Accessory Nerve Function After Modified Radical and Lateral Neck Dissections


  • This study was carried out in Gazi University Medical Center, 06510 Besevler, Ankara, Turkey.


Objectives: Evaluate preoperative and postoperative electrophysiological changes related to the accessory nerve with reference to dissection technique, modified radical neck dissection, and lateral neck dissection.

Study Design: Prospective electrophysiological analysis of accessory nerve function in a total of 20 laryngeal carcinoma patients after neck dissection, 12 being lateral neck dissection (4 bilateral) and 8 being modified radical neck dissection.

Methods: Distal latencies, compound muscle action potentials, and electromyography findings were investigated before surgery and, in early and late postoperative periods in 20 laryngeal carcinoma patients. Results were evaluated by Student t test and χ2 test for intragroup and intergroup differences.

Results: In the lateral neck dissection group, postoperative distal latencies were longer, without statistical significance, whereas in the modified radical neck dissection group postoperative latencies were statistically longer. Postoperative compound muscle action potentials were significantly lower in both groups. Electromyographic work-up showed deterioration in early postoperative periods and improvement in late postoperative periods. When intergroup differences were compared, both postoperative compound muscle action potential and electromyographic findings were worse in the lateral neck dissection group.

Conclusions: The accessory nerve function after modified radical neck dissection is better than function after lateral neck dissection because of increased stress applied to the nerve during retraction of the sternocleidomastoid muscle for achievement of a better exposed surgical field in lateral neck dissection.


Radical neck dissection, first described by Crile at the beginning of this century, has been an important milestone in the surgical treatment of head and neck cancer with cervical metastasis. In this procedure the sternocleidomastoid muscle, internal jugular vein, and accessory nerve are removed along with the lymph atics. Excision of these three nonlymphatic structures, however, usually results in significant functional loss and an increase in postoperative morbidity. Functional neck dissection preserving the three nonlymphatic structures is said to yield the same oncologic treatment results as radical neck dissection.1 This observation has led to application of modified radical neck dissection in such cases, and sparing the accessory nerve even in clinically positive necks does not cause any deleterious effect on survival and neck recurrence.2-4

Sacrificing the accessory nerve during radical neck dissection results in dysfunction of the trapezius muscle. Ewing and Martin defined the classical symptoms related to accessory nerve-sacrificing dissections in 19525: 1) drooping of the shoulder, 2) limited forward flexion of the shoulder, 3) limited lateral abduction, 4) rotation of the scapula, and 5) electromyographic changes. These symptoms may lead to significant restrictions in the patient's professional and social life.6 In addition to resection of the accessory nerve, the use of pectoralis major myocutaneous flaps together with postoperative radiotherapy increases adverse effects on shoulder movements and postoperative morbidity.7 In contrast, resection of the accessory nerve during radical neck dissection does not always result in disability of shoulder function.8 In some radical neck dissection operations, shoulder function remained normal or close to normal in 25% of patients, as shown by electromyography (EMG),9 and in 67% of patients who were questioned for symptoms related to loss of trapezius muscle function.8 This finding has been ascribed to a subfascial innervation pattern of the plexus, which has a deep course in the neck.10,11 On the other hand, the probability of functional impairment of the shoulder joint after modified radical neck dissection that spares the accessory nerve is approximately 20% to 30%.8,12,13 In fact, this functional impairment is an unexpected finding because the trapezius muscle has both superficial and subfascial anastomosis and it may suggest that preserving anatomical integrity of the accessory nerve may not guarantee absolute adequacy of its function. In this study, lateral neck dissection and modified radical neck dissection with preservation of the accessory nerve were compared in terms of accessory nerve function.


