Magnetic resonance imaging to evaluate the functional behavior of parotid salivary glands in normal subjects and radiotherapy patients

Abstract

Xerostomia is often a major cause of decreased quality of life for head and neck cancer survivors. To further understand dose-volume factors in post-RT parotid salivary gland function, we investigated the potential of magnetic resonance imaging (MRI) to evaluate the functional behavior and dynamics of parotid salivary glands. T2-weighted and diffusion-weighted MRI images were obtained in five normal subjects and two post-RT patients using a custom echo-planar imaging (EPI) pulse sequence. Signal was received using a commercial coil placed adjacent to the parotid glands. For each patient, EPI multi-slice image sets were obtained at 30 second intervals. A 3 cc volume of lemon juice was delivered to each subject via plastic tubing after 3 ’resting’ frames. Images were analyzed by averaging multi-slice regions-of-interest within the parotid glands or the cerebellum. In 4 of 5 normal subjects, the T2-weighted signal intensity decreased by approximately 5–15% over 1–3 mins. after parotid stimulation. The control patient who was given distilled water did not show such a decrease. The background (cerebellum) signal also did not show any decrease. Diffusion-weighted images (analyzed for one patient only) did not show any significant change between pre-and post-injection. One patient, 6 years post-RT with grade I xerostomia, and with RT mean parotid doses of 69 and 19 Gy, showed no change in either parotid gland signal with stimulation. Results for a second patient at 1.5 years post-RT, with grade II xerostomia, and RT mean parotid doses of 67 and 14 Gy, are shown in the Figure. Following stimulation, a significant decrease (∼20%) in parotid signal was noted on the less irradiated right gland while the heavily irradiated left gland did not show any decrease in signal. A consistent time-dependent change in T2-weighted image intensity was demonstrated corresponding to parotid stimulation. This T2-weighted effect may be caused by a decrease in spin density and/or T2 relaxation time due to the transport of fluid out of the glands and/or ducts. A reduced or absent signal was observed for heavily irradiated glands. This is the first report of an MRI technique that is sensitive to dynamic parotid salivary function. Potentially, this technique could be used to probe the relationship between dose and post-RT parotid gland function.