Shoulder disability after different selective neck dissections (levels II-IV versus levels II-V): a comparative study

OBJECTIVES/HYPOTHESIS: The objective was to compare the results of clinical and electrophysiological investigations of shoulder function in patients affected by head and neck carcinoma treated with concomitant surgery on the primary and the neck with different selective neck dissections. STUDY DESIGN: Retrospective study of 40 patients managed at the Department of Otolaryngology, University of Brescia (Brescia, Italy) between January 1999 and December 2001. METHODS: Two groups of 20 patients each matched for gender and age were selected according to the type of neck dissection received: patients in group A had selective neck dissection involving clearance of levels II-IV, and patients in group B had clearance of levels II-V. The inclusion criteria were as follows: no preoperative signs of myopathy or neuropathy, no postoperative radiotherapy, and absence of locoregional recurrence. At least 1 year after surgery, patients underwent evaluation of shoulder function by means of a questionnaire, clinical inspection, strength and motion tests, electromyography of the upper trapezius and sternocleidomastoid muscles, and electroneurography of the spinal accessory nerve. Statistical comparisons of the clinical data were obtained using the contingency tables with Fisher's Exact test. Electrophysiological data were analyzed by means of Fisher's Exact test, and electromyography results by Kruskal-Wallis test. RESULTS: A slight strength impairment of the upper limb, slight motor deficit of the shoulder, and shoulder pain were observed in 0%, 5%, and 15% of patients in group A and in 20%, 15%, and 15% of patients in group B, respectively. On inspection, in group B, shoulder droop, shoulder protraction, and scapular flaring were present in 30%, 15%, and 5% of patients, respectively. One patient (5%) in group A showed shoulder droop as the only significant finding. In group B, muscle strength and arm movement impairment were found in 25% of patients, 25% showed limited shoulder flexion, and 50% had abnormalities of shoulder abduction with contralateral head rotation. In contrast, only one patient (5%) in group A presented slight arm abduction impairment. Electromyographic abnormalities were less frequently found in group A than in group B (40% vs. 85% [P = .003]), and the distribution of abnormalities recorded in the upper trapezius muscle and sternocleidomastoid muscle was quite different: 20% and 40% in group A versus 85% and 45% in group B, respectively. Only one case of total upper trapezius muscle denervation was observed in group B. In both groups, electroneurographic data from the side of the neck treated showed a statistically significant increase in latency (P = .001) and decrease in amplitude (P = .008) compared with the contralateral side. There was no significant difference in electroneurographic data from the side with and the side without dissection in either group. Even though a high number of abnormalities was found on electrophysiological testing, only a limited number of patients, mostly in group B, displayed shoulder function disability affecting daily activities. CONCLUSION: The study data confirm that clearance of the posterior triangle of the neck increases shoulder morbidity. However, subclinical nerve impairment can be observed even after selective neck dissection (levels II-IV) if the submuscular recess is routinely dissected.