A Review of Multi-Lumen Breast Brachytherapy for Savi With Acuros Software

Purpose: To compare new commercially available software called Acuros to conventional TG43 formalism dose calculations for patients undergoing accelerated partial breast irradiation with SAVI. Materials and Methods: Currently available software, based on TG43 formalism, calculates brachytherapy dose distributions assuming the surrounding media is equivalent to water. This approximation does not take into account tissue-air interfaces, which could erroneously increase the dose distributions near the air boundaries. Recent papers have analyzed tissue-air interfaces with MCNP calculations and in-vitro measurements. The results as stated in the papers confirm that tissue-air interfaces play a large role in the buildup of dose seen at these interfaces. In this research, patients who received treatments with SAVI interstitial devices were calculated on Varian Brachyvision utilizing both TG 43 and Acuros software. All patients were treated at Arizona Center for Cancer Care, Peoria Arizona. For both algorithms, clinically significant statistics from the RTOG 0413 APBI study were compared. Structures that are defined from RTOG 0413 protocol are as follows: balloon, breast tissue, and PTV_EVAL (which is defined as 1 cm expansion of the balloon but contained between 5mm from skin surface and the chest wall). Our study compared the results of the two algorithms for SAVI patients that were treated with varying volumes of PTV_EVAL and skin distances. Results: The data shows that in all cases of small skin distances!2 mm to large skin distances of 10 mm that the maximum skin doses were lower for all Acuros calculations when compared to the TG43 formalism. The results were the same regardless of the volume of the PTV_EVAL and the coume of air cavity that remains insdie the multi-lumen struts as seen by CT evaluation. Conclusions: The Acuros calculations showed a noticeable decrease in the maximum dose to the skin when compared to the TG43 formalism. This result is primarily due to the lack of full backscatter that would otherwise have been assumed when using the TG43 calculations. While the air cavities inside the device would have predicted a greater contribution at a distance the tissue-air interface reduced the effect.