiveness is selective — and nobody gets to choose which side of that line their program lands on.

Looking ahead — Thursday, July 30th, the Cellular, Tissue, and Gene Therapies panel votes on RP1, with the FDA verdict committed by August 2nd. Three days from vote to decision. That's compressed even by adcomm standards.

Watch the voting questions, not just the vote. When the agency has already written that a dataset is not interpretable, how it frames the question tells you whether it wants cover for a rejection or a route to a narrow label. Those are very different documents.

Also today, IQVIA reported second quarter results that sparked a significant beat, with shares climbing strongly in pre-market trading and the company maintaining its position as a large-cap industry player. When the biggest clinical research services provider gets rewarded on earnings, that's a read on trial spend across the whole industry, and it's worth more as a sentiment indicator than most single-company prints.

Which is today's tension in one sentence. Sponsors are having a brutal week at the agency, and the company that runs their trials just beat.

And that is your Pharma Closeout for Tuesday, July 28th — Replimune walking into Thursday's panel with the agency's position already on paper, Gossamer pulling seralutinib back from Chiesi ahead of a September filing, and IQVIA printing a beat into all of it. Thursday is the one that matters. Follow wherever you listen and tomorrow's briefing lands on its own.

I'll be reading the CTGTAC voting questions the minute they post — that document will tell you more about Thursday's outcome than anything either side says between now and then. Have a good evening.