Segmental architecture of the spinal accessory nerve

Abstract

Nicolas de Tribolet, MD Editor-in-Chief Acta Neurochirurgica The spinal accessory nerve has become a workhorse for brachial plexus reconstruction following traumatic peripheral nerve injury. It is frequently used as an extraplexal motor donor for transfer to injured nerves, including the suprascapular nerve, musculocutaneous nerve, and others [1, 10]. The spinal accessory nerve’s large motor fiber content [3] and infrequent injury from brachial plexus trauma [2] has yielded a high success rate for these transfers [9, 11]. The concept of nerve transfers for peripheral nerve injuries has been extended to apply to spinal cord injury patients [4]— transferring supralesional nerves into infralesional nerves to enable cortical control of the targeted muscle. A 36-year-old man presented to our clinic for consideration of reconstructive surgery for a high cervical cord injury from a surfing accident 14 years previously. His injury consisted of cervical fractures from C1 through C6, with burst fracture and dislocation of C4 and C5. His current lowest level of function was C4 (ASIA impairment scale A); his motor exam consisted of full strength shoulder shrug bilaterally, but without motor control of any muscle of the left upper extremity. On the right he had weak 4/5 (MRC scale) shoulder abduction and 2/5 external rotation, 3/5 pectoralis major muscle, 4/5 elbow flexion, and trace flexor carpi radialis contraction. Our goals were to identify transferable nerves to achieve elbow flexion on the left; the only donor source would be the spinal accessory nerve. Needle EMG was performed to assess for evidence of denervation of the trapezius muscle that would indicate fewer healthy axons for transfer. Very mild chronic denervation was observed in the middle trapezius, whereas there was moderate chronic denervation in the lower trapezius, including polyphasic potentials, rapid firing, and large motor unit potentials (Fig. 1). The pattern of denervation suggested correlation with his spinal cord injury: an MRI demonstrated near-total loss of spinal cord parenchyma from mid-cervical three to rostral cervical six vertebral bodies which form the C4 through C6 nerves. A segmental pattern of innervation of the trapezius muscle has not been previously defined. Several animal models have shown consistent innervation of the trapezius muscle by the C1–C7 spinal cord segments [5], with the predominant innervation deriving from C2–C6 [6, 8, 12]. The bilateral EMG findings suggest to us that the trapezius muscle is segmentally innervated, with rostral muscle coordinated by rostral cervical spinal segments and caudal muscle coordinated by caudal cervical spinal segments. The salient message, along with the previously undetected pattern of innervation, is that the concepts of nerve transfers for brachial plexus injury cannot be uniformly applied to nerve transfers for spinal cord injury. Although the spinal accessory nerve is a cranial nerve by its dural exit, it is a spinal nerve by origin of its motor neuron pool [7]. This segmental architecture is susceptible to high cervical spine injury, reducing the ability of the spinal accessory nerve to serve as a source for transfer of healthy motor neurons. Preoperative electrodiagnostic evaluation of the quality of the innervation of the trapezius muscle is essential for consideration of the M. A. Mahan (*) Division of Neurological Surgery, Barrow Neurological Institute, St. Joseph’s Hospital and Medical Center, 350 W. Thomas Road, Phoenix, AZ 85013, USA e-mail: Neuropub@dignityhealth.org