Skip to content
clinicaltrialshub.com
Clinical Trials Hub

Clinical trials, drug development and research news

Rare Diseases

Spinal injection cut death risk by 55% in severe Niemann-Pick C

The FDA approved Lembaz (adrabetadex) on 17 November 2026 for infantile-onset Niemann-Pick C, after 36 treated patients had a 55% lower risk of death than matched untreated controls.

The East Tower of Rush University Medical Center in Chicago, where the expanded access programme's lead investigator is based.
Photo: Nathan Hicks, via Wikimedia Commons · CC BY 3.0
Contents
  1. What is Niemann-Pick disease type C?
  2. How the survival analysis was done
  3. What the survival results showed
  4. A worked example with the numbers
  5. The earlier trial that did not meet its goal
  6. What about side effects?
  7. The case for caution
  8. Can I get this drug in the UK?
  9. What to ask if this disease is in your family

A twice-monthly injection into the spinal fluid cut the risk of death by 55% in people with the severe, early form of Niemann-Pick disease type C (NPC), according to the analysis that led the US Food and Drug Administration to approve Lembaz (adrabetadex) on 17 November 2026.

The approval covers adults and children who weigh at least 7 kg, about 15 pounds, and whose neurological symptoms began before the age of six. Doctors call that group infantile-onset NPC. The drug is given as an intrathecal injection, meaning a trained clinician injects it into the fluid around the spinal cord, every two weeks.

Median survival from the start of the comparison was 8.5 years in 36 treated patients and 4.9 years in a matched group of untreated patients drawn from a natural history study. The FDA said this was the first treatment shown to help people with this form of the disease live longer. Two other NPC medicines were already approved in the United States, but neither has been shown to reduce the risk of death, and both have age or weight limits that leave some of the youngest patients out.

The evidence is not a conventional randomised survival trial. It is a comparison of treated patients against matched external controls, which the FDA judged adequate to support approval. That design needs careful reading, and so does an earlier sham-controlled trial of the same drug that did not meet its neurological goals.

What is Niemann-Pick disease type C?

NPC is a rare inherited lysosomal storage disorder. Faulty NPC1 or NPC2 proteins mean cells cannot move cholesterol and other fats to where they are needed. The fats build up, especially in the brain. Over time, people lose movement, coordination and the ability to swallow.

The FDA estimates that NPC affects about 1 in 150,000 people. Beren Therapeutics, the public benefit corporation behind Lembaz, says about half of those living with NPC in the United States, some 475 children, have the infantile-onset form. The company also gives mean ages at death of about 5.6 years when neurological signs start before the age of two, and about 13.4 years when they start between two and six. Those figures come from the company, not from the approval analysis.

Because lost neurological function is not regained, doctors and families have pressed for earlier diagnosis and for a treatment that can be started in very young children. Lembaz is a formulation of 2-hydroxypropyl-beta-cyclodextrin. The prescribing information says adrabetadex binds cholesterol and helps move it out of the lysosome, but it also says the way the drug reduces the risk of death "has not been fully characterized." The drug is injected into the spinal fluid, not given as a tablet, so that it can reach the brain and nervous system.

The FDA granted approval to Mandos, LLC, a subsidiary of Beren Therapeutics, which is based in Thousand Oaks, California. Lembaz received Breakthrough Therapy designation, and Beren received a Priority Review Voucher. The company said commercial treatment sites in the United States should start to open in early December 2026.

How the survival analysis was done

The effectiveness claim rests on an integrated survival analysis described in section 14 of the label. It included only patients whose neurological signs began before the age of six. The safety and effectiveness of Lembaz for reducing deaths in people whose neurological signs began at six or older have not been established, and that group is outside the approved use.

Thirty-six treated patients were drawn from three clinical studies and an expanded access programme:

  • Trial 1 had a 52-week randomised, double-blind, sham-controlled period in patients aged 4 to 21 whose neurological signs began before 15 (NCT02534844), plus an open-label extension (NCT04958642).
  • Trial 2 (NCT01747135) was a 78-week open-label, single-centre dose-escalation study in 14 patients aged 4 to 23.
  • Trial 3 (NCT03643562) was an open-label study in 27 patients aged 4 to 43.
  • The expanded access programme included 85 patients aged 1 to 43, treated for a mean of 44 months.

Of the 36 patients in the survival analysis, 11 first started Lembaz in Trial 1, two in Trial 2, two in Trial 3 and 21 in the expanded access programme. Elizabeth M. Berry-Kravis of Rush University Medical Center in Chicago was the lead investigator for the expanded access programme.

They were matched to untreated patients from Study 1, a prospective observational natural history study. Forty-four people in that study whose neurological signs began before the age of six were eligible. Matching used disease subtype (early versus late infantile-onset), age at neurological symptom onset within three months, and whether the patient was already taking miglustat, a tablet often used off-label in NPC. A control patient also had to be alive at the "index age", the age at which their matched treated patient started Lembaz.

