Sarepta Therapeutics confirmed Thursday that it will permanently close enrollment in ENVISION, the Phase 3 confirmatory trial that underpinned FDA’s accelerated approval of Elevidys (delandistrogene moxeparvovec-rokl) for non-ambulatory Duchenne muscular dystrophy patients. The announcement came during the company’s third-quarter 2026 earnings call — one day after TechTimes reported that Novartis had confirmed its own AAV gene therapy, Zolgensma, had caused tumors in two children via vector integration.

For the families of non-ambulatory DMD patients, Sarepta’s decision closes a chapter that began with rare hope and ended in regulatory limbo. Access to Elevidys for this population — the most severely affected boys and young men living with the disease — has been blocked since November 2025, when FDA stripped the non-ambulatory indication from Elevidys’ label after three patients died from acute liver failure. With ENVISION now formally closed, the only pathway back to access runs through a 25-patient open-label safety study that is not designed to demonstrate clinical benefit, and whose data will need to persuade an FDA that has shown little patience for confirmatory evidence shortcuts in recent years.

What Just Happened, and Why It Matters

Duchenne muscular dystrophy is a rare, X-linked genetic disease caused by mutations or deletions in the dystrophin gene on chromosome Xp21. Without functional dystrophin — a structural protein that acts as a shock absorber in muscle fibers during contraction — muscles are progressively destroyed by ordinary movement. The disease is relentless: most patients lose the ability to walk in their early to mid-teens, and even with modern respiratory and cardiac care, median life expectancy remains around 28–30 years. It affects roughly 1 in every 3,500 to 5,000 males born worldwide, with an estimated 15,000–20,000 patients in the United States.

Elevidys, the first gene therapy ever approved for Duchenne, works by delivering a synthetic micro-dystrophin gene directly into muscle cells using an adeno-associated virus (AAVrh74) as the delivery vehicle. Because the full dystrophin gene is too large to fit inside any viral vector, Sarepta’s researchers — led internally by Louise Rodino-Klapac, the company’s president of research and development, who co-developed the micro-dystrophin construct during her earlier career at Nationwide Children’s Hospital in Ohio — engineered a shortened but functional version. After infusion, the vector particles enter muscle cells, travel to the nucleus, and begin expressing micro-dystrophin protein. Because AAV does not integrate into the genome in most cases, the gene does not permanently alter the cell’s DNA — but it also cannot be re-dosed, because the immune system retains memory of the viral capsid and will attack subsequent infusions.

The FDA approved Elevidys in June 2023 under its accelerated approval pathway — a track created in 1992 to speed access to treatments for serious conditions with unmet need. Accelerated approval rests on surrogate endpoints: markers thought to predict clinical benefit rather than direct evidence of that benefit. For Elevidys, the surrogate was micro-dystrophin expression in skeletal muscle. That made post-marketing confirmatory evidence a legal requirement, not a suggestion. ENVISION (Study SRP-9001-303) was that confirmatory trial — a global, randomized, double-blind, placebo-controlled Phase 3 study designed to enroll roughly 150 patients across multiple countries and demonstrate that micro-dystrophin expression translated into measurable clinical benefit for non-ambulatory patients and older ambulatory patients.

It never got there.

A Year of Deaths and a Trial Frozen in Place

Before ENVISION could reach full enrollment, three patients died.

In March 2025, Sarepta disclosed the first death: a young man who died from acute liver failure following an Elevidys infusion. The company noted he had also experienced a recent cytomegalovirus (CMV) infection, which may have contributed to the severity of his response, and emphasized that acute liver injury was an already-labeled risk of AAV-based gene therapies. The European Medicines Agency moved quickly, requesting temporary holds on three Elevidys trials in EU countries. Sarepta maintained that the risk-benefit profile remained favorable.

Three months later, in June 2025, a second non-ambulatory patient — a 15-year-old enrolled in ENVISION — died from acute liver failure. Sarepta suspended commercial Elevidys shipments to non-ambulatory patients and voluntarily paused all ENVISION dosing worldwide. Then, in July 2025, a third death prompted the FDA to formally request a halt to all Elevidys distribution and place clinical trials on hold — a step that sent the company’s stock into freefall, with shares declining roughly 40% at the peak of the crisis.

