The $12 Billion Question: What Novartis's DM1 Failure Reveals About the Limits of Acquisition-Driven Drug Development

On September 8, 2026, Novartis announced that del-desiran, the antibody-oligonucleotide conjugate it acquired when it bought Avidity Biosciences for $12 billion in October 2025, had failed its pivotal Phase 3 HARBOR trial in myotonic dystrophy type 1. The drug did not demonstrate statistically significant improvement over placebo on the primary endpoint of video hand opening time, a measure of the delayed muscle relaxation that defines one of DM1's most disabling features. Novartis shares fell approximately 13% in pre-market trading. The company's stock had already dropped sharply days earlier when pelacarsen, its Lp(a)-lowering cardiovascular drug, also failed its Phase 3 outcomes trial. The two failures, arriving within days of each other and following the suspension of multiple CAR-T trials after three patient deaths, represent a concentrated period of clinical and commercial damage that is unusual even by the standards of an industry accustomed to setbacks.

The del-desiran failure is worth examining carefully, not just as a corporate event but as a case study in what happens when the logic of large-scale pharmaceutical acquisitions collides with the genuine difficulty of rare disease drug development.

What Del-Desiran Was Supposed to Do

Myotonic dystrophy type 1 is caused by an expansion of CTG repeats in the DMPK gene, which leads to the production of toxic mRNA that sequesters muscleblind-like proteins in the nucleus of muscle cells. The downstream consequences are wide-ranging: progressive muscle weakness, myotonia, cardiac involvement, cognitive impairment, and a multisystem deterioration that affects roughly 80,000 people in the United States and Europe. There are no approved disease-modifying treatments. Patients manage symptoms. Nothing touches the underlying biology.

Del-desiran was designed to change that. It is an antibody-oligonucleotide conjugate, a platform technology that pairs a muscle-targeting monoclonal antibody binding to the transferrin receptor 1 with a small interfering RNA payload designed to degrade the toxic DMPK mRNA at its source. The approach is conceptually elegant: use the antibody to deliver the RNA silencing agent directly to muscle tissue, where the pathology resides. Phase 1/2 data from the MARINA study, published in the New England Journal of Medicine in 2026, showed that the drug could reduce DMPK mRNA levels in muscle and improve molecular markers of disease. The FDA granted it Orphan Drug, Fast Track, and Breakthrough Therapy designations. The scientific case appeared credible enough to justify a $12 billion acquisition.

The Phase 3 HARBOR trial enrolled approximately 159 patients, randomized to del-desiran or placebo every eight weeks for 54 weeks. The primary endpoint was vHOT, a video-based measure of how quickly patients can open their hand after forming a fist. It is a functional measure of myotonia, one of the most characteristic and disabling features of DM1. The drug did not beat placebo on that measure. Novartis noted evidence of clinical activity in secondary endpoints and exploratory analyses, but without the primary endpoint, those signals carry limited regulatory weight.

The Acquisition Logic and Its Limits

When Novartis announced the Avidity acquisition in October 2025, external analysts projected peak annual del-desiran revenues as high as $6 billion, making DM1 the largest single commercial opportunity in the deal. The acquisition also brought two other AOC programs: del-zota for Duchenne muscular dystrophy, which has received FDA priority review for accelerated approval, and del-brax for facioscapulohumeral muscular dystrophy, which is in earlier development. Those programs remain active. But del-desiran was the centerpiece, the asset that justified the price and the timeline.

The failure raises a question that the pharmaceutical industry rarely asks directly: how much of the confidence embedded in a large acquisition price reflects genuine scientific conviction, and how much reflects competitive pressure to secure assets before rivals do? The AOC platform was genuinely novel, and the Phase 1/2 data were real. But Phase 1/2 data in rare neuromuscular disease have a history of generating optimism that Phase 3 trials do not always sustain. The gap between molecular biomarker improvement and functional clinical benefit is one of the most persistent challenges in the field, and DM1 is a particularly complex disease to measure, given its multisystem nature and the heterogeneity of its presentation across patients.

Guggenheim Securities analysts had described the HARBOR readout as "crucial" for validating the Avidity deal. Oppenheimer called del-desiran a "$12 billion disappointment." Those characterizations are accurate as far as they go, but they also reflect a broader pattern in pharmaceutical M&A: the tendency to price acquisitions as if Phase 3 success is the base case, when the historical probability of Phase 3 success in rare neuromuscular disease is considerably lower than that framing implies.

What This Means for the DM1 Field

The HARBOR failure does not invalidate the AOC platform or the broader scientific hypothesis that silencing DMPK mRNA can produce clinical benefit in DM1. It raises questions about the specific design choices in the HARBOR trial, including the choice of vHOT as the primary endpoint, the dose and dosing interval, and whether the patient population enrolled was optimally selected to detect a treatment effect. Novartis has said it will analyze the full dataset and engage with health authorities to determine next steps. Whether that process leads to a redesigned trial or a decision to discontinue the program will depend on what the secondary and exploratory data show.

The competitive landscape in DM1 has shifted. Jefferies analysts noted that Dyne Therapeutics, which is developing z-basivarsen, an antisense oligonucleotide candidate with what analysts describe as stronger tissue penetration and better biodistribution than del-desiran, now "claims frontrunner status" in the indication. Dyne's Phase 1/2 study is ongoing, with topline data expected in early 2027. The del-desiran failure may actually benefit Dyne by removing a well-resourced competitor from the near-term regulatory pathway, while also providing the field with data about what endpoint design and patient selection challenges need to be addressed.

The Broader Signal for Rare Disease Development

The timing of the del-desiran failure, arriving alongside the pelacarsen cardiovascular miss and the CAR-T safety events, has focused attention on Novartis's pipeline in a way that is uncomfortable for the company. But the deeper lesson extends beyond any single organization. The pharmaceutical industry has spent the past several years paying historically high prices for late-stage rare disease assets, driven by the logic that unmet need, regulatory incentives, and platform technology justify premium valuations even before Phase 3 data exist.

That logic is not wrong in principle. The rare disease space has produced genuine breakthroughs, and the regulatory environment has evolved to support development in small patient populations. But the HARBOR failure is a reminder that the distance between a compelling Phase 1/2 dataset and a Phase 3 win is real, and that the complexity of measuring functional benefit in multisystem diseases like DM1 is not fully resolved by molecular evidence of target engagement. For the roughly 80,000 patients in the United States and Europe living with DM1, the failure means another delay in the arrival of a disease-modifying treatment. For the field, it is a data point that will inform how the next generation of DM1 trials is designed, what endpoints are chosen, and how the gap between molecular and functional benefit is bridged. That process is slower and more uncertain than acquisition announcements suggest. It is also the only way the science actually moves forward.