Modeling study for optimization of skin dose for partial breast irradiation using Xoft Axxent electronic...

Balloon brachytherapy with the MammoSite system (Hologic Inc., Bedford, MA) is a widely used approach for accelerated partial breast irradiation. Inherent to this approach, high skin doses can occur if the balloon to skin distance is small. This has been associated with late skin toxicity, particularly telangiectasia. The Xoft Axxent electronic brachytherapy balloon applicator (Xoft, Fremont, CA) is a novel device for accelerated partial breast irradiation. It is unique in that it uses an electronic 50-kV source. This source has a pronounced anisotropy with constriction of isodose distribution at the proximal end of the catheter. This anisotropy can be considered as an advantage to optimize skin dose when the cavity to skin distance is small. In this study, we simulated various balloon-insertion orientations to optimized skin surface dose.

Breast phantoms were constructed of tissue-equivalent material. Xoft Axxent balloon catheters were inserted at a distance of 6mm from the surface. The catheter was placed at three different catheter to surface orientations: (1) perpendicular to the surface, (2) oblique to the surface (45 degrees), and (3) parallel to the surface. Three-dimensional treatment planning was then performed using Nucletron's Plato planning system (Nucletron, Columbia, MD). Multiple dwell positions were used, and the dose was optimized to the target volume. The target volume was defined as volume from the balloon surface to 1-cm distance from the balloon surface or to the phantom surface (if less then 1cm from the balloon surface). Target volume coverage was compared between plans using dose-volume histograms. Surface doses were compared using isodose line distribution and surface point doses. Plato planned surface doses were then verified by direct measurement using Landauer Dot InLight dosimeters (Landauer, Glenwood, IL).

Excellent target coverage was obtained for all three catheter orientations with a D(95) of > or =95%. Surface dose was lowest for the perpendicular orientation with a calculated dose of 99%. The parallel orientation had the highest surface dose of 164%. The oblique orientation showed intermediate results with a surface dose of 117%. Measured surface doses were reproducible and correlated well with calculated values.

Optimized Xoft Axxent balloon catheter orientation using source anisotropy and multiple dwell positions can be used to minimize excessive skin dose and yet maintain optimal tumor cavity coverage when the cavity to skin distance is small. This has the potential to decrease skin late effects and improve cosmetic outcome. Further clinical study is warranted.