Introduction
Outcome measures that are responsive to the biological process of muscle reinnervation and correlate with indices of motor function would be useful for clinical assessment following peripheral nerve injury. This study aimed to interrogate the validity of Motor Unit Number Estimation (MUNE) as an outcome measure in a controlled animal model of muscle reinnervation. The primary objective was to compare MUNE to histological markers of motor axons (choline acetyltransferase (ChAT)) and muscle reinnervation (alpha-bungarotoxin (α-BTX) and neurofilament). The secondary objective was to determine the relationship between MUNE and an index of motor function (Static Sciatic Index (SSI)).

Methods
Eighteen female Sprague-Dawley rats underwent sciatic nerve crush (10mm distal to the hip joint). The number of motor axons innervating the tibialis anterior muscle was determined using a modified multipoint stimulation (MPS) MUNE technique and ChAT counts of the nerve segment to the tibialis anterior 7 (n=3), 21 (n=5), 31 (n=5) and 42 (n=5) days post-injury. The reinnervation of neuromuscular junctions (NMJs) was monitored by quantifying the proportion of α-BTX motor endplates co-stained with neurofilament. The mean number of motor axons innervating each motor endplate was measured. SSI was recorded before the crush, 1 and 7 days post-injury and then every subsequent 3-4 days.

Results
In uninjured, contralateral tibialis anterior muscles the mean MUNE value was 80.8 (±12) which was not significantly different compared to the 74 (±9.5) ChAT positive axons in the nerve segment to tibialis anterior. With regards to injured muscles, MUNE significantly under-estimated the number of motor axons detected using histology 21 and 31 days post-injury (p<0.001). By 42 days post-injury, there was no statistically significant difference between MUNE (64.9 ±19) and ChAT (71.4 ±6.2) count. The recovery of CMAP as a percentage of the pre-crush value demonstrated moderate-good correlation with the proportion of reinnervated NMJs (R2=0.63, p<0.001). There were approximately two motor axons per motor endplate 21 days post-injury which corresponded to Single Motor Unit Potentials (SMUPs) that were approximately two fold larger compared to uninjured muscles. This regressed towards uninjured levels by 42 days post-injury. Finally, MUNE correlated with improvements in SSI post-injury.

Conclusions
MUNE corresponds with histological features of muscle reinnervation and an index of motor function (SSI) in a rodent nerve crush model. Future studies should address whether MUNE can detect a meaningful clinical response in clinical trials of therapies which hope to enhance human peripheral nerve repair.