Development of vigabatrin-induced lesions in the rat brain studied by magnetic resonance imaging, histology, and immunocytochemistry.

Abstract

Vigabatrin, the gamma-aminobutyric acid transaminase (GABA-T)-inhibiting anticonvulsant drug, was given orally at a dose of 275 mg/kg/day to rats (n = 6) in their feed for a period of 12 weeks, during which T2-weighted magnetic resonance images (MRIs) and diffusion-weighted MRIs (DWIs) were collected at weeks 1, 3, 6, 9, and 12. Half the rats (n = 3; and half their age-matched littermate controls; n = 3) were then killed for histopathological confirmation of the observed VGB-induced cerebellar and cortical white-matter lesions. VGB was removed from the diet and additional MRIs of the remaining rats taken at weeks 14, 17, 20, and 24, at which time they (n = 3), along with remaining controls (n = 3), were also killed for histopathology. The T2-weighted MRIs acquired were used to compute T2 relaxation time maps. Statistically significant VGB-induced T2 increases were observed in the frontal and occipital cortices and in the cerebellar white matter (CWM). The cerebellar lesions were more clearly discerned by eye in the DWIs than by T2-contrast alone. During the recovery period the VGB-treatment group CWM-T2 and CWM-DWI hyperintensity greatly decreased as the reversible lesion disappeared. As expected, histological and immunocytochemical examinations demonstrated the presence of intra-myelinic edema, microvacuolation, and reactive astrocytosis in the CWM and cortex after 12 weeks VGB-treatment. In the remaining animals microvacuolation of the white matter had not completely resolved during the 12-week recovery phase. The data show that quantitative MRI T2-relaxometry can be used to detect VGB-induced CNS pathology, and also suggest that DWI is particularly sensitive to the cerebellar lesion. The reversible neurotoxicity of global GABA-elevation in experimental animals is discussed.