記住我
ORIGINAL ARTICLE
Year : 2022 | Volume
: 55
| Issue : 4 | Page : 140-146
Can sensory protection improve the functional outcome in delay repaired rat brachial plexus injury?
Tommy Nai-Jen Chang1, Wei-Chuan Hsieh1, Jo-Chun Hsiao2, Bassem W Daniel3, David Chwei-Chin Chuang1, Johnny Chuieng-Yi Lu1
1 Department of Plastic and Reconstructive Surgery, Chang Gung Memorial Hospital, Chang Gung Medical College and Chang Gung University, Taoyuan, Taiwan
2 Department of Plastic and Reconstructive Surgery, Chang Gung Memorial Hospital, Chiayi, Taiwan
3 Department of Plastic and Reconstructive Surgery, Chang Gung Memorial Hospital, Chang Gung Medical College and Chang Gung University, Taoyuan; Department of Plastic and Hand Surgery, Freiburg University Medical Center, Freiburg, Germany
Correspondence Address:
Johnny Chuieng-Yi Lu
No. 5, Fu-Hsing St. Kwei-Shan, Taoyuan
Taiwan
Source of Support: None, Conflict of Interest: None
DOI: 10.4103/fjs.fjs_233_21
Background: Reconstruction of brachial plexus injuries (BPIs) at a delayed time point may prolong the denervation of target muscles and jeopardize the outcome. Sensory protection has been hailed as a promising technique that may help preserve muscle mass and restore functional outcome. We utilize the rat brachial plexus model to investigate the difference between early and delay repair, and evaluate if sensory protection of distal nerves can assist in delayed repair.
Materials and Methods: Forty-eight Lewis rats were randomly assigned to four groups (n = 12 in each group, including one positive control group). All the rats were transected at the upper, middle, and lower trunk levels with a 2-cm gap. Group I underwent immediate reconstruction from the upper trunk to the median; Group II underwent the same reconstruction but at 4 months after the initial transection; Group III was same as Group II and additional sensory protection to the median nerve via a nerve graft from the lower trunk. The final outcome was studied and analyzed 16 weeks postoperatively.
Results: Group I (immediate repair) showed the best functional results in muscle contraction force, muscle action potential, and muscle weight, in addition to higher axon counts. Groups II and III (delayed repair) both showed inferior results to Group I, and sensory protection did not show any significant improvements in outcome.
Conclusion: Delayed repair still shows inferior outcomes to acute repair in BPIs. There is no sufficient evidence to support the use of sensory protection in delayed repair.
Keywords: Early and delay repair, median nerve, rat brachial plexus injury, sensory protection
Reconstruction for adult brachial plexus injuries (BPIs) has undergone advances in surgical strategy to improve outcome. Even so, it still remains a daunting challenge to achieve consistent outcome given the unique nature of each patient's presentation. Most BPI patients suffer from multiple complicated comorbidities that required prolonged hospitalization and stabilization of the general condition. Furthermore, paralysis of the injured limb requires a 3–5-month observation period to wait for any spontaneous recovery in incomplete nerve injuries.[1] This distance between the location of the nerve injury and the target muscles is quite long given the length of the extremity. As such, prolonged denervation leads to the muscle apoptosis and fibrosis, resulting in an irreversible functional deficit.[2],[3] Since “time is muscle,” there is a search for additional measures that can help preserve the muscle integrity, such as “sensory protection” through the use of an anatomical nearby sensory nerve to co-innervate the target muscle.
Sensory nerves are different from motor nerve, as they have not been shown to build motor endplates. The physiological mediators are also different (e.g., neurotrophin).[4],[5] A possible “babysitter” effect by sensory nerve on motor nerves has been postulated for many years, but the results became more encouraging in recent years due to new animal studies.[6],[7] In 1992, Ochi used grafted dorsal root ganglia in a sciatic nerve model, which indicated that sensory axons could delay the weakening and atrophy of muscles after denervation.[8] Bain proved that sensory protection has a significant effect on denervated rat skeletal muscles, especially if the denervation or sensory reinnervation periods were >3 months.[9] Terzis used a rat upper biceps brachii model with sensory protection of the motor nerve, and the grooming test at 2 weeks postoperatively showed favorable outcomes.[10] Some recent studies found adjuvant electrical muscle stimulation to synergistically augment sensory protection effects[11],[12],[13] In 2015, Placheta et al. used supraorbital nerve to enhance a motor cross-face nerve graft in an end-to-side coaptation.[14] Fluorescent imaging proved that sensory protection of cross-face nerve grafts adds to nerve regeneration and can counteract degeneration in long nerve grafts.
