SU‐E‐T‐656: Are We Ready for Biological Planning? A Dosimetric Study

Abstract

Purpose: To evaluate the feasibility of using a commercially available IMRT optimization and plan evaluation tool based on radio‐biological models. Methods: A retrospective dose optimization study was performed for 5 lung cases treated using IMRT to evaluate the biological modules available in Pinnacle system. A hybrid objective function consisting of physical dose constraints, for PTVs, and EUD constraints, for organs at risk, was used for optimization. The OARˈs were: spinal cord, esophagus, cardiac ventricle, and both lungs. The “a” value, used in EUD definition, was a variable to allow the optimizer to match the physical dose optimization solution. A biological optimized plan, (BOP) was deemed acceptable if matches or exceeds the quality of a physical dose optimized plan (PDOP). TCP, NTCP, or P+ values were not included in the optimization however; they were used during plan evaluation. The TCP and P+ are based on Poisson model with linear quadratic cell survival for each individual voxel. NTCP module is based on LKB model. For model parameters we choose a combination of published data and our clinical standards. Results: For serial structures the Max EUD constraint was able to bring the dose to levels obtained during physical dose optimization, however it was unable to easily control the magnitude of maximum dose. Values for parameter “a” are 15–20 for serial structures and a=1.0 for parallel structures. Global values for TCP, NTCP, P+ were 56.8% versus 53.4%, 7.4% versus 6.4%, and 49.4% versus 46.9% for BOP and PDOP, respectively. Conclusions: We have performed a retrospective treatment planning study to investigate the feasibility of using biological models for optimization purposes. Although, results between PDOP and BOP plans are comparable it remains to be seen if the clinicians will be willing to accept these concepts in clinical practice without the existence of clinical trials.