$IRON: The Drug Roche Threw Away Just Swept Phase 3, and the FDA Clock Restarts in December

On Monday, Disc Medicine (NASDAQ: IRON) closed at $78.93, up 38.9% on 8.5 million shares, roughly fourteen times its recent daily volume. The trigger was the cleanest kind of biotech catalyst: APOLLO, the pivotal Phase 3 trial of bitopertin in erythropoietic protoporphyria (EPP), met both co-primary endpoints and every secondary endpoint. The company called it "a potential breakthrough in the science and treatment of EPP." Then, after the close, it did what well-advised biotechs do on a 38% up day: it launched a $350 million stock offering.

That combination, a clean Phase 3 sweep plus immediate dilution, is why $IRON is at the top of the trending lists this morning. The stream is treating it as a straightforward good-news story. The more interesting frame is the one almost nobody is running: eight months ago the FDA rejected this exact drug, the stock lost 22% in a day, and Monday's close is only about 10% above where IRON traded the day before that rejection. The market is pricing the Phase 3 win as if it merely repairs the February damage. Whether that is conservative or correct depends on resubmission mechanics, dilution math, and a commercial question the trial data just reframed.

The Drug Roche Gave Up On

Bitopertin has one of the stranger second acts in drug development. Roche originally developed it as a schizophrenia treatment and ran it through a large late-stage program that failed on efficacy, leaving behind a safety database the company says now spans more than 4,000 trial participants. In 2021, Disc Medicine licensed the molecule in a deal reported at roughly $200 million in total value, a castoff price, on a different thesis entirely: bitopertin inhibits GlyT1, a glycine transporter, and glycine is a rate-limiting input to heme synthesis. Throttle the pathway and you throttle production of protoporphyrin IX (PPIX), the phototoxic molecule that accumulates in EPP patients.

EPP is a rare genetic disease in which PPIX buildup makes ordinary light exposure cause burning pain, often within minutes. Patients organize their lives around avoiding daylight. There is exactly one approved therapy anywhere: Clinuvel's Scenesse (afamelanotide), a subcutaneous implant that increases skin pigmentation, approved in the US, EU, Australia, Israel, and, as of July, Canada. Clinuvel reported total revenues of A$105.3 million for its fiscal year ended June 2025, up 10%, substantially driven by Scenesse. Bitopertin, a once-daily oral pill that attacks the disease mechanism rather than tanning the skin around it, would be the first disease-modifying treatment and the first oral option. Disc's investor materials put the addressable EPP market at more than $2 billion; that is a company estimate, but the Scenesse revenue base confirms real, paying demand in a disease most investors have never heard of.

What APOLLO Actually Showed

APOLLO randomized 183 adults and adolescents (12 and older) with EPP or the related XLP 1:1 to bitopertin 60 mg daily or placebo for six months, across the US, Canada, Europe, the UK, and Australia. Per the October 19 topline release:

• Co-primary 1: whole-blood PPIX fell 41.4% versus placebo (p<0.0001) — the biomarker endpoint
• Co-primary 2: average monthly time in light without pain increased 12.2 hours versus placebo in month six (p=0.0010) — the clinical endpoint
• Secondaries: all statistically significant, including a 71% reduction in the incidence rate of phototoxic reactions (p<0.0001) and improvements on both patient global impression scales
• Safety: no Grade 3+ treatment-emergent adverse events; the notable tolerability item is dizziness in 50% of treated patients versus 11% on placebo, described as mild-to-moderate and transient

For a rare-disease trial, this is about as clean as topline gets: both the mechanism marker and the lived-experience endpoint hit, with the kind of p-values that do not invite re-analysis. The dizziness rate is the one number worth filing away for the label discussion, but with no serious adverse event signal it is unlikely to be an approval question.

Briefings like this land in members' inboxes before the market prices them in. Join AlphaBriefing free.

The February Rejection That Set This Up

Here is the context that makes Monday's move legible. Disc originally asked the FDA to approve bitopertin under accelerated approval based on Phase 2 biomarker data, PPIX reduction, before APOLLO read out. In February 2026 the agency said no. The Complete Response Letter, as described in Disc's own offering documents, questioned whether the Phase 2 PPIX data were "reasonably likely to predict clinical benefit." On February 13 the stock fell from $71.65 to $55.95, down 21.9% in a session, and it spent the next eight months grinding sideways in the mid-$50s to mid-$60s.

Read carefully, the CRL was a demand, not a verdict: show us patients doing better, not just a biomarker moving. APOLLO is the direct answer. The trial was designed with the clinical endpoint (pain-free light exposure) as a co-primary precisely because that is what the agency said was missing, and after a Type A meeting the FDA agreed a successful APOLLO could serve as the basis for the CRL response. The company now plans to resubmit the NDA by the end of 2026, with an FDA decision anticipated by mid-2027.

So the question the market is actually pricing: with the specific deficiency the FDA named now answered by a statistically unambiguous Phase 3, what is left between this drug and approval — and what do a $350 million raise, the share-count math, and the mid-2027 calendar do to the risk/reward from $79?

The rest of this briefing is for paid members: the dilution math on the $350 million offering against the actual share count, the pro forma cash position and what it signals about launch intent, the resubmission calendar mechanics, the Scenesse displacement question, and scenario-by-scenario price zones anchored to levels this stock has actually traded.

AlphaBriefing Paid gets you every investment thesis, scenario framework, and catalyst brief we publish — the analysis private intel clients pay four figures for, at a fraction of that.

Unlock the full briefing →