Radiation Dose Sensitivity of Subregions of the Larynx to Patient-Reported Swallowing Outcomes

Abstract

PurposeAssess any correlation between swallowing dysfunction and radiation dose to five subregions of the larynx. Methods and MaterialsA cohort of 136 head and neck cancer patients, treated with either photon or proton radiotherapy, was assessed using an endpoint of patient-reported swallowing scores, evaluated with the EORTC QLQ-H&N35 survey, within one month following treatment. Five subregions of the larynx were contoured and dosimetric metrics were extracted for each subregion as well as the total larynx. Univariate and multivariate logistic regression statistical analyses were used to determine statistical correlation with the dose metrics and clinical variables. Univariate regression models were statistically compared using a non-nested model test. ResultsUnder univariate analysis, unilateral vs. bilateral nodal irradiation (p=0.004), concurrent chemotherapy (p=0.007), and surgery (p=0.015) were statistically significant predictors of poor swallowing score. Unilateral vs. bilateral irradiation was statistically significant under multivariate analysis (p=0.039). The epiglottis was the most predictive subregion of swallowing score with a majority (21 out of 25) of dosimetric variables being identified as statistically significant. The maximum dose to the epiglottis was the most significant dosimetric variable tested for poor swallowing score in both univariate (p=0.003) and multivariate (p=0.051) analyses. Comparison of univariate models indicated a general preference for epiglottic variables with the mean dose to the epiglottis being preferred at a statistically significant level in many cases. ConclusionsThese results show the relatively increased sensitivity of the epiglottis compared to the rest of the larynx when considering patient-reported decrements in quality-of-life swallowing score and support both the inclusion of the epiglottis in standard larynx contours and the assessment of the epiglottis dose during plan evaluation. Our data suggests that keeping the mean and max doses to the epiglottis <20–37 Gy and <53–60 Gy, respectively, will reduce swallowing difficulties.