Cervical lymphadenectomy of level II encompasses lymph nodes associated with the upper internal jugular vein and the spinal accessory nerve (SAN). Removal of tissue superior to the SAN (submuscular recess-(SMR)) was recently shown to be unwarranted in selected cases of squamous-cell cancer. Thirtyfive patients with non-squamous-cell cancer (SCC) of the head and neck treated with cervical lymphadenectomy were prospectively evaluated. Thirty-seven neck dissection specimens were histologically analysed for the number of lymph nodes involved with cancer. At the time of surgery, level II was separated into the supraspinal accessory nerve component (IIa) and the component anterior to the SAN (IIb).

Neck dissections were most commonly performed for cancer of the thyroid gland (19) followed in frequency by the parotid gland (seven), skin: melanoma (five), basal-cell cancer (two), and other sites (four). Twenty-five neck dissections were modified-selective procedures and 12 were either radical or modified radical neck dissection. Twenty-nine necks were clinically N+ and eight N0. Histological staging was pathologically N+ in 32 neck dissection specimens. Level IIb contained an average of 12 nodes and the IIa component contained a mean of 5.0 nodes. Level II contained metastatic disease in 28 of 32 histologically node-positive specimens (87 per cent). Level IIa was involved with cancer in six cases (16 per cent), five of which were pre-operatively staged as clinically N+. All cases (100 per cent) with level IIa involvement had level IIb positive nodes. Three of the level IIa positive cases were cancer of the parotid gland comprising 43 per cent of this sub-group of patients.

Incidence of involvement of SMR in non-SCC cases is not uncommon. The additional time required and morbidity associated with dissection of the supraspinal accessory nerve component of level II are probably justified when performing neck dissection in cancer of the thyroid gland. The SMR should be excised in cancer of the parotid gland. Large-scale prospective controlled studies with long-term follow-up periods are necessary to support resection of level IIb only.