All three deaths occurred in non-ambulatory patients. The pattern is not incidental: non-ambulatory DMD patients are at a more advanced stage of the disease, with progressive systemic involvement including dilated cardiomyopathy (affecting roughly half of 18-year-old DMD patients), compromised respiratory function, and potentially altered metabolic reserves — all of which may reduce the liver’s resilience when it comes under immune attack.

The biological mechanism driving these deaths is now reasonably well understood. AAV gene therapy works by hijacking the machinery of a naturally occurring virus to deliver genetic cargo to human cells. The AAV capsid — the protein shell surrounding the viral vector — is recognized as foreign by the immune system. CD8+ cytotoxic T lymphocytes mount a response against cells that display AAV capsid peptides on their surface, including hepatocytes (liver cells) that the vector infects in transit. In most younger, healthier patients, this immune response is manageable. In a subset of non-ambulatory patients — possibly those with prior immune history from CMV or other viral infections, possibly those with compromised hepatic reserve from disease progression — the response becomes overwhelming. The result is acute liver failure: rapid loss of roughly 80–90% of hepatic cell function, coagulopathy, and neurological deterioration.

By November 2025, FDA approved a revised Elevidys label that removed the non-ambulatory indication entirely and added a boxed warning for acute liver failure — the FDA’s most prominent safety alert, reserved for risks so serious they can outweigh benefit. Ambulatory patients were not affected; their label retained traditional approval status. But for non-ambulatory patients and their families, November 2025 was the moment the door closed.

ENVISION Is Finished — Here Is What That Means for the FDA

On Thursday’s earnings call, Sarepta CEO Michael Severino — a physician-scientist with a biochemistry degree from the University of Maryland and an M.D. from Johns Hopkins, appointed to the role in July 2026 after prior stints leading R&D at AbbVie and running gene-writing startup Tessera Therapeutics — announced the formal end of ENVISION.

“After careful consideration of our clinical development strategy, we’ve decided not to resume enrollment of the ENVISION study,” Severino said. “ENVISION has been on hold for over a year, and we have recruited approximately half of its intended sample size.”

The company cited operational and technical barriers accumulated during the extended hold as key factors in the decision. Sarepta said it will analyze the blinded data already collected from Part 1 of the study — representing roughly 90 participants enrolled before the freeze — and share findings with the Duchenne community. The Parent Project Muscular Dystrophy, one of the leading DMD advocacy organizations, called the closure “disappointing news for the individuals and families who have been following and participating in the trial.”

What the ENVISION closure does to the regulatory situation is more consequential than the company’s framing suggests.

The FDA’s accelerated approval framework was explicitly designed around confirmatory trials: not safety studies, not open-label cohorts, not retrospective analyses — controlled, adequately powered trials that demonstrate clinical benefit directly. When Congress passed the Omnibus legislation in 2022, it gave FDA clearer and faster authority to withdraw accelerated approvals when confirmatory evidence fails to materialize on schedule. The agency used that authority aggressively in oncology, withdrawing a series of cancer indications in 2022 and 2023 where confirmatory data underdelivered.

ENVISION was Elevidys’ confirmatory trial for the non-ambulatory indication. It is now closed without full enrollment. This creates a problem the FDA has not previously faced in precisely this form: a trial that could not be completed because completing it in the same form would require treating patients in a population where three prior recipients died.

The company’s argument is that ENVISION was never the near-term vehicle for restoring non-ambulatory access anyway. “ENVISION would not have provided data in time to inform that decision under any case,” Severino said, pointing investors toward a first-quarter 2027 FDA meeting as the operative regulatory milestone. That is true as far as it goes — a fully enrolled and completed Phase 3 trial would take years, while the sirolimus safety data could be in hand by early 2027. But it sidesteps the question the FDA must answer: whether Phase 1b open-label safety data from roughly 25 patients is an adequate evidentiary basis for sustaining an accelerated approval, or whether the non-ambulatory indication should be formally withdrawn.

