← Back
NeuroTrials.ai
Neurology Clinical Trial Database

T3CAD

Tocotrienol-rich vitamin E complex in CADASIL (T3CAD): a randomised, double-blind, placebo-controlled trial

Year of Publication: 2026

Authors: Chabriat H, Biard L, Guey S, ..., Hervé D

Journal: eClinicalMedicine

Citation: eClinicalMedicine 2026;94:103853. Published Online 30 March 2026.

Link: https://doi.org/10.1016/j.eclinm.2026.103853


Clinical Question

Does a palm-derived tocotrienol-rich vitamin E complex (HOV12020) reduce clinical progression (stroke, disability, or cognitive decline) over 24 months in patients with CADASIL?

Bottom Line

In patients with symptomatic CADASIL, 24 months of tocotrienol-rich vitamin E (HOV12020) did not reduce stroke, disability, or cognitive decline; Bayesian analysis confirmed futility. The compound is unlikely to confer benefit in advanced CADASIL, though it was very well tolerated.

Major Points

  • First randomised, placebo-controlled preventive intervention trial in CADASIL
  • Treatment failure (composite of stroke, disability progression, or cognitive decline) occurred in 68.0% on HOV12020 vs 61.5% on placebo over 24 months
  • Bayesian analysis met predefined futility criteria — low posterior probability of treatment benefit
  • Placebo event rate (61.5%) far exceeded the 40% initially expected, indicating rapid progression in symptomatic CADASIL
  • No significant differences in secondary clinical endpoints or exploratory MRI markers
  • Excellent safety and tolerability — no excess adverse events vs placebo
  • Demonstrates feasibility of Bayesian small-sample designs in rare, slowly progressive cerebrovascular diseases
  • Findings suggest future trials should target earlier disease stages in CADASIL

Design

Study Type: Single-centre, randomised, double-blind, placebo-controlled, pilot/phase 2 trial with Bayesian analysis framework

Randomization: 1

Blinding: Double-blind — participants, investigators, outcome assessors, and data analysts blinded; placebo identical in appearance and administration schedule to active treatment

Allocation: 1:1 permuted block randomisation (undisclosed block size, no stratification), centrally generated by independent statistician using SAS v9.4; allocation concealment via central CRO implementation

Enrollment Period: 15 January 2021 to 11 February 2022

Follow-up Duration: 24 months

Centers: 1

Countries: France

Sample Size: 51

Analyzed: 51

Analysis: Bayesian framework for primary endpoint to estimate posterior probability of treatment benefit; predefined futility criteria; quantile regression with Wilcoxon rank-sum sensitivity tests for exploratory MRI endpoints

Power Calculation: Designed assuming 40% placebo treatment-failure event rate; historical French CADASIL cohort data (2013–2016) used to define eligibility and characterize short-term clinical progression

Registration: EudraCT 2019-002867-23; ClinicalTrials.gov NCT04658823


Inclusion Criteria

  • Age 45–75 years inclusive at time of informed consent
  • Confirmed CADASIL diagnosis: pathogenic NOTCH3 mutation with odd number of cysteine residues OR positive skin biopsy showing granular osmiophilic material on electron microscopy
  • Presence of at least one prevalent lacune on MRI (3D T1 or FLAIR)
  • Confluent white matter hyperintensities (Fazekas grade 2 or 3) on T2-weighted or FLAIR MRI
  • Mini-Mental State Examination (MMSE) score ≥15
  • Modified Rankin Scale score 0–3
  • Women of childbearing potential: not pregnant, not breastfeeding, willing to use contraception during treatment and for 28 days thereafter
  • Ability to provide signed informed consent

