Disc's sunlight-pain pill passes Phase 3; company launches $350M stock sale

Disc Medicine is a clinical-stage drug developer, meaning it has no approved products and no product sales. Its work centers on how the body builds heme, a part of red blood cells, and how it handles iron.¹ It has three drugs in testing: bitopertin for EPP, selcodebart for anemia in the bone-marrow cancer myelofibrosis, and DISC-3405 for polycythemia vera, or PV.²

Disc says liver damage strikes 20-30% of EPP patients, there is no cure, and the only approved therapy is Scenesse, a hormone implanted under the skin to boost melanin. Disc's annual report counts about 14,000 diagnosed patients in the United States, and its new slides put the American EPP market at more than $2 billion, a company estimate.

Disc licensed bitopertin from Roche in May 2021, and Roche had already tested it in more than 4,000 people for other uses. Disc filed for accelerated approval in September 2025. The FDA accepted the file in November 2025, then sent a rejection, called a complete response letter, in February 2026.¹ The reason was that the earlier trials did not show that a drop in PPIX, the pigment that builds up in EPP, went with more time in sunlight. Before the rejection, Reuters reported that reviewers had delayed the review two weeks over efficacy and abuse concerns. A law firm seeking clients put the stock's fall at 21.9% on February 13, 2026; the 10-Q reports no material legal proceedings in the first half of 2026.²

At a meeting in the second quarter of 2026, the FDA said APOLLO, if successful, could be the basis for Disc's response and could potentially support a traditional approval. APOLLO is a randomized, double-blind Phase 3 trial against placebo in patients 12 and older.¹ It enrolled 183 patients in the United States, Canada, Europe and Australia, split almost evenly between 60 mg of bitopertin once a day and placebo for six months.²

The stock slid before the APOLLO news, from $69.49 on September 18 to $55.72 on October 14. A Wall Street Journal market page showed a one-year range of $40.00 to $99.50.

Fidelity's parent, FMR, reported a passive 12.8% stake as of June 30, 2026, and insiders and major holders together owned about 31% of the shares.¹ Insiders were selling before the readout too: director Kevin Bitterman exercised options for 27,136 shares on September 11, 2026 and sold every one of them at $70.11 that same day, taking that holding line to zero.² His earlier Form 4, which reported his last sales on September 3, still showed 646,212 shares held across three holding lines after those sales.³ CEO John Quisel's August 17 notice of sale covered 33,000 shares worth $2,601,720 and listed earlier sales under a Rule 10b5-1 plan, a trading schedule set in advance, adopted in March 2026: 33,000 shares in July for $2,473,779 and 34,000 in June for $2,387,296, or about 100,000 shares across June, July and August.⁴

Disc's last big share sale, in October 2025, was at $84.00 a share, with some existing investors buying.¹ Insiders sold before the readout: CEO John Quisel exercised 33,000 options and sold the same 33,000 shares at $78.11 to $79.84 on August 17, director Kevin Bitterman sold 54,496 shares at $79.95 to $80.98 over the first three days of September and on September 11 exercised options for 27,136 shares and sold every one at $70.11, and four Atlas Venture funds sold 125,949 shares for about $10.1 million gross over the same three days, with none of those filings saying whether the sales ran through Rule 10b5-1 plans.² Lots of 2,810 and 2,339 shares in Bitterman's September 3 filing match Atlas fund sales that day, so his sales and the funds' may be the same shares rather than extra ones.³ All of these sales came before the stock slid to $55.72 on October 14. Quisel's August prices and Bitterman's early-September prices, about $78 to $81 a share, were close to the $78.93 close after the news.⁴

Both main goals were met. Whole-blood PPIX fell 41.4% more than on placebo, with a p-value under 0.0001, which means chance is a very unlikely explanation.¹ Average monthly time in light without pain during month six rose 12.2 hours more than on placebo, with a p-value of 0.0010. At the start, patients logged 11.4 hours on the bitopertin side and 10.7 hours on the placebo side over a 14-day screening period.²

Every secondary goal was met too. Painful light reactions came 71% less often than on placebo over six months, and total pain-free time over the six months rose, though the slides give no hours for that gain.¹ Time in sunlight before the first warning symptoms rose 56 minutes against placebo, but that goal was measured in adults only.²

Side effects are the main question. Dizziness hit 50% on bitopertin against 11% on placebo, nausea 20% against 11%, and fatigue 17% against 7%.¹ Disc says the dizziness was all mild or moderate and generally short-lived, and that no side effect was worse than Grade 3. Grade 3 events, the most severe level seen, hit 7% on the drug against 3% on placebo, and serious events numbered 1 on the drug against 2 on placebo. No one on bitopertin left the study for any reason.²

Disc plans to resubmit its application by the end of 2026 and expects an FDA decision by mid-2027. One slide reads "Targeting launch Nov." with no year beside it, which does not fit a mid-2027 decision, so we treat it as unclear. Disc will show fuller data at a future medical meeting.

