Elective neck dissection in patients with salivary gland carcinoma: A systematic review and meta-analysis.

OBJECTIVE

Elective neck dissection in patients with salivary gland carcinoma and clinically negative lymph nodes is controversial. Reported proportion of occult nodal metastases vary with histological subtype, tumour classification and preoperative diagnostic methods. This is a systematic review and meta-analysis on the role of END in salivary gland carcinoma.

METHODS

A search in PubMed, Embase and Cochrane was performed. Original articles in English with data on tumour characteristics, clinical and pathological N-classification, and neck dissection were included. Reporting Items for Systematic Reviews and Meta-analyses were followed. Random effect modelling was performed to pool the data. Meta-analysis of proportions was performed for occult metastases overall, for T3/T4 versus T1/T2 tumours and for tumours with high-grade versus low-grade histology. Heterogeneity across studies was assessed with I-squared statistics.

RESULTS

We included 22 articles in the qualitative synthesis and meta-analysis. The pooled proportion of occult metastases was 21%. In patients with T3/T4 tumour, the pooled proportion of occult metastases was 36%, and in patients with high-grade histology, it was 34%. Most studies concluded that END should be performed in patients with advanced T-classification and high-grade histology tumours. Nine studies assessed occult metastases per level.

CONCLUSION

The overall occult metastases proportion does not require END in all patients with salivary gland carcinoma. We recommend END in patients with high-grade or unknown histology or T3/T4 tumours. END should involve level II and III, and level I should be included in tumours in the submandibular gland, sublingual gland and minor oral salivary glands.