Matthew Edwardson, MD

Favre I, Zeffiro TA, Detante O, Krainik A, Hommel M, and Jaillard A. Upper Limb Recovery After Stroke Is Associated With Ipsilesional PrimaryMotor Cortical Activity: A Meta-Analysis. Stroke. 2014

Interpreting the results of functional imaging studies to explore post-stroke plasticity can be challenging. Invariably these studies include only a small sample of subjects. As a result, many functional imaging studies conflict on the exact location of increased brain activity in the recovery period. In this article, Favre and colleagues bring clarity to post-stroke brain remodeling by combining the results of multiple fMRI and PET studies. In subjects with good functional recovery they found increased activity in the ipsilesional M1 and medial-premotor cortex (PMC). Subjects with poor recovery had increased activity in the cerebellar vermis. 


The authors performed activation likelihood estimation to assess cortical activation patterns in 24 prior fMRI and PET studies. Subjects had motor impairments of the upper limb with either partial or complete recovery. 89% had subcortical strokes. Studies were separated into acute (<35 days post-stroke) and chronic (>3 mo. post-stroke) stages. The authors found an increase in cortical activation in contralesional dorsolateral PMC in the acute stage that faded with time. This acute contralesional shift did not correlate with recovery. In contrast, during the chronic stage there was a ventral shift in ipsilesional M1 activation that occurred in those with good recovery.

This meta-analysis consolidates the evidence from prior functional imaging studies and suggests that functional imaging may be a useful biomarker for stroke recovery. Many prior studies show a contralesional shift in cortical activation in the acute phase post-stroke. Whether this shift is adaptive or maladaptive remains unclear; the present meta-analysis found no association with recovery. What is more clear is that later in the recovery process a shift in activation patterns back to the lesional side and particularly a ventral shift in M1 occurs in those with good recovery. The authors also discovered increased activity in the cerebellar vermis in those with poor recovery – a novel finding that requires further confirmation. Studies like this increase our confidence in using functional imaging as a biomarker for stroke recovery. Time to return of ipsilesional activation and ventral shift in ipsilesional M1 may emerge as important endpoints in future stroke rehabilitation trials.