Twenty laryngeal carcinoma patients without previous surgical intervention were included in this study. Besides the surgical procedure directed to the primary site of involvement, 12 patients underwent 16 lateral neck dissection procedures (4 of them were bilateral) in which the accessory nerve was preserved and levels II, III, and IV were dissected. This group comprising N0 patients was group I. Eight patients with N1 neck underwent modified radical neck dissection, again sparing the accessory nerve (group II). All of the patients recruited for the study were men. The mean age was 54.2 years in the lateral neck dissection group (group I), and 48.6 years in the modified radical neck dissection group (group II).

Electrophysiological studies were performed before surgery, during the third postoperative week, and 3 months thereafter to compare the effects of these procedures on the accessory nerve and the trapezius muscle. Accessory nerve conduction was studied utilizing the Cherrington method.14 Surface electrodes were used for recording. An active electrode was placed on the upper part of the trapezius muscle, at the level of C7 spinous process, at an approximate lateral distance of 5 cm. A reference electrode was placed on the tendon of the muscle. Electrical stimulation was applied superficially between the clavicle and the mastoid protuberance, and behind the sternocleidomastoid muscle when present. Stimulus intensity and duration were increased until the compound muscle action potential (CMAP) amplitude was elicited maximally. The distance between the recording and the stimulus electrodes was measured by a tape. The distance was the same for all 3 tests in each patient. Distal latency was measured as the time from stimulus onset to the first negative deflection of the CMAP. CMAP amplitude was measured peak to peak in mV. Needle EMG of the trapezius muscle was performed using concentric needle electrodes. Recordings were obtained during rest and mild and maximal contractions. Denervation potentials at rest, presence of motor unit potentials, during voluntary contraction, and recruitment pattern during maximal contraction were investigated.

Results were analyzed for intragroup and intergroup differences. Accessory nerve conduction findings were studied using Student t test. Fisher's exact χ2 test was used for the interpretation of the needle EMG findings. A P value <.05 was considered significant.


There was no statistically significant preoperative difference in the accessory nerve distal latency between groups I and II (P > .5). In early postoperative phase, no CMAP was recorded in seven patients (43.75%) in group I and in three patients (37.5%) in group II. No statistically significant difference was observed between the groups (P > .05). In patients with recordable CMAPs, no significant difference was found in early postoperative distal latencies of the two groups (P > .5). In the late phase all patients had recordable CMAPs, again without statistical significance in the distal latencies between the two groups (P > .05). Although there was no difference between the amplitudes in the preoperative and postoperative late periods (P > .5, P > .05 respectively), those of the early postoperative period were found to be significantly lower in the selective neck dissection group when compared with the modified radical neck dissection group (P < .05) (Table I).

Table TABLE I.. Accessory Nerve Conduction Findings.
original image

In group I, distal latency prolongation was observed in the postoperative early and late phases with no significant difference in between (P > .05, P > .5, respectively). In these patients, CMAP amplitude was significantly lower in the early and late postoperative phases (P = .000, P = .000). There was a significant increase in the amplitudes during the third month compared with the early phase (P < .05) (Table II).

Table TABLE II.. Electrophysiological Findings in Selective Neck Dissection.
original image

Patients who underwent modified radical neck dissection had significantly abnormal distal motor latencies in the postoperative early and late phases compared with the preoperative period (P < .05, P < .05, respectively). Three months after surgical intervention distal latencies improved compared with the early postoperative period without being statistically significant (P > .05). In this group CMAP values were significantly low in the early and late postoperative recordings (P < .05 and P < .005, respectively), but the early and late CMAP amplitudes did not show any significant difference (P > .05) (Table III).

Table TABLE III.. Electrophysiological Findings in Modified Radical Neck Dissection.
original image

Needle EMG of the trapezius muscle revealed denervation potentials in all 16 (100%) patients in group I and in 6 (75%) patients in group II in the early postoperative phase. There was no significant difference between the groups (P > .05). In the late phase denervation potentials were noted in four (25%) patients in group I. Spontaneous activity was no longer present in the group with modified radical neck dissection. Motor unit potential loss in the early phase was observed in 10 (62.5%) patients in group I and 3 (37.5%) patients in group II without a statistically significant difference in between (P > .05). Needle EMG was repeated 3 months later and demonstrated motor unit potentials in all patients. During voluntary maximal contraction, recruitment pattern was abnormal in the early phase in 16 (100%) patients in group I and 6 (75%) patients in group II, and in the late phase in 4 patients (25%) and 1 patient (12.5%), respectively (P > .05 and P > .5) (Table IV).

