Purpose: Planning objectives used in radiotherapy plan optimization are driven by population-based studies, however patient-to-patient variation in tumor and normal tissue response is well established. Various patient characteristics have been explored as potential biomarkers to personalize treatment planning for this variance. The objective of this study was to evaluate if HPV status is a significant, independent factor affecting normal tissue toxicities for head-and-neck cancer patients treated with radiotherapy.
Methods: HPV status, radiation treatment doses to organs-at-risk, and reported early and late dysphagia were retrospectively obtained for 101 patients treated at our clinic. Mean doses were extracted for the entire pharynx, superior/middle/inferior pharyngeal constrictor muscles (PCM), and cricopharyngeus. Early and late dysphagia grades were collected on the 0-5 CTCAE scale and binarized to 0-2 versus 3-5. T-tests were used to evaluate relationships between outcome and dose, and chi-squared tests to evaluate associations between outcome and HPV. Multivariate logistic regression models were built to predict outcome as a function of HPV status, sex, age, and dose metrics. P-values<0.05 were considered significant.
Results: Significant dose metrics for early dysphagia included mean dose to the Pharynx(p=0.032), middle PGM(p=0.020), inferior PGM(p=0.015), and cricopharyngeus(p=0.022) with respective increased dose differences of 7.0, 7.1, 12.1, and 11.4 Gy. No dose metrics were significantly associated with late dysphagia; however, HPV-negative status was significantly associated with late dysphagia (p=0.028). HPV-negative status was not significantly associated with early dysphagia. In logistic models, results were consistent: all four organ mean doses remained significant covariates for early dysphagia while HPV was the only significant covariate for late dysphagia.
Conclusion: HPV status was a significant predictive covariate for late dysphagia in our cohort. While further investigation is warranted to evaluate the root cause of this association, these results may be useful for assessing an individual patient’s toxicity risk and personalizing their treatment or follow-up.
Funding Support, Disclosures, and Conflict of Interest: X Ray discloses a lab services agreement with Varian. P Sanghvi discloses past relationships with Varian and VisionRT.