Author: Eric Aliotta, Michalis Aristophanous, Laura I. Cervino, Yu-Chi Hu, Nancy Y Lee, Anyi Li, Jung Hun Oh, Teeradon Treechairusame, Pengpeng Zhang 👨🔬
Affiliation: Department of Radiation Oncology, Memorial Sloan Kettering Cancer Center, Department of Medical Physics, Memorial Sloan Kettering Cancer Center, Memorial Sloan Kettering Cancer Center 🌍
Purpose: To automatically monitor weekly anatomical changes of parotid glands during adaptive radiotherapy of head and neck cancers, systematically assess the relationship between parotid gland shrinkage and acute xerostomia, and facilitate the decision-making in treatment adaptation.
Methods: We established an Automated Watchdog in Adaptive Radiotherapy Environment (AWARE) to track volumetric changes in tumors and OAR observed on weekly CBCTs. GTV and OAR contours were propagated from planning CT to weekly CBCTs via deformable registration and reviewed by clinicians. AWARE triggers a re-simulation around week 3 of radiotherapy when GTV or external body contour changes significantly. Adaptive plans were delivered around week 4 to improve dosimetric quality. We retrospectively analyzed 164 AWARE patients treated between 2021 and 2024, evaluated adverse events using Common Terminology Criteria for Adverse Events (V.5), and identified 45 patients who developed acute xerostomia (grade >=2). The median follow-up time was 15.5 (8-23) months.
Results: Among patients who developed acute xerostomia, the parotid gland shrank 13±10%, 18±11%, and 24±12% in the first, second, and third week, respectively, significantly larger (p-value < 0.0005, t-test) than 8±11%, 11±13%, and 17±11% among patients without xerostomia. The area under the receiver operating characteristic curve was 0.66, 0.69, and 0.72 for predicting xerostomia by parotid gland shrinkage in the first, second, and third weeks, respectively. Rates of adaptation were 84% vs. 71%, and the planning mean dose of parotid gland through week 3 was 11.5±4.4Gy vs. 10.4±3.0Gy, with or without xerostomia, not statistically significant under t-test.
Conclusion: This outcome study indicates that parotid gland shrinkage can be predictive of xerostomia as early as the first week of radiotherapy. Frequent imaging surveillance and timely plan adaptation triggered by parotid gland shrinkage at the early stage of radiotherapy could benefit patients more than mid-late course adaptation when the parotid gland already receives substantial radiation doses.