Can point doses predict volumetric dose to rectum and bladder: a CT-based planning study in high dose rate intracavitary brachytherapy of cervical carcinoma?

Abstract

Point doses, as defined by the International Commission on Radiation Units and Measurements (ICRU), are classically used to evaluate doses to the rectum and bladder in high dose rate intracavitary brachytherapy (HDR-ICBT) in cervical cancer. Several studies have shown good correlation between the ICRU point doses and the volumetric doses to these organs. In the present study we attempted to evaluate whether this correlation could be used to predict the volumetric doses to these organs. A total of 150 HDR-ICBT insertions performed between December 2006 and June 2008 were randomly divided into two groups. Group A (n=50) was used to derive the correlation between the point and volumetric doses using regression analysis. This was tested in Group B (n=100) insertions using studentised residuals and Bland-Altman plots. Significant correlations were obtained for all volumetric doses and ICRU point doses for rectum and bladder in Group A insertions. The strongest correlation was found for the dose to 2 cc volumes (D(2cc)). The correlation coefficients for bladder and rectal D(2cc) versus the respective ICRU point doses were 0.82 and 0.77, respectively (p<0.001). Statistical validation of equations generated in Group B showed mean studentised residual values of 0.001 and 0.000 for the bladder and rectum. However, Bland-Altman analysis showed that the error range for these equations for bladder and rectum were ±64% and ±41% of the point A dose, respectively, which makes these equations unreliable for clinical use. Volumetric imaging is essential to obtain proper information about volumetric doses.