Dose-response analysis of parotid gland function: what is the best measure of xerostomia?

Abstract

Purpose: To describe the dose-response relationships for the different measures of salivary gland recovery following radical radiotherapy for locally advanced head and neck squamous cell cancers (LA-HNSCC).

Methods and materials: Dosimetric analysis of data from the PARSPORT trial, a Phase III study of conventional RT (RT) and intensity modulated radiotherapy (IMRT) for LA-HNSCC was undertaken to determine the relationship between parotid gland mean dose and toxicity endpoints: high-grade subjective and objective xerostomia and xerostomia-related quality of life scores. LKB-NTCP parameters (TD50, m and n) were generated and tolerance doses (D50) reported using non-linear logistic regression analysis.

Results: Data were available on 63 patients from the PARSPORT trial. Parotid saliva flow rate provided the strongest association between mean dose and recovery, D50=23.4 Gy (20.6-26.2) and k=3.2 (1.9-4.5), R(2)=0.85. Corresponding LKB parameters were TD50=26.3 Gy (95% CI: 24.0-30.1), m=0.25 (0.18-1.0 and n=1). LENTSOMA subjective xerostomia also demonstrated a strong association D50=33.3 Gy (26.7-39.8), k=2.8 (91.4-4.4), R(2)=0.77).

Conclusion: We recommend using the LENT SOMA subjective xerostomia score to predict recovery of salivation due to its strong association with dosimetry and ease of recording.