Evaluation of patient residual deviation and its impact on dose distribution for proton radiotherapy

The residual deviations after final patient repositioning based on bony anatomy and the impact of such deviations on the proton dose distributions was investigated. Digitally reconstructed radiographs (DRRs) and kilovoltage (kV) "portal verification" images from 10 patients treated with passively scattered proton radiotherapy was used to estimate the residual deviation. These changes were then applied to the location of isocenter points that, in effect, moved the isocenter relative to the apertures and compensators. A composite verification plan was obtained and compared with the original clinical treatment plan to evaluate any changes in dose distributions. The residual deviations were fitted to a Gaussian distribution with μ = -0.9 ± 0.1 mm and σ = 2.55 ± 0.07 mm. The dose distribution showed under- and overcovered dose spots with complex dose distributions both in the target volumes and in the organs at risk. In some cases, this amounts to 63.5% above the intended clinical plan. Although patient positioning is carefully verified before treatment delivery and setup uncertainties are accounted for by using compensator smearing and aperture margins, a residual shift in a patient's position can considerably affect the dose distribution.