Even so, clinical application of sensory protection remains limited, since it is difficult to validate its sole contribution to nerve regeneration. Another reminder is that most outcomes have been presented in the animal models using lower extremities.[15] From our clinical experiences, we still think that sensory protection may still be worthy to apply in facial, upper, and lower extremity neuropathy cases since the donor site morbidity is limited. If so, this can definitely consolidate more consistent outcomes. In this study, we mimic the clinical scenario of total BPI to investigate the effect of sensory protection for finger flexion reinnervation, with attention directed specifically at the time point of repair.
Methods and MaterialsMale Lewis rats (8–10 weeks old) were used in accordance with established principles of care and animal research of the Chang Gung Memorial Hospital Animal Care Committee. All the animals were maintained on standard rat chow and water ad libitum with a 12-h light-dark cycle in accordance with the rules and regulations governing Chang Gung Memorial Hospital Animal Care Committee and the Ministry of Health and Education, Taiwan (Approval Number 2013031303).
Forty-eight Lewis rats were randomly assigned to four groups (three study groups and one control group, n = 12 in each group). All the rats received transection at the upper, middle, and lower trunk levels with a 2-cm gap. Group I underwent immediate reconstruction with a 2.5-cm autograft from the upper trunk (UT) to the median; Group II underwent the same reconstruction method but at 4 months after the initial transection; Group III also was reconstructed at 4 months after the initial transection but received additional sensory protection from the pectoral nerve to the median nerve in end-to-side fashion. An additional sham group was used as a normal control [Figure 1].
Figure 1: Surgical group illustration. In all the groups, the segmental resection of upper, middle, and lower trunk (distal to pectoral nerve) performed. The upper trunk coapts to lower trunk with nerve graft as well. In Group 1, the nerve graft performed immediately, and in Group 2, the nerve graft performed 4 months later. No sensory co-innervation in both groups. The procedure in Group 3 was the same in Group 2 plus the sensory branch co-innervation via the sensory branch of pectoral nerveAnimal model
The right brachial plexus was the experimental site. The median nerve innervated forearm flexor muscles were the target muscles for evaluation. Sixteen weeks was the endpoint to assess the outcome. The outcomes in the target muscles were evaluated using (1) electrophysiological assessments (muscle weight, muscle contraction force, and electromyogram); (2) histomorphometry of the median nerve, distal to the nerve coaptation site, including histology, axon count, and diameter; and (3) retrograde labeling (dorsal root ganglion [sensory neuron] and spinal cord [motor neuron]). Since electrodiagnostic testing and retrograde labeling cannot be performed on the same animal, we designed two groups with 12 rats in each.
Surgical procedure
All surgical procedures were performed using sterile conditions. All surgical procedures were performed aseptically under inhalational anesthesia using isoflurane (FORANE®, Baxter, P. O. BOX, San Juan, USA). The animals were placed in prone position, shaved, and washed with antiseptic solution. Under an operating microscope, the right brachial plexus was exposed via a vertical incision in the paravertebral region, and the nerve trunks were identified using a nerve stimulator (Vari-Stim Hand-Held Nerve Locator/Stimulator, Medtronic Xomed, Inc., Minneapolis, Minn., USA). For all the groups, the upper, middle, and lower trunks were transected to create a 2-cm gap and the distal stumps were inserted into local muscles to avoid uncontrolled regeneration. The animals were then placed in supine position to expose the brachial plexus in the axillary region. Nerve reconstruction was performed as previously described [Figure 2]. Closure of the wounds was carried out with nylon 4-0 sutures. Postoperatively, animals were rewarmed and returned to their cages.
