Rerouting Peripheral Nerves to Restore Function
Nerve to Nerve Grafting
Nerve to nerve grafts. Plastic surgeons have long used peripheral nerve to nerve bridging to restore function to denervated skin or muscle. For example, Bell's palsy is a condition where the facial nerve has been damaged, usually on one side. Both sensory and motor loss occurs on one side of the face. Although most people recover substantially from Bell's palsy, disfiguring residual deficits may remain. To correct this condition, surgeons use a nerve bridge from the intact facial nerve on the other side to the other side. Motor axons from the intact facial nerve will grow into the distal stump of the damaged facial nerve, innervate muscles, and return a smile to the face.
Peripheral nerve bridge from above the injury site to muscle. Nerves above the spinal cord injury site can be rerouted to muscle or bladder below the injury site. As illustrated below in figure 3, injury to the cord disrupts ascending and descending axons. A peripheral nerve bridge from a spinal root or nerve above the injury site to a muscle below the injury site may lead to functional innervation of the muscle. In this rerouting procedure, the nerve to the muscle is cut. Because axons grow slowly (<1 mm/day), depending on the distance, regeneration may take months or years, by which time the muscle may have undergone atrophy. The patient must learn to use motor and sensory circuits above the injury site to control and sense different muscles. Finally, a peripheral nerve may have to be sacrificed for the bridge. The sural nerve, a sensory nerve in the leg, is usually used for this purpose.
Professor Giorgio Brunelli of the University of Brescia carried out nerve-to-nerve grafting in several patients. He dissected out a branch of the ulnar nerve (a nerve to the hand) and tunnelled it subcutaneously (under the skin) to the hip where he connected it to the peroneal branch of the sciatic nerve. Two patients who had this nerve-to-nerve grafting procedure recovered some ability to activate muscles in the legs voluntarily. Perani, et al. (2001) reported functional MRI studies of these patients, showing activation of the arm regions of the sensorimotor cortex when these patients moved their legs. This surgery, however, has some limitations and drawbacks. First, it requires a paraplegic who still has intact motor and sensory function in the hands. Second, it sacrifices a nerve from the arm. One advantage of this procedure, however, is the length of the ulnar nerve. It can reach all the way to the peroneal nerve without requiring a bridge graft. This should reduce the distance that the axons have to regenerate.
Nerve to Spinal Cord Grafting
In 1981, David & Aguayo showed that spinal axons will grow into peripheral nerves that are inserted into the spinal cord. This discovery, confirmed many times in animal studies, is beginning to be used to repair of brachial or lumbosacral plexus injury. These plexi are networks of peripheral nerves emanating from the cervical and lumbar cord respectively. Trauma frequently tears, stretches, compresses, or even pull out the spinal roots. If an injured dorsal root is reinserted into the spinal cord, the axons do not grow into the spinal cord well. However, if a ventral root is inserted into the spinal cord, axons will grow out of the spinal cord into the root and regenerate all the way to the muscle.
Ventral root to cord grafts. Carlstedt, et al. (2000) at the Royal National Orthopedic Hospital at Stanmore in London recently reported 10 cases in which they re-inserted ventral nerve roots back into the spinal cord. In most of the cases, they found evidence for regeneration of motor axons through the ventral root to muscle. At 9-12 months after injury, 3 of the 10 cases recovered useful (grade 3 by Medical Research Council standard) muscle power. In some of the patients, they observed co-contraction of agonist and antagonist muscles, something that is not normally seen, suggesting that some of the axons grew into the wrong muscles. They also found a surprising degree of sensory function, including both thermal and mechanical induced sensations, as well joint position. The mechanism of the sensory recovery is not well understood. At recent meetings, Carlstedt has reported larger series where he inserted the peripheral nerve directly into the spinal cord and observed motor recovery.
Nerve to cord grafts. Nerve-to-cord grafts, however, differs from root-to-cord in several important respects. First, the dorsal root sensory ganglia that contains the sensory neurons are usually situated outside the dura. Insertion of the nerve into the spinal cord may allow axons from neurons and axons from the cord to grow into the nerve but not the dorsal root sensory ganglion. Therefore, sensory recovery should be limited. Second, axons that are not from motor neurons may grow into the peripheral nerve graft. Since motor axons use the neurotransmitter acetylcholine to activate muscles and sensory axons do not form muscle endplates, it is not clear what happens when non-acetylcholinergic axons innervate muscles.
Brunelli (2001) recently reported bridging a peripheral nerve (peroneal) to the T6 spinal cord in a woman with a T8 spinal cord injury. Over a period of 2 years, the woman recovered some voluntary movement in the muscles innervated by the bridged nerve. While the recovery is not yet sufficient for functional locomotion, the procedure suggests strongly that the muscles were innervated by axons that directly came from the spinal cord. The figure below illustrates some of the potential mechanisms by which this recovery might have occurred, as well as other potential explanations. In earlier studies of monkeys, Brunelli & Brunelli (1996) had shown that corticospinal tract axons can grow out of the peripheral nerves that have been inserted into the spinal cord. Some of the recovery from the patient, however, cannot be readily explained by the grafting. For example, Brunelli reported in a recent meeting Brescia that she was recovering some bladder function. This would of course suggest that she may have been incomplete to begin with and is now recovering additional function.
