I went into this approval expecting to cheer, and the first question I caught myself asking was an awkward one. How does a drug whose only randomized trial found nothing end up as the first medicine the FDA says keeps children alive? Adrabetadex is about the plainest molecule you could put in a vial, a little barrel of glucose units. On November 17, 2026, the FDA approved it as Lembaz to reduce the risk of death in children and adults with infantile-onset Niemann-Pick disease type C (NPC), and it got there on a survival comparison against children who were never in the study.
NPC is inherited and recessive, caused by broken NPC1 or NPC2 genes. Cholesterol that should be shipped out of the cell’s recycling compartment, the lysosome, gets stuck there and piles up, worst of all in the brain. The FDA puts prevalence at roughly 1 in 150,000, and Beren Therapeutics, the parent of approval holder Mandos, estimates about 475 American children have the infantile-onset form. NeurologyLive reports a mean age at death of about 5.6 years when neurological symptoms start before age 2, and 13.4 years when they start between 2 and 6. Elizabeth Berry-Kravis of Rush University, who leads the drug’s expanded access program, put the clinical problem bluntly to Contemporary Pediatrics: “neurological function that is lost cannot be regained.” The two NPC drugs approved in 2024, arimoclomol and levacetylleucine, have never been shown to reduce deaths, and the FDA notes both carry age and weight restrictions.
Why would a sugar fix a fat problem?
My first reaction was that someone had the chemistry backwards. The answer turns out to be the molecule’s shape. Hydroxypropyl-beta-cyclodextrin is a ring whose outside loves water and whose hollow core repels it, and cholesterol sits comfortably in a pocket like that. The leading explanation is that the ring gives trapped lysosomal cholesterol a way out, which is why Berry-Kravis describes Lembaz as “a way to address the underlying cholesterol trafficking defect.” The catch is built into the same chemistry. The ring is big and water-loving and doesn’t get from the bloodstream into the brain, so a trained provider injects it into the spinal canal every two weeks, for patients who weigh at least 7 kilograms and whose symptoms began before age 6.
Nobody learned that delivery lesson in a corporate lab. Chris Hempel’s twin daughters, Addi and Cassi, were diagnosed with NPC in October 2007. In April 2009 the FDA allowed the girls weekly intravenous infusions of cyclodextrin through a chest port, under their physician, Caroline Hastings of Children’s Hospital Oakland. The researcher David Begley then found the compound wasn’t crossing from the blood into the brain, and animal work in cats and mice suggested direct brain delivery extended life. In 2010 the FDA cleared Hastings to inject it into the twins’ spinal fluid, with plans for ports feeding it into the brain’s ventricles, which is close to how Lembaz is given today.
- 2009IV infusions for the Hempel twins
- 2017NIH-led trial published
- 2018randomized trial misses
- 2021Mandos buys the program
- 2026FDA approval
The National Institutes of Health built the first formal human evidence. An NICHD-led phase 1/2a trial, published in The Lancet in 2017, gave intrathecal cyclodextrin monthly to 14 patients aged 4 to 23 at the NIH Clinical Center for 12 to 18 months, plus three more every two weeks at Rush for 18 months. Compared with 21 patients from an earlier natural-history study, treated patients scored lower on neurological severity for cognition and speech. The work ran under a cooperative research agreement with Sucampo Pharmaceuticals, with drug supplied by Janssen, and parent-run charities chipped in: the Ara Parseghian Medical Research Foundation, Dana’s Angels Research Trust, Hope for Haley, Samantha’s Search for the Cure Foundation, the National Niemann-Pick Disease Foundation and Support of Accelerated Research for NPC Disease. Most participants also lost more hearing, which hearing aids compensated for.
Four owners and a trial that found nothing
By then the molecule, called VTS-270, had passed from Vtesse to Sucampo and on to Mallinckrodt. In November 2018 Mallinckrodt reported that its double-blind phase 2b/3 trial in 51 patients, which measured change in the NPC Clinical Severity Score at week 52, showed “no significant separation from placebo.” Children in both arms unexpectedly showed no disease progression over the year. Chief science officer Steve Romano pointed to inconsistencies in the study population, Patient Worthy reported, including age and the severity of the children’s cases. Contemporary Pediatrics describes it as the drug’s only randomized trial. The FDA had granted breakthrough therapy designation in 2016, rescinded it on that trial’s data, and granted it again in 2025. In between, Mallinckrodt went into Chapter 11, and Mandos closed its purchase of adrabetadex on July 1, 2021, promising to keep the expanded access programs running and reanalyze the data.
