Testing upper motor neuron function in amyotrophic lateral sclerosis: the most difficult task of neurophysiology.

Clinical signs of upper motor neuron involvement are an essential observation to support the diagnosis of amyotrophic lateral sclerosis. However, clinical signs of upper motor neuron can be difficult to elicit in patients with motor neuron disease. One postulated reason for this problem is the presence of marked limb weakness and amyotrophy in motor neuron disease. This has been observed in patients with genetic mutations and clear-cut pathological evidence of upper and lower motor neuron degeneration. Less commonly, it has been recognized that the pattern of upper motor neuron lesion in amyotrophic lateral sclerosis is rather different from other conditions, in which there is damage to other descending motor fibres from extra-Rolandic motor cortical areas (Swash, 2012). In particular, the impact of the concomitant and spinal motor neuron loss and of spinal interneuron degeneration on the expression of the usual signs of upper motor neuron lesion are not well known (Swash, 2012). Probably, other clinical signs suggesting upper motor neuron dysfunction, such as mirror movements or the presence of apraxia, should be explored to support corticomotor involvement in motor neuron disease. Over the years, different techniques have been developed that help the neurologist to identify, measure and understand the upper motor neuron lesion in amyotrophic lateral sclerosis. Many transcranial magnetic stimulation methods have been used in this context to investigate integrity of the motor cortex and its descending pathways. A number of results reflecting delayed central conduction time and reduced motor-evoked potential amplitude indicate corticospinal tract damage (de Carvalho et al., 2003). Nonetheless, in amyotrophic lateral sclerosis, there is strong evidence for increased cortical excitability in early phases of the disease progression, suggesting either an initial phase in which the lower motor neuron demise is compensated or representing the initial pathogenic disturbance leading to lower motor neuron dysfunction (Vucic and Kiernan, 2006). Nevertheless, the combination of positive (increased cortical excitability) and negative (delayed central conduction time and reduced motor response) findings make transcranial magnetic stimulation a difficult technique with which to document serial upper motor neuron degeneration. Many groups have devoted much time and effort seeking the most sensitive and simple neuroimaging technique for detection of the upper motor neuron lesion in motor neuron disease. Voxel-based morphometry has been inconsistent in measuring the motor area; surface-based morphometry reveals cortical thinning in the precentral gyrus, but with a poor correlation with clinical findings; fractional anisotropy as evaluated by diffusion tensor imaging is one promising approach, but shows non-specific changes, and its usefulness to quantify progression is not clear; magnetic resonance spectroscopy is sensitive to motor cortex pathology, but the studies using this method have been inconclusive; functional MRI with blood oxygenation level-dependent contrast is potentially very effective for exploring neuronal interconnection dysfunction in amyotrophic lateral sclerosis, but still needs more investigation; and novel neuroinflammatory and inhibitory positron emission tomography ligands might have utility in the future (Turner, 2012). However, expense and practical issues limit the use of these sophisticated imaging techniques to a few highly specialized centres. Thus far, therefore, no method to investigate upper motor neuron function has proved useful and applicable as a measure of efficacy in clinical trials, despite some enthusiasm for the threshold tracking transcranial magnetic stimulation as a marker of early diagnosis. EMG is also not the preferred method for assessing upper motor neuron dysfunction, although some F-waves and H-reflex measurements and firing-rate analysis (de Carvalho et al., 2012) represent interesting developments. However, an exciting new window for EMG as a way to test upper motor neuron function has been opened by coherence analysis. In the past, a number of studies have investigated oscillatory activity of 15–30 Hz in the primary motor cortex both in humans (Conway et al., 1995) and monkeys using local field potential recordings (Murthy and Fetz, 1996). Indeed, cortically driven muscle contraction originates beta frequency oscillation synchronization between cortex and contralateral muscles, as well as between individual muscles (intermuscular coherence), as identified many years ago in animal studies (Baker et al., 1997). Experimental evidence shows that this oscillation is generated by Layer V of the motor cortex (Roopun et al., 2006). Cross-correlation analysis to study the firing pattern of the motor units showed that patients with stroke and spinal cord injury have abnormal motor unit synchronization (Farmer et al., 1993). In this issue of Brain, Fisher and co-authors describe their experience investigating intermuscular coherence in the 15–30-Hz range in controls and in patients with progressive muscle atrophy or primary lateral sclerosis. In summary, significant beta-band coherence was observed in all control subjects and all patients with progressive muscular atrophy tested, but not in those with primary lateral sclerosis. The authors conclude that intermuscular coherence in the 15–30-Hz range is dependent on an intact corticospinal tract, whereas it is not substantially altered by loss of anterior horn cells. These results are very stimulating and open a new avenue for investigation. Although this report has some limitations—including the small number of subjects, unilateral recording and the absence of systematic longitudinal evaluation (limited to one case)—it is evident that the study will have great impact on the field. It will now be possible inexpensively to investigate upper motor neuron function using standard EMG equipment and the appropriate software in any neurological centre. To complement this information by investigating corticomuscular coherence with EEG recording is an exciting possibility (Mima and Hallet, 1999). Scientific Commentaries Brain 2012: 135; 2579–2584 | 2581