Should the injured and intact hemispheres be treated differently during the early phases of physical restorative therapy in experimental stroke or parkinsonism?

Over a century ago the intact cortex was proposed to contribute to recovery from unilateral brain injury, but its possible role in functional outcome has become more appreciated in recent years as a result of anatomic, metabolic and behavioral studies. Although use of the contralesional limb is naturally impaired after sensorimotor cortex injury, neural and astrocytic events in the intact hemisphere may give rise to, and may be influenced by, an enhanced ability to compensate for lost motor function. The debate is still open as to whether the neural changes are generally compensatory in nature, with activity in the homotopic cortex leading to greater capability in the nonimpaired limb, or whether they are actually a matter of reorganization in the homotopic cortex leading to connections to denervated targets in the opposite hemisphere, thus allowing the homotopic cortex to control motor programs there. Although both phenomena may occur to some degree, there is mounting evidence in support of the former view. Careful behavioral techniques have been developed that can expose compensatory tricks, and the time course of these behaviors correlates well with anatomic data. Moreover, if the intact cortex sustains a second lesion after recovery from the first, forelimb sensorimotor function specific to the first-impaired side of the body is not worsened. Partial denervation of callosal fibers coming from the injured hemisphere, plus preferential use of the good forelimb caused by a cortical injury, may increase trophic factors in the intact hemisphere. These and related events seem to provide a growth-favorable environment there that permits motor learning in the intact forelimb at a level of skill exceeding that which a normal animal can attain in the same period of time. There are anecdotal cases in human neurologic patients that are consistent with these findings. For example, a colleague of the authors who sustained a unilateral infarction that rendered his dominant right hand severely impaired noticed that soon after the stroke he was able to use his left hand for writing and computers as well as he had ever used his right hand. Cross-midline placing tests also indicate that the structural events observed in the intact cortex may potentiate projections to the damaged hemisphere. These changes may help restore the capacity of tactile information projecting to the intact hemisphere to control limb placing in the impaired forelimb. Neural events in the injured hemisphere can be affected by behavior differently than the neural events in the intact hemisphere. Different therapeutic strategies might well be used on opposing limbs at different times after unilateral sensorimotor cortex injury to optimize recovery (and, indeed, to avoid exaggerating the insult). Finally, the details of reorganization in both hemispheres differ greatly depending on the type of brain injury sustained (eg, in stroke versus Parkinson's disease), suggesting that an approach that considers the role of both hemispheres is likely to be beneficial in research on a broad variety of brain pathologies.