Treatment planning in 3 and 4 dimensions and the role of Monte Carlo techniques

Purpose: Monte Carlo dose calculation is considered to be the most accurate method to simulate doses delivered in radiation therapy. It is rarely used in clinical routine because of usually long computing times. Given the fact that dose distributions are more conformal in proton therapy than with conventional modalities, the potential impact of Monte Carlo dose calculation is presumably bigger in proton therapy. The goal of this work was to facilitate the clinical use of Monte Carlo dose calculation to support routine treatment planning and delivery at the Francis H Burr Proton Therapy Center at Massachusetts General Hospital (NIGH). Further, we aimed at incorporating time-dependent patient information, as obtained from 4D CT data, into the dose calculation process. Results: Proton Monte Carlo dose calculation is clinically implemented at NIGH. To do so, the treatment head and patient geometries had to be modeled in detail, the beam had to be well characterized and the dose engine had to be linked to patient database and hardware control parameters. So far, dose calculations have been done for many cases in radiosurgery and for breast, prostate, nasopharynx, paranasal sinus, paraspinal, and lung malignancies. Differences to the commercial pencil-beam algorithm could be identified in cases of frequent soft-tissue/bone/air interfaces and in particular for patient with metallic implants. The resulting Monte Carlo dose distributions can be imported into the planning system for analysis. Based on a simulation of the ionization chamber in the treatment head, the Monte Carlo generated doses are given in absolute units (Gy per proton fluence). To incorporate motion information for thoracic tumors we developed a fourdimensional Monte Carlo dose calculation technique, which allows the consideration of 4D CT data sets. Conclusions: Proton Monte Carlo dose calculation can be done efficiently for treatment planning support. Many steps on the Monte Carlo dose calculation level as well as in linking the dose engine to the patient database and the hardware control system were required to accomplish this.