CHMP Rejects Arimoclomol Again; Zevra Seeks Re-Examination

EMA's human medicines committee said no to arimoclomol for the second time. At its 20–23 July 2026 meeting, the Committee for Medicinal Products for Human Use (CHMP) adopted a negative opinion on Zevra Therapeutics' Marketing Authorisation Application for the drug in Niemann-Pick disease type C (NPC), a rare and fatal neurodegenerative disorder. The rejection turned on the evidence, not the disease: CHMP's stated concern was insufficient demonstration of efficacy, with uncertainties in data handling and analytical approach undermining confidence in the pivotal trial's results. Zevra says it will fight the opinion — and the drug is already approved and in use an ocean away.

What CHMP said was missing

CHMP's opinions are decisions about evidence, and the committee was specific about where this one fell short: not the existence of a clinical effect, but the reliability of how it was shown. The stated concern was that uncertainties in data handling and analytical approach in the pivotal trial undermined confidence in its results — a methodology objection rather than a judgment that the drug does nothing. For a rare-disease application, where small pivotal populations already strain statistical power, an analytical or data-handling gap can be enough to tip a borderline efficacy case into a negative opinion, independent of what a differently run analysis of the same data might have shown.

A negative opinion is not the end — procedurally

EU centralized-procedure rules give a sponsor a formal way to contest a CHMP opinion before it becomes final: a request for re-examination, filed within 15 days of the opinion, laying out the grounds for reconsideration. Zevra has said it will use that window. If granted, re-examination sends the file back to CHMP (often with a scientific advisory group's input) before a final recommendation goes to the European Commission. It is a real second look, not a formality — but it is also not unusual to invoke, and it does not change what CHMP will be looking for the second time: the same data, argued again.

Same drug, opposite outcome, and it isn't the first time

• Approved in the US: arimoclomol, combined with miglustat and marketed as Miplyffa, was approved by FDA on 20 September 2024 for neurological symptoms of NPC in patients 2 years and older — the first FDA-approved treatment for the disease, with priority review, orphan drug, rare pediatric disease, fast track, and breakthrough therapy designations along the way.
• Rejected in the EU, twice: a first Miplyffa application was withdrawn before a final opinion in March 2022; this CHMP cycle is the second attempt, and it ended in a negative opinion rather than a withdrawal.
• What that divergence usually turns on: FDA and EMA can and do reach different conclusions from overlapping evidence packages when their reviewers weigh trial design, endpoints, or statistical handling differently — NPC's small patient population makes any single pivotal trial's analytical choices unusually consequential in either review.

The same 20–23 July meeting that rejected arimoclomol also recommended a dozen new medicines for approval, including the EU's first refillable eye implant. Reading the two stories side by side is the more complete picture of what the committee does in a single sitting: it is not only approving, and a rare-disease sponsor with a drug already on the market in another jurisdiction is not guaranteed a second one just because the first one worked.