Knowledge-based automatic optimization of adaptive early-regression-guided VMAT for rectal cancer

Area of Science:

• Radiation Oncology
• Medical Physics
• Cancer Treatment

Background:

• Rectal cancer treatment often involves neoadjuvant radio-chemotherapy.
• Adaptive radiotherapy (ART) aims to personalize treatment based on daily anatomical changes.
• Knowledge-based (KB) planning can improve the efficiency and quality of radiotherapy plans.

Purpose of the Study:

• To implement a knowledge-based optimization strategy within an adaptive radiotherapy (ART) technique for rectal cancer.
• To evaluate the effectiveness of a scaled KB model for the ART phase of treatment.

Main Methods:

• A two-phase protocol was used, including an initial phase and an adaptive phase with dose escalation.
• A knowledge-based model was generated using RapidPlan for the initial phase and scaled for the ART phase.
• Automatic plans (RP) were compared against original manual plans (RA) for organ at risk (OAR) and planning target volume (PTV) parameters.

Main Results:

• Automatically generated plans were generally equivalent or superior to manually created plans.
• Significant improvements in bladder and bowel dose metrics (V30Gy, V40Gy) were observed with RP plans.
• Generalized Equivalent Uniform Dose (gEUD) analysis indicated improved OAR sparing with the KB-ART approach.

Conclusions:

• The knowledge-based model proved robust and adaptable for the adaptive radiotherapy phase.
• Automatically generated plans using the KB-ART strategy demonstrated comparable or slightly better performance than manual plans.
• This approach offers a promising method for optimizing neoadjuvant radio-chemotherapy for rectal cancer.