Brachial plexus injury was one of the severe injuries to the upper limb. It was reported that 10–20% of peripheral nerve injuries were brachial plexus injuries,16 especially the total brachial plexus root avulsion, which was difficult to repair. At present, nerve transfer is the primary method for total brachial plexus nerve root avulsion. It is widely accepted that the extraplexal nerve transfer is a feasible way for functional restoration of the affected limb,19,22 which includes intercostal nerve transfer,21 spinal accessory nerve transfer,1 phrenic nerve transfer10 and cC7 transfer11. In patients with total brachial plexus injury, cC7 transfer is usually used to repair the median nerve. Clinically, simultaneous repair of two recipient nerves by cC7 transfer has been reported to achieve a certain curative effect.5 Although the spinal accessory nerve was often transferred to the suprascapular nerve for shoulder abduction reconstruction,14 the recovery of shoulder abduction was not satisfying.8 Recent studies showed that the repair of both axillary and suprascapular nerves could achieve effective shoulder abduction.12 So this study was designed to explore the efficacy of cC7 in repairing the median and axillary nerves at the same time.
In the study, behavioral tests, EMG examination, mean muscle fiber cross-sectional area, nerve fiber count and gene expression related to muscle atrophy were used to demonstrate the efficacy of cC7 transfer. EMG and the number of nerve fibers represented nerve regeneration after nerve repair. The cross-sectional area of muscle fiber reflected the degree of muscle atrophy and regeneration. Based on previous studies, the expressions of MAFBOX and MURF1 were positively correlated with muscle atrophy.6 All the results showed there were no significant differences in the efficiency of median nerve recovery between Group A and C, while both of the two Groups improved the median nerve regeneration, compared with Group D. The above results indicated that cC7 transfer to both median and axillary nerves would not affect the recovery of median nerve, compared with cC7 transfer to median nerve alone. Anatomical reports showed 17000–40000 myelinated nerve fibers in cC7 nerve root4. The average number of myelinated nerve fibers was about 1480020 in the median nerve, while it was 2704 in axillary nerve.26 The number of nerve fibers in cC7 nerve root was more than that of any other recipient nerve and it had the capacity to provide enough dynamic nerve fibers to both of median and axillary nerves in theoretically, which was consistent with our results.
Many studies recommended the suprascapular and axillary nerves should be repaired simultaneously to obtain better shoulder joint function, because the repair of suprascapular nerve alone was not enough for the reconstruction of shoulder abduction.12,15 In aspect of the number of axillary nerve fibers, the results in Group B and C were statistically better than that in Group D, while there was no statistical difference between Group B and C. The results of shoulder abduction and EMG also showed deltoid recovery in Group B and C, while no statistical difference existed between the two groups. All these results above indicated that the axillary nerve could be repaired no matter by cC7 transfer to axillary nerve alone or by cC7 transfer to axillary and median nerves.
As for the cross-sectional area of DEL fibers, although there were no statistical differences among Group B, C and D, the mean cross-sectional areas of DEL fibers decreased successively form Group B to C to D. The results demonstrated muscle atrophy was related to denervation and nerve transfer made innervation which could reduce the muscle atrophy. There was functional impairment in Group D compared with Group B and C, while no statistical demonstration was showed in muscle fiber count of DEL. Because the completion of an action requires sufficient muscle fibers for contraction, when there are not enough muscle fibers regeneration, the functional differences are revealed. But at the same time, in the aspect of cross-sectional area of the muscle fiber, statistics do not necessarily show differences in cross-sectional area. The cross-sectional area of DEL successively decreased from Group B to C to D, which coincided with the result of functional test. Another reason might be the time from operation to biopsy, which was not long enough to make the muscle completely atrophy in Group D. It induced no statistical differences of the cross-sectional area of DEL to occur among groups.
Bodine SC et al.2 reported only a small subset of genes was universal in all atrophy models. Two of these genes encode ubiquitin ligases: Muscle RING Finger 1 (MuRF1) and Muscle Atrophy F-box (MAFbx). Overexpression of MAFbx in myotubes produced atrophy, whereas mice deficient in either MAFbx or MuRF1 were found to be resistant to atrophy. The models were the rats with denervation or immobilization or unweighting, which were a little different from our models. In our study, there were four groups, including one denervation group and three nerve regeneration groups. From our research, the negative data of MAFbx and MURF1 in DEL coincided with the result of the mean cross-sectional areas of DEL, while the data in gene expression in FCR was not so consistent with the mean cross-sectional areas of FCR. We know that muscle regeneration takes time. When innervation to the muscle occurred, the time course of inducing the change of MAFbx and MURF1 was unknown yet, which is an intriguing question. Maybe the time from operation to biopsy was not long enough to show the different change of MAFbx and MURF1 between innervation and denervation groups. We think the innervation process with the gene regulation might not just be the opposite process of denervation and its correlation with these two genes needs a further study.
The proportion of the recovery of axillary nerve was less than that of median nerve in terms of behavior tests, EMG and the number of nerve fibers in the study. The main reason was the anatomy of axillary nerve, which passed through the quadrilateral foramen. This nerve was easily compressed in the quadrilateral foramen, especially during the process of nerve growth, which influenced the growth of the nerve fibers. In addition, the axillary nerve was smaller in diameter than the median nerve and most of the cC7 nerve fibers would grow into the median nerve.
Previous meta-analysis showed that the average effective rate of cC7 transfer to median nerve was 50%, which was close to our result (70%).13 There were also some clinical and experimental reports of cC7 transfer to two recipient nerves simultaneously. Gao et al.7 reported cC7 transfer to the median nerve and biceps branch or the median nerve and triceps branch. The recovery rate of motor function of median nerve was 68.2%, while those of biceps branch and triceps branch were 66.7% and 20%, respectively, which suggested that cC7 could achieve exact results by repairing two recipient nerves at the same time. Pan et al. reported an animal experiment of cC7 transfer to the median and musculocutaneous nerves. The target muscles innervated by the median and musculocutaneous nerves had regeneration,18 but the details about the recovery rate of muscle strength were not reported. Chuang et al.5 also found that cC7 could repair the median and musculocutaneous nerves at the same time. The recovery rates of finger and elbow flexion were 39% and 82.6%, respectively. Terzis et al.23 repaired axillary nerve with hemi-cC7 nerve root and muscle strength of DEL achieved M3 or above was 20%. In our experiment research, the axillary nerve was repaired with the total root of cC7 and effective rate was 60% in rats. There were two probable reasons for the difference. One was the difference of nerve growth speed between human beings and rats. The other was the difference in the number of donor nerve fibers. There was no animal experimental report on cC7 transfer to both of the median and axillary nerves before.
There were some limitations in this study. First, there was a lack of a simple and accurate rat behavior test for shoulder external rotation. Secondly, the muscle tension measurement was an important method for judging the recovery of muscle function, while we did not use it in our experiment. Third, the number of samples in the research is small, which made the results not so convincing. Afterwards, we’ll increase the number of animals in the future research. In both the functional tests and EMG results, we compared the percentages of effectiveness (positive) among groups. Due to the small number of samples, the statistical analysis seemed to lower the power of the test. The amplitude and latency of CMAP of the muscles could be used for evaluation. In the future research, we’ll use amplitude and latency as indicators for nerve regeneration evaluation. In addition, there was a remodeling process of cortical plasticity after cC7 transfer and its mechanism would be the focus in the future research.