Closed Brachial Plexus Trauma
A 22-year-old, right-handed man was riding a snow-mobile when he hit a jump off balance and landed on his right neck and shoulder. He did not lose consciousness but immediately experienced pain and numbness on the lateral aspect of his right shoulder and upper arm as well as complete paralysis in not being able to either abduct his right arm at the shoulder or flex his arm at the elbow. He was taken to the emergency department, where he was seen by a neurologist who scheduled him for a follow-up exam and electromyography/nerve conduction velocity (EMG/NCV) and somatosensory evoked potentials (SSEPs) study in 3 weeks. This study demonstrated 4 + denervation changes in his right supraspinatus, infraspinatus, deltoid, and biceps muscles with no voluntary units. He had an absent median nerve sensory nerve action potential and somatosensory evoked cortical response over his contralateral scalp. Other changes were minimal.
The exam, as described, with no right arm abduction at the shoulder and flexion across the elbow, with numbness along the lateral shoulder and upper arm, is most consistent with an injury to which of the following?
B. The patient may require surgery in the next several months if functional recovery does not occur, both to determine the grade of nerve injury, using electrophysiological techniques and to select the type of nerve repair that would be of greatest benefit.
Most traumatic peripheral nerve injuries, particularly those involving the brachial plexus that represent the nerves traveling between the spinal cord and upper extremity in the neck, are closed and involve stretching and compression of nerves. The amount, duration, and distribution of force applied to nerves will determine the biological grade and extent of nerve injury, which are important determinants of prognosis, treatment, and clinical outcome. There are three basic biological grades of nerve injury that involve progressive involvement of the three major structural and functional components of a nerve (Fig. 103.1).1
Fig. 103.1 This schematic illustrates the three distinct biological grades of nerve injury and relates them to the three basic structural components of a peripheral nerve. A neurapraxic grade of injury involves the Schwann cells, which insulate axons, and thereby produces a reduction or complete blockage of nerve conduction across the damaged demyelinated nerve segment. Because the axons remain intact, the target muscles do not atrophy or show denervation changes on electromyographic (EMG) studies. The next two grades of nerve injury are progressively more severe and do involve interruption of axons, with loss of sensory and motor function and the development of muscle denervation changes on EMG testing. An axonotmetic grade of injury preserves the supporting structures surrounding axons, which can serve as highways upon which regenerating axons can grow to reach their target structures. The most severe neurotmetic grade of nerve injury disrupts the highways, either through intraneural fibrosis or an actual loss of continuity with production of a gap, to the point where axonal regeneration cannot occur. These neurotmetic injuries require a surgical repair with or without a graft for axonal regeneration to occur.
A neurapraxic grade of injury involves injury primarily to the myelin insulation of axons formed by the Schwann cells. A neurapraxic grade of injury results in reduced or absent nerve conduction across the injured segment of nerve. However, because the axons are preserved, stimulation of the nerve distal, but not proximal, to the injured nerve segment will give rise to a normal nerve conduction response, with contraction of muscles innervated by the nerve. An electromyographic (EMG) study will not show significant muscle denervation changes, such as spontaneous fibrillations. Voluntary motor units may be present if there is a partial and not complete nerve conduction block. Neurapraxic injuries almost always result in recovery of full function over a period of weeks to several months and therefore do not require a surgical repair.
An axonotmetic grade of injury involves more severe nerve trauma resulting in damage to axons to the point where distal Wallerian degeneration occurs. A nerve conduction response will not occur, and the EMG will show muscle denervation changes and no voluntary units. However, the extracellular matrix constituting the bands of Büngner, which represent highways upon which axons can grow, are preserved, and therefore recovery through axonal regeneration is possible. The degree of functional recovery through successful axonal regeneration with reinnervation of muscle and sensory organs depends on several factors.2 Important factors are the distance that axons must travel to reach their target structures, the complexity of the nerve in terms of its branching pattern, and whether it is a mixed nerve or only a pure sensory or motor nerve. In general, the shorter the distance that needs to be traversed and the simpler the nerve, the more likely it is that recovery of function will occur. Axonotmetic injuries therefore do not require a surgical repair. However, because axonotmetic cannot always be distinguished from the more severe neurotmetic grade of nerve injury, it is often necessary to wait several months, and, if there is no clinical or electrodiagnostic evidence of nerve recovery, to perform a surgical exploration to determine if axons are regenerating past the segment of nerve damage using intraoperative electrophysiological stimulation and recording techniques. If a nerve conduction response along the damaged nerve pathway can be recorded, then axonal regeneration is occurring, and no repair is performed beyond an external neurolysis, in which the nerve is separated from surrounding tissue, which is often scar.3
The neurotmetic grade of nerve injury is the most severe type and produces disruption of nerve pathways to the point where axonal regeneration cannot occur. One type of neurotmetic injury involves actual cutting of the nerve so that it is no longer in continuity. A second, more common type involves damage to the nerve sufficient to produce enough intraneural fibrosis to block axons from regenerating. This type is called a neurotmetic grade of nerve injury that is in continuity. Like the axonotmetic grade of injury early on, there are no axons to conduct a nerve response, and the muscles show denervation changes on the EMG. Neurotmetic injuries that are not in continuity require identifying and then trimming away scar tissue at the proximal and distal ends of the nerve. If the two nerve ends can be brought together without producing tension, then a primary nerve suture repair can be performed with several small-caliber monofilament sutures supplemented with fibrin glue. If the two ends cannot be approximated without producing tension, then nerve grafting is performed using either small segments of sensory nerve, usually harvested from the sural nerve in the lower leg, or bioabsorbable synthetic nerve tubes if the gap is < 4 cm in length.
