Integrating prompt gamma range verification and relative biological effectiveness in multicriterial proton treatment planning

质子放射治疗正成为一种越来越广泛的癌症治疗选择，它允许将高剂量区域严格限制在肿瘤靶体积和有危险的备用器官，因为最大剂量沉积在范围的末端，即所谓的布拉格峰。然而，两个方面使质子治疗的计划和交付具有挑战性：1)布拉格峰位置对解剖变化和组织停止特性知识的不确定性敏感；2)质子的相对生物有效性（RBE）是可变的。目前在临床实践中，布拉格峰定位的问题是通过避免光束方向直接将远端剂量下降直接放在关键结构前面，通过增加安全边际，并通过稳健的优化方法，以牺牲一些剂量一致性为代价来解决。此外，在研究环境中，由辐照引起的伽马（PG）辐射被用于体内范围验证，从而能够实时监测剂量传递。变量RBE目前在临床实践中被忽略，假设恒定值为1.1，但变量RBE方案的可能含义越来越受到关注。该项目的总体目标是开发模型和算法，将PG范围验证和线性能量转移（LET）和RBE作为基于多准则优化（MCO）的质子规划系统研究的新标准。该MCO系统非常适合于研究相互冲突的计划目标之间的权衡，它将用于获得新的见解，了解将PG范围验证和LET/RBE考虑添加到目标覆盖和正常组织保留的标准剂量指标中的临床潜力。PG监测的优化包括识别和增强最适合的铅笔光束点，以及预测最终的PG图像质量，包括目前正在开发的几种PG相机的响应模型。用于PG监测和LET/RBE量化的目标函数将集成到MCO求解器中，并通过生成计划数据库来逼近帕累托前沿。帕累托前端将通过使用图形用户界面进行探索，如有必要，包括非凸导航。将模拟局部LET/RBE的变化，并开发具体增加或减少子区域LET/RBE的工具。最后，在一项计划研究中，新系统被用于研究肿瘤剂量、危险器官保留、LET/RBE变化和PG监测适用性之间的权衡，最终旨在更好地利用质子放疗的优势特性来改善癌症患者的临床结果。

Proton radiotherapy is becoming an increasingly widespread treatment option against cancer which allows to tightly confine the high dose area to the tumor target volume and spare organs at risk, because the dose maximum is deposited at the end of the range, the so-called Bragg peak. Two aspects however make the planning and delivery for proton therapy challenging: 1) The Bragg peak position is sensitive to anatomical variations and uncertainties in the knowledge of the tissue stopping properties, and 2) the relative biological effectiveness (RBE) of protons is variable. The issue of Bragg peak positioning is currently addressed in clinical practice by avoiding beam directions that would place the sharp distal dose fall-off directly in front of critical structures, by adding safety margins, and by robust optimization approaches, at the cost of sacrificing some dose conformity. Moreover, in research settings prompt gamma (PG) emissions induced by the irradiation are used for in-vivo range verification, enabling a monitoring of the dose delivery ideally in real-time. The variable RBE is currently disregarded in clinical practice by assuming a constant value of 1.1, but possible implications of a variable RBE scheme are subject of increasing attention. The overarching goal of this project is to develop models and algorithms which incorporate PG range verification and biological effectiveness in terms of linear energy transfer (LET) and RBE as new criteria in a research proton planning system based on multicriteria optimization (MCO). This MCO system is ideally suited to study the trade-offs of conflicting planning goals, and it will be used to gain novel insights into the clinical potential of adding PG range verification and LET/RBE considerations to standard dose metrics of target coverage and normal tissue sparing. Optimization of PG monitoring includes the identification and boosting of the most suited pencil beam spots, and predicting the final PG image quality including the response models of several PG cameras currently under development. Objective functions for PG monitoring and LET/RBE quantification will be integrated in an MCO solver, and the Pareto front will be approximated by generating a database of plans. The Pareto front will be explored by using a graphical user interface including, if necessary, non-convex navigation. Local LET/RBE variations will be modeled, and tools to specifically increase or decrease the LET/RBE of subareas will be developed. Finally, in a planning study the new system is used for investigating the trade-offs between tumor dose, organ at risk sparing, LET/RBE variations, and suitability for PG monitoring, which ultimately aims at better utilization of the advantageous properties of proton radiotherapy to improve the clinical outcomes for cancer patients.