Abstract
OBJECTIVE: To optimize the protection of organs at risk (OARs) in left breast cancer radiotherapy, this study investigated how physical parameter adjustments affect the performance of a Rapidplan-based dose-volume histogram (DVH) prediction model. METHODS: Twenty patients who underwent left breast-conserving surgery were enrolled. Partial arc volumetric modulated arc therapy (VMAT) plans were designed per patient, with X-direction field width set to half-beam and right breast (Breast-R) contoured as an avoidance structure to generate Rapidplan model. The model was used to predict and generate three plans: AP_partial arc (avoidance structure prioritized), RP_partial arc (no avoidance structure), and FP_partial arc (expanded field width). Dosimetric comparisons against the original plan evaluated the impact of parameter selection. RESULTS: AP_partial arc reduced mean doses of Breast-R, Heart, Lung-L, and Lung-R by 7.7 cGy, 9.8 cGy, 16.7 cGy, and 1.1 cGy, respectively (p < 0.05). Conversely, RP_partial arc increased mean dose of Breast-R by 66.3 cGy (p < 0.05). FP_partial arc raised V5 of Lung-L, V5 of Heart, and mean dose of Lung-L by 4.01%, 2.25%, and 36 cGy (p < 0.05). CONCLUSION: The knowledge-based partial arc model for rapid planning of left breast cancer accurately predicts the DVH of OARs. However, before performing dose prediction, physical parameters such as radiation field width and planned avoidance structures should be considered to reduce the risk of low-dose exposure volume to OARs and secondary cancer.