Who merits a neck dissection after definitive chemoradiotherapy for N2-N3 squamous cell head and neck cancer?

The role of neck dissection (ND) after definitive chemoradiotherapy for squamous cell head and neck cancer is incompletely defined. We retrospectively reviewed 109 patients with N2-N3 disease treated with chemoradiotherapy to identify predictors of a clinical complete response in the neck (CCR-neck), pathologic complete response after ND (PCR-neck), and regional failure.

All patients were given 4-day continuous infusions of 5-fluorouracil (1000 mg/m2/d) and cisplatin (20 mg/m2/d) during the first and fourth weeks of either once daily (n = 68) or twice daily (n = 41) radiation therapy. ND was considered for all patients after completion of chemoradiotherapy and was performed in 32 of the 65 patients achieving a CCR-neck after chemoradiotherapy and in all 44 patients with residual clinical evidence of neck disease. CCR-neck, PCR-neck, and regional failure were then correlated with potential predictors, including T, N, largest lymph node size (<3 cm, > or =3 cm), primary tumor site, and radiation fractionation schedule.

Achievement of a CCR-neck was predicted by N, N2 vs N3 (53 of 80 vs 12 of 29, p =.019) and by largest lymph node size, <3 cm vs > or =3 cm (19 of 25 vs 46 of 84, p =.06). Achievement of a PCR-neck could not be predicted by any clinical parameter. Regional failure occurred both in patients undergoing ND and those not dissected (5 of 76 vs 4 of 33, p =.33) and proved more likely only in the ND patients with residual positive pathology compared with those achieving a PCR-neck (5 of 25 vs 0 of 51, p <.001). Primary site was not a useful predictor of CCR-neck, PCR-neck, or regional failure. Most importantly, CCR-neck (vs

After chemoradiotherapy, clinical parameters do not identify those patients with residual neck node disease or those at risk for regional failure, suggesting that ND be considered for all N2-N3 patients.