That reanalysis is what got approved. The FDA counted 36 infantile-onset patients treated across three clinical studies and an expanded access program, and compared them with a matched external control group of children who never got the drug. Median survival was 8.5 years in the treated children and 4.9 years in the controls, and the agency reports a 55 percent reduction in the risk of death, with the benefit holding whether or not patients also took miglustat, an older drug used off-label. The FDA called that comparison “the adequate and well-controlled study supporting approval,” the evidence the law requires.
A different set of numbers had circulated first. Survival results titled “Substantial survival benefit with adrabetadex treatment in individuals with infantile-onset Niemann-Pick disease type C” appeared in Molecular Genetics and Metabolism, and an analysis presented in March 2026 compared 72 treated children with 119 matched controls, reporting a 71 percent lower risk of death and five-year survival of 84 percent against 42 percent. Contemporary Pediatrics notes that analysis differs from the one behind the approval, and no public account explains why the treated group in the FDA’s version is half the size.
The questions that decide whether this survival gap belongs to the drug are still open in public. Which children made up the control group, from which years, matched on what, and starting the survival clock when? Beren hasn’t reported follow-up length or confidence intervals, and Contemporary Pediatrics flags the obvious risk that historical patients differ from treated ones in ways matching can’t fully fix. I started out suspicious of a borrowed control group, and I still am. What shifted me is the 2018 trial itself: a year in which neither arm got worse on a symptom score says almost nothing about which children are alive five years later, so the randomized miss and the survival claim don’t cancel each other out. They measured different things, and only one of them was randomized.
The price Beren hasn’t announced
The children pay part of the cost in hearing. Lembaz carries a boxed warning that hearing loss can be permanent and range from mild to profound, including in some children whose hearing was normal before treatment. According to Contemporary Pediatrics, the label requires baseline audiology, retesting 1 to 2 weeks after the first and second doses at each dose level, then monthly until three stable results, then yearly, and the dose isn’t raised when loss exceeds set limits. Nobody knows whether lowering or continuing the dose changes how much more hearing goes. Ataxia and fatigue in the first days after a dose have put some patients in the hospital.
The rest of the cost is money, and none of the coverage of the approval reports a price. The drug should reach treatment sites in early December, and Pharmacy Times reports a family access program offering out-of-pocket help for eligible patients. With the approval, Beren also received a priority review voucher, NeurologyLive reports, the kind of asset companies sell to one another. CEO Jason Camm thanked the agency for “the close collaboration with the FDA over the last 5 years,” the stretch after Mallinckrodt’s failed trial. The years before that, which produced the delivery route, the first patients and the first controlled human data, belonged to a family pushing for their daughters, an Oakland physician, NIH scientists and charities funded by grieving parents, with Sucampo as the commercial partner along the way.
If my toddler had this diagnosis, I would start the injections and book the audiologist the same week, and I would not sign anything with Beren until it said, on the record, what it plans to charge for a molecule a family first fought to get into their daughters’ veins.
Sources
- FDA: Lembaz (adrabetadex) approval announcement (Nov. 17, 2026)
- Contemporary Pediatrics: FDA approves adrabetadex for infantile-onset Niemann-Pick type C
- NeurologyLive: FDA approves adrabetadex as first therapy to improve survival in infantile-onset NPC
- Molecular Genetics and Metabolism: Substantial survival benefit with adrabetadex treatment in infantile-onset NPC
- NICHD: Experimental treatment for Niemann-Pick disease type C1 appears safe, effective (Aug. 2017)
- BioCentury: Mallinckrodt Niemann-Pick therapy fails pivotal trial (Nov. 2018)
- Patient Worthy: Niemann-Pick type C drug hopeful disappoints (Nov. 2018)
- Business Wire: Mandos completes acquisition of adrabetadex from Mallinckrodt (July 2021)
- Pharmacy Times: Adrabetadex becomes first therapy shown to extend survival in infantile-onset NPC
- Global Genes: Children’s Hospital Oakland receives FDA clearance for cyclodextrin delivery to the twins’ brains
- Nature Reviews Drug Discovery: FDA approves first two drugs for rare Niemann-Pick disease (2024)