References of the abstract :
Introduction
In 1991, I undertook a thesis in electrophysiology and I chose Professor Delwaide as my thesis director. It was decided with Professor Delwaide that I would focus on motor units (MU) counting. Indeed, at the beginning of the 1990s, there was a renewed interest in these counting techniques, with Macro EMG, the statistical method, the computerized version of the initial method proposed by McComas in 1971, the multiple-point stimulation method, the spike-triggered averaging technique, and the F-wave method. The general principle of MU counting was quite simple. The first step was to estimate the average MU size from a sample of 10 to 20 MU potentials. Then, the size of the maximum motor response (CMAP) was divided by this value to obtain an estimate of the number of motor units (MUNE). Thus, ultimately, most counting techniques were distinguished by the method used to estimate the average MU size. To achieve this, McComas was the first, in 1971, to propose a new type of nerve stimulation, known as incremental stimulation. This method involved percutaneous nerve stimulation with a very short stimulation duration (50 μs), where the intensity was gradually increased in 0.1 mA increments. This incremental stimulation allowed the individual and sequential activation of motor axons. When nerve stimulation was applied in increments of 0.1 mA at a single stimulation point, it became possible to obtain ten successive increments of the CMAP amplitude, corresponding to the activation of ten MUs. This made it possible to estimate the average MU size and, subsequently, the MUNE value.
Adapted Multiple-Point Stimulation (AMPS) method
Professor Delwaide sent me to Canada to meet Alan McComas at McMaster University to learn more about this technique and begin to define the design of my future experiment. From the beginning, McComas was aware that his technique was affected by a methodological bias he called “alternation.” The alternation was linked to the fact that the activation of a MU, in response to percutaneous electrical stimulation, did not follow an all-or-nothing law. Indeed, the probability of activation fluctuated from 0% to 100% depending on the intensity of the nerve stimulation, following a sigmoid curve. This bias was responsible for an underestimation of the average MU size and an overestimation of MUNE. To eliminate the bias associated with alternation, Brown & Milner-Brown, then Doherty, proposed the multiple-point stimulation method (MPS). The nerve was stimulated at ten distinct points, and at each stimulation point, a single MU was activated. The challenge of this technique was that it was not easy to locate ten stimulation points along the nerve to evoke ten different MUs, particularly in pathology, when the number of MUs was reduced. This led us to develop and validate an intermediate technique between incremental stimulation and MPS, which we called the adapted multiple-point stimulation method (AMPS). With this technique, 4 stimulation sites were generally sufficient, 3 in the distal forearm region and one at the elbow, with 2 to 3 units recruited at each point. To minimize alternation, MU potentials had to be activated in an all-or-nothing manner and motor axons recruited with distinct thresholds, in an orderly and reproducible manner, without any fractionation of the motor response to successive stimuli. To ensure that MUs activated at different stimulation points were distinct from each other, the morphology of each MU potential was reconstructed using a trace subtraction program. By adding the amplitude of the four maximal motor responses obtained at the four stimulation points and dividing this sum by ten, the estimate of the average MU size was obtained. By dividing the size of the maximal CMAP by this value, we obtained the MUNE.
Results
To validate the AMPS, we measured the correlation of the results obtained using it, with those obtained from an already validated method, the F-response method, in 54 control subjects. We also evaluated the test-retest reliability, which was found to be excellent. Following this, we established normative values on a sample of 59 healthy subjects according to age. We calculated a loss of 1.4% of MU per year. Subsequently, we applied our technique to patients with amyotrophic lateral sclerosis (ALS). We demonstrated a reduction, often significant, in the number of MUs in most of our patients, as well as an increase in the mean MU size, which we attributed to collateral reinnervation. As a result, the reduction in the maximal CMAP size was often less pronounced and the values frequently remained within normal limits. We also showed a negative correlation between the number of MUs and the average MU size, again linked to collateral reinnervation, except in four patients whose disease progression was particularly rapid, indicating failure of collateral reinnervation in these patients. Also in ALS patients, we demonstrated that the percentage reduction in MUNE at the 4th month compared to the baseline assessment could predict the percentage reduction in MUNE at the 12th month compared to the baseline assessment. This allowed us to distinguish patients with rapid disease progression from those with slower progression, which could be of interest when using the MUNE technique in therapeutic trials. Later, Pierre Bouche opened the doors of his department to me at the Pitié-Salpêtrière hospital, allowing me to expand my recruitment of patients among those suffering from motor neuron degeneration. We demonstrated that ALS patients with a survival of less than three years lost their MU more quickly than those with a survival of more than three years. This period also allowed me to apply our counting technique to patients with primary lateral sclerosis and patients with Kennedy's disease. A canonical discriminant analysis allowed us to distinguish the three groups of patients by introducing variables such as the decrement value during stimulation at 3 Hz, the maximum amplitude of the motor response divided by the disease duration, the area of the motor response evoked in the tibialis anterior muscle and the product of the F waves persistance and their maximum amplitude, which we called the F index. And then came the Motor Unit Number Index, MUNIX. In this method, the maximum CMAP was compared to the interference pattern detected by surface electrodes in terms of amplitude, area and power to derive an index of the number of MUs. After having published a comparative study between our counting method and the MUNIX procedure, both in healthy subjects and in patients suffering from motor neuron degeneration, I had the opportunity to participate in a large multicenter study initiated by Professor Attarian from La Timone hospital in Marseille. This study demonstrated, among other things, that for the purposes of longitudinal monitoring of patients, a variation of more than 20% in the MUNIX sum score could be interpreted as a significant change.
Conclusions
MUNE remains a relevant parameter, especially for assessing the effectiveness of new treatments, for instance in spinal muscular atrophy. Our counting technique, AMPS, has numerous advantages: it's non-invasive, reliable, fast, and applicable to any patient regardless of the reduction in MU.