The analysis used many-to-one matching. Twenty-six unique control patients were weighted so that the control group counted as 36, the same as the treated group. Survival was measured from the index age to death or last follow-up. All of the studies allowed people to take miglustat at the same time.

In both groups the mean index age was 5.7 years (range 1 to 19). Most patients were White (94%). Sixty-seven percent had early infantile-onset disease, and 50% were on miglustat at the start. Sex was not part of the matching: 58% of treated patients were male, compared with 33% of the controls. Median follow-up from the index age was 5.6 years in the treated group and 4.2 years in the controls. Treated patients received Lembaz for a median of 3.6 years, at a median final dose of 750 mg (the maximum recommended dose is 900 mg every two weeks).

What the survival results showed

Treated patients lived longer than the matched controls. The hazard ratio for death was 0.45 (95% confidence interval 0.21 to 0.95), which the FDA described as a 55% reduction in the risk of death. The label says this longer survival was seen whether or not patients were already taking miglustat.

Deaths were 12 of 36 treated patients (33.3%) and a weighted 24 deaths in the control group, reported as 66.9%. Median survival from the index age was 102 months with Lembaz and 59 months in the controls, or 8.5 years versus 4.9 years. Those medians were estimated with a weighted Kaplan-Meier method, truncated at 110 months, the longest follow-up in the treated group. The hazard ratio came from a weighted Cox model that included treatment and baseline miglustat use.

Acting FDA Commissioner Kyle Diamantas said: "For the first time, there is a treatment shown to help adults and children with this serious disease live longer, and we are committed to continuing to support the development of therapies for the rare disease community."

A worked example with the numbers

It helps to separate the relative result from the absolute one. Our guide to reading trial results covers this distinction in more detail.

The hazard ratio of 0.45 means that, at any point in the follow-up, the estimated instantaneous risk of death in the treated group was 45% of the risk in the matched controls, or 55% lower. That is a relative figure. In absolute terms, 33.3% of the 36 treated patients had died by the end of follow-up, against a weighted 66.9% of the controls. Subtracting those two percentages gives a difference of 33.6 percentage points. In a group of 36 people, that is about 12 fewer deaths than in the weighted control group, which is another way of looking at the 12 versus 24 deaths already reported. This arithmetic uses the label's rounded percentages and is a check on them, not a new trial result.

The median gap is 102 minus 59 months, which is 43 months, or 3 years and 7 months of extra median survival from the index age (43 divided by 12 is 3.58 years; the FDA's 8.5 minus 4.9 years is 3.6 years, the same gap after rounding).

The confidence interval still matters. The upper bound of 0.95 is close to 1, which would mean almost no difference in risk. The result crosses the usual threshold for statistical significance, but the interval is wide because the treated sample is only 36 people. A modest true effect, or a larger one, would both fit the data.

The earlier trial that did not meet its goal

Trial 1 is the same programme that, in a broader group of patients, did not show a neurological benefit against sham treatment.

The UK Health Research Authority's public summary of the study says that on 20 January 2021 the then sponsor announced there was no clear evidence that adrabetadex was beneficial in NPC1, and that the risk was higher than the chance of benefit. The sponsor stopped the trial. In Parts A and B, the summary says, adrabetadex was no better than sham on the NPC severity scale (NPC-SS) and a clinician global impression of change (CGIC) after 52 weeks. The reason for the lack of difference was not known.

That randomised period enrolled treatment-naive patients aged 4 to 21 whose neurological signs had begun before the age of 15, a wider group than the later approval. The FDA label lists 56 patients in Parts A and B. The HRA summary gives 58, a small discrepancy we cannot resolve from the two documents. In the label's safety subset of patients who received up to 900 mg, only 56% had neurological onset before the age of six.

So the approval does not rest on that trial having met its original neurological endpoints. It rests on a later survival comparison, in a narrower infantile-onset group, against matched people who were not in the same randomised trial. Our guide to what it means when a clinical trial fails is a useful companion to that history. The randomised portion was blinded and sham-controlled; much of the survival data, including the expanded access programme, was open-label.

What about side effects?

Lembaz carries a boxed warning for hearing impairment, including permanent hearing loss. The loss has ranged from mild to profound, and it has occurred in some people whose hearing was normal before treatment. Higher frequencies (2,000 to 8,000 Hz) were affected most often, but mid and lower frequencies were affected too. The label says there are not enough data to know whether staying on the same dose increases the risk of further loss, or whether lowering the dose reduces that risk.

