Relative biological effectiveness and dose-averaged linear energy transfer studies for proton therapy Monte Carlo treatment planning

• Main Author: Koh, Calvin Wei Yang
• Other Authors: Lew Wen Siang
• Format: Thesis-Doctor of Philosophy
• Language: English
• Published: Nanyang Technological University 2022
• Subjects: Science::Physics
• Online Access: https://hdl.handle.net/10356/155462
• Institution: Nanyang Technological University

Proton Therapy shows dosimetric advantages over conventional X-Ray Therapy in terms of the control of doses to normal tissues. In fact, it should be the preferred treatment in most paediatrics and local recurrent cancer when available. This is due to the physical characteristics of proton interactions where they have a finite range as they traversed through a medium. This gives rise to a steep increase in dose distribution with a sharp fall-off known as the Bragg Peak. Despite the dosimetric advantage, physical and radiobiological uncertainties are a concern in Proton Therapy. In other words, it is not sufficient to know it stops, instead, it is important to know where it stops and the biological effects that follow. This thesis aims to address the issue of radiobiological uncertainties with Relative Biological Effectiveness (RBE) for treatment planning which is defined as a scalar quantity Proton Therapy. Currently, a constant RBE of 1.1 is used clinically to include the differential biological effect of protons as compared to photons during treatment planning. However, the counter-argument against the use of a constant RBE of 1.1 stems from the possibility of under-or over-dosing in the target volume. As suggested by AAPM-TG-2561, there is a need to understand the spatial variations of RBE within and outside the target volume. It is also recommended for the use of variable RBE models for certain clinical situations such as target volumes that are close to Organ-At-Risks (OARs). Numerous RBE models were developed to account for the uncertainties arising from dependent quantities of patient radio-sensitivity alpha-beta ratio ( ) , linear energy transfer (LET), cell lines used in experiments and experimental set-up. The biological uncertainties are investigated via Monte Carlo (MC) Simulation through three projects using Linear Quadratic (LQ)-based RBE models, where the model is dependent on proton dose ( ), dose-averaged LET ( ) and alpha-beta ratio ( / ) ratio. The first project investigated the estimation of in these RBE model and the dosimetric impact of uncertainty on a clinical case. The second project investigated the MC simulation parameters for simulating and its mathematical functions when calculating . The final project investigated the systematic dose variation from different RBE models in the clinical case where OARs are located near the target. The results from these studies showed the importance of calculating via MC simulation and how it can lead to an increase in biological uncertainties in Proton Therapy. The dosimetric impact of simulating when there is insufficient knowledge of cellular composition could lead to huge uncertainty during the simulation and this uncertainty could propagate down to the resulting RBE models. The results show that using cellular materials instead of water during MC simulation is important under low ( / ) and low dose (2 ) conditions. In addition, a standard protocol is proposed for sampling in MC simulation which is required as this would affect secondary electrons production and will improve the accuracy of the dose distribution and . Therefore, the MC simulation protocols and scoring method is defined and standardized to facilitate future cross-institutional studies. Lastly, based on the two previous projects, it was established that RBE values are bound to be associated with large uncertainties due to variations in biological experiments and calculations reported in the literature. It is thus challenging to select a single RBE model based on experimental data. Instead, in our last project, we focus our effort on arriving at an RBE model-agnostic approach treatment planning with Multi-Field Optimization (MFO) vs Single-Field Optimization (SFO) to minimize the systematic dose variation between different RBE models. In this study, brain tumor cases are used. MFO planning technique showed a better option for overlapping brain tumors with OARs in eliminating RBE-weighted dose uncertainties. In conclusion, this thesis examined the uncertainties in RBE arising from different experimental set-up and calculations from different MC simulation parameters. The studies have concluded that RBE uncertainties are still challenging and the choice of MFO treatment planning technique may possibly yield an RBE-model agnostic dose distribution. Standadization for calculations are important for MC simulation. The knowledge gained from these studies will be beneficial for the future development of RBE-based treatment planning system for proton beam therapy using MC simulation.