Radiobiological optimization including consideration of secondary cancer risk: A treatment modality comparison study for pediatric medulloblastoma

Purpose: To estimate and compare normal tissue toxicity and secondary cancer (SC) risks from 3D conformal radiotherapy (3D CRT), arc radiotherapy (ART) and intensity modulated proton therapy (IMPT) for treatment of pediatric medulloblastoma (MB). The purpose was also to investigate the possibility of optimizing radiotherapy treatment with the aim of reducing normal tissue toxicity as well as the risk of SC. Materials and methods: Treatment plans were generated with 3D CRT, ART as of the RapidArc implementation and spot-scanned IMPT. The patient material consisted of treatment records including CTand MRI-scans of four males and six females aged 4 to 15 years old. These patients were treated with craniospinal irradiation (CSI) for MB at Copenhagen University Hospital between 2007-2009. The treatment regimen studied was CSI to a prescribed dose of 30.6 Gy followed by a boost to the posterior fossa to 54 Gy. In order to optimize the plans with the aim of reducing normal tissue toxicity a thorough literature review was conducted to find organ specific dose-volume vs. toxicity data relevant to children treated with RT. SC cancer risks were estimated using organ-equivalent dose models based on the combined SC data from the atomic bomb survivors and patients receiving RT for Hodgkin’s disease. Using the sitespecific SC incidence provided by the Life Span Study (LSS) the RapidArc plans were re-optimized with the aim of reducing SC risk. A robustness analysis regarding the reoptimization concept was conducted by estimating SC risks with different models. Results: The results show that the risk of developing a solid SC from inversely optimized arc therapy can be reduced to the same level as for 3D CRT. With regards to normal tissue toxicity, the risks for several severe complications were considerably lower with RapidArc than with 3D CRT. As for treatment with 3D CRT it was shown that the choice of spinal field width has a considerable impact on SC risk as well as normal tissue toxicity. The risks of normal tissue toxicity as well as SC were substantially lower for IMPT compared to the photon techniques. Conclusions: This study demonstrates the possibility of reducing SC risk from inversely planned arc therapy by including it in the optimization process. As SC risk calculations are subject to large uncertainty, absolute values should be treated with some reservation. Comparisons between different techniques can however be considered more reliable. The risk of severe complications such as heart failure was shown to be substantially higher for 3D CRT compared to intensity-modulated therapy. The ability to reduce the SC risk related to ART along with the lower risk of normal tissue toxicity favors RapidArc over 3D CRT for treating MB. The potential benefit of treating children with spot-scanned proton therapy compared to photon techniques is clearly illustrated in this study.