MAGNETIC-RESONANCE-IMAGING IN HEREDITARY AND IDIOPATHIC ATAXIA

We used magnetic resonance imaging (MRI) to study brain and spinal cord morphology in hereditary and idiopathic ataxia. Our interest was in whether the classical neuropathologic categories-cerebellar cortical atrophy (CCA), olivopontocerebellar atrophy (OPCA), and spinal atrophy (SA)-could be identified in vivo and which clinical phenotype corresponded to which morphologic category. To this end, we measured the size of the cerebellar vermis, cerebellar hemispheres, fourth ventricle, middle cerebellar peduncles, basis pontis, medulla oblongata, and cervical spinal cord on T1-weighted images of 61 patients and 24 healthy controls. Five patients with Friedreich's ataxia (n = 7) and all with late-onset Friedreich's ataxia (n = 3) had SA without major involvement of the brainstem or cerebellum. Morphologic findings in patients with early-onset cerebellar ataxia with retained tendon reflexes (n = 11) were heterogeneous: six patients had MRI findings compatible with CCA, and two patients had a combination of SA and CCA. The three remaining patients had an atypical pattern of atrophy. Similarly, the morphologic changes in patients with autosomal-dominant cerebellar ataxia with additional noncerebellar symptoms (ADCA-I; n = 13) were nonuniform: atrophic changes typical for CCA, OPCA, or SA were each present in one case, four patients had a combination of OPCA and SA, and the remaining patients could not be assigned to one of the morphologic categories. In autosomal-dominant cerebellar ataxia with a pure cerebellar syndrome (ADCA-III; n = 6), all patients except one had CCA. Patients with idiopathic cerebellar ataxia (IDCA) fell into two clinically and morphologicly distinct groups: the majority of patients with additional noncerebellar symptoms (IDCA-P; n = 14) had OPCA, whereas almost all patients with a pure cerebellar syndrome (IDCA-C; n = 7) had CCA.