Paraplegia (I 995)n. 359-361 © 1995 International Medical Society of Paraplegia All rights reserved 0031.1758/95$12.00 Repairing spinal roots after brachial plexus injuries MA Glasby! and TEl Hems2 1 Department of Anatomy, University of Edinburgh Medical School, Teviot Place, Edinburgh EH8 9AG; 2Nuffield Orthopaedic Centre, Headington, Oxford, UK The problems of repairing spinal roots after brachial plexus avulsion injuries are discussed in the light of current surgical diagnosis and treatment. An advancing understanding of the cellular mechanisms of nerve regeneration and progress in surgical technology indicates a possibility for the repair at least of ventral roots with grafts which may be of neural or non-neural origin. Enhancement of the regenerative properties may further be made possible by the application of neurotrophic factors at the repair site or centrally. The short and long-term implications of current research into these methods are discussed. Keywords: spinal root repair; nerve regeneration; ventral root implantation; neurotrophic factors The present rarely regained in adults but appears to have a slightly better prognosis in young children.4,5 Over 350 patients suffer severe damage from closed traction injuries of the brachial plexus each year in the United Kingdom.1 Many of these are young adults involved in motorcycle accidents.2 Over recent years, The future surgical exploration has been increasingly recom­ mended if the mechanism of injury suggests that a The typical traction injury sustained by any spinal root major disruption is likely and if examination reveals a consists of rupture of all neural elements whilst the complete absence of function of all or any part of the integrity of the fibrous epineurium is maintained plexus.1,3-5 Operation should be carried out as soon as (Sunderland type IV).9 Repair is by excision of the possible after injury. damaged segment and interposition of a nerve auto­ For the individual patient preoperative investigations graft between the two nerve ends. Normal nerve are directed at distinguishing between pre- and post­ anatomy at least in the macroscopic sense must be ganglionic lesions.6 CT scanning with contrast enhance­ identified at each nerve end and this usually means that ment is probably the best investigation, but no method relatively long nerve grafts are required.5 Technically is 100% accurate.7 long grafts offer no particular problems but the At operation it is important to explore the whole eventual success of any nerve grafting technique is plexus since lesions can occur at any level. Nerve inversely proportional to graft length. In the future it ruptures are required with conventional nerve auto­ may be possible to increase the speed of regeneration grafts or with a vascularised ulnar nerve fraft if the C8 by the use of nerve cell rescue factorslO but the ultimate and Tl roots are irreparably damaged. ,5 At present problem is one of increasing the specificity of regenera­ pre-ganglionic damage is not thought to be amenable to tion so that individual pioneering axons become chem­ repair. However useful function can be restored to ically directed to their proper end organs as they some muscles using nerve transfers (neurotisation) traverse the inevitable geometrical chaos imposed by from the accessory nerve, cervical plexus or intercostal the repair site. nerves.s These transfers appear to have the added A further factor which must affect the eventual bonus of relief of pain in many patients. outcome of any nerve repair is the stretch reflex. 11,12 Results from these procedures are better if repair is Even when after nerve repair, good motor power and carried out within 3 months of injury. If surgery is protective sensation have been restored, the patient performed within days of injury, before Wallerian still complains of poor coordination. It is statistically degeneration has occurred, intraoperative electro­ highly unlikely that any monosynaptic reflex arc will be physiological testing allows recognition of the predomi­ restored after nerve injury and repair and this must be nantly motor bundles of the major nerves, thus expected to have a serious effect upon the servo control facilitating appropriate connection of grafts and trans­ of movement. Barker13,14 in his elegant experiments on fers. Repair of ruptures within the posterior triangle muscle receptor reinnervation has demonstrated that will achieve worthwhile function of the shoulder and both muscle spindles and Golgi tendon organs can elbow in 60% of patients. Useful function of the hand is become reinnervated with each other's homonymous Repairingspinal rootsafter brachial plexus injuries MA Glasbyand TEJ Hems 360 nerve fibres and since these receptors signal quite repair the dorsal roots because their central processes opposite information about muscle movement, the supported by oligodendrocytes are unable to regener­ likelihood of reestablishing useful proprioceptive con­ ate. However one of the interesting possibilities sug­ trol is remote. gested by in vitro studies is that