The FDA’s answer will set a precedent that extends well beyond Sarepta and Elevidys. If the agency accepts ENDEAVOR Cohort 8 data as a confirmatory substitute, it establishes that rare disease gene therapy developers can replace failed Phase 3 trials with smaller safety cohorts when safety-related halts make Phase 3 continuation infeasible. That could be a reasonable and humane accommodation for a genuinely novel category of therapeutic challenge. If the FDA declines, it effectively tells the rare disease gene therapy field that confirmatory evidence failures — including those forced by safety crises rather than by efficacy misses — will be treated the same way: approval withdrawn.

The Sirolimus Bet: What ENDEAVOR Cohort 8 Is Actually Testing

Sarepta’s alternative path runs through Cohort 8 of its ENDEAVOR study (SRP-9001-103), an open-label Phase 1b trial that the FDA approved for dosing in November 2025.

The cohort is enrolling approximately 25 non-ambulatory patients in the United States. Under the protocol, participants receive sirolimus — a macrolide compound that inhibits mTOR (mammalian target of rapamycin) kinase, suppressing T and B cell activation — for 14 days before Elevidys infusion and continue for 12 weeks afterward. The drug was originally isolated in 1972 from soil bacteria on Easter Island (its name derives from Rapa Nui) and has been FDA-approved since 1999 for kidney transplant rejection prophylaxis; it also coats drug-eluting coronary stents to prevent restenosis.

The rationale for adding sirolimus to the Elevidys protocol is mechanistic: if the deaths were caused by an overactive CD8+ T-cell response to AAV capsid proteins, then dampening that response before the vector is infused — and sustaining the suppression while micro-dystrophin expression is established — should reduce or eliminate the hepatotoxic cascade. The 14-day lead time is designed to achieve sustained mTOR inhibition before the immune system encounters the AAV vector.

The scientific rationale received real-world support from a retroactive analysis presented at the 2026 World Muscle Society Annual Congress in which researchers examined outcomes in patients who had received prophylactic sirolimus ahead of their Elevidys infusion. No acute liver failure events were observed in that sirolimus-pretreated group — a finding that is preliminary and from a small dataset, but that provided Sarepta and the FDA enough confidence to proceed with ENDEAVOR Cohort 8 as an FDA-approved study.

Rodino-Klapac, whose career has been built around developing the science behind Elevidys, was direct about what is at stake: “Notwithstanding our decision to not resume enrollment in ENVISION, our commitment to serve non-ambulatory individuals with Duchenne remains steadfast. This population continues to have fewer treatment options.”

Sarepta said enrollment in Cohort 8 is on track to complete by the end of 2026, with FDA engagement planned for the first quarter of 2027 once 12-week safety and efficacy data are in hand. The primary endpoints are twofold: the incidence of acute liver injury (the safety question) and micro-dystrophin expression at 12 weeks (the surrogate efficacy question). It is worth noting that micro-dystrophin expression is the same surrogate endpoint that underpinned the original 2023 accelerated approval — the trial is not designed to show clinical benefit directly, only that the safety problem has been solved.

Sirolimus is not without its own risks. The drug carries FDA safety warnings for decreased renal function, increased infection susceptibility, and a 2009 notice of increased mortality in stable liver transplant patients switched to sirolimus. Rodino-Klapac has acknowledged that sirolimus carries its own risks, including infection — a material concern in a population with already-compromised immune and muscle function.

Commercial Reality: Stabilizing From a Low Floor

The backdrop to this clinical drama is a battered franchise trying to find its footing. Elevidys generated $83.5 million net product revenue in the third quarter of 2026, against a total net product revenue figure for Sarepta of $323.8 million — with $240.3 million coming from the company’s older antisense oligonucleotide (PMO) exon-skipping therapies (AMONDYS 45, VYONDYS 53, and EXONDYS 51). Year-over-year, total net product revenue fell from $370.0 million in Q3 2025 to $323.8 million in Q3 2026, a decline of roughly 12.5% driven almost entirely by the collapse in Elevidys sales as the safety crisis eroded physician and family confidence.

