Neck dissection, either elective or therapeutic, is part of the surgical treatment of most patients with oral cavity cancer. An often overlooked morbidity of neck dissection is the shoulder dysfunction syndrome caused by handling of the spinal accessory nerve. Maintaining the anatomical continuity of the nerve does not necessarily translate to maintaining the physiologic function of the nerve.
To compare the shoulder dysfunction between baseline and 3 months in patients undergoing modified neck dissection (MND) subjectively by Range Of Motion score (ROM) and Neck Dissection Impairment Index score (NDII), and objectively by Electromyography study (EMG).
Materials and Methods:
All patients underwent preoperatively and at three months postoperatively shoulder function assessment which included the ROM, NDII and EMG studies. Fourty-seven necks were subjectively and objectively assessed in 45 consecutive treatment naïve histologically confirmed oral cavity cancer patients. Primaries reconstructed with pedicled pectoralis major myocutaneous flap were excluded.
Thirty four patients (75.6%) were farmers by occupation, for whom the normal shoulder function is vital. There were 39 males and 6 females with an age range of 29-91 years. The most common primary subsite was the tongue (21 patients) followed by buccal mucosa (11 patients). All had MND with preservation of all three non-lymphatic structures. Twenty seven patients (60%) were histologically node negative. Eleven patients (23.4%) had level IIA involved, 2 patients (4.3%) had Level IIB involved and none had Level V involved histologically. A median of 7 and 5 nodes was dissected in level IIA and IIB nodal stations respectively.
The study showed that in spite of 50% of the patients having a painful ROM, the degree of ROM was almost the same pre and postoperatively. This was not affected by nodal positivity status nor the positivity at level II.
37/47 (78.7%) shoulders had normal to slight dysfunction at NDII assessment at 1 month, which improved to 44/47 (93.6%) shoulders at 3 months. There was a significant improvement in the QOL scores over 3 months irrespective of the nodal status (p < 0.05).
The median baseline NDII scores at 1 month (38 score) compared to NDII scores at 3 months (46 score) showed a statistically significant correlation indicating better QOL scores and less morbidity. The median standardized NDII score was 70 at 1 month that improved to 90 at 3 months.
At 3 months, even though there was a severe to moderately- severe abnormal EMG finding, the patients had ROM & NDII score of near normal values with no functional deficit seen in the shoulder function clinically.
EMG with NDII correlation showed a weakly negative correlation with Rho value - 0.367 that was statistically significant whereas kappa correlation showed slight level of agreement that was not statistically significant. EMG on correlation with ROM showed a weakly negative correlation with Rho value of - 0.244 that was not statistically significant.
The NDII and ROM are reliable predictors of subjective assessment of shoulder dysfunction. Both showed a positive correlation in terms of prediction of shoulder dysfunction.
EMG did not show any correlation with NDII or ROM. Although, this discordance is also reported in literature between objective and subjective assessment for shoulder dysfuction, almost all the patients went back to their occupation with little effect on their quality of life..