A special type of injury occurs when spinal nerve roots are pulled out of, or avulsed from, the spinal cord. In such cases, the motor axons will degenerate and produce muscle denervation with no voluntary motor units. However, if the injury is purely preganglionic (between the dorsal root ganglion and the spinal cord), and not postganglionic (damage to the nerve distal to the dorsal root ganglion), then a sensory nerve conduction response will be recordable. However, the patient will experience no sensation and an SSEP mediated by the avulsed dorsal root will not be present.
In this case, the patient suffered a severe, complete injury involving the upper trunk of his right brachial plexus. Magnetic resonance neurography (MRN) showed no evidence of spinal nerve root avulsion, especially well seen on Fast Imaging Employing Steady-State Acquisition (FIESTA, GE Healthcare, Waukesha, WI) pulse sequence images. MRN can be done instead of a cervical computed tomographic myelography study. Both can visualize meningeal diverticula and other abnormalities involving the spinal nerve roots that are highly suggestive of an avulsion type of injury. The MRN study done did show increased signal on pulse sequences sensitive to water and focal enlargement and contrast enhancement of his left upper trunk. A repeat clinical exam and electrodiagnostic evaluation 4 months after his trauma did not show any evidence of recovery of function in either his right shoulder abductor or his elbow flexor muscles. Given his lack of recovery, documented both clinically and electrodiagnostically, it was elected to take him to surgery for an exploration and possible repair of the upper trunk of his brachial plexus guided by intraoperative findings, both anatomical and electrophysiological.4
Preoperatively, the patient was told that he might require a nerve graft repair using either his sural nerve or a bioabsorbable tube if the gap were < 4 cm. He was referred to a medical center with expertise in dealing with complex peripheral nerve injuries and the application of intraoperative electrophysiological stimulation and recording techniques. A preoperative chest X-ray with inspiration and expiration showed good up and down movement of his diaphragm, indicating that his phrenic nerve was working.
At surgery, he was positioned supine and administered general anesthesia after being intubated. A Foley catheter was also placed. The anesthesiologists were told that no long-acting paralytic agents should be used that might interfere with the physiological assessment of his nerves. His head was placed on donut-shaped padding and turned to the left to expose his right neck, which was prepped along with his entire right upper extremity and lower leg for possible harvesting of the sural nerve. Fine needle sterile electrodes were placed in the right upper extremity muscles as well as his contralateral left scalp for monitoring motor and sensory responses. Baseline values were obtained before infiltrating his incision (above the right clavicle in a skin crease) with local anesthetic with epinephrine. No motor-evoked responses were obtained in his right supraspinatus, infraspinatus, deltoid, and biceps muscles, and there was no median nerve somatosensory response recordable over his scalp, confirming a complete right upper trunk injury.