Exclusion Criteria

  • Stroke with persistent neurological deficit within 6 months before screening
  • Other neurodegenerative disorders (Parkinson's disease, Alzheimer's disease)
  • Uncontrolled seizures or untreated hypertension
  • Significant hematologic, cardiac, pulmonary, metabolic, neurological, or psychiatric disorder
  • Recent major depression (within 6 months); history of suicide attempt within 6 months; positive C-SSRS item 4 or 5 at screening/baseline
  • Myocardial infarction within 3 months, active angina, symptomatic heart failure
  • Hemophilia or bleeding-risk disorder; indication for anticoagulant therapy
  • Cancer within 5 years (except basal cell, squamous cell, or stage 1 prostate)
  • Drug or alcohol abuse/dependence; HIV, hepatitis B or C
  • Severe allergy/intolerance to vitamin E (tocopherols or tocotrienols), sensitivity to polyoxyl castor oil, soybean or peanut allergy
  • Contraindication to MRI
  • ALT, AST, amylase, or lipase >2× ULN; total bilirubin >1.5× ULN; creatinine clearance <60 mL/min
  • Tocotrienol supplementation within 3 months before screening; vitamin E other than study drug; unstable cognitive enhancer (e.g., donepezil)
  • Participation in another interventional trial within 30 days before screening
  • Pregnant or breastfeeding women

Baseline Characteristics

Total N: 51

Median Age: 58 years (range 43–73)

Male: n=24 (47%)

Confirmation: Genetically confirmed CADASIL (NOTCH3 mutation or positive skin biopsy)


Arms

FieldHOV12020 (tocotrienol-rich vitamin E complex)Control
N2526
InterventionPalm-derived tocotrienol-rich vitamin E complex — one capsule containing 285 mg total tocotrienols and tocopherols twice daily (after morning and evening meals), total daily dose 570 mgVitamin-E–stripped soybean oil capsules with only trace/undetectable isoflavone and no biologically relevant activity, identical appearance and administration schedule
Duration24 months24 months

Outcomes

OutcomeTypeControlInterventionHR / OR / RRP-value
Treatment failure within 24 months, defined as occurrence of at least one of: (1) centrally adjudicated stroke event; (2) increase in modified Rankin Scale ≥1 point to a score ≥2 attributable to CADASIL; or (3) cognitive decline defined as ≥4-point increase in VADAS-Cog scorePrimary61.5% (placebo)68.0% (HOV12020)Bayesian analysis showed low posterior probability of benefit, meeting predefined futility criteria
Time to first failure eventSecondaryNo significant difference between groups
Time to first stroke eventSecondaryNo significant difference reported
Time to first increase in disability (mRS)SecondaryNo significant difference reported
Time to first cognitive decline (VADAS-Cog)SecondaryNo significant difference reported
Longitudinal change in clinical scores: CDR-SB, MDRS, MDRS-I/P subscale, TMT-A and TMT-B times, SPPB scoreSecondaryNo significant differences observed
Patient-reported outcomes: Short Form 36 and Disability Assessment for DementiaSecondaryNo significant differences observed
Exploratory MRI markers of disease progressionSecondaryNo significant differences observed on quantile regression with Wilcoxon rank-sum sensitivity testing
Number and type of serious adverse events and total AEs over 24 monthsSafetyExcellent safety profile; no excess of adverse events with HOV12020 vs placebo; well tolerated
OverallAdverseHOV12020 was well tolerated with an excellent safety profile and no excess of adverse events compared to placebo

Criticisms

  • Small single-centre sample (n=51) limits power for definitive efficacy conclusions outside the Bayesian futility framework
  • Placebo event rate (61.5%) far exceeded the 40% expected, suggesting enrollment of more advanced disease than anticipated — limits ability to detect benefit
  • Authors acknowledge findings apply to advanced-stage CADASIL; earlier disease stages may still respond
  • Single-centre French cohort may limit generalisability
  • 24-month duration may be insufficient for a slowly progressive disease

Funding

Hovid Berhad (study sponsor)

Based on: T3CAD (eClinicalMedicine, 2026)

Authors: Chabriat H, Biard L, Guey S, ..., Hervé D

Citation: eClinicalMedicine 2026;94:103853. Published Online 30 March 2026.

Content summarized and formatted by NeuroTrials.ai.