According to a recent report by Larry Johnston in Paraplegia News, Dr. Shaocheng Zhang in Shanghai has rerouted the intercostal nerve to spinal cord nerve roots below the injury. Specifically, after microsurgically releasing and decompressing the cord, he bridged intercostal nerves to the distal stumps of spinal roots that controlled various functions (e.g., muscle function, bladder control, or sensation). In over 30 patients that he followed for an average of 2.5 years, some regained lower extremity muscular control and could stand up and walk a short distance with crutches and braces. Several had improved bowel and bladder control and proprioception. It is important that these reports of recovery be carefully and critically documented because recovery may occur for many reasons besides the grafting.
Peripheral nerve re-routing poses several risks that people should understand before they consent to the procedure. The first and most important risk is that one or more peripheral nerves will be sacrificed. For example, in the case of grafting an ulnar nerve to the peroneal nerve, the person loses the function of the ulnar nerve. In addition, the peroneal nerve which innervates several of the most important muscles of the leg will be cut in the process. The muscles innervated by the peroneal nerve will undergo some atrophy. If the nerve graft does not result in regeneration and reinnervation of those muscles, the muscles may become severely atrophic. These muscles are not only critical for eventual walking but represent padding that prevents decubiti. Finally, when a sural nerve graft is used as a bridge graft, the cutaneous sensory functions served by that nerve will be lost as well.
A second risk is damage to the spinal cord when the nerves are implanted into the spinal cord. Some white matter and gray matter will be damaged in the process. Any damage to the spinal cord above the injury site is worrisome, particularly in people who have cervical spinal cord injury. Great care, therefore, must be taken to ensure that the insertion of the nerve into the spinal cord be done at a site that does not contain crucial neurons and spinal tracts. Given the likelihood of regenerative therapies becoming available in the future, damage to the spinal cord below the injury site should also be avoided as much as possible.
A third risk stems from the use of neurons that are not normally involved in locomotion to activate lower limb muscles. While several groups have now shown that peripheral nerve rerouting can result in voluntary activation of leg muscles, the neurons that are inducing the muscle activation are not part of the neuronal network coordinated by the lumbar locomotor center. It is not yet clear that people can learn to use upper spinal cord circuits to achieve coordinated locomotion. For example, thoracic or cervical neurons may not be well-coordinated with lumbar sensory input and activity of other muscles in the legs. These risks are not trivial because the rerouting will be difficult to reverse.
In summary, surgeons have long used peripheral nerve re-routing to restore function denervated areas of the body, taking advantage of the ability of peripheral nerves to support long-distance axonal growth. For example, it is frequently used to reinnervated denervated muscles or skin resulting from peripheral nerve or cranial nerve damage. Over 20 years ago, David & Aguayo reported that if a peripheral nerve were inserted into the spinal cord, axons from neurons in the spinal cord will grow into the nerve. This has been now tried in several people with spinal cord injury injury. For example, Brunelli described a recent case of a young woman who received a nerve bridge from the peroneal nerve to the spinal cord above the injury site. Although there was some recovery in this case, the mechanisms of the recovery are still not well understood and there is a need to demonstrate that the procedure can induce coordinated locomotor recovery. People should understand the risks of these procedures, including the sacrifice of one or more peripheral nerves, possible damage to the spinal cord from the implantation of nerves into the spinal cord, and the unproven possibility that the rerouted circuits can participate in coordinated motor function.
Brunelli GA (2001). Direct neurotization of muscles by presynaptic motoneurons. J Reconstr Microsurg. 17 (8): 631-6. Summary: The spinal cord cannot heal after severance because the central nervous system is "non-permissive" to the advancement of axons that regrow from presynaptic motoneurons. With the aim of overcoming paraplegia, the author has carried out extensive experimental research since 1980, first in rats and subsequently in monkeys, severing the cord and connecting its cephalad stump with the muscular nerve branches by means of peripheral-nerve grafts, and using various surgical protocols. Functional connections were established, ascertained by physical, electrophysiologic, and histologic examinations. In this reported study, it is demonstrated that presynaptic motoneurons are also able to reconstruct the cytoskeleton of peripheral neurons, as well as motor end-plates. The possibility of elongation of the axons of presynaptic motoneurons into the peripheral nerve up to the muscle had not previously even been hypotheized. This possibility, which has now been validated, can open the door to new surgical techniques for spinal-cord lesions. In addition, the author presents preliminary results from a single human case, utilizing the surgical procedures of the preceding animal experiments. Department of Orthopaedics, University of Brescia, Italy.
