Factors predictive of outcome following primary total laryngectomy for advanced squamous cell carcinoma

Abstract

The increasing use of primary chemoradiation (CRT) for laryngeal squamous cell carcinoma (SCC) means that historical surgical data sets are not representative of the modern laryngectomy patient. We analyse a contemporary total laryngectomy (TL) cohort to identify factors predictive of outcome. This is a retrospective consecutive case note review in a UK tertiary referral centre. Demographic, staging, treatment and outcome data were collected. Oncological outcomes are expressed using the Kaplan-Meier method. The log-rank test was used for univariate analysis and cox regression for multivariate analysis. Sixty consecutive patients between 2003 and 2010 underwent primary TL, 28 including partial pharyngectomy. Median age was 61 years and mean follow-up was 24 months (1-78 months). Thirty six patients died during the study period, 24 of their disease. Of the disease-specific deaths, two occurred peri-operatively, four from local, two from regional and 18 from distant disease [two patients had simultaneous local and distant recurrence (DR)]. Five-year overall survival, disease-specific survival, loco-regional recurrence-free survival and distant recurrence-free survival (DRFS) were 36, 51, 87, 62 %, respectively. Of 17 parameters analysed, pN-stage, extra-capsular spread, a non-cohesive tumour front, thyroid infiltration and involvement of level 6 were significant predictors of disease-specific survival (DSS) on univariate analysis. pN > 1 and the presence of adverse histological features were found to be independent predictors of DSS and DRFS on multivariate analysis. Neither was significantly associated with loco-regional recurrence-free survival. Around half of patients who undergo TL for stage IV SCC will die of disease within 5 years, with most deaths attributable to DR. Surgery provides excellent loco-regional control but patients, especially those with advanced nodal disease and/or adverse histological features, should be thoroughly screened for occult distant disease. Level of evidence 4.