Chronic radiation-associated dysphagia (RAD) is a dreaded complication following head and neck radiotherapy. Different factors likely contribute to the development of late RAD. We hypothesize that age may be a significant factor predisposing patients to develop long term swallow dysfunction. We sought to model chronic RAD in patients treated with IMRT for oropharyngeal cancer as a function of age at time of treatment and dose to non-target swallowing muscle regions of interest (ROIs). We examined a group of 300 patients with T1-T4N0-3M0 oropharyngeal cancer patients who received definitive IMRT and concurrent chemotherapy at the University of Texas MD Anderson Cancer Center between 2002 - 2013. Chronic RAD was defined as any of the following ≥ 12 months post-IMRT: videoflouroscopy/endoscopy detected aspiration or stricture, gastrostomy tube, and/or aspiration pneumonia. DICOM-RT plan data were autosegmented using a validated region-of-interest (ROI) library and included inferior, middle and superior constrictors (IPC, MPC, and SPC), anterior and posterior digastrics (ADM, PDM), mylo/geniohyoid complex (MGM), intrinsic tongue muscles (ITM), and gengioglossus (GGM). Age was binned by decade of life, <50, 50-59, 60-69, and >=70 years of age. The probability of chronic RAD as a function of mean dose (to the above ROIs), stratified by age, was estimated by using logistic probability models and subsequent unsupervised nonlinear curve. Of 300 patients, 34 patients developed chronic RAD (11%). Age as a continuous and discrete variable was a significant correlate of chronic RAD, both independently and in concert with continuous dose for all muscle groups examined. Substantively distinct muscle-specific dose-response profiles were observed as a function of age (e.g. a patient in their 50s who received 60 Gy to the SPC had an observed 5% chronic-RAD rate, while a 70-year-old cohort (at the same dose level) exhibited a 20% rate.) Serial muscle-specific dose-response analysis demonstrated the effect to be stable across all observed muscle ROIs, with a FDR-corrected p<0.05 for all dose/muscle/age models, suggesting that an inclusion of dose as a covariate improves modeling of chronic RAD. Age at time of treatment appears to moderate dose-response probability of chronic RAD after chemo-IMRT for OPSCC. Our data demonstrate aging muscles show lower dose thresholds toward the development of potentially meaningful clinically apparent late RAD. Further investigation into age-specific dose constraints and age-adjusted dysphagia prophylaxis models appears warranted; however, our data suggest that uniform monolithic dose constraints may fail to accurately prognosticate or preclude clinical toxicity in aged populations.
Read full abstract