Surface optimization for the mammosite applicator

Abstract

We present a technique to optimize the dwell times and positions of a HDR Ir-192 source in brachytherapy procedures using the MammoSite applicator. Due to balloon deformation and the anisotropy of the Ir-192 source, the portion of the PTV along the source axis receives less dose compared to more central portions of the PTV if a single dwell position is used. The surface optimization method attempts to conform the 100% isodose line to the surface of the PTV. The surface optimization technique is described and the dosimetric characteristics of this method are compared to the single dwell position technique based on 20 patients treated at our institution. The study population consists of 20 patients treated using the MammoSite device between October 18, 2002 and February 13, 2004. Treatment was delivered in 10 fractions of 3.4 Gy, b.i.d. Each patient was planned using two techniques based solely on CT images. To improve the accuracy of the balloon catheter reconstruction and placement of the prescription point, the original CT dataset was resliced such that the entire catheter is visible on a single slice. The single dwell position technique uses a single source in the center of the balloon to deliver 3.4 Gy to the prescription point 1cm from the balloon surface along a line perpendicular to the source axis. The surface optimization technique uses optimization points on the surface of the PTV and optimizes the dwell times to equalize the dose at the optimization points. The resulting dwell times are then scaled to deliver 340 cGy to the single dwell position prescription point The surface optimization technique increased the percentage of the PTV covered by the prescription dose compared to the single dwell position technique from a mean of 85% to 94%. The mean percentage of the PTV receiving 150% of the prescription dose was 37% for the surface optimization technique and 29% for the single dwell position technique. The dose homogeneity index, DHI, and full width at half maximum (FWHM) of the differential PTV dose volume histogram were used to evaluate dose uniformity. The surface optimization had a mean DHI = 0.61 and the single dwell position had a mean DHI = 0.67. The surface optimization method had a mean FWHM = 204 and the single dwell position technique had a mean FWHM = 202. Compared to the single dwell position technique, the surface technique corrects for the balloon elongation and source anisotropy and provides greater coverage of the PTV. The DHI indicates that the single dwell position technique had a more homogenous plan. However, DHI fails to account for the portion of the PTV receiving less then 100% coverage and the single dwell position method had a larger DHI because it is normalized to a smaller V100. The FWHM values are equivalent for the two techniques and thus indicate equivalent uniformity(TABLE 1).