Current methods of motor unit number estimation (MUNE) suffer from subjectivity and/or extrapolation from biased samples of units.  To avoid these criticisms a new MUNE method was recently described which fits a model to a detailed stimulus-response curve (CMAP scan)(Bostock, 2016).  The aims of the present study were to examine the validity of CMAP Scan MUNE (MScan) with regard to inter-rater and intra-rater variability in the assessment of patients with amyotrophic lateral sclerosis (ALS) and healthy controls, and to compare the new method with multipoint stimulation MUNE (MPS) and motor unit number index (MUNIX).

            METHODS: A prospective study was undertaken on 20 ALS patients and 20 sex- and age-matched healthy controls.  Baseline examinations with the 3 methods to be performed twice each by two blinded physicians (ABJ and HT).  Follow-up examinations to be done by one physician (ABJ) after 4 and 8 months. The preliminary baseline data presented here are from 17 ALS patients and 10 controls.  Reproducibility of MUNE values was  assessed by the inter-rater, intra-rater and total coefficients of variation (CV).

            RESULTS:  In all categories (inter- and intra-rater, patients and controls), the results for MScan MUNE were more reproducible (lower CV) than for either MPS or MUNIX.  Overall the median CV for MScan MUNE was 8.8%, which was significantly lower than for MPS (26.9%, P<10^-7 by Mann-Whitney) or for MUNIX (14.9%, P<0.01).  MScan MUNE values for patients were similar to, and highly correlated with those for MPS (R=0.95). MUNIX values were higher than MPS or MScan and less well correlated (R=0.85 with MPS, 0.82 with MScan).

            CONCLUSIONS:  MScan provides more consistent MUNE values than MPS or MUNIX and because they are based on all motor units, MScan MUNE values are less likely to be affected by non-representative sampling than other methods. This promising new method may therefore be suitable for assessing disease progress and treatments of neuromuscular disorders.