Prognostic factors and selection criteria in the retreatment of head and neck cancers.

ABSTRACT

OBJECTIVES: To determine predictors of treatment selection, outcome, and survival, we examined a cohort of previously irradiated head and neck squamous cell carcinoma (HNSCC) patients. MATERIALS AND METHODS: We retrospectively analyzed 100 patients at our institution who were treated for recurrent or second primary (RSP) HNSCC, focusing on subgroups receiving reirradiation (ReRT) alone and those undergoing surgical salvage (SS) with or without post-operative reirradiation therapy (POReRT). Logistic regression modeling was performed to identify factors predictive of retreatment modality. Cox regression modeling was used to determine prognostic factors for progression free survival (PFS) and overall survival (OS). RESULTS: ReRT alone was less likely in current smokers and neck recurrences, with reirradiation more likely in primary site recurrences. POReRT was significantly more likely in patients with positive surgical margins (PSM), neck dissection, or organ dysfunction. POReRT omission negatively impacted PFS when PSM (HR 8.894, 95% CI 1.742-45.403) and perineural invasion (PNI) (HR 3.391, 95% CI 1.140-10.089) were present. Tracheostomy was associated with worse OS, but ReRT alone and POReRT improved OS. PSM correlated with worse OS, regardless of whether POReRT was given (HR 14.260, 95% CI 2.064-98.547). CONCLUSION: This analysis confirms known factors for predicting outcome and shows nonsmoking status and primary site recurrence as predictors for ReRT alone. POReRT for PSM and PNI improves PFS. Tracheostomy patients are more likely to have ReRT due to acute toxicity not limiting treatment and POReRT improves OS compared to surgery alone. The presence of PSM negatively impacts survival which cannot be overcome by POReRT.