Session Item

Xerostomia remains a common side effect of radiotherapy (RT) for patients with head and neck (H&N) cancer despite advancements in treatment planning and delivery. Molecular PET imaging of the prostate specific membrane antigen (using ⁶⁸Ga/¹⁸F-PSMA) is highly specific to prostate cancer, but uptake in salivary glands reflects the presence of generally abundantly PSMA-positive secretory cells. We aimed to objectively quantify the dose-response of salivary glands using PSMA PET.

30 H&N cancer patients were included in a prospective study. They received RT with 70 Gy in 35 fractions over 7 weeks. PSMA PET/CT was acquired in treatment position at baseline and at 6-months post-RT. The PET scans were deformably registered to the planning CT and the associated dose distribution. Dose, pre-RT SUV and post-RT SUV were extracted for every voxel inside each delineated parotid salivary gland. The data was analysed using a generalised linear mixed effects model using a log link. 

Fig. 1 shows the dose distribution, and baseline and post-RT PSMA PET images of one patient. The baseline SUV was observed to moderate the effect of dose on the post-RT SUV. The population and patient-specific dose-response curves for the parotid glands, can be seen in Fig. 2. The model fit the data well with an R² of 0.79. The D₅₀ of the population-curve is 34 Gy. The population curve indicates that for a 1 Gy increase in dose, the post-treatment SUV decreases by 1.8%, for a baseline SUV of 10. 

The PSMA PET response in salivary glands after RT demonstrates a strong relationship between dose and loss of secretory cells. Differences in patient sensitivity can also be observed, which the model can account for. The population curve could potentially be used in dose planning, using a planning objective to maximize the predicted post-treatment SUV. This could be improved upon further by using a pre-treatment PSMA scan to get patient-specific curves.