Brunelli GA and Brunelli GR (1996). Experimental surgery in spinal cord lesions by connecting upper motoneurons directly to peripheral targets. J Peripher Nerv Syst. 1 (2): 111-8. Summary: This research was aimed at assessing the possibility to connect central motoneurons with skeletal muscles through PNS segments bypassing a lesion of the spinal cord. The investigation was performed in 20 non-human primates (Macaca fascicularis). The surgical paradigm consisted of anastomosing the lateral bundle of the spinal cord directly with the sciatic nerve of the right hindlimb, using the peroneal nerve as a graft. The animals were followed-up clinically for 18 months; at the end of this observation period, they underwent electrophysiological examinations before being killed. Specimens were taken from the spinal cord, graft, sciatic nerve and potentially reinnervated muscles, and processed for routine light microscopy and immunohistochemistry. Postoperative mortality was fairly high (six monkeys), yet the overall outcome was regarded as very good because the animals were neither restrained nor intensively cared for. Five of the surviving monkeys showed clinical, electrophysiological and histological evidence of successful reinnervation. This research demonstrated that upper CNS motoneurons are potentially capable of elongating neuritic processes into the endoneural tubes of a connecting graft, up to reaching a peripheral nerve (sciatic), and restoring functional connections with the relevant skeletal muscles. Department of Orthopedics, University of Brescia Medical School, Italy.
Carlstedt T, Anand P, Hallin R, Misra PV, Noren G and Seferlis T (2000). Spinal nerve root repair and reimplantation of avulsed ventral roots into the spinal cord after brachial plexus injury [see comments]. J Neurosurg. 93 (2 Suppl): 237-47. Summary: OBJECT: The authors review the first series of 10 cases in which injured intraspinal brachial plexus were surgically repaired. They describe the technique of spinal cord implantation or repair of ruptured nerve roots, as well as patient outcome. METHODS: Spinal root repair/implantation was performed from 10 days to 9 months postinjury. There were nine male patients and one female patient. Postoperatively in most cases, regeneration of motor neurons from the spinal cord to denervated muscles could be demonstrated. The first signs of regeneration were noted approximately 9 to 12 months postoperatively. Useful function with muscle power of at least Medical Research Council Grade 3 occurred in three of 10 cases. Magnetic brain stimulation studies revealed a normal amplitude and latency from the cortex to reinnervated muscles on surgically treated and control sides. A certain degree of cocontraction between antagonistic muscles (for example, biceps-triceps) compromised function. With time there was a reduction of cocontractions, probably due to spinal cord plasticity. In these cases there was also, surprisingly, a return of sensory function, although the mechanism by which this occurred is uncertain. Sensory stimulation (thermal and mechanical) within the avulsed dermatomes was perceived abnormally and/or experienced at remote sites. There was some return of patients' sense of joint position. CONCLUSIONS: A short time lag between the accident and the surgery was recognized as a significant factor for a successful outcome. Reimplantation of avulsed nerve roots may be combined with other procedures such as nerve transfers in severe cases of brachial plexus injury. Department of Orthopaedics, Karolinska Hospital, Stockholm, Sweden. firstname.lastname@example.org.
David S and Aguayo AJ (1981). Axonal elongation into peripheral nervous system "bridges" after central nervous system injury in adult rats. Science. 214 (4523): 931-3. Summary: The origin, termination, and length of axonal growth after focal central nervous system injury was examined in adult rats by means of a new experimental model. When peripheral nerve segments were used as "bridges" between the medulla and spinal cord, axons from neurons at both these levels grew approximately 30 millimeters. The regenerative potential of these central neurons seems to be expressed when the central nervous system glial environment is changed to that of the peripheral nervous system.
Perani D, Brunelli GA, Tettamanti M, Scifo P, Tecchio F, Rossini PM and Fazio F (2001). Remodelling of sensorimotor maps in paraplegia: a functional magnetic resonance imaging study after a surgical nerve transfer. Neurosci Lett. 303 (1): 62-6. Summary: The adult mammalian brain has the capacity of reorganising its neural connections in response to lesions/modifications of the peripheral and central nervous system. We show in vivo, using functional magnetic resonance imaging (fMRI), that in paraplegics the lower-limb sensorimotor cortex is invaded by the arm representation. This functional reshaping appears to be reversible. Indeed, surgical transfer of the ulnar nerve to the ipsilateral quadriceps and hip muscles allowed their contraction in a paraplegic patient. During fMRI, these voluntary movements activated the hip and thigh representation in sensorimotor cortex. We suggest that the functional recovery of the lower-limb functional maps might have been driven by the restored somatosensory inputs from the reactivated periphery. The voluntary movements of the lower-limbs are regained through the 're-awakening' of the corresponding sensorimotor cortex. Institute of Neuroscience and Bioimaging-CNR, University Vita-Salute HSR, Milano Via Olgettina 60, 20132 Milan, Italy. email@example.com.