FSHD, a disease with no treatment, draws Epicrispr's $90M bet

Facioscapulohumeral muscular dystrophy (FSHD) has spent decades as a footnote in the rare-disease catalog. The genetic cause has been mapped since the early 1990s, the patient population is small, and no drug has ever been approved. That situation shifted in the last twelve months as four biotechs began testing four different mechanisms against the same disease. FierceBiotech's interim Phase 1/2 readout from Epicrispr is the first data point in that contest.

Epicrispr's lead asset, EPI-321, is an AAV-delivered epigenetic editor that turns down the DUX4 gene in muscle cells without rewriting DNA. The mechanism works like a dimmer switch rather than a cut-and-paste: a viral vector carries an editor that silences a toxic protein long linked to FSHD's muscle loss. The first three patients treated with a single intravenous infusion had gained 0.5 to 1.3 pounds of lean muscle mass by six months, per FierceBiotech. The number is small, the cohort is single-arm, and the endpoint is a surrogate for the function patients want back.

Epicrispr closed a $90 million Series C co-led by Octagon Capital and Janus Henderson Investors, with Sanofi Ventures participating. The company had raised $68 million roughly sixteen months earlier. Epicrispr was a 2023 Fierce 15 honoree and is based in San Francisco.

Novartis closed its $12 billion takeover of Avidity Biosciences in February 2026, picking up delpacibart braxlosiran, an antibody-oligonucleotide conjugate now in Phase 3, according to FierceBiotech. Sarepta has an siRNA candidate in Phase 1/2. Scholar Rock recently started its own Phase 2 trial. Each program attacks FSHD from a different angle: Avidity's approach uses an antibody to carry a DNA-cutting payload into muscle, Sarepta's uses RNA interference, Scholar Rock's targets myostatin to free up muscle growth, and Epicrispr's tunes the upstream DUX4 signal itself.

Epicrispr's first three patients represent among the earliest published human tests of DUX4 suppression as a treatment approach for FSHD. Six months and three patients are early enough that natural history, dosing, and the surrogate endpoint could all confound the signal. The readout confirms that an AAV-delivered epigenetic editor can be administered to FSHD patients and move a muscle-mass number. Whether that translates into function is the question the next cohorts will have to answer.

The closest comparable FSHD data point is delpacibart braxlosiran, the antibody-oligonucleotide conjugate that Novartis picked up in its $12 billion Avidity takeover in February 2026. It is the only FSHD program in Phase 3. Epicrispr's lean-mass signal is the only public mechanism proof-of-concept for DUX4 suppression. If delpacibart braxlosiran reads out first and shows a function benefit, it will set the bar Epicrispr's later trials have to clear, on a disease with no approved therapy in its entire clinical history.