Comparison of dose statistics for bladder wall and rectum wall vs whole organs for VMAT prostate treatment

Abstract

Dose-wall histograms (DWHs) have been used as alternatives to dose-volume histograms (DVHs) for hollow organs, with the rationale that the dose delivered to the interior of a hollow organ would be unrelated to the level of radiation damage. The purpose of this study is to conduct a statistical comparison of dose statistics for both walled and solid structure contours for both bladder and rectum in the treatment of intermediate risk prostate cancer with volumetric arc therapy (VMAT). Ten intermediate risk prostate cases were randomly selected. Rectum and bladder were first contoured as solid structures, and then the corresponding wall structures were generated using either a slice-by-slice cropping (2D method), or with a full 3D cropping tool (3D method). Each case was then inverse planned using a 2-arc VMAT technique. Two plans per case were created, 1 with a hypofractionated treatment and 1 with a standard fractionated treatment. DVHs were calculated for solid structure contours, and DWHs were calculated for the walled structure contours generated using 2D and 3D contouring tools. A nonparametric Spearman statistic correlation test was used to compare a large number of relevant dose histogram points, and to establish the relationship between dose statistics for walled and solid structures. Several notable relationships were observed. Maximum rectal dose was strongly correlated between the solid structure and both the 2D-generated (Spearman's correlation r_s = 0.988, p < 0.01) and 3D-generated (r_s = 0.952 p< 0.01) wall structures. This indicates that the rectal hot spot occurred in or near the wall for all cases, suggesting that both structure types give similar maximum dose information for rectum. Maximum bladder dose was not significantly correlated between solid structures and the 2D (r_s = 0.596, p= 0.069) and 3D-generated (r_s = 0.681, p= 0.03) counterparts. This suggests that the maximum dose is not consistently in or near the bladder wall. This favors the use of bladder wall contours when considering bladder toxicity, with the maximum dose to the wall potentially being more relevant radiobiologically. This analysis was extended to many other relevant points on the rectum and bladder histogram curves. Where correlations are strong, equations of best-fit are presented. This work establishes several statistically-significant relationships between bladder and rectum DVHs and DWHs for VMAT irradiation of intermediate-risk prostate cancer. This information may be used to inform contouring requirements for clinical trial design as well as for standard patient care.

Keywords: Bladder wall; DWH; PROFIT; Rectum wall; VMAT.