P185] Comparison of two inverse planing algorithms HIPO and IPSA in terms of their suitability for breast brachytherapy planning

Abstract

Purpose The aim of the study was to compare two algorithms of inverse brachytherapy planning Hybrid Inverse Planning Optimization (HIPO) and Inverse Planning Simulated Annealing (IPSA) in terms of their suitability for breast brachytherapy planning. The values of dose distribution parameters was evaluated for target volumes and critical organs. Methods The analysis included 40 treatment plans prepared for 20 breast cancer patients treated with brachytherapy boost using rigid interstitial applicators. For each patient CTV and organs at risk were contoured. Next two treatment plans were performed using HIPO and IPSA algorithms, with individual constrains selected to be fully comparable. The source step was set at 3 mm and dwell time gradient restriction (DTGR) at 0.5 level. Obtained treatment plans were not supposed to fulfill clinical criteria, but only to show differences between the analyzed algorithms. The collected data were analyzed statistically using the Wilcoxon test for non-parametric variables. Results For CTV the following average values were obtained: V100 71.26% and 82.35%, V150 13.85% and 19.11% and V200 6.31% and 8.39% for the IPSA and HIPO algorithms respectively. Similarly, COIN and DNR values were higher using the HIPO algorithm and were 0.65 and 0.93 and 0.2 and 0.22. In case of organs at risk, the following values were obtained: for chestwall D1cc 49.60% and 53.80% and D0.1 cc 55.62% and 60.56%, for skin D1cc 41.91% and 54.72% and D0,2 cc 55.66% and 62.1% respectively for IPSA and HIPO algorithms. The mean dose in the lung was also higher for the HIPO algorithm and reached 15.19% and 21.48% respectively. All differences were statistically significant at p Conclusions Presented results indicate the predominance of the IPSA algorithm, especially in the area of lowering doses for critical organs. However the coverage parameter COIN clearly indicates the advantage of the HIPO algorithm. This is directly due to much more uniform dwell times generated by HIPO algorithm. The presented comparison may be used as a valuable hint for choosing the initial optimization method in planning the BT boost for breast cancer patients.