How a Modified Leucine Derivative Unlocked Ataxia-Telangiectasia’s First FDA Clearance
IntraBio’s Aqneursa cleared the FDA as the first treatment for ataxia-telangiectasia on a -1.88 SARA point crossover win, treating downstream neurodegeneration rather than an intractable 305 kDa kinase defect.
On September 18, 2026, the FDA approved Aqneursa (levacetylleucine) oral suspension for ataxia in adult and pediatric patients with ataxia-telangiectasia who weigh at least 15 kg, about 33 pounds. IntraBio Inc. received Priority Review and Orphan Drug designation. Until that Friday, A-T had zero approved treatments worldwide. Roughly 1 in 40,000 people inherit this autosomal-recessive disease from biallelic ATM mutations. They lose cerebellar Purkinje cells, then gait, balance, speech, and fine motor control, while immunodeficiency and hematologic cancers close in. The new drug does not restore ATM. It still shifted a 40-point ataxia scale by nearly two points in 12 weeks.
A crossover that made the signal hard to dismiss
Phase 3 IB1001-303 (NCT06673056) was a multinational, randomized, double-blind, placebo-controlled crossover at 10 centers in the United States, United Kingdom, Germany, Slovakia, Spain, and Switzerland. The Lancet Neurology published the July 2026 analysis. Seventy-three patients enrolled, 26 adults and 47 children, median age 13, range 4 to 50. Each received 12 weeks of Aqneursa then 12 weeks of placebo, or the reverse. Seventy patients, 96 percent, finished both periods. Dosing was 4 g per day for patients at or above 35 kg, split 2 g morning, 1 g afternoon, and 1 g evening, or weight-tiered near 0.1 g/kg/day.
The primary endpoint was the Scale for the Assessment and Rating of Ataxia, scored 0 to 40. A linear mixed model found a treatment effect of -1.88 points (95% CI -2.70 to -1.06; p < 0.0001). Mean change was -1.92 on Aqneursa versus -0.14 on placebo. The clinically meaningful threshold on SARA is 1.0 to 1.5 points, so the point estimate cleared that bar. Sequence 1 patients who switched from drug to placebo worsened by 1.81 SARA points in period 2, drifting back toward baseline. That washout argued for a true, reversible symptomatic effect rather than a lucky trajectory.
FDA’s primary measure was Functional SARA, a 0 to 16 scale covering gait, sitting, stance, and speech. The treatment difference was -0.57 points (95% CI -0.92 to -0.23; p = 0.001), with mean change of -0.65 on drug versus -0.10 on placebo. ICARS improved by -2.84 points (p = 0.003). Investigator CGI-I moved as well (p = 0.02). In exit interviews, 81 percent of respondents reported symptom improvement and 66 percent reported meaningful daily-life change. Safety was quiet: no drug-related serious adverse events and no treatment-emergent deaths. Falls (4 percent), skin laceration (3 percent), and urinary tract infection led the list. Animal data flagged potential fetal harm, so pregnancy is contraindicated. N-acetyl-D-leucine competes for monocarboxylate transporter uptake and should not be used at the same time. Aqneursa inhibits P-gp, so monitoring for related adverse reactions with P-gp substrates is required.
Why ATM itself was never the drug target
ATM is a 305 kDa serine/threonine kinase that polices DNA damage and mitochondrial-lysosomal homeostasis. That mass has blocked AAV gene replacement. IntraBio left the gene alone. Levacetylleucine is the pure L-enantiomer of N-acetyl-leucine, distinct from the racemic Tanganil mixture sold for vertigo in France since 1957. It crosses the blood-brain barrier on monocarboxylate transporters. Its exact molecular target remains unknown. Downstream, it appears to restore ATP production, normalize lysosomal trafficking through TFEB, reset membrane potential in hyperpolarized or depolarized cerebellar neurons, and dampen neuroinflammation. The company had already won a September 2024 clearance for neurological manifestations of Niemann-Pick type C, another lysosomal storage disease. A pivotal Phase 3 study, IB1001-304, is underway in CACNA1A-related disorders. The A-T program is the same wager: treat metabolic collapse below a genetic lesion that chemistry cannot repair.
Ataxia moved. Cancer risk did not.
That split is the limit of the label. Aqneursa is indicated for ataxia, not for ATM’s tumor-suppressor job. Patients still face lymphomas, leukemias, and immune failure. A 1.88-point SARA gain can change gait and speech in a clinic visit. It does not rewrite DNA repair. Families will feel the difference in daily function while oncology and immunology remain untouched. Counseling, pricing, and payer reviews will have to treat an oral suspension as a functional therapy, not a cure for the kinase defect that defines the disease.
An oral first, and a neuroprotection question still open
For a condition that waited decades for any labeled option, a weight-tiered oral suspension is a practical win. The crossover design, the washout reversal, and a primary effect that beat the 1.0-point SARA threshold give clinicians something they can measure. Durability is the next test. If the mitochondrial-lysosomal axis is only propped up while drug is on board, Purkinje cells may keep dying once the bottle stops. The ongoing extension study has to show whether 12-week functional gains hold, deepen, or stall. Until those data land, IntraBio has the first A-T therapy and a mechanism that still refuses to name its target.