Changes in propriospinally mediated excitation of upper limb motoneurons in stroke patients.

It has been argued that, in humans, a part of the descending command to upper limb motoneurons is transmitted through cervical propriospinal pre-motoneurons. We explored whether excitation of these putative propriospinal neurons projecting onto extensor carpi radialis (ECR) motoneurons was modified in patients recovering from stroke. Suppression of the voluntary on-going ECR EMG activity by stimulation of cutaneous afferents in the superficial radial nerve was used to estimate the component of the descending command passing through the propriospinal relay. The degree of suppression was assessed on both sides of 30 stroke patients (divided into two groups, whether recovery of wrist extension was poor or good by the time of the investigation) and of 34 age-matched controls. Single cutaneous volleys elicited a suppression which was symmetrical and of the same degree in patients and controls. In contrast, the amount of on-going EMG suppression produced by a train, which was symmetrical in normal subjects, was asymmetrical in most stroke patients: it indeed was significantly greater on the affected side of stroke patients with poor recovery of wrist extension than (i) in their non-affected side; (ii) in controls; and (iii) in the affected side of patients with good recovery. Cutaneous suppression of the H reflex, the motor evoked potential (MEP) and the on-going EMG was compared in three patients with poor recovery by the time of the first test; there was a small suppression of the H reflex on the affected side, but the asymmetry was much less than that of the on-going EMG and the MEP. In patients explored twice during the course of recovery, the asymmetry in the suppression of the on-going EMG tended to disappear, while recovery of wrist extension improved. This suggests that, when patients have not yet recovered, a relatively greater component of the descending command is mediated through the propriospinal relay. The findings are consistent with transiently increased efficacy of descending (possibly reticulospinal) projections onto propriospinal neurons, due to hyperexcitability of these neurons or unmasking and/or reorganization of the projections to them.