Before treatment, patients need a full hearing evaluation, including tests such as pure tone audiometry, speech recognition and tympanometry. Hearing is rechecked 1 to 2 weeks after the first dose at each dose level and again after the second dose, then more often until the results are stable, and at least every 12 months once a stable dose is reached. The dose is not increased if hearing drops beyond set limits.

The main safety table comes from Trial 1 Parts A and B, in 32 patients who received up to 900 mg every two weeks and 18 who received sham treatment, followed for 52 weeks. This group is broader than the approved infantile-onset indication. Adverse reactions reported in at least 25% of Lembaz patients, and more often than on sham, were hearing impairment (71.9% on Lembaz, 44.4% on sham), ataxia including problems with balance, gait, falls or coordination (68.8% versus 27.8%), fatigue (56.3% versus 27.8%), vomiting (53.1% versus 16.7%), back pain (50.0% versus 16.7%), headache (43.8% versus 16.7%), fever (40.6% versus 16.7%), diarrhoea (31.3% versus 5.6%), nausea (25.0% versus 5.6%) and cough (25.0% versus 22.2%).

Most ataxia events (61%) and almost all fatigue events (91%) started within three days of a dose. The median duration of each was two days, though ataxia lasted as long as 476 days in at least one case and fatigue as long as 76 days. One patient on 900 mg needed hospital care for severe ataxia, and one needed hospital care for severe fatigue on the day of treatment. In studies that titrated the dose with hearing tests, 37% (22 of 59) of patients had hearing impairment.

Aspiration pneumonia occurred in 9% (3 of 32) on Lembaz and 6% (1 of 18) on sham in Parts A and B. In the open-label extension, one patient had a fatal aspiration pneumonia after receiving Lembaz with sedation. The label says it is unclear whether that event was due to the drug, the sedation, the underlying disease, or some mix of those.

The recommended starting dose is 100 mg every two weeks under the age of six months, 200 mg from six months to under four years, and 400 mg from four years of age, in people who weigh at least 7 kg. After at least two doses at each level, the dose can be raised in 100 mg to 200 mg steps, up to 600 mg in people who weigh 7 kg to under 12 kg, or 900 mg in people who weigh 12 kg or more. The injection is given over one to two minutes by, or under the supervision of, a clinician experienced in lumbar puncture.

The case for caution

Several limits sit alongside the survival result.

The comparison is not randomised. Matching tried to balance subtype, age at onset and miglustat use, and it required controls to be alive at the index age, which reduces one obvious source of bias. It did not balance sex, and the male share differs by 25 percentage points. Other unmeasured differences between people who got into trials or expanded access and people followed in a natural history study could still remain. The control group is 26 people, stretched by weighting to match 36 treated patients.

The treated sample is small, and the confidence interval around the hazard ratio is wide. The FDA's 55% figure is a point estimate, not a precise count of how much longer every child will live.

The original randomised trial in a broader NPC group did not show a neurological benefit on its planned scales. The approval does not fill that gap. Families should not read the new survival analysis as proof that the earlier neurological scores were wrong, or that lost skills will come back. The label does not report a randomised effect on those scales in the infantile-onset subgroup either.

Hearing loss is common and can be permanent, including in people who started with normal hearing. Fortnightly lumbar punctures, often for years, are also a substantial burden, especially for young children.

Lembaz has not been shown to reduce deaths in people whose neurological symptoms began at six years of age or later.

Can I get this drug in the UK?

As of 18 November 2026, we found no MHRA marketing authorisation and no NICE appraisal for Lembaz or adrabetadex. The United States approval does not by itself make the drug available on the NHS. UK centres did take part in the earlier VTS-270 trial, according to the Health Research Authority summary, but that programme was stopped in 2021.

Two other NPC medicines already have a foothold in this disease, with tighter limits than Lembaz. A 2025 therapeutics bulletin from the American College of Medical Genetics and Genomics describes them as arimoclomol (sold in the United States as Miplyffa) used with miglustat in people aged two years and older, and levacetylleucine (Aqneursa) for people who weigh at least 15 kg. Both were approved by the FDA in September 2024. The FDA's Lembaz announcement says that neither of the two earlier approved NPC therapies has been shown to reduce the risk of death.

In Europe the picture is mixed. The EMA has authorised Aqneursa for neurological symptoms of NPC in people aged six years and older who weigh at least 20 kg, either with miglustat or alone if miglustat cannot be used. At its 9 to 12 November 2026 meeting, the EMA's CHMP confirmed its refusal of Meplyffa (arimoclomol citrate) for NPC in people aged two years and older, after a re-examination. NICE's appraisal of arimoclomol for NPC remains listed as in development, without a published recommendation.

Our guide to how new medicines get approved explains the usual FDA, MHRA and NICE steps. Other recent rare neurological programmes we have covered include AMT-130 in Huntington's disease, ulefnersen in FUS-ALS and deramiocel in Duchenne muscular dystrophy.

