Servier's $1.55 billion bet on Becker muscular dystrophy failed its first major test

The first large placebo-controlled test of sevasemten, an oral drug designed to keep damaged muscle fibers from injuring themselves when they contract, found that the medicine did not slow Becker muscular dystrophy, a rare genetic condition that progressively weakens the muscles of the arms and legs.

The drug came from Edgewise Therapeutics, which sold its muscular dystrophy business to Servier for $1.55 billion in cash in a deal announced in June and closed in July. Servier disclosed the topline result on November 3: the study missed its primary endpoint, the secondary endpoints were missed as well, and full results will be presented in the future. Servier's announcement Servier published no effect size, no failed secondary endpoint and no adverse-event figures. It said the trial showed a continued favorable safety and tolerability profile, and that the topline results are not aligned with certain earlier findings. It is conducting a comprehensive analysis of the GRAND CANYON dataset together with relevant prior clinical data, and no decision on the program has been announced. AllSci's summary

The primary endpoint was change from baseline in the North Star Ambulatory Assessment, a 17-item test of everyday movement that includes a four-stair climb and a 10-meter walk or run. After 18 months, adults taking sevasemten showed no statistically significant advantage over adults taking placebo. The ClinicalTrials.gov record The drug works by inhibiting fast skeletal myosin, the contractile protein in the fibers most easily injured in dystrophy, so it preserves muscle that already exists rather than correcting the genetic defect. Protected muscle did not translate into slower weakness.

Nothing changed in the clinic for the roughly 12,000 people estimated to live with Becker muscular dystrophy in the United States, Europe and Japan. There is still no approved treatment, and the population figure is Edgewise's own estimate rather than independently validated prevalence data. Edgewise's 2024 topline release

Servier agreed on June 1 to buy Edgewise's muscular dystrophy business for up to $2.65 billion: $1.55 billion in cash up front, plus up to $1.1 billion in regulatory and commercial milestones. The acquisition was announced in June and completed on July 13. At closing, Edgewise said the trial that would decide the price was fully enrolled with 175 participants and powered at greater than 98% to show a statistically significant difference versus placebo, with topline data expected in the fourth quarter of 2026. The up-front proceeds also fully funded EDG-7500, Edgewise's lead cardiovascular program in hypertrophic cardiomyopathy, a condition that thickens the heart muscle, through potential approval; that Phase 3 was also due to report in the fourth quarter of 2026. Edgewise's closing filing Servier reported 16.2% revenue growth for the fiscal year ended September 2025 and has named rare neurology as its next growth driver. Servier's fiscal-year results

CANYON, Edgewise's 2024 Phase 2 trial in 40 adults and 29 adolescents, met its primary endpoint by reducing creatine kinase, an enzyme that leaks out of injured muscle, by 28% more than placebo on average over months six through twelve (p=0.02). A second damage marker, TNNI2, fell 77% against placebo (p<0.001). Those were the numbers behind the purchase price, and they described a marker of muscle injury rather than a slower disease. Edgewise was careful about the rest: the study was not powered for the functional endpoints. On the same 17-item scale the later trial was built around, the drug-placebo difference at month 12 was 1.1 points in favor of sevasemten, p=0.16, in an adult placebo arm of 12 people, and the treated group started with more advanced disease on every functional measure. Edgewise's CANYON topline release

Between the 2024 biomarker result and the placebo-controlled readout, the purchase price rested on an untested assumption: that a drug which protects muscle from injury would slow the disease. Nothing Edgewise published in that stretch tested the second half of that sentence in a randomized comparison. ARCH and the MESA open-label extension enrolled people who had already taken sevasemten and people who switched onto it from placebo, with nobody assigned a concurrent placebo to compare against; as of the March 2025 data cut, 99% of eligible participants, 85 people, had enrolled in MESA. CANYON participants continuing in MESA improved 0.8 points on the functional scale over 18 months, and placebo switchers improved 0.2 points since starting the drug. Edgewise's 2025 program update The scale is ordinal, so points do not carry cleanly between studies, and in a slowly progressive disease a within-patient comparison cannot separate the drug from time.

After a Type C meeting, the process in which the FDA reviews a development plan before a sponsor files, Edgewise reported that the agency considered the CANYON data alone insufficient for accelerated approval, reiterated that the NSAA is a clinically meaningful endpoint for traditional approval, encouraged natural-history modeling, and supported the coming study as "a single adequate well-controlled study." Edgewise's 2025 program update No agency meeting record has been published; that account comes from the sponsor. One more well-controlled trial, and by Edgewise's own account of the meeting, the agency had said one was enough.

GRAND CANYON, the pivotal cohort inside the larger EDG-5506-201 protocol, randomized about 175 adults two to one to 10 mg of sevasemten or placebo over 18 months. The registry entry The labels do not line up, and it is worth naming who uses which. The registry lists protocol EDG-5506-201 under the phase PHASE2, with an official title beginning "A Phase 2 Randomized, Double-blind, Placebo-controlled Study," while its brief summary calls GRAND CANYON the protocol's "pivotal cohort." Servier's November 3 release calls it "the pivotal GRAND CANYON study" and labels CANYON "the Phase 2 CANYON study." Edgewise's June 2025 update called it "the ongoing global pivotal placebo-controlled cohort." AllSci's November 4 brief uses "phase iii" in the headline while its body describes GRAND CANYON as "the pivotal, placebo-controlled cohort within the Phase II CANYON program." A decisive failure reads as a verdict on the drug. A failure in a study the registry files as mid-stage reads as a verdict on a program that was still expected to produce more studies.

Parent Project Muscular Dystrophy, a patient advocacy organization, called the result "deeply disappointing," asked Servier for a webinar, and said it will keep following the program. The advocacy group's statement The statement is a position from an interested party rather than independent validation, and it names the constituency behind the trade press coverage: people with a disease that still has no approved treatment.

Servier's Duchenne trials, LYNX and FOX, remain active though no longer recruiting, with primary completion expected in January and March 2027, and the company has not said whether the Becker readout changes them. The sevasemten trial records The full GRAND CANYON dataset has no publication date attached. When it arrives, the numbers worth reading are the NSAA change in both arms and the placebo arm's own 18-month trajectory, because those were what the $1.55 billion was supposed to predict.