Image-guided intracavitary high-dose-rate brachytherapy for cervix cancer: A single institutional experience with three-dimensional CT-based planning

Abstract

Purpose: To evaluate and report volumetric dose specification of clinical target volume (CTV) and organs at risk with three-dimensional CT-based brachytherapy. In this study, we analyzed CTV volumes and correlated the dose specification from CT-based volumes with doses at classical point A and International Commission on Radiation Units and Measurements (ICRU) points.

Methods and materials: Ten patients who underwent definitive high-dose-rate brachytherapy for cervical cancer between May 2006 and March 2007 were retrospectively identified for this study. Each patient underwent five intracavitary insertions with CT-compatible ring and tandem applicators using a universal cervical Smit sleeve. Dose of 6.0Gy per fraction was prescribed to the 100% isodose line. The dose distribution was modified using the feature of "geometry optimization" to achieve maximum CTV coverage and to spare the organs at risk. The minimal doses for most irradiated 2, 1, 0.1cm(3) of bladder (D(BV2) , D(BV1), and D(BV0.1)) and rectum (D(RV2), D(RV1), and D(RV0.1)) were determined from dose-volume histograms and were compared with the doses estimated at the ICRU reference points.

Results: The mean CTV of the 10 patients had a shrinkage trend over the five fractions, with a mean of 77.4cm(3) from the first fractions and a mean of 65.5cm(3) from the last fractions (r=-0.911, p=0.031). CTV volumes directly correlated with dose to point A (r=0.785, p=0.007). Eight of 10 patients achieved an average dose received by at least 90% of volume (D(90)) >/=6.0Gy. For bladder, the doses determined from the 3-dimensional (3D) plan correlated significantly with the doses to the ICRU reference bladder point, for example, D(BV2) (r=0.668, p<0.001), D(BV1) (r=0.666, p<0.001), and D(BV0.1) (r=0.655, p<0.001). However, for rectum, the estimated doses to the ICRU reference rectal point did not correlate significantly with doses determined from 3D plan, for example, D(RV2) (r=0.251, p=0.079), D(RV1) (r=0.279, p=0.049), and D(BV0.1) (r=0.282, p=0.047).

Conclusions: Our experience showed that excellent dose coverage of CTV can be achieved with image-guided CT-based planning with geometric optimization although maximal sparing of rectum was not achieved. Careful dose constraints and standardization of D(90) should be considered when optimizing doses to target tissues such that normal tissue constraints can be met.