Some signals suggest the floor has been reached, if not yet rebuilt. Sarepta reported that enrollment form activity for Elevidys — a leading indicator of near-term dosing — was in line with second-quarter levels. The company raised the low end of its full-year 2026 total net product revenue guidance to between $1.25 billion and $1.3 billion, bolstered by its broader portfolio. Per William Blair analysts, the business is “appearing to stabilize.” Jefferies analysts have pointed to a potential “turnaround story” built on a runway of 2027 catalysts.

For Q4 2026, management guided that Elevidys revenue would trend in line with the third quarter, subject to the normal variability inherent in a one-time gene therapy — where quarterly figures are lumpy because each dose is a single six-figure transaction rather than a recurring prescription.

Sarepta also faces a separate near-term regulatory event: a PDUFA date of February 28, 2027 for FDA’s decision on traditional approval of AMONDYS 45 and VYONDYS 53, following a Phase 3 study whose top-line result — announced earlier in 2026 — did not meet its primary stair-climbing endpoint. The company is making a statistical case for traditional approval based on the data pattern when pandemic-affected portions of the nine-year study are excluded; the FDA will decide whether that argument holds.

A supplemental biologics license application to expand Elevidys to children under four years of age is planned for mid-2027 filing, based on data from earlier clinical work. This indication, if approved, would be separate from the non-ambulatory question and would represent a modest commercial expansion for the ambulatory indication.

A Community in Waiting

The closure of ENVISION is acutely felt in the Duchenne community precisely because Elevidys was never an ordinary drug. It was the first gene therapy ever approved for DMD — the first attempt to address the disease at its molecular root rather than manage its symptoms — and the non-ambulatory indication represented a rare and hard-fought hope for the boys and young men who had already lost the ability to walk.

Since the safety crisis of 2025, the extended hold on the non-ambulatory indication, and now the formal closure of the confirmatory trial, that hope has been deferred rather than extinguished. The sirolimus path is real. But it is narrow: 25 patients in one open-label cohort, without a concurrent control arm, targeting the same surrogate endpoint that the FDA has already approved once — and whose clinical significance has been a matter of ongoing scientific and regulatory debate.

The community letter released Thursday, in which Sarepta attempted to hold the trust of families who had enrolled patients in ENVISION or who were following the program in hope of future access, acknowledged the weight of the moment. The company restated its commitment to sharing ENVISION Part 1 blinded data findings. It reaffirmed the sirolimus path. It did not make promises about timelines it cannot control.

What it could not offer was certainty — and in a disease where time matters in ways that most medical conditions cannot match, the absence of certainty carries its own weight.

What to Watch

Q1 2027 (target): FDA engagement on ENDEAVOR Cohort 8 data. This is the operative milestone. The 12-week safety and efficacy data from roughly 25 patients will be presented to FDA; the agency’s response — whether it views this data as sufficient to inform a path back to non-ambulatory access — will determine whether the accelerated approval survives in some form or is effectively abandoned.

End of 2026 (target): Cohort 8 enrollment completion. Sarepta has said enrollment is on track. If that holds, primary endpoint data should be available in Q1 2027.

February 28, 2027: FDA PDUFA date for AMONDYS 45 and VYONDYS 53. Traditional approval of the PMO franchise would provide commercial stability for the broader Sarepta portfolio regardless of the Elevidys non-ambulatory outcome.

Mid-2027: Planned sBLA submission for Elevidys in children under four. A separate expansion of the ambulatory indication that does not depend on resolution of the non-ambulatory regulatory question.

Ongoing: ENVISION Part 1 blinded data analysis. Sarepta has committed to sharing findings from the data already collected before the freeze. These findings may or may not provide useful signal on efficacy in the enrolled population; their regulatory weight will be limited given the absence of a completed, fully enrolled trial.

The question hanging over all of these milestones is not purely commercial. Sarepta’s bet on sirolimus is simultaneously a scientific hypothesis, a regulatory gamble, and a commitment to a community of patients who have already waited long enough. Whether the FDA treats ENDEAVOR Cohort 8 as adequate confirmation or as evidence insufficient for a provisional approval to survive may shape not just Elevidys’ future, but the future of how the agency approves — and maintains — accelerated approvals for rare disease gene therapies where the standard confirmatory trial framework collides with medical reality.

