[Impact of neck dissection in N2-3 oropharyngeal squamous cell carcinomas treated with definitive chemoradiotherapy: An observational real-life study].

PURPOSE: The purpose of this study was to assess the efficacy in terms of neck failure of an initial neck dissection before definitive chemoradiotherapy in N2-3 oropharyngeal squamous cell carcinomas, as well as the dosimetric impact and the acute and delayed morbidity of this approach.

MATERIALS AND METHODS: All patients consecutively treated between 2009 and 2018 with definitive chemoradiotherapy using intensity-modulated conformal radiotherapy (IMRT) for a histologically proven N2-3 oropharyngeal squamous cell carcinomas were retrospectively included. The therapeutic approach consisted of induction chemotherapy, followed by cisplatine-based chemoradiotherapy preceded or not by neck dissection. Neck dissection was discussed on a case-by-case basis in a dedicated multidisciplinary tumour board for patients with a dissociated response to induction chemotherapy, defined as a better response on the primary than on the node. Chemoradiotherapy without neck dissection was systematically performed in case of a major lymph node response to induction chemotherapy (decrease in size of 90% or more). Intensity-modulated radiotherapy using a simultaneous-integrated boost delivered 70Gy in 35 fractions on macroscopic tumour volumes, 63Gy on intermediate-risk levels or extra-nodal extension and 54Gy on prophylactic lymph node areas.

RESULTS: Two groups were constituted: 47 patients without an initial neck dissection (62.7%), and 28 patients with a neck dissection prior to definitive chemoradiotherapy (37.3%). Initial patient characteristics were not statistically different between the two groups. The median follow-up was 60.1months (range: 3.2-119months). Incidence of neck failure was higher in patients without neck dissection (P=0.015). The neck failure rate at 5years was 19.8% (95% confidence interval: 7.4-30.6%; P=0.015) without neck dissection versus 0% following neck dissection. All lymph node failures occurred in the planned target volume at 70Gy. Upfront neck dissection suggested a decrease in the mean dose received by the homolateral parotid gland (P=0.01), mandible (P=0.02), and thyroid gland (P=0.02). Acute toxicity of chemoradiotherapy after neck dissection suggested a reduction in grade≥3 adverse events (P=0.04), early discontinuation of concomitant chemotherapy (P=0.009) and feeding tube-dependence (P=0.008) in univariate analysis. During follow-up, there was no difference between the two groups in terms of xerostomia, dysgeusia, dysphagia or gastrostomy dependence in univariate analysis.

CONCLUSION: Neck dissection prior to definitive chemoradiotherapy in N2-3 oropharyngeal squamous cell carcinoma was associated with high neck control without additional mid and long-term morbidity.