Duchenne cell therapy hints at heart benefit before FDA verdict

Deramiocel, a cell therapy made from donor heart cells, slowed the spread of scar tissue in the hearts of boys and young men with Duchenne muscular dystrophy in the phase 3 HOPE-3 trial. The new cardiac MRI results were presented at the American Heart Association (AHA) Scientific Sessions in Chicago on 7 November 2026, where they won the meeting's Outstanding Research Award in Pediatric Cardiology. The US Food and Drug Administration is due to decide whether to approve the treatment by 22 November.

The heart findings come with real caveats. The trial's main cardiac test, a change in pumping strength across all patients, missed statistical significance. The scarring result rests on 22 patients with usable scans. And the FDA is now being asked to approve deramiocel for arm function, a different claim from the heart benefit the company first sought.

What was presented at AHA 2026?

Jonathan Soslow, a paediatric cardiologist at Vanderbilt University Medical Center, presented the cardiac results of HOPE-3 on 7 November. According to Capricor's announcement, the talk added two things to what was already published: cardiac MRI data on myocardial fibrosis (scarring of the heart muscle), and a comparison with untreated boys from a natural history database run by the Duchenne Muscular Dystrophy Cardiac Care Consortium.

Scarring was measured with LGE imaging, a cardiac MRI method in which a gadolinium contrast dye lingers in scarred tissue, so damaged parts of the heart light up. Cardiologists split the left ventricle into segments and count how many show scar. In Duchenne that count tends to rise year after year.

Over 12 months, the number of newly scarred segments grew by 3.0 fewer in the deramiocel group than in the placebo group (p=0.02). Against the natural history patients the gap was 2.9 segments (p=0.01). Soslow said in the company statement that cardiologists "watch the scar increase and the heart function decrease year after year in patients with Duchenne", and that seeing signals of stabilisation for both was encouraging.

How did the heart results look in numbers?

Here are the cardiac figures Capricor released, with the number of patients behind each comparison. LVEF is left ventricular ejection fraction, the share of blood the main pumping chamber pushes out with each beat.

Only the first row was a key secondary endpoint, the one the trial was set up to test formally after the primary endpoint. It did not reach significance. Under the trial's fixed testing order, every result below it is described as nominal, meaning the p-value was not adjusted for the number of comparisons. Capricor's release says this plainly, and it also notes that the natural history comparison was post hoc (planned after the data were in).

The scarring row deserves the most caution. Nine treated patients were compared with 13 on placebo. A difference in so few people can be real, but it can also shift a lot when more patients are added. Our guide to placebo and blinding explains why the randomised column carries more weight than the natural history one.

Why did the main heart result change from p=0.04 to p=0.09?

When Capricor first announced topline HOPE-3 results on 3 December 2025, it said LVEF had improved by 2.4 percentage points against placebo with p=0.041, a statistically significant win. That is no longer the official figure.

In its second-quarter 2026 update, the company said it found "an issue with the statistical model" in the clinical study report during peer review and went back to the statistical analysis plan written before unblinding. Re-run that way, the LVEF difference became 1.8 points with p=0.09. Capricor said the cardiomyopathy subgroup result (p=0.02) and the primary endpoint were unchanged.

That correction matters for anyone reading older coverage. Headlines from December 2025 that said the trial "met" its cardiac endpoint describe an analysis that has since been withdrawn.

What did HOPE-3 find for arm and hand function?

The primary endpoint was arm function, measured with the Performance of the Upper Limb 2.0 scale (PUL 2.0). It scores tasks such as lifting the arms, reaching and picking up objects, and it matters for boys who no longer walk, because arm use decides whether someone can feed himself, use a phone or drive a powered wheelchair.

The full results were published in The Lancet on 29 July 2026, led by Craig McDonald of UC Davis Health. At 12 months, the percentage change in total PUL 2.0 score favoured deramiocel by 4.55% (95% CI 0.47% to 8.63%; p=0.029). On the elbow-level part of the scale the difference was 8.12% (p=0.008), and a filmed test of independent eating, the Duchenne Video Assessment Eat 10 Bites task, also favoured treatment (p=0.0176), according to a summary of the paper by Rare Disease Advisor.

Capricor describes the arm result as a 54% slowing of decline. That kind of figure is easy to misread, so here is how it works. The company defines percent slowing as the treatment difference divided by the decline in the placebo group. The reported absolute difference was 1.2 points on the PUL 2.0 scale. If 1.2 points is 54% of the placebo decline, then the placebo group lost about 1.2 / 0.54 = 2.2 points over the year, and the deramiocel group lost about 2.2 minus 1.2 = 1.0 point. Those two derived figures are our arithmetic from the published numbers, rounded. Both groups still got worse. The treated group got worse more slowly.

That pattern is what a disease-slowing treatment looks like in Duchenne. Nobody in HOPE-3 regained lost strength on average. The question regulators face is whether a one-point difference on a scale that runs to 42 is enough to change daily life, and whether it lasts.

Who took part in the trial?

HOPE-3 (NCT05126758) enrolled 106 boys and young men aged 10 or older at 20 US sites between June 2022 and May 2024. Out of 139 people screened, 54 were randomised to deramiocel and 52 to placebo. Capricor's topline release put the average age at about 15, and everyone stayed on a stable dose of corticosteroids, the standard Duchenne treatment. About 90% were taking heart medicines at the start.

