has been consistently shown to be superior in experimental studies. One limitation is a lack in consistency in outcome assessments that are selected for studying functional recovery. In addition, functional and behavioral results and other tests of regeneration, such as axon counts, often do not correlate with one another. We hypothesize that there is a key difference between nerve "regeneration" and "functional recovery." The purpose of this study was to compare five techniques for primary

more »

... e repair in a rodent model, and identify potential correlations between functional and histomorphometric assessments. METHODS: Sciatic nerve transection was performed in male Lewis rats. The nerve was repaired using one of five techniques: interrupted epineural, running epineural, grouped fascicular, epineural with collagen wrap, and tension. A sham group was also included. Following recovery, a combination of behavioral (rotarod and ladder rung), electrophysiologic (compound muscle action potential (CMAP), nerve conduction velocity (NCV)), advanced imaging, and histomorphometric assessments were performed. Statistical analysis was performed using one-way analysis of variance and Spearman's rank correlation coefficient. To create a single measurement of aggregate performance, the individual assessment scores were standardized and converted to z-scores. Z-scores reflect the number of standard deviations from the mean; this approach more fairly weights each of the assessments when evaluating overall aptitude. RESULTS: In all animals (n=48), functional testing was performed at week 13, followed by electrophysiology and imaging at week 14. The sham group (n=7) performed the best on both behavioral assays (p<0.001) and also demonstrated the highest mean CMAP amplitude (p<0.001) and fastest NCV (p<0.001). Behavioral and electrophysiologic results were poorest in the tension group (n=9). Axons in all repair groups were more numerous and smaller compared to those in the controls (p=0.01). Qualitative inspection revealed that aberrant axon sprouting was greatest in the tension group (n=9). There was a significant difference in total z-scores, with the sham group performing best and tension group performing worst when outcomes on all assessments were totaled (p=0.04). When all test results were compared between groups, few significant correlations were observed. Ladder performance correlated positively with rotarod duration (p=0.001) and degree of myelination (p=0.02). NCV positively correlated with both degree of myelination (p=0.01) and mean axon fiber area (p=0.02). CONCLUSION: No single tension-free repair technique was significantly different from another, supporting traditional beliefs that techniques that minimize tension lead to equivocal outcomes. The Rotarod and ladder rung behavioral assays allowed for differentiation between satisfactory and suboptimal functional results. A relative lack of correlation between outcomes assessments supports further study of the differences between assessments of nerve regrowth and functional recovery.