Adaptive objective function and weights for inverse radiotherapy planning: A versatile strategy for complex cases

A gap in radiotherapy planning optimization has existed: there is no clear systematic guidance for manually adjusting complex plans that elude automated solutions or standard protocols. While extensive research has advanced automated planning, little attention has been given to optimizing the inevitable human intervention required for challenging scenarios. This study aimed to develop an efficient optimization strategy for radiotherapy planning applicable to different tumor types and treatment planning systems (TPSs), particularly for complex cases and patient-specific requirements. We established an objective value-based optimization strategy and validated it using a cohort of 120 previously treated patients from two cancer centers representing diverse tumor sites and complexity levels. The strategy was applied to reoptimize treatment plans created with four commercial treatment planning systems: Pinnacle, UIH, Eclipse, and Monaco. A comprehensive comparison between reoptimized plans (RPs) and clinical plans (CPs) was performed, evaluating dosimetric parameters, monitor units (MUs), and optimization efficiency. The RPs demonstrated overall quality improvements on Pinnacle, UIH and Eclipse while maintaining equivalent performance on Monaco. Specifically, the conformity index (CI) significantly improved for Pinnacle, UIH and Eclipse RPs (all p < 0.05). Organ-at-risk (OAR) protection showed significant improvements in 80 % cases (96/120), with system-specific success rates of 83 % (Pinnacle), 90 % (UIH), and 93 % (Eclipse) versus 53 % (Monaco). Planning efficiency was excellent, with an average of (5 ± 4) loops for the optimization process. In complex cases, RPs exhibited particularly notable improvements: significantly better CI for Pinnacle, UIH, and Eclipse (all p < 0.05), and OAR sparing improved in Pinnacle (80 %), UIH (80 %), Eclipse (90 %) but deteriorated in Monaco (50 %). We successfully developed an objective-based function and weights adaptive strategy for inverse radiation planning, quantitatively guiding optimization across four commercial TPSs. The robustness of this strategy in complex cases and patient-specific scenarios suggests its potential to streamline and improve radiotherapy planning processes.