Most participants were in the late walking or non-walking stage of the disease, but still able to bring a hand to the mouth. Of the 106, 64 had documented cardiomyopathy at baseline. Deramiocel was given as an intravenous infusion of 150 million cells every three months, four doses in all, in outpatient clinics.

This group is older and more disabled than most Duchenne trial populations. Many recent drugs were tested in boys who could still walk, so results for people who use a wheelchair are scarce. McDonald said in the December 2025 release that HOPE-3 was the first phase 3 trial in a largely non-walking Duchenne population to meet its primary endpoint.

What is deramiocel and how is it meant to work?

Deramiocel, also known as CAP-1002, is made of cardiosphere-derived cells grown from donated human heart tissue. Allogeneic means the cells come from a donor, so one batch can treat many patients and no surgery is needed to collect a patient's own cells. The treatment works the same way whatever genetic mutation causes a patient's Duchenne, which separates it from exon-skipping drugs and gene therapies that suit only some patients.

Capricor says the cells act mainly by releasing tiny packages called exosomes. These are thought to switch macrophages, a type of immune cell, from an inflammatory state to a healing one, and to reduce scarring. Duchenne itself is caused by missing dystrophin, a protein that holds muscle cells together, so deramiocel does not fix the underlying fault. The aim is to slow the damage that follows.

What side effects were reported?

According to the Lancet paper as summarised by Rare Disease Advisor, adverse events of any kind affected 94.3% of the deramiocel group and 82.7% of the placebo group. Almost all (98%) were mild or moderate. Hypersensitivity reactions, such as allergic-type reactions around infusion, were the clear difference: 41.5% with deramiocel against 15.4% with placebo. Serious adverse events were more common in the placebo group, and nobody died during the study.

Capricor reports that it has given about 1,300 infusions of deramiocel to more than 200 people with Duchenne across three trials, with some in open-label extensions for over five years.

Where does the FDA review stand?

The road to this decision has been long:

• 11 July 2025: the FDA declined to approve Capricor's first application, which sought a heart indication based on earlier data.
• 3 December 2025: HOPE-3 topline results were announced, and the company later resubmitted with the new trial data.
• 29 July 2026: an FDA advisory committee voted 9 to 3 that the evidence did not support effectiveness for Duchenne cardiomyopathy. The panel did not vote on arm function but, by Capricor's account, its discussion of that endpoint was "directionally supportive".
• 24 August 2026: the FDA extended the review to 22 November after Capricor submitted 24-month open-label data and asked for an indication focused on upper limb function.

Capricor's August update also disclosed that an FDA inspection of trial records in July ended with a Form 483 citing one observation, which the company has answered. Advisory committee votes are not binding, and the narrower request means the FDA is now weighing a different question from the one the panel voted on. Our explainer on how new medicines get approved covers what each of these steps means.

If deramiocel is approved, the rare paediatric disease designation it holds may earn Capricor a priority review voucher, which companies can sell. That gives the company a financial stake in approval beyond sales, something worth keeping in mind when reading its statements.

Could deramiocel reach patients in the UK?

Not soon. HOPE-3 ran only in the US, and we found no application to the MHRA or the European Medicines Agency. Deramiocel has orphan drug and advanced therapy designations in Europe, and Capricor said in August that it had begun talking to regulators in Europe and Japan. That is an early step, a long way from a licence.

The UK context is useful, though. In May 2026 NICE recommended givinostat, sold as Duvyzat, for people aged 6 and over who can walk or stand at the start of treatment, in line with who was studied in its trials. That leaves out many of the older, non-walking patients who made up most of HOPE-3. Muscular Dystrophy UK keeps a list of currently available treatments that families can check.

What should families ask their Duchenne team?

If you or your son has Duchenne and you are following this story, these questions can help at the next clinic visit:

• What does his latest cardiac MRI show for ejection fraction and scarring, and how fast has it changed?
• Is he on the heart medicines that guidelines recommend for Duchenne, such as ACE inhibitors and drugs that block aldosterone?
• Which trials near us accept boys who no longer walk?
• If deramiocel is approved in the US, would it ever be possible to get it through a trial or an early access scheme here?

Our guide to finding and joining a clinical trial explains how to search UK and international registries.

This article is for general information and is not medical advice. Talk to your own Duchenne team before changing any treatment.

How should this result be read?

Two things can be true at once. HOPE-3 met its primary endpoint in a group of patients that trials have mostly ignored, and its safety record looks acceptable. Its heart results, which the AHA chose to honour, are suggestive but not proof, because the formal test missed and the scarring data come from 22 people.

Other heart drugs for rare diseases have faced the same gap between promising signals and hard outcomes. The CARDIO-TTRansform trial of eplontersen is a recent example of how a strong biological effect does not always translate into fewer heart events. If you want a refresher on what a missed endpoint does and does not mean, see what it means when a clinical trial fails.

The next fixed date is 22 November 2026, when the FDA's decision on arm function is due. A separate trial built to test heart outcomes in Duchenne would be the way to settle the cardiac question.