Impact Of Low-Dose Parotid Gland Planning Objectives In Head And Neck Cancer Radiation Therapy

Low-dose planning constraints for the combined parotid glands may help to reduce the severity of xerostomia in irradiated head and neck cancer patients when used in addition to mean dose constraints. A low-dose objective for the combined parotid glands has been included at our institution since the beginning of 2014, being utilized only in cases in which standard objectives were not compromised. This pragmatic controlled study examines prospectively acquired dose and toxicity data from the Oncospace Learning Health System to determine the role of low-dose parotid exposure in acute xerostomia outcomes. We used an integrated analytic relational database that systematically captures patient outcomes and all aspects of radiation therapy treatment plans. From a total of 642 head and neck cancer patients prospectively added to this knowledge base, 230 patients treated from 2008 to 2013 (Group A) and 50 patients treated in 2014 (Group B) were identified having Common Terminology Criteria for Adverse Events version 4.0 xerostomia assessed 3 months posttreatment. Baseline xerostomia scores were subtracted. The following dose features were also queried: minimum, mean, and maximum dose (Dmin, Dmean, Dmax), and dose to 20%, 50%, and 80% of the combined parotid volume (D20, D50, D80). Significant reductions in each dose point were assessed using the Welch 2-sample t test. Xerostomia severity was modeled by logistic regression using stepwise selection from significant dose points (including interaction terms). Mean baseline-adjusted xerostomia scores were significantly lower (P < .001) for Group B compared to Group A, with mean and standard deviations of 0.46 ± 0.79 and 0.89 ± 0.82, respectively. Group B also demonstrated significantly lower doses (P < .005) to the combined parotid glands for Dmean (-14.6%), D50 (-21.9%), and D80 (-21.3%). Of the logistic models incorporating these doses, backward stepwise selection revealed that Dmean was most predictive of xerostomia. Low doses to the parotid glands did not serve as a superior predictor of acute xerostomia. However, constraining the low-dose exposure to parotid tissue may reduce the mean dose, which may exert an indirect influence on xerostomia severity. Further prospective, multivariate evaluation is needed to assess the impact of low-dose objectives on xerostomia scores, target coverage, and sparing of adjacent normal structures.