At 16:01 Eastern on October 19, after the close, Disc launched a $350 million sale of new shares and pre-funded warrants. A pre-funded warrant is a share bought almost in full up front, with $0.0001 left to pay, and buyers can set caps such as 4.99% or 9.99% on how much stock it lets them own.¹ The price, share count and fees are blank in the preliminary prospectus. Underwriters led by Jefferies, Leerink Partners, Morgan Stanley and Cantor get a 30-day option to buy $52.5 million more.²

What Disc gets is money to prepare a launch: the release says the cash will help support the potential commercialization of bitopertin. Disc assumes about $328 million of net proceeds and says that, with its existing cash, it can then fund operations and debt payments into 2030, against into 2029 without the sale.¹ Neither runway counts any money from bitopertin sales.²

What holders give is ownership. At the $78.93 close, $350 million would buy about 4.4 million shares, about 12% of the 38.4 million shares outstanding in July, before the underwriters' extra option; the real price is not set yet.¹ The company, its directors and its officers agreed to a 45-day lock-up, but insiders may still sell under trading plans set up before it. Disc also has $189.9 million left on an at-the-market program, which lets it sell stock through Cantor over time.²

Disc burns real cash. In the first half of 2026 it used $106.4 million in operations, about $53 million a quarter, and posted a net loss of $123.0 million that includes $25.9 million of stock pay, a non-cash cost.¹ In the June quarter alone it lost $59.5 million, with research at $46.9 million and selling and overhead at $18.1 million. Disc expects selling and overhead costs to rise substantially as it prepares to sell a drug.² Cash and securities fell from $717.7 million at June 30 to about $682.8 million at September 30, a drop of $34.9 million. That later figure is unaudited, and the filings do not split it into spending and other moves.³

Debt is small next to the cash. Disc owes $60.0 million to Hercules Capital at the greater of 8.25% or the prime rate plus 1.75%, pays interest only through November 2028, and must repay by December 1, 2029, plus a $4.1 million end-of-term fee.¹ June-quarter loan interest was $0.9 million, against $6.5 million of interest income. The lender holds a first lien on everything except intellectual property, and the loan lets it act on its own judgment of a material adverse change.²

A minimum-cash test starts July 1, 2028, or July 1, 2029 if a milestone is met, once the loan tops $50.0 million, and the test is waived while market value is at least $1.0 billion. Disc can also draw another $50.0 million at its option until April 30, 2027.¹ Net of the loan, Disc had about $657.7 million of net cash at June 30. With no earnings, ratios like debt to earnings do not apply.²

Counting options, restricted stock units and pre-funded warrants, the share count at June 30 was about 43.8 million fully diluted, against 38.3 million common shares. The options carry an average exercise price of $44.95, well below the current price.¹ Disc's shelf prospectus lists all 10,000,000 authorized preferred shares as undesignated, and the board can issue preferred stock without a holder vote. The secured lender is paid ahead of common holders.²

At $78.93, the market value is about $3.03 billion. Less the June 30 net cash, investors pay roughly $2.37 billion for the drugs, close to Disc's own estimate of a more than $2 billion American EPP market.¹ Approval has costs too: Disc owes Roche a $15.0 million milestone on American approval and royalties from the high single digits to the high teens on sales.²

The bull case: APOLLO met every goal in a trial the FDA agreed could support the response, and Roche had already tested the drug in more than 4,000 people. No one on bitopertin quit the study, and 97% of patients rolled into the long-term extension.¹ EPP has no cure and one implanted rival, and orphan drug status brings seven years of American market exclusivity if approved. With the sale, Disc says its cash lasts into 2030 without any bitopertin revenue.²

The bear case, strongest point first. The FDA has said no once, and approval still rests on its review of a drug that made half of patients dizzy.¹ The time-in-light result, at a p-value of 0.0010, is weaker than the PPIX result; one secondary goal was tested in adults only, and the six-month total gain has no number yet. Roche's main compound patent on bitopertin expires in 2026, so protection leans on Disc's own patents, expected to expire between 2041 and 2045, and on orphan exclusivity.² And the $350 million sale dilutes holders by about 12% while selling costs are set to rise.³

Disc's own goals, which are plans and not promises: refile by the end of 2026, an FDA decision by mid-2027, and fuller APOLLO data at a medical meeting. By the end of 2026 it also expects more polycythemia vera data and first sickle cell data for DISC-3405, plus more myelofibrosis data and Phase 3 plans for selcodebart.

What would prove the bull case wrong: a refiling that slips past the end of 2026, detailed data showing dizziness that lasts, or a second FDA rejection by mid-2027. The first test is the price of the share sale, the next is the refiling date, and the last is the decision.¹ Our dated call, below, asks whether IRON holds above a level set between the $56.94 close before the news and the $78.93 close after it, a month after the planned refiling.²

The price, the number of shares, the fees, the net proceeds and the dilution per share for the $350 million sale are all blank in the preliminary prospectus, so the exact dilution is not yet known.

The slides say total pain-free time over the six months rose against placebo but give only the p-value, not the hours.

No filing says whether the FDA will ask for abuse-risk studies or a controlled-drug label for bitopertin.

Disc has given no list price. A slide reading "Targeting launch Nov." has no clear year and sits ahead of a decision expected by mid-2027.

Disc calls the dizziness generally short-lived but gives no figure for how long it lasted.

Disc has given only an unaudited cash figure for September 30, so the third quarter's operating burn and loss wait for its next quarterly report.