Framework for Dosimetry and Treatment Planning of TTFields

Abstract

A post-hoc analysis study establishing the connection between TTFields dose at the tumor bed and overall survival in 340 newly diagnosed Glioblastoma patients from the pivotal EF-14 clinical trial (Stupp R, et al. JAMA. 2017;318:2306-2316) was published with our co-investigators in 2019 [Ballo M, et al. Int J Radiat Oncol Biol Phys. 2019;104(5):1106-1113]. More recently, utilizing patient data from that same trial, we were able to establish a connection between tumor progression patterns and TTFields dose distribution in TTFields treated patients. The results from these studies provide rationale for developing a framework for TTFields dosimetry and treatment planning. Utilizing this framework, physicians could maximize delivery of TTFields to target regions, and manipulate the dose as needed during the course of disease progression to address specific patterns of failure. Here we present a first effort to establish such a framework utilizing computational simulations, automated optimization algorithms and sophisticated visualization techniques.The framework we developed comprises the following steps. (a) Creation of a patient-specific computational algorithm: Semi-automatic and automatic algorithms are utilized to segment a patient MRI into different tissue types, and conductivity values are assigned to each tissue type. During this step the user is required to denote the target region in which dose should be optimized, as well as avoidance areas on the skin where it would be preferable not to place the transducer arrays. (b) Array placement optimization: During this phase virtual transducer arrays are placed on the computational model and an iterative optimization algorithm is utilized to find transducer arrays layouts that maximize TTFields dose in the target region. (c) Treatment plan selection: Several different array layouts identified by the optimization algorithm are presented to the user, so that an effective treatment plan for the patient can be empirically determined. Colormaps and isodose lines enable the user to compare field distributions generated by the different layouts. Tables and Dose Volume Histograms enable the user to quantitatively compare TTFields dose delivered to regions of interest.This treatment planning framework has been tested using over 20 image sets of glioblastoma patients. Our experience indicates this is a practical framework that can potentially increase TTFields dose delivered to targeted tumor regions significantly.We believe that this framework will assist the physicians with optimizing TTFields treatment by providing qualitative and quantitative tools for planning and also by normalizing the process of individualized treatment. Additionally, it will provide tools for adaptive treatment planning following any clinical need (e.g., tumor progression).