Enhancing dosimetric practices through knowledge-based predictive models: a case study on VMAT prostate irradiation

Area of Science:

• Radiation Oncology
• Medical Physics

Background:

• Radiation therapy planning requires extensive dosimetric knowledge, which is challenging to acquire.
• Knowledge-based planning (KBP) offers a methodology to standardize and enhance treatment planning.
• Prostate cancer treatment planning is complex due to proximity of organs at risk.

Purpose of the Study:

• To assess the impact of empirical predictive models based on KBP for prostate cancer treatment.
• To detect suboptimal treatment plans and homogenize planning practices.
• To evaluate the dosimetric effects of implementing KBP models in routine clinical practice.

Main Methods:

• Analysis of 25 VMAT prostate treatment plans using KBP to correlate dose indicators with patient geometry.
• Linking dose indicators to geometric characteristics like PTV-OAR distance and OAR volume.
• Prospective implementation of KBP in a cohort of 25 patients, compared to 41 prior plans.

Main Results:

• Strong geometric predictive factors for organs at risk identified (R² > 0.8).
• Implementation of KBP led to significant dose reductions in evaluated organs at risk, particularly for the peritoneal cavity.
• Reduced the percentage of plans exceeding clinical recommendations for organs at risk from 19% to 8%.

Conclusions:

• KBP methodology provides a robust, personalized predictive model for organ at risk dose estimation in prostate cancer.
• Model implementation improved organ sparing and established a foundation for harmonizing dosimetric practices.
• KBP aids in identifying suboptimal plans and advancing dosimetric knowledge.