of blocking the At present there are really no options available to the inhibitory tendencies of these cells by.means of specific surgeon for treating avulsed spinal rootS.15 The theoret­ antibodies. Today two major groups are involved in the ical possibility of reimplanting avulsed spinal roots was experimental evaluation of motor root repair and one recognised in the first years of this century by Basil of these groups has a very limited human clinical Kilvingtonl6.17 whose very considerable body of work experience. In Edinburgh, Glasby and Hems have used on peripheral nerve repair has been largely forgotten, freeze-thawed muscle autografts to bridge the short only to be rediscovered by others. Kilvington's goal was defect between the anterior grey horn of the spinal cord reinnervation of the bladder after spinal cord injury by and the motor root of mixed peripheral nerves con­ the rerouting of functioning but unimportant spinal tributing to the brachial plexus and the cauda equina. motor roots originating above the lesion into the distal The experimental animal has been the sheep. Success­ processes of nerves from below the transection. This ful motor reinnervation was demonstrated by these goal still remains to be achieved. workers after transection and after avulsing the ventral In brachial plexus injuries, there has been a resurg­ rootS.19-21 In Stockholm, Carlstedt and colleagues have ence of interest in the reimplant at ion of motor roots performed similar operations in primates in which the after avulsion injury. The nature of the injury may vary avulsed ventral roots have been reimplanted into the 2 - 4 in each case but there is a strong probability that some anterior funiculus of the spinal cord directly. 2 2 There motoneuron cell bodies remain intact in the anterior was a clear and significant reinnervation of muscles grey horn of the injured spinal segment. Theoretically supplied by the injured segment that had shown total axons may regenerate out of the spinal cord into ventral denervation immediately after injury. In a single roots if these are repaired. The environment of the human case, Carlstedt et at (personal communication) peripheral nervous system is able to support regenera­ used sural nerve grafts to effect the intradural repair of tion and there is no reason why fibres should not grow motor nerve roots of the brachial plexus injured by distally and reinnervate target muscles, as occurs after severe traction. Within 3 years these workers observed repair of any peripheral nerve. The appropriateness of recovery of function in all of the muscles supplied by the environment is all-important and implies a similar­ the damaged roots with the exception of the intrinsic ity to peripheral nerve containing Schwann cells and muscles of the hand. This important case indicates that basement membrane rich in neurotrophic and neuro­ surgery was effective in the recovery of function. topic matrix proteins rather than the central nervous However the whole issue clearly needs more extensive system environment which has been shown to inhibit evaluation both clinically and in the research labora­ neurite regrowth. Functional recovery after repair of tory. dorsal roots is not possible, at least in adult animals, as The question of whether a role exists for nerve this would require the central processes of peripheral growth factors in surgical nerve repair remains to be sensory neurons to regenerate into the CNS and make answered. These factors, perhaps more accurately connections. However, reinnervation of the cord from described as 'nerve cell rescue factors' have variously dorsal roots may be possible in neonates. been shown to enhance nerve cell growth in vitro and Early work by BonneyI5 and Jamieson and EamesI8 some dramatic claims and predictions have been made supported this theoretical possibility. These workers by individuals clearly unfamiliar with the practicalities undertook some preliminary studies which indicated of treating nerve injuries. It seems likely that these the possibility of useful repair of ventral but not dorsal substances might, if used appropriately, have a great spinal roots in dogs. In 1977 a single human case deal to offer in enhancing the regenerative capacity of 0, showed that the operation was technically possible and injured nerves. 1 25 However at present there is no clear produced no harmful effects but unfortunately no evidence that any particular factor is of immediate use recovery was seen. As a result of this important early in the clinical situation or in any realistic model of the work a number of points were identified. These have clinical situation. The reason for this is that, besides the set the scene for future experimental and clinical work. need to demonstrate an effectiveness in making cells Because torn spinal roots become rapidly fixed to distal grow, for these factors to be clinically useful, they must structures, it is possible to contemplate successful be shown to be capable of localised delivery at the reattachment only in the first 48 h at most.