What to ask if this disease is in your family

If you care for someone with NPC, these questions can help a specialist put the new approval in context:

  • Did neurological symptoms begin before the age of six, and does my child weigh at least 7 kg?
  • Is this person already on miglustat, Miplyffa or Aqneursa, and how would Lembaz fit with those medicines?
  • What is their current hearing, and how would we monitor it if treatment started?
  • How close is the nearest centre that can give a fortnightly intrathecal injection, and what does a lumbar puncture involve at this age?
  • If we are in the UK, is there a named-patient, trial or early-access route, or only a wait for an MHRA and NICE review?

You can search for NPC studies using our guide to finding and joining a clinical trial.

This article reports an FDA approval and trial results for general information. It is not medical advice, and treatment decisions should be made with a specialist who knows the individual patient.

Frequently asked questions

Who can receive Lembaz in the United States?

The FDA approved Lembaz to reduce the risk of death in adults and children who weigh at least 7 kg and have Niemann-Pick disease type C with neurological symptoms that began before the age of six. That group is called infantile-onset NPC. The drug has not been shown to reduce deaths in people whose neurological symptoms began at six or older, and that use is not approved.

How is Lembaz given?

A clinician experienced in lumbar puncture injects it into the spinal fluid over one to two minutes, every two weeks. The starting dose depends on age (100, 200 or 400 mg), and the dose can later be raised, as tolerated, up to 600 mg or 900 mg depending on weight. Hearing is tested before treatment and at set points after each dose increase.

Did a randomised trial prove that Lembaz extends survival?

No. The approval is based on 36 treated patients from several studies and an expanded access programme, compared with matched untreated patients from a natural history study. An earlier randomised, sham-controlled trial in a broader NPC group did not show a benefit on planned neurological scores after 52 weeks, and that study was stopped in 2021.

What are the main side effects of Lembaz?

The label carries a boxed warning for hearing impairment, including permanent hearing loss, which has occurred even in people with normal hearing at the start. In the sham-controlled trial, hearing problems, ataxia, fatigue, vomiting, back pain, headache, fever, diarrhoea, nausea and cough were all common. Ataxia and fatigue often begin within three days of a dose.

Is Lembaz available on the NHS?

Not as of 18 November 2026. We found no MHRA licence and no NICE recommendation for Lembaz or adrabetadex. The US approval does not automatically apply in the UK. Families in the UK should ask a specialist whether any trial, named-patient or early-access option exists, and watch for a future MHRA and NICE review.

Sources

  1. FDA Approves First Drug Shown to Help Patients with Infantile-Onset Niemann-Pick Disease Type C Live Longer. US Food and Drug Administration, 2026-11-17
  2. LEMBAZ (adrabetadex) injection, for intrathecal use, Prescribing Information. US Food and Drug Administration / Mandos, LLC, 2026-11
  3. Beren Therapeutics Announces U.S. FDA Approval of LEMBAZ (adrabetadex), First & Only Medicine to Improve Survival in Infantile-Onset Niemann-Pick Disease, Type C. Beren Therapeutics, 2026-11-17
  4. A Phase 2b/3 Trial of VTS270 in Subjects with Niemann-Pick Type C1 Disease (NCT02534844 / NCT04958642). ClinicalTrials.gov, 2026-11
  5. VTS270 in Subjects with Niemann-Pick Type C1 Disease (trial summary). UK Health Research Authority, 2022
  6. Adrabetadex in Niemann-Pick Type C, Trial 2 (NCT01747135). ClinicalTrials.gov, 2026-11
  7. Adrabetadex in Niemann-Pick Type C, Trial 3 (NCT03643562). ClinicalTrials.gov, 2026-11
  8. Arimoclomol and levacetylleucine for the treatment of neurologic manifestations of Niemann-Pick disease type C: A therapeutics bulletin of the ACMG. Genetics in Medicine Open, 2025-12-23
  9. Meeting highlights from the CHMP, 9-12 November 2026. European Medicines Agency, 2026-11-13
  10. Aqneursa (levacetylleucine) EPAR. European Medicines Agency, 2026
  11. Meplyffa (arimoclomol) EPAR. European Medicines Agency, 2026-11
  12. Arimoclomol for treating Niemann-Pick disease Type C [ID1312]. NICE, 2026-11
  13. FDA Approves Adrabetadex as First Therapy Shown to Improve Survival in Infantile-Onset Niemann-Pick Disease Type C. Neurology Live, 2026-11-17

Medical disclaimer: For information only, not medical advice. Speak to a qualified healthcare professional about your care.

The Briefing

Trial results that matter, every Friday.

One email. The week's readouts, approvals and research, explained without hype.

Related reading