Assessment of Knowledge-Based Planning for Prostate Intensity Modulated Proton Therapy

Area of Science:

• Radiation Oncology
• Medical Physics
• Computational Biology

Background:

• Intensity-modulated proton therapy (IMPT) offers precise radiation delivery.
• Prostate cancer treatment requires careful planning to spare organs at risk.
• Knowledge-based planning (KBP) models can potentially automate and improve treatment planning.

Purpose of the Study:

• To evaluate a proton-specific knowledge-based planning (KBPP) model for creating robustly optimized IMPT plans in prostate cancer.
• To assess the quality and accuracy of KBPP-generated plans compared to expert-generated plans.

Main Methods:

• Developed a KBPP model using data from 30 prostate cancer patients previously treated with VMAT.
• Re-planned 45 localized prostate cancer patients using robustly optimized IMPT.
• Validated the KBPP model using the remaining 15 patients' data.
• Assessed plan quality by comparing KBPP and expert plans for target coverage and organ-at-risk sparing.

Main Results:

• The KBPP model demonstrated good quality with high R² (0.87) and acceptable χ² values.
• KBPP-generated IMPT plans were clinically acceptable and comparable to expert plans.
• KBPP plans showed marginally higher rectal V65Gy but lower bladder V65Gy, lower hotspot, and lower integral dose.
• KBPP plans achieved slightly better clinical target volume V95 and a lower homogeneity index.

Conclusions:

• The KBPP model effectively generates high-quality, robustly optimized IMPT plans for prostate cancer.
• KBPP-generated plans are comparable to, and in some aspects superior to, expert-generated plans.
• Further validation is needed for the KBPP model in more complex treatment sites.