The US Food and Drug Administration approved an expanded use of Wellcovorin, or leucovorin calcium, on March 10, 2026, for adults and children with a rare genetic disorder that prevents enough folate from reaching the brain.
The indication is narrow: patients must have cerebral folate transport deficiency with a confirmed variant in the FOLR1 gene, abbreviated FOLR1-CFTD. The decision does not authorize leucovorin as a treatment for autism or for every condition described broadly as cerebral folate deficiency.
No clinical trial tested leucovorin's efficacy and safety in people with FOLR1 variants. The FDA instead conducted a structured review of published patient cases, compared their reported course with the disorder's natural history and used biological evidence about folate restoration to support the clinical findings.
The Approval Covers a Genetically Defined Disorder
FOLR1-CFTD is an ultra-rare neurological syndrome. Variants in the folate receptor 1 gene disrupt transport of folate into the central nervous system. Patients can develop severe developmental delay, seizures, movement disorders and other neurological complications as the untreated disease progresses.
Leucovorin is converted to 5-methyltetrahydrofolate, or 5-MTHF, an active form of folate. It has long been used for other approved purposes, including reducing the toxicity of folate-antagonist medicines such as methotrexate. The 2026 action added the genetically confirmed transport disorder to Wellcovorin's label through a supplemental application under section 505(b)(2) of federal drug law.
The distinction between the product and the molecule also matters. FDA approval went to GSK as the holder of the original Wellcovorin application, but the current label says GSK does not manufacture or market its 5-milligram or 25-milligram tablets. STAT reported that the company did not intend to restart production. An approved label therefore does not by itself establish product availability, price or insurance coverage.
Forty-Six Published Cases Formed the Clinical Record
The FDA identified 46 unique patients with genetically confirmed FOLR1 variants, leucovorin exposure and available outcome information in 26 case reports and case reviews published through 2024. Thirty of those patients appeared in more than one publication, so reviewers de-duplicated the literature rather than counting each paper as a separate person.
Patient-level information was extracted for seizure control, motor function, development and behavioral features. The review categorized outcomes as complete recovery, substantial improvement, partial improvement or no improvement. FDA staff reported some clinical benefit in 40 of 46 patients, including substantial improvement or complete recovery in 30.
Those percentages come from retrospective publications, not randomized groups. The comparison was an estimated untreated trajectory derived from descriptions of the disease before therapy and from patients who did not receive leucovorin. Reviewers argued that FOLR1-CFTD is relentlessly progressive and that the magnitude and consistency of the reported changes were unlikely to reflect spontaneous improvement.
The approved label presents a more focused oral-treatment subset. Twenty-seven patients were reported to have received leucovorin only by mouth. Twenty-four had improvement in neurological symptoms, no change or no progression. Treatment duration was limited, dosing information was incomplete and measurements were not collected on a uniform schedule.
Mechanistic Evidence Supported but Did Not Replace the Cases
FDA reviewers also examined 5-MTHF in cerebrospinal fluid, a direct measure connected to the disorder's folate-transport pathway. Among seven patients in the oral-only group with measurements both before and after treatment, all had increased 5-MTHF and five reached a level above 40 nanomoles per liter.
In the broader review, 15 patients had post-treatment measurements suitable for analysis; 12 reached levels above that threshold. The agency considered the biomarker pattern confirmatory evidence that leucovorin was affecting the pathway responsible for the disease. It was not treated as proof that every neurological symptom would improve in every patient.
The safety assessment combined published experience in the 46 patients with decades of information from leucovorin's other uses. No deaths or treatment-related serious adverse events were documented in the reviewed FOLR1-CFTD cases, but systematic adverse-event reporting appeared in only three cases. That limitation makes absence of a reported problem weaker than active safety monitoring in a prospective trial.
The label warns about hypersensitivity reactions, including itching, rash, hives, breathing difficulty, chills, impaired temperature regulation and potentially fatal anaphylaxis. The approval is not a reason to self-treat, assume folate-related symptoms establish the genetic disorder or use the medicine for an unapproved condition without clinical evaluation.
The Evidence Exception Must Remain Auditable
This decision is neither a conventional randomized trial approval nor a regulatory rubber stamp. The FDA documented a specific evidentiary argument: an ultra-rare, progressive, genetically defined disease; large reported changes across de-duplicated cases; a natural-history comparison; and a biomarker tied to the known mechanism.
That argument can be examined. Retrospective case reports are vulnerable to selective publication, inconsistent outcome recording, missing safety data and differences in treatment timing. Natural-history controls do not balance unknown confounders the way randomization can. The FDA concluded that the observed effect was large enough to overcome those weaknesses, but the weaknesses did not disappear.
The defensible precedent is therefore narrower than “old drugs no longer need trials.” It is that literature-based evidence may support approval when the population is exceptionally small, the disorder is well defined, the untreated course is severe and predictable, and the clinical changes align with strong mechanistic evidence.
Regulatory credibility depends on enforcing that entire test, not just invoking rarity or compassion. Here, the public review identifies the patients, the response categories, the missing data and the biological measurements. Future sponsors should have to meet the same level of specificity. Access without transparency lowers the standard; access built on an auditable evidentiary case does not.