Using gEUD based plan analysis method to evaluate proton vs. photon plans for lung cancer radiation therapy.

The goal of this study was to exam the efficacy of current DVH based clinical guidelines draw from photon experience for lung cancer radiation therapy on proton therapy. Comparison proton plans and IMRT plans were generated for 10 lung patients treated in our proton facility. A gEUD based plan evaluation method was developed for plan evaluation. This evaluation method used normal lung gEUD(a) curve in which the model parameter "a" was sampled from the literature reported value. For all patients, the proton plans delivered lower normal lung V5 Gy with similar V20 Gy and similar target coverage. Based on current clinical guidelines, proton plans were ranked superior to IMRT plans for all 10 patients. However, the proton and IMRT normal lung gEUD(a) curves crossed for 8 patients within the tested range of "a", which means there was a possibility that proton plan would be worse than IMRT plan for lung sparing. A concept of deficiency index (DI) was introduced to quantify the probability of proton plans doing worse than IMRT plans. By applying threshold on DI, four patients' proton plan was ranked inferior to the IMRT plan. Meanwhile if a threshold to the location of curve crossing was applied, 6 patients' proton plan was ranked inferior to the IMRT plan. The contradictory ranking results between the current clinical guidelines and the gEUD(a) curve analysis demonstrated there is potential pitfalls by applying photon experience directly to the proton world. A comprehensive plan evaluation based on radio-biological models should be carried out to decide if a lung patient would really be benefit from proton therapy.

Investigation on using high-energy proton beam for total body irradiation (TBI).

This work investigated the possibility of using proton beam for total body irradia-tion (TBI). We hypothesized the broad-slow-rising entrance dose from a monoen-ergetic proton beam can deliver a uniform dose to patient with varied thickness. Comparing to photon-based TBI, it would not require any patient-specific com-pensator or beam spoiler. The hypothesis was first tested by simulating 250 MeV, 275 MeV, and 300 MeV protons irradiating a wedge-shaped water phantom in a paired opposing arrangement using Monte Carlo (MC) method. To allow ± 7.5% dose variation, the maximum water equivalent thickness (WET) of a treatable patient separation was 29 cm for 250 MeV proton, and > 40 cm for 275 MeV and 300 MeV proton. The compared 6 MV photon can only treat patients with up to 15.5 cm water-equivalent separation. In the second step, we simulated the dose deposition from the same beams on a patient's whole-body CT scan. The maximum patient separation in WET was 23 cm. The calculated whole-body dose variations were ± 8.9%, ± 9.0%, ± 9.6%, and ± 14% for 250 MeV proton, 275 MeV proton, 300 MeV proton, and 6 MV photon. At last, we tested the current machine capability to deliver a monoenergetic proton beam with a large uniform field. Experiments were performed on a compact double scattering single-gantry proton system. With its C-shaped gantry design, the source-to-surface distance (SSD) reached 7 m. The measured dose deposition curve had 22 cm relatively flat entrance region. The full width half maximum field size was measured 105 cm. The current scatter filter had to be redesigned to produce a uniform intensity at such treatment distance. In con-clusion, this work demonstrated the possibility of using proton beam for TBI. The current commercially available proton machines would soon be ready for such task.