TCP-Driven Biological Planning for High-Dose-Rate Brachytherapy

Abstract

Purpose/Objective(s): Tumor control probability (TCP) measures the probability that no malignant cells are left in the affected organ. It is thus an important clinical metric for measuring treatment success. Zaider and Minerbo provide a complex yet important TCP formalism which links the treatment duration, dose deposit, radiobiology of tumor cells and cell life characteristics. We perform feasibility tests on TCP-driven biological planning for high dose rate brachytherapy (HDR) where the objective maximizes TCP of the resulting plans. Robustness of the planning is tested, and plan quality and potential outcome significance are evaluated. Materials/Methods: CT and 18F-fluorodeoxyglucose based-PET images are obtained for 15 cervical cancer patients. All received prior external beam radiation. The CTV prescribed dose is 5 Gy for 4 fractionations, with the PET-identified pockets prescribed an escalated dose ranging from 7.59.0 Gy per fraction. The treatment model determines the dwell time and seed location that maximizes TCP while constraining PTV coverage, the lower/upper dose, and dose-volume shape for organs-at-risk and PTV. Two alternative plans are contrasted: standard HDR plan and PET-pocket escalated plan. Results: TCP for standard plans ranges from 48-63% whereas TCP for PET-guided escalated plans ranges from 82-99%. The increase is most significant with the smallest PET-identified pockets. Uniformly, bladder and rectum receive 5-8% reduced dose. There is a marginal difference in PTV dose between the standard and escalated plans. All resulting TCPdriven plans are clinically acceptable. Conclusions: TCP can be a very important objective for treatment plan optimization. With advances in biological/functional imaging, there is an urgent need to incorporate radiobiological parameters and TCP within the planning process. This study marks the first use of TCP as the driving objective for treatment planning. Although the optimization problem is very difficult to solve and requires computational breakthroughs, the resulting plans improve significantly the overall local tumor control, and also reduce dose to organs at risk. Rush University began clinical studies in July 2011. A long term outcome study must be carried out to gauge the overall impact. Author Disclosure: E.K. Lee: None. F. Yuan: None. A. Templeton: None. R. Yao: None. K. Kiel: None. J. Chu: None.