Frequently Asked Questions

Can non-ambulatory DMD patients currently receive Elevidys?

No. The non-ambulatory indication was removed from Elevidys’ FDA label in November 2025 following three deaths from acute liver failure. As of this writing, only ambulatory patients aged four and older can receive Elevidys commercially. Non-ambulatory patients may be eligible to participate in ENDEAVOR Cohort 8, the clinical trial testing sirolimus prophylaxis, but trial participation is not the same as commercial access.

What is ENVISION and why was it important?

ENVISION (Study SRP-9001-303) was the Phase 3 randomized, placebo-controlled confirmatory trial that the FDA required Sarepta to complete as a condition of the June 2024 accelerated approval of Elevidys for non-ambulatory patients. Under FDA’s accelerated approval framework, approval based on a surrogate endpoint (in this case, micro-dystrophin expression in muscle) is conditional on completing a confirmatory trial that demonstrates actual clinical benefit. ENVISION was that trial. Its permanent closure means Sarepta must find an alternative way to satisfy — or persuade FDA to waive — that confirmatory evidence requirement.

What is ENDEAVOR Cohort 8 and what does it aim to show?

ENDEAVOR Cohort 8 is an open-label Phase 1b clinical trial enrolling approximately 25 non-ambulatory US patients. It tests whether adding sirolimus (an immunosuppressant) to the Elevidys protocol — 14 days before infusion and 12 weeks after — can prevent the acute liver failure that killed three prior recipients. Its primary endpoints are the incidence of acute liver injury and micro-dystrophin expression at 12 weeks. It is a safety study, not a Phase 3 efficacy trial, and it is significantly smaller than ENVISION. Whether FDA will accept its data as an adequate basis for restoring the non-ambulatory indication is the central regulatory question of 2027 for the Duchenne community.

Why does sirolimus help prevent liver injury in gene therapy patients?

Acute liver failure in AAV gene therapy is believed to be caused by an immune response: the immune system’s CD8+ cytotoxic T lymphocytes recognize AAV capsid proteins on the surface of transduced liver cells and attack them. Sirolimus inhibits mTOR, a kinase that is required for T and B cell activation and proliferation. By dampening this immune response before the vector is infused, sirolimus may prevent the cascade of immune-mediated liver damage before it becomes irreversible. No acute liver failure was observed in a small retroactive analysis of patients who received sirolimus prophylaxis prior to Elevidys infusion, which provided the scientific rationale for ENDEAVOR Cohort 8.

Why are non-ambulatory DMD patients at higher risk of liver complications than ambulatory patients?

No single explanation has been confirmed, but the available evidence points to disease progression as a likely factor. Non-ambulatory DMD patients are further along in the disease course, with more extensive cardiac involvement (dilated cardiomyopathy affects roughly half of 18-year-old DMD patients), compromised respiratory function, and reduced metabolic reserve. All three fatalities occurred in this group, while ambulatory patients have not shown the same risk profile at the same rate. It is also possible that pre-existing viral immune history (such as prior CMV infection, which was noted in the first death) plays a role in predisposing certain patients to a more severe hepatic immune response.

Is Elevidys still available for ambulatory patients?

Yes. The ambulatory patient indication holds traditional FDA approval (not accelerated approval), which means it is not at risk of withdrawal due to the confirmatory trial situation. Elevidys remains commercially available for ambulatory DMD patients aged four and older, and Sarepta is planning to expand this indication to children under four via a supplemental application expected to be filed in mid-2027.

What does this mean for the broader gene therapy field?

The Elevidys situation — alongside Novartis’s separate disclosure that Zolgensma caused tumors in two children via vector integration — has focused attention on the safety profile of AAV vectors at a critical moment for the field. How FDA responds to Sarepta’s request to treat Phase 1b safety data as a confirmatory substitute will shape whether rare disease gene therapy developers can maintain approvals when safety-related trial disruptions make standard confirmatory timelines impossible. The gene therapy field, regulators, and rare disease patient communities are watching the Q1 2027 FDA engagement closely.