Non-invasive electrophysiological assessment of the corticospinal tract in health and disease

Abstract

To date, no candidate markers of upper motor neuron (UMN) function have performed sufficiently well to enter widespread clinical use, and the lack of such markers impedes both the diagnostic process and clinical trials in motor neuron disease (MND). We studied 15-30Hz intermuscular coherence (IMC), a novel marker of UMN function, and central motor conduction time (CMCT), an established marker of UMN function based on transcranial magnetic stimulation (TMS), in healthy volunteers and patients newly diagnosed with MND. To clarify the relative contributions of different parts of the motor system to IMC generation, we examined IMC in patients with longstanding diagnoses of hereditary spastic paraparesis (HSP), multifocal motor neuropathy (MMN) and inclusion body myositis (IBM). Previous studies reported conflicting results for the relationship between CMCT and predictors such as age and height. We only found a significant correlation between lower limb CMCT and height. IMC did not vary significantly with age, allowing data from healthy subjects across all ages to be pooled into a single normative dataset. The variability of IMC between subjects was considerable, and within a given subject variability was greater between than within recording sessions; potential contributors are discussed. Anodal transcranial direct current stimulation (tDCS) caused a significant increase in IMC, but interindividual variability was substantial, which might hinder its future use as an adjunct to IMC. To compare individual disease groups to the normal cohort, we evaluated the area under the receiver-operating characteristic curve (AUC). IMC generally matched or exceeded the performance of CMCT in discriminating patients with MND from normal, achieving AUCs of 0.83 in the upper and 0.79 in the lower limb. Previous evidence suggests that IMC abnormalities are primarily attributable to corticospinal tract (CST) dysfunction. In line with this, most patients with HSP exhibited diminished IMC. However, patients with MMN also showed decreased IMC, suggesting either that subclinical CST involvement was present or that dysfunction of lower motor neurons (LMNs) may affect IMC; clarification through computational modelling is suggested. In iii IBM, IMC was generally increased, which might reflect that the altered motor unit discharge pattern makes synchronisation more readily detectable. IMC appears to be a promising marker of CST function. It remains to be clarified how strongly it is influenced by LMN lesions, and optimisation of methods should help to minimise the variability of results. Since IMC is non-invasive and can be measured using commonly available EMG equipment, wider dissemination should prove straightforward.