Parotid gland stem cell sparing radiotherapy for head and neck cancer patients: a double-blind randomized controlled trial.

BACKGROUND Radiotherapy for head and neck cancer (HNC) frequently leads to salivary gland damage and subsequent xerostomia. The radiation response of parotid glands of rats, mice, and patients critically depends on dose to its stem cells, mainly located in the gland's main ducts (stem cell rich (SCR) region). Therefore, this double-blind randomized controlled trial aimed to test the hypothesis that parotid gland stem cell sparing radiotherapy preserves parotid gland function better than currently-used whole parotid gland sparing radiotherapy. METHODS HNC patients (n=102) treated with definitive radiotherapy were randomized between standard parotid sparing and stem cell sparing (SCS) techniques. The primary endpoint was >75% reduction in parotid gland saliva production compared to pretreatment production (FLOW12M). Secondary endpoints were several aspects of xerostomia 12 months after treatment. RESULTS Fifty-four patients were assigned to the standard arm and 48 to the SCS arm. Only dose to the SCR regions (contralateral 16 and 11 Gy (p=0.004) and ipsilateral 26 and 16 Gy (p=0.001), standard and SCS arm respectively) and pretreatment patient-rated daytime xerostomia (35% and 13% (p=0.01), standard and SCS arm respectively) differed significantly between the arms. In the SCS arm, 1 patient (2.8%) experienced FLOW12M compared to 2 (4.9%) in the standard arm (p=1.00). However, a trend towards better relative parotid gland salivary function in favor of SCS radiotherapy was shown. Moreover, multivariable analysis showed that mean contralateral SCR region dose was the strongest dosimetric predictor for moderate-to-severe patient-rated daytime xerostomia and grade ≥2 physician-rated xerostomia, the latter including complaints of alteration in diet. CONCLUSIONS No significant better parotid function was observed in SCS radiotherapy. However, additional multivariable analysis showed that dose to the SCR region was more predictive for development of parotid gland function related xerostomia endpoints, than dose to the entire parotid gland.