Figure 2: The surgical model in Groups 2 and 3 showing the chronic denervated surgical site and how to reconstruct the defect via the nerve graft as well as the sensory co-innervationElectromyogram
Under general anesthesia, the median nerve and forearm flexor muscle were exposed through the previous incision. About 10 mm of the nerve length from the forearm flexor muscle was exposed. A hook electrode was placed into the distal muscle, while the ground electrode was placed subcutaneously. A two stimulating hook electrode with 2 mm apart was placed around the median nerve. Stimulation was delivered by an electrical stimulator (Biopac System, BSL Software Installation Package, Windows, Goleta, California) and fixed at 1 msond at constant current between 10 mA and 10 A, while the compound muscle action potentials were recorded.
Tetanic muscle contraction force measurement
The force of tetanic muscle contraction was assessed according to a previously described protocol.[12] First, the resting length of the forearm flexors was determined. Then, the distal muscle insertion was detached from volar wrist level and attached to the force displacement transducer (FT03 Force Displacement Transducers, Grass Instruments, Quincy, Massachusetts) at resting length. In this position, shoulder, elbow, and wrist were immobilized with pins to prevent motion artifacts. A bipolar platinum electrode was used to deliver a stimulating current to the median nerve at the same location as described for electromyography. The threshold stimulus was determined as a stimulation that produced a noticeable twitch of the forearm flexor muscles. Nerve stimulation was performed at different thresholds (1–10 times of the initial threshold stimulus) using different voltages and frequencies (0.6–1.2 V and 1.0–60 Hz, respectively). The maximum tetanic contraction strength was measured at 1 and 60 Hz and recorded as grams/weight. The mean maximal isometric muscle contraction of the repeated muscle contraction forces at 5 times with pulse duration of 1.0 msond was determined. The data for tetanic muscle contraction forces were collected, controlled, and analyzed using MacLab systems (ADInstruments, Colorado Springs, Colorado).
Forearm flexor muscle weight
After the above measurements, animals were euthanized by intracardiac injection of pentobarbital. Under an operating microscope, the entire left and right forearm flexor muscles were harvested by dividing its origins with bone. The muscle was weighed immediately, and the results were expressed as left/right muscle weight ratios.
Nerve morphology study
The flexor muscles were embedded in optimal cutting temperature compound and snap frozen in liquid nitrogen. Segments of median nerve, 5 mm long, were obtained bilaterally just before its entry into the flexor muscles for recipient nerve study. The subsequent protocol for processing and analyzing nerve morphology was the same as described in the previous section.
Axon count
Nerve specimens (about 3–5 mm in length) were obtained from target nerves. Samples will be fixed in 2.5% glutaraldehyde and postfixed in 2% osmium tetroxide. Each nerve will be embedded in 100% Epon. One-micrometer-thick transverse sections were made from the nerve to obtain successive sections in 1-mm intervals, which will be stained with 2% toluidine blue and photographed under a light microscope with ×400 magnification. Myelinated axons will be counted manually with ×1000 magnification.
Retrograde labeling of dorsal ganglion roots
After removal of the forearm flexor muscles, the median nerve at the level of the muscle entry was covered by a cap filled with FluoroGold (Fluorochrome, LLC, Denver, Colo.) dye, but one suture was added to fix the nerve to the cap, and a few drops of fibrin glue (Tisseel; Baxter Healthcare, Baxter, Ill.) were used to complete the seal.[16] Seven days later, the surgical site was revisited to see whether the cap was still in situ. After confirmation, the rats were perfused with 200 ml of 4% paraformaldehyde followed by 200 ml of normal saline. The spinal cord from the C5-T1 level and the bilateral C5-6 dorsal root ganglia were harvested and fixed in 4% paraformaldehyde for 24 h. Dorsal root ganglia were then cryoprotected in 30% sucrose in 4% paraformaldehyde for 1 day and the spinal cords were cryoprotected for 4 days. All samples were embedded in tissue freezing medium (General Data Company, Inc., Cincinnati, Ohio) and frozen over dry ice.[17],[18] The segment of the spinal cord was examined under the electron microscope to check the part with uptake of dye. Further, the highlighted part (dorsal ganglion C5 and C6) was sent for detailed confocal study, and the number of neurons was counted and documented.
