Submandibular Gland Transfer for the Prevention of Radiation Induced Xerostomia in Head and Neck Cancer — Dosimetric Impact With Intensity Modulated Radiotherapy

Abstract

Post-radiation xerostomia is an important toxicity of head and neck radiotherapy, contributing to decreased quality of life and other adverse effects. Submandibular gland (SMG) transfer prior to radiation previously demonstrated a statistically significant benefit in xerostomia reduction in the 2/3 dimensional treatment era. In the procedure, the SMG is transposed anteriorly to the submental space. In patients receiving radiation to the level II but not level IB neck, this displacement removes the SMG from the treatment portal, thereby preserving gland function. Herein we evaluate the dosimetric implications of SMG transfer on intensity modulated radiotherapy (IMRT) treatment plans.We conducted a retrospective review of oropharynx cancer patients who underwent SMG transfer surgery followed by curative IMRT at our institution. Pre-surgery computed tomography (CT) scans had previously been fused with the planning CT. Separate scans were fused to the planning CT for 5 patients who had not had previous fusions. Pre-surgery SMG glands were contoured. Dose to the pre-surgery SMG was assessed on the treated clinical plan (TCP). A new plan (NP) was then generated, optimizing based on the pre-surgery SMG. Dosimetry attempted to meet all standard institutional constraints and match the salivary gland mean doses of the TCPs. Mean salivary gland, oral cavity, and larynx doses were collected. We evaluated the differences in plan outcomes using t-tests.18 patients with median age 61.5 years (range 25-74) underwent SMG transfer between 2019 and 2021 (94.4% male and 94.4% received concurrent cisplatin). Mean SMG dose and mean total parotid and SMG gland (TPSG) dose post-transfer was 27.5 Gy and 26.1 Gy (standard deviation (SD) 6.9 and 3.3), respectively; pre-transfer SMG and TPSG mean dose was 40.4 Gy and 27.8 Gy (SD 9.4 and 3.5) and 29.6 Gy and 26.4 Gy (SD 7.7 and 3.9) in TCPs and NPs, respectively. Differences between pre- and post-transfer SMG and TPSG doses in TCPs were statistically significant, as were differences between pre-transfer SMG and TPSG doses in TCPs versus NPs (all P < 0.001). Differences between pre-transfer NP and post-transfer TCP doses were not statistically significant (P > 0.1). Mean oral cavity and larynx doses were 32.9 Gy and 31.4 Gy (SD 6.8 and 9.9) and 33.5 Gy and 30.8 Gy (SD 7.5 and 9.6) in TCPs and NPs, respectively (all P > 0.1).SMG mean dose was lower by ∼2 Gy in TCPs relative to the non-transferred SMG mean dose in reoptimized NPs, although this difference did not reach statistical significance. Inverse planning likely decreases the benefit of SMG transfer to reduce the risk of xerostomia. More investigation is needed to evaluate if patient subsets would benefit more than others based on tumor location (tongue base versus tonsil) and treatment modality (proton versus photon).