Table TABLE IV.. Postoperative Electromyographic Findings of the Trapezius Muscle.
original image


One of the primary goals of modification of radical neck dissections is the preservation of the accessory nerve and thereby the decrease of postoperative morbidity. The relationship between preservation of the accessory nerve and postoperative morbidity was investigated by several authors and shoulder function was found to be better in any nerve-sparing dissection.7,8,14,15 Although functional superiority of accessory nerve-sparing neck dissection over accessory nerve-sacrificing neck dissection is more or less established, there are few studies comparing accessory nerve functional changes between various types of nerve-sparing neck dissections.13,16

Leipzig et al.13 and Sobol et al.16 commented on differences between nerve-sparing dissections. Leipzig et al.13 found lesser degree of shoulder dysfunction (30%) when the nerve, muscle, and vein were preserved during two types of neck dissection-supraomohyoid and functional-than those with nerve preservation only-modified radical neck dissection (50%) and classical radical neck dissection (60%). They concluded that there might be a functional disability associated with any type of neck dissection in which the accessory nerve is preserved. This effect was considered to be related to some degree of traction on the accessory nerve and this was more prominent for modified radical neck than the supraomohyoid and functional neck dissection group. Sobol et al.16 compared functional results of three types of neck dissections, namely classical radical neck dissection, modified radical neck dissection, and supraomohyoid neck dissection, on the basis of electromyographic and range of motion assessments of spinal accessory nerve function. Those patients who underwent supraomohyoid neck dissection had significantly better results than patients who underwent either modified radical or classical radical neck dissection. This difference between supraomohyoid neck dissection with others was attributed to lack of disturbance of the posterior neck during supraomohyoid type, which results in less damage to the accessory nerve.

In a study by Remmler et al.,14 however, shoulder disability of patients who had undergone nerve-sparing dissections (supraomohyoid, posterior, functional, modified radical) were compared with that of patients who had undergone nerve-sacrificing neck dissections (radical neck dissection), but intragroup differences of the nerve-sparing group were not mentioned in their paper. They reported that nerve-sparing dissections were followed by a significant, but temporary and reversible phase of shoulder dysfunction, whereas radical neck dissection was followed by profound and permanent shoulder morbidity.

There is paucity of knowledge on accessory nerve function after lateral neck dissection and modified radical neck dissections and our study was designed to clarify this distinction. Slowing or total loss of nerve conduction in the early postoperative period was due to segmental demyelinization (neuropraxia) induced by traction of the nerve during surgery. In this study, CMAP amplitude decrease was statistically significant in early and late postoperative phases in patients with lateral neck dissection, but prolongation of the distal latencies was not significant during the same periods. In the group with modified radical neck dissection, significant prolongation in the distal latency and decrease of the CMAP amplitudes were observed compared with the preoperative period. When two groups were compared, it was found that CMAP amplitudes in group I in the early postoperative period were significantly lower, and needle EMG of the trapezius muscle yielded a higher rate of pathological findings in the same group. This was compatible with a more pronounced axonal injury compared with modified radical neck dissection technique. The presence of denervation potential in the trapezius muscle was a sign of partial axonal injury (axonotmesis), which was observed in some patients. Serial nerve conduction studies and EMG recordings are important in predicting prognosis of nerve injury. In total nerve lesions, voluntary motor unit potentials disappear on the first day of injury. On the third day, the nerve becomes inexcitable. Positive sharp waves appear on the eighth day, and fibrillation potentials on the fourteenth.17 For this reason, we interpreted our findings on two separate periods after surgery. The electrophysiological parameters during the 14th and 21st days were studied by testing both nerve conduction changes and pathological needle EMG findings on the denervated muscle. Regeneration of the nerve was evaluated on the third month. In all patients, nerve excitability and CMAP amplitudes were increased; denervation potentials were decreased and more motor unit potentials were recruited. It was concluded that the function of the accessory nerve improved after both procedures.

