A longitudinal fMRI study: in recovering and then in clinically stable subcortical stroke patients

The aim of this 1-year longitudinal fMRI study was to compare hand motor activation patterns between cerebrovascular paretic patients with a subcortical infarction and healthy elderly subjects and to evaluate the changes between the subacute phase and the chronic phase of recovery. We studied eight right-handed patients with pure motor hemiparesis due to a single ischemic infarct of the corticospinal tract. Each patient underwent a first fMRI (E1) 20 +/- 9 days after stroke, a second (E2) after 4 months and a third (E3) 12 months after stroke. During each fMRI session, the patients performed an active motor task consisting of audio-paced (1 Hz) finger flexion-extension of the paretic hand and underwent a passive motor task consisting of flexion-extension of the paretic hand performed by an examiner. Data were analyzed with SPM99 (random effect analyses). Patients had recovered at E2, were stable between E2 and E3, but still experienced a hand weakness. Displacement of activation maxima coordinates in patients compared to healthy subjects suggested an early reorganization within the SMA and a secondary reorganization within the ipsilesional S1M1 at E2. The main differences between patients and healthy subjects were (1) recruitment of the posterior part of the cingulate cortex and SMA, (2) a general hyperactivation (except in the deefferented primary motor cortex) and (3) an evolution in the S1M1 activation from an early (20 days after stroke) contralesional hyperactivation to a later (4 months after stroke) ipsilesional hyperactivation concomitant to recovery. Changes in activation were confirmed by the passive task that involved no effort and little attention. Despite clinical stability, changes in brain processing seemed to occur between E2 and E3 corresponding to a normalization of ipsilesional S1M1 activation, a decrease of bilateral cerebellar activation, and a progressive increase in SII-BA 40 activity suggesting evolving compensatory networks to sustain recovery. (C) 2004 Elsevier Inc. All rights reserved.