INTENSITY MODULATED RADIOTHERAPY--COMPUTATIONAL STUDIES

Washington University, Saint Louis MO

Investigators

Linked publications & trials

Abstract

Proposed intensity modulated radiation therapy (IMRT) techniques aim to provide higher tumor dose irradiation, and thereby better tumor control, while reducing dose to nearby important normal tissues. The main goal of this project is to use computer simulations to understand the potential benefits of a range of proposed photon and proton IMRT techniques relative to conventional techniques. In the first phase, technical issues relating to IMRT treatment planning will be investigated, including: global vs. local optimization, effects of tumor control probability (TCP) model assumptions, the effect of treatment margins and geometrical uncertainties, number and angles of incident fields needed for optimal irradiation, and type of intensity modulation (full resolution vs. segmental). These questions will be studied in detail using a suite of two- dimensional (2D) test problems, derived from patient CT data sets, which will cover a wide range of potential IMRT planning problems. Results from this phase of 2D simulations will be tested in more limited three-dimensional (3D) simulations. In the second phase, 2D and 3D simulations will be made comparing several proposed IMRT techniques, including: few- and many-field intensity modulated coplanar photon delivery, intensity-modulated arc therapy, and a novel intensity modulated proton beam method. Each plan will be optimized with alternative objective functions, maximum TCP and maximum minimum target dose, but constrained to the same normal tissue dose-volume limits. Those plans will be compared with conventional photon technique plans. At least four different sites will be studied in full 3D for all the IMRT techniques: lung, prostate, head and neck, and abdominal tumors. We will investigate the anatomical and geometrical factors which may indicate a dosimetric superiority of one IMRT technique over another. Proposed IMRT techniques will be difficult to compare directly in clinical trials except on a limited scale. The approach in this project is to conduct computational comparisons of a range of proposed IMRT techniques, using the same radiobiological and dosimetric assumptions, in order to help understand the relative potential benefits of each technique compared with conventional treatment. We hypothesize that IMRT will require more than 3-5 fields, using radiobiologically realistic and fully 3D planning methods, to reach the level of TCP possible using many-field delivery.