Temporary dysfunction of the accessory nerve shown in our study had been observed by Remmler et al.14 in a similar way. They evaluated the strength of parts of the trapezius muscle that were found to be decreased significantly at 1 and 3 months postoperatively and then recovered at 6 and 12 months when compared with radical neck dissection. Electrodiagnostic tests also showed improvement from 49% at 1 and 3 months to 73% and 92% for the sixth and twelfth postoperative months.

In the present study, we observed functional disturbance of the accessory nerve with both types of neck dissections, modified radical and lateral, in which the nerve was preserved. Relatively worse findings encountered in lateral neck dissection patients when compared with modified radical neck dissection patients were attributed to varying degrees of traction of the nerve during each procedure. In lateral neck dissection, to expose and dissect the lymphatics along the jugular vein, the sternocleidomastoid muscle has to be retracted almost until the end of the operation. This traction is even more attenuated for dissection of the triangle posterosuperior to the accessory nerve. In contrast, in modified radical neck dissection, the nerve is identified and dissected free during the early stages of the procedure and separated from surrounding structures in a better exposed field with less traction achieved by transection of the sternocleidomastoid muscle along the course of the accessory nerve.

Better functional results obtained by Leipzig et al.13 and Sobol et al.16 may appear to contrast our results. In our opinion, this diversion might arise primarily from application of different subtypes of selective neck dissections and comparing their results with those of modified radical neck dissection and secondarily from some changes in the surgical technique. Leipzig et al.13 considered supraomohyoid and functional neck dissection patients as one group and compared their findings with the findings of modified radical neck dissection patients. Sobol et al.16 compared physical dysfunction between supraomohyoid neck dissection and modified radical neck dissection. Our study involved comparison of results obtained after lateral neck dissection and modified radical neck dissection.

Failure of full recovery from shoulder dysfunction after neck dissection despite electrophysiological improvement was explained by Patten and Hillel5 as transient limitation of shoulder movement due to adhesive capsulitis secondary to neck dissections. The thickened and contracted joint capsule is strongly attached to the humeral head and sticks to it like a plaster, deserving the term “adhesive capsulitis,” which may also be encountered in the immobilized upper extremity, hemiplegia, diabetes mellitus, myocardial infarction, and cervical disc disease.18,19 A sedentary lifestyle after surgery could be a predisposing factor in the emergence of shoulder symptoms, although it has been stated that the incidence of adhesive capsulitis is not very high after radical neck dissection.12,13 Although there is a need for controlled studies for the beneficial effect of physiotherapy after different types of neck dissection, early commencement of physiotherapy may have a contributory role in decreasing the possibility of shoulder complaints that may occur in the ensuing months.


After modified radical and lateral neck dissections there may be a dysfunction of the accessory nerve unrelated to the preferred dissection technique during the first 3 months, but the dysfunction may be more severe and sustained in procedures that induce an increased tension on the nerve, and caution must be exercised to avoid traction on the nerve at time of dissection. With time and as experience is accumulated, morbidity related to accessory nerve dysfunction can be minimized or eliminated in all types of nerve-sparing dissections by improvements in surgical and instrumental techniques.

Appendix: VISIT

The Laryngoscope Interactive Case Studies

on the World Wide Web

Review current case studies, written by an expert on the topic, and discover useful techniques and procedures. Our interactive site allows you to share your own experiences and interact with colleagues to examine the advantages and disadvantages of various approaches. A new and exciting case study will be posted every month. Past case studies will always be available to view at the site.