Vigabatrin-associated brain abnormalities on MRI (VABAM) in children with epileptic spasms.

backgroundThe aim of this study was to identify risk factors and pattern of brain magnetic resonance imaging (MRI) changes in patients with vigabatrin-associated brain abnormalities on MRI (VABAM).

methodsThis retrospective observational study included patients who were treated with vigabatrin for epileptic spasms (ES). Patients who underwent at least one MRI during vigabatrin treatment were included. For each patient, we defined the MRI abnormalities and their anatomic location.

resultsSeventy-six patients (57.6%) underwent MRI during vigabatrin therapy. Thirty-four out of the 76 patients (44.7%) had VABAM. In 27 out of the 34 cases (79.4%), T2/FLAIR abnormalities were accompanied by restricted diffusion on the diffusion-weighted imaging (DWI), with restricted diffusion most commonly seen in the pons (100%). Several potential risk factors were compared between the MRI-positive and MRI-negative groups. Initiating vigabatrin at a young age (p = 0.028), patient age <24 months (p = 0.031), and a higher peak vigabatrin dose (p = 0.040) were significantly associated with VABAM. Duration of vigabatrin exposure, ES etiology, presence of microcephaly, developmental delays, being underweight, concurrent use of more than three antiseizure medications (ASMs), concurrent use of oral steroids, delayed brain myelination, and occurrence of infection during vigabatrin exposure did not appear to predispose patients to VABAM. MRI reports of patients with VABAM prior to enrollment in our trial were reviewed. 41.2% (14/34) of MRIs were reported as normal, 8.8% (3/34) of MRIs were reported to have findings of uncertain significance, and 29.4% (10/34) MRIs were reported with a differential diagnosis other than VABAM. Only 20.6% (7/34) of MRIs were reported as having VABAM.

conclusionsAvailability of MRI data during vigabatrin exposure, feedback from expert pediatric neuroradiologists knowledgeable about VABAM, and information from well-designed prospective studies could help improve our understanding of the prevalence of VABAM. IMPACT: VABAM are common in children with epileptic spasms, especially in those under 2 years old and on higher doses, but they are often reversible and clinically silent. This study reveals a higher-than-expected prevalence of VABAM (44.7%) and identifies younger age and higher vigabatrin doses as key risk factors. It also uncovers frequent misdiagnosis or under recognition on MRI. The study emphasizes the importance of increased awareness of VABAM among clinicians and radiologists to prevent misdiagnosis and guide safer use of vigabatrin.