Statistical analysis
The statistical analysis was performed using ANOVA with Tukey's post hoc test for all comparisons except of each group regarding muscle force contraction, muscle weight, and electromyography, where unpaired t-tests were applied. All analyses were performed with GraphPad Prism (GraphPad Software. Inc., La Jolla, CA 92037, USA). P < 0.05 was considered statistically significant.
ResultsMuscle
As for the muscle contraction, Group I showed higher force which was very close to the normal control without statistical significance, Groups 2 and 3 both showed lower contractility but without statistical significance in between [Figure 3]a. However, if we looked into the details of Groups 2 and 3, we found that also the sensory protection procedure can increase the muscle weight, however, the decrease in electromyography was noted, which makes the overall contractility similar [Figure 3]b and [Figure 3]c.
Figure 3: (a) Comparison of muscle contraction force between the three experimental groups, the contralateral side served as the positive control. There is no significant difference between the control group and the immediate reconstructive group (Group 1), but early reconstruction has much better outcome than the delayed groups (Groups 2 and 3), and there was no statistical significance between these two groups. (b) The muscle weight showed significant difference (P < 0.05) in early (Group 1) and delayed reconstruction (Groups 2 and 3), and Group 3 is better than Group 2. Furthermore, all the experimental groups were worse than the normal control group. (c) The electromyography showed no evidence of improvement in the sensory co-innervation group (Group 3), even worse than the nonsensory co-innervation group (Group 2)Nerve
As for the nerve morphology, compared the immediate (Group 1) and delay repaired (Group 2) groups, the re-innervated axon counts decreased a lot after 4-month waiting (Group 1:2 = 7877 ± 1401 vs. 2365 ± 936.7, P < 0.05), and some more axons were noted in Group 3 after the sensory supercharge (Group 2:3 = 2365 ± 936.7 vs. 5955 ± 3014, P < 0.05). As for the axon diameter, in all the groups, the axon diameters were much smaller than the control group, which is the presentation of the regenerative axons, but there was no statistical significance in between [Figure 4]a and [Figure 4]b. In terms of the fiber diameter and myelin thickness, all the experimental groups were statistically smaller than the controlled group. The immediate repaired group (Group 1) revealed similar fiber thickness than Group 2 (0.773 ± 0.066 vs. 0.737 ± 0.061, P = 0.3918), however, Group 1 and 2 were statistically larger than Group 3 which the sensory component involved (0.578 ± 0.091, P < 0.05). As for the fiber diameter, all the experimental groups were lower than the control group without statistical significance [Figure 4]c and [Figure 4]d, the nerve morphology between the groups showed in [Figure 4]e.
Figure 4: (a and b) The axon count and the axon diameter between the three experimental groups, the contralateral side served as the positive control. The immediate reconstructive group (Group 1) showed more axon counts than the positive control group, but the axon diameter was relatively small. Group 3 showed bigger axon count than Group 2. (c and d) The myelin thickness and diameter of the control group were significantly greater in the control group. Between the experimental groups, the immediate reconstructive group (Group 1) revealed thicker myelin sheath than the delayed groups but no differences between the fiber diameter. (e) The cross section of the nerve in different groups, the normal control showed good size and regular axons, on the other hand, the experimental groups showed the irregular regenerative axons with smaller in diameter but more in total numbersRetrograde labeling
The retrograde labeling of the sensory (dorsal root ganglion of C5 and C6) and the motor neurons (at spinal cord) showed no significant difference between all motor neurons, however, Groups 1 and 3 (392.23 ± 105.33 vs. 366.02 ± 90.78) showed more sensory neurons compared with Group 2 (392.23 ± 105.33 vs. 237.38 ± 71.36), with statistical significance [Figure 5]a and [Figure 5]b.
