Objective: Avoiding injury to the motor nerves of the levator scapulae muscle during neck dissection surgery should reduce postoperative shoulder dysfunction. Understanding the surgical anatomic relationships of these motor nerves is imperative for reducing this morbidity. This study was undertaken to elucidate this relevant anatomy, because the surgical and anatomic literature lack adequate description.

Methods: Cervical (C3, C4) and brachial (C5 via dorsal scapular nerve) plexi contributions of the levator scapulae were assessed with respect to posterior triangle landmarks in 37 human cadaveric necks.

Results: An average of approximately two (actual 1.92) nerves from the cervical plexus (range, one to four nerves) emerged from beneath the posterior border of the sternocleidomastoid to innervate the levator scapulae. The two most superior cervical plexus contributions to the levator scapulae emerged from the posterior border of the sternocleidomastoid on average 21.94 mm and 25.68 mm, respectively, caudal to the emergence of cranial nerve XI. These same cervical plexus nerves on average emerged from the posterior border of the sternocleidomastoid 10.32 mm and 14.64 mm, respectively, caudal to the punctum nervosum. An average of approximately two (actual, 1.94) nerves from the cervical plexus (range, one to three nerves) crossed the anterior border of the levator scapulae to either innervate this muscle on its superficial surface or just at its anterior border. Cervical plexus contributions crossed the anterior border of the levator scapulae in a superior to inferior progression. The two most superior contributions crossed the anterior border of the levator scapulae on average 15.03 mm. and 21.50 mm. respectively inferior to this muscle's intersection with the sternocleidomastoid. In any given neck specimen, cervical plexus nerves emerging from the posterior border of the sternocleidomastoid in route to the levator scapulae could branch or come together, which explains the difference in the average number of nerves that crossed the anterior border of the levator scapulae compared with the average number that emerged from the posterior border of the sternocleidomastoid. On average, the dorsal scapular nerve after piercing scalenus medius crossed deep to the anterior border of the levator scapulae 43.23 mm inferior to this muscle's intersection with the sternocleidomastoid. All innervation to the levator was deep to the prevertebral fascia. Among study parameters, statistically significant (p <0.05) differences were not encountered between right and left necks.

Conclusions: The levator scapulae receives predictable motor supply from the cervical and brachial plexi. Our data elucidate surgical anatomy useful to head and neck surgeons.