A surgical dissection was performed in which the following steps were performed: several superficial sensory nerves of the cervical plexus were dissected free and spared, the fat pad in the posterior triangle of the neck was dissected and retracted laterally; several external jugular vein branches were ligated and cut, and the omohyoid muscle was cut for exposure; the phrenic nerve running along the anteromedial surface of the anterior scalene muscle was carefully isolated and encircled with a small vessel loop for protection; the anterior scalene muscle was cut; and the upper trunk was identified with a small branch coming off it proximally, contributing to the phrenic nerve. The upper trunk was found to be extremely scarred over a 3 cm segment, and direct stimulation of it did not elicit any muscle contraction at stimulus thresholds well above those necessary to elicit muscle contraction when stimulating other intact nerves, such as the phrenic nerve and adjacent middle trunk. Under magnification, the scarred upper trunk was circumferentially externally neurolysed from surrounding scar tissue by first working from normal anatomy proximal and distal to the scarred nerve segment: proximally the right C5 and C6 spinal nerves were dissected free and stimulation of both of these elements gave rise to a centrally recordable SSEP response over the left scalp but no distal response resulting in muscle contraction; distally, the anterior and posterior divisions as well as the suprascapular nerve arising from the distal upper trunk had a normal anatomical appearance, though stimulation resulted in no nerve response propagating either proximally or distally, indicating an absence of axons. Once the scarred segment of upper trunk was circumferentially neurolysed from the surrounding tissues, electrical stimulation proximal and distal to this segment produced no nerve action potential indicating a lack of axonal regeneration through the scarred region and indicating a neurotmetic grade of nerve injury in continuity. The scarred segment of upper trunk was resected, and the proximal and distal nerve stumps were carefully trimmed back to normal-appearing fascicular structure. Stimulation of the proximal stump gave rise to a centrally propagating SSEP response, indicating the presence of viable axons. Once the scarred segment of upper trunk was resected, this left a gap of 2 cm, which was too long to bring the proximal and distal nerve stumps together without causing tension and thereby necessitated placing a nerve graft, either sural nerve or a bioabsorbable tube. It was elected to harvest a 10 cm segment from the patient′s right lateral lower leg above the ankle, which was cut into four 2.5 cm segments that were used to bridge the gap and have a little excess length to minimize the chance of being dislodged by tension from stretching postures. Multiple grafts were required because the diameter of the sural nerve was considerably smaller than the diameter of the upper trunk. The four grafts were secured in place with a single 7–0 monofilament suture at each end, followed by wrapping each end with a small piece of surgicel, followed by the application of several mL of fibrin glue to secure the grafts in place.5
The patient′s right arm was placed in an arm sling postoperatively for 2 weeks to avoid placing tension on the grafts as they healed in place. After 2 weeks, he was instructed to begin gentle range of motion exercises about his right shoulder and elbow on a daily basis on his own and under the supervision of a physiatrist on a weekly basis for several months. Six months after the surgery, he began to notice some contraction of his supraspinatus muscle, with less subluxation of his proximal upper arm within the shoulder joint, along with EMG confirmation of reinnervation of his supraspinatus and infraspinatus muscles, with a few voluntary units seen in his right biceps muscle but none in his right deltoid. One year after the surgery, he could flex his arm at the elbow against gravity (4/5 motor strength), and he noted some contraction of his right deltoid muscle. Two years after his surgery, he had strength as follows: right supraspinatus and infraspinatus 4 +, biceps 4, and deltoid 3.
Closed traumatic injuries cause stretching and compression of nerves and often involve the brachial plexus. These injuries can result in avulsion of spinal nerve roots from the spinal cord but otherwise usually leave nerves in continuity. Neurapraxic injuries cause demyelination but leave the axons intact and usually result in recovery of full function over a period of weeks to several months. More severe injuries cause axonal degeneration. If the extracellular substrate upon which axons can grow remains intact, called an axonotmetic grade of injury, then recovery can occur over many months depending on how far the axons must regenerate to reach their target muscles and sensory structures. In more severe nerve injuries, called a neurotmetic grade, the substrate for axonal growth is blocked by intraneural fibrosis, when the nerve remains in continuity, or is absent when the nerve loses its continuity. Neurotmetic injuries require a surgical repair to remove blocking scar tissue and reapproximate the two ends of the nerve, either directly or with a graft when there is a gap of sufficient length to prevent bringing the two ends together without causing tension. In both axonotmetic and neurotmetic grades of injury, the degree and timing of recovery depend on both the distance and complexity of the pathways that axons must traverse to reach their motor and sensory target structures.6
This review article describes the three different biological grades of nerve injury (neurapraxic, axonotmetic, and neurotmetic) in relation to treatment, both medical and surgical, and clinical outcomes.
2. Filler AG, Kline DG. General principles in evaluating and treating peripheral nerve pathology, injuries, and entrapments and their historical context. In: Youmans Neurological Surgery. Vol 3. 6th ed. Philadelphia, PA: Elsevier Saunders; 2011:2361–2367
This textbook chapter describes the intraoperative electro-physiological techniques necessary to distinguish the different grades of nerve injury and help determine the type of nerve repair to be performed.