Patterns of lymph node spread of cutaneous squamous cell carcinoma of the head and neck

BACKGROUND: Among patients with cutaneous squamous cell carcinoma (SCC) of the head and neck, recent studies have shown that those with involvement of the parotid gland also have a high incidence of neck node involvement. Treatment of the neck by either surgery or radiotherapy is therefore recommended among patients with parotid SCC, even if clinical examination is negative. The aim of this study was first to analyze patterns of metastatic spread in the parotid and cervical lymph nodes and then to correlate the pattern of involved nodes with the primary cutaneous site in order to guide the appropriate extent of surgery, should neck dissection be used to treat the neck in patients with parotid SCC. METHODS: A cohort of 209 patients with cutaneous SCC of the head and neck and clinically evident regional metastatic disease was reviewed retrospectively from 3 Australian institutions. The distribution of involved nodes was obtained from pathology reports; the anatomic sites of primary cutaneous cancers were then correlated with these findings. RESULTS: Among 209 patients, 171 (82%) had clinical parotid involvement. Of these, 28 had clinical neck disease, whereas 143 had parotid disease alone. Thirty-eight (18%) patients had neck disease only. A total of 199 patients were treated surgically, whereas 10 received radiotherapy alone. Surgery included 172 parotidectomies and 151 neck dissections (93 of which were elective). Primary sites were cheek (21.7%), pinna (20.4%), temple (15.8%), forehead (15.8%), postauricular region (5.9%), neck (5.3%), anterior scalp (5.3%), posterior scalp (3.3%), periorbital (3.3%), nose (2.6%), and chin (0.6%). Among pathologically positive necks, level II was most frequently involved (79%). Level IV (13%) and level V (17%) were only involved in extensive lymph node disease, the exception being for isolated level V metastases from the posterior scalp. CONCLUSIONS: Primary sites were mainly localized to the lateral aspect of the head. Among patients with cutaneous SCC involving the parotid and neck, level II was the most commonly involved neck level. The distribution of involved nodes suggests that in a patient with parotid involvement and a clinically negative neck with an anterolateral primary, a supraomohyoid neck dissection, always including the external jugular lymph node(s) would be appropriate. In the case of a posterior primary, level V should be dissected as well. In patients with parotid SCC and a clinically positive neck, a comprehensive neck dissection is recommended.