Figure 5: (a) The retrograde labeling study of the number of motor neurons stained from the spinal cord and the sensory neuron stained from the dorsal root ganglion of C5 and C6. In all the groups, the number of motor neurons was no statistical difference, but Groups 1 and 3 showed more sensory neurons than Group 2 (P < 0.05). (b) The stain figures showed the cross section of the spinal cord segment from C5-T1 level for the motor neuron dye of the three groups (S-1 to S-3), and 577 the sensory neuron on dorsal root ganglion (C5-1-3, C6-1-3). In motor neurons (S-1 to S-3), Group 1 was more than Groups 2 and 3 (delayed repair, without/with sensory neurotization); in sensory neurons (C5-1-3, C6-1-3), 580 the stained neuron of the immediate repair group also the highest, but the number of Group 3 was also higher, contributed by the sensory co-innervation Discussion The most crucial aspect during peripheral nerve reconstruction is the selected time point of surgical intervention. Usually, the earlier the better, but for high-energy traction injury in closed BPIs, intervention at 3 to 5 months postinjury is considered more appropriate as the better time point for surgery.[19],[20] Furthermore, patients with other devastating injuries may need prolonged hospitalization to recover and stabilize. If patients do present at a suboptimal time point after acute injury, there are risks of irreversible limb palsy if the golden time for nerve surgery becomes elusive. In this study, we compared the immediate repair group from the delayed repaired groups to see how the functional recovery was compromised. In addition, we added a third group of sensory protection to the delayed repair group to see if the outcome can be improved.[21],[22],[23]
In the first part of the study, we compared the results of normal control, immediate repair, and delayed repair. From the result of action potential, tetanic force, and the muscle weight, not surprisingly, the immediate repaired group had significantly better outcome than the delayed repaired group. In the second part of the study, we added sensory protection to see if it can decelerate the muscle denervation. Although it can really increase the total number of axons and the muscle weight, the electromyography decreases; therefore, the overall contractility was not changed. This anticipated positive effect was not observed. Based on the result, we cannot conclude that the sensory protection leads to better outcome in this rat model.
The chronic denervated muscle displayed the irreversible structural damage resulting from the combination of fiber necrosis, connective tissue hyperplasia, and exhaustion of satellite cell regeneration.[2],[11] Following prolonged denervation, the muscle deteriorates by downregulating the neurotransmitters/neurotrophic factors and decreasing the number of motor endplates. Histologically, the sensory nerve is not equally comparable to the motor nerves, but it is able to keep the muscular metabolism activates, which might have a protective role as well.[5],[10],[24] By persistently activating neurotrophic signal pathways and secreting the growth factors, the sensory nerve might stimulate the muscle fibers and accelerate the healing of injured native nerves.[25] So far, the effect of sensory protection has been reported mainly for the sciatic or the tibial nerves,[10],[13] while other studies have proposed a modified babysitter procedure of the distal median nerve injury, which implicates the protective effect through a higher muscle weight and stronger grasping force in comparison to the control group.[23] In our study, higher axon count with larger diameter and thinner myelin sheath were observed in Group 3, which could explain the increase of the muscle weight, but no statistical difference in the muscle contraction force and the muscle action potential. The result indicated that the sensory co-innervation may change muscle morphology but is still not powerful enough to augment the functional outcome. Possible explanations may be the timing of our approach and the target muscle we chose. In the delayed groups, the repair was 4 months after the transection, which may be too late to reverse the atrophy of denervated muscle and recover functionally.[9],[11],[25] Li et al.[26] demonstrated a similar model for chronic degeneration of gastrocnemius muscle in rats. After transecting the tibial nerve, sensory nerve protection from the sural nerve in end-to-side or end-to-end fashion to the tibial nerve at a delayed time point resulted in preserved muscle fiber. Another reason for the less-than-ideal outcome in Group 3 might be due to long distance between the site of coaptation and the flexor muscles.
The limitation of the study includes (1) the surgical model cannot mimic the actual clinical scenario – the avulsion injury, since the design of the reconstructive method was more compromised; (2) the result of the sensory protection in the immediate reconstruction group was not studied; and (3) the other surgical models need to be considered as well, such as elbow flexion model – closer to the target muscles.
ConclusionThe proposed sensory co-innervation model showed no strong evidence to improve regeneration of motor nerve in the rat BPI chronic repaired model for finger flexion. Future studies may focus on shorter denervated time or/and shorten distance of the target muscle from the nerve coaptation site.
Financial support and sponsorship
This study was supported by the grant of the Chang Gung Memorial Hospital, Linkou, Taiwan (CMRPG3G0111 and CMRPG3E0851).
Conflicts of interest
There are no conflicts of interest.
References
留言 (0)