Role of FDG-PET/CT in detecting metabolic changes in parotid glands following photon versus proton therapy

Xerostomia is a disabling side-effect of head and neck cancer (HNC) radiation therapy (RT) that significantly impacts quality of life. While proton therapy offers dosimetric advantages over photon therapy, metabolic evidence of tissue-sparing effects remains limited. This retrospective study assessed ¹⁸F-FDG PET/CT for evaluating radiation-induced metabolic changes in parotid glands following proton versus photon RT in 50 HNC patients with base-of-tongue tumors treated with photon (n=14, intensity-modulated radiation therapy (IMRT)) or proton (n=36, pencil-beam scanning intensity-modulated proton therapy (IMPT)) with concurrent chemotherapy. All patients received ¹⁸F-FDG PET/CT before and approximately 3 months post-treatment. Bilateral parotid glands were manually delineated and averaged per patient. Both groups demonstrated significant total mean standardized uptake value (SUV_mean) increases (photon: 1.13±0.23 to 1.45±0.38, P<0.001; proton: 1.30±0.27 to 1.41±0.21, P=0.017), with significantly greater increases in the photon group for both ΔSUV_mean (0.32 vs. 0.11, P=0.017) and ΔSUV_max (0.44 vs. 0.07, P=0.011; Mann-Whitney U tests). Between-group effect sizes were large (Cohen's d=0.86 and 0.95, respectively), and multivariable regression confirmed the treatment group effect (ΔSUV_mean β=-0.25, P=0.017, adjusting for baseline SUV and disease stage; ΔSUV_max β=-0.30, P=0.015, adjusting for baseline SUV). These findings suggest that proton therapy produces significantly lower increases in parotid gland ¹⁸F-FDG uptake compared to photon therapy, supporting ¹⁸F-FDG PET/CT as a quantitative biomarker for radiation-induced parotid gland injury. Future prospective studies correlating metabolic changes with functional outcomes are warranted.