The Association of Canonical Speech- and Swallow-Related Dosimetric Constraints with Patient-Reported Quality of Life Outcomes: Results from a Phase II Trial of Dose De-Escalated Chemoradiation for HPV-Associated, Locally-Advanced Oropharyngeal Carcinoma

Abstract

Radiation dose de-escalation for patients with human papillomavirus (HPV)-related oropharyngeal carcinoma may improve quality of life. We conducted an exploratory exercise to examine potential relationships between canonical dosimetric parameters and speech- and swallow-related and overall patient-reported quality of life outcomes on a subset of patients treated on a phase II trial of de-escalated chemoradiation for HPV-related oropharyngeal carcinoma. An analysis was performed of 24 patients treated at a single-institution on a phase II, prospective study of induction chemotherapy followed by dose de-escalated chemoradiation for HPV-related, locally-advanced oropharyngeal carcinoma. Following two cycles of induction carboplatin/paclitaxel, reduced-dose chemoradiation with concurrent paclitaxel was administered to 54-60 Gy in 27-30 fractions. Intensity-modulated radiotherapy was used for all patients. Univariable logistic regression analysis was used to evaluate potential associations between multiple speech-and swallow-related dosimetric constraints and quality of life domains at 3 and 12 months post-treatment involving swallowing, chewing, speech, taste, and saliva, as well as overall quality of life, using the University of Washington Quality of Life instrument. Canonical models were constructed to analyze dosimetry to the temporomandibular joints, oral cavity, lips, parotid glands, submandibular glands, larynx, pharynx, cricopharyngeal inlet, and cervical esophagus, as analyzed by maximum, mean, and/or partial-volume parameters. The post-chemotherapy gross tumor volume, planning target volume (PTV) receiving the high, intermediate, and low dose levels, and combined volume of all PTV dose levels were also analyzed. Twenty-five percent of patients had T3-T4 tumors, and 13% had N2c nodal disease. Thirteen percent had >10 pack-year smoking history. At 12 months, a decreased taste domain score was significantly associated with a higher volume of high-dose level PTV (p=0.028). Also, a decreased health-related quality of life score at 12 months was significantly associated with a higher contralateral parotid gland mean dose (p=0.037), higher contralateral parotid gland V30Gy (p=0.046), and higher cricopharyngeal inlet mean dose (p=0.036). No dosimetric parameters were significantly associated with quality of life differences at 3 months (p>0.05, for all). For patients treated by de-escalated chemoradiation, canonical dosimetric constraints for conventional chemoradiation are poorly correlated with speech- and swallow-related and overall quality of life outcomes. While a smaller high-dose PTV, lower contralateral parotid gland dose, and lower cricopharyngeal inlet dose appear to be related to better quality of life outcomes, these require additional validation. Future studies of de-escalated chemoradiation may reveal new dosimetric constraints to help guide treatment planning.