Evaluation of HDR interstitial breast implants planned by conventional and optimized CT-based dosimetry systems with respect to dose homogeneity and conformality.

BACKGROUND AND PURPOSE: Recently, the use of brachytherapy for partial breast irradiation has increased significantly. The aim of this study was to make dosimetric comparisons between conventional (CONV) and CT-based optimized dosimetry systems applied to breast implants.

PATIENTS AND METHODS: 17 patients treated with high-dose-rate (HDR) interstitial brachytherapy were selected for the study. Two patients had two-plane and 15 three-plane implants. Treatment planning was based on conventional two isocentric radiographs and dose point optimization (CONV). For each patient postimplant CT scans were taken, and the target volume (lumpectomy cavity with 1 cm margin) was outlined in all axial slices. The treatment planning was repeated using CT images. The dose distributions were analyzed by dose-volume histograms. To quantify the dose distributions, volume (V90, V100, V150, V200) and dose (D90, D(min), mean central dose [MCD]) parameters, along with the dose nonuniformity ratio (DNR), dose homogeneity index (DHI), external volume index (EI) and conformal index (COIN) were used. For each implant, three more virtual treatment plans were created using the Paris dosimetry system (PDS), geometrically optimized system (GOS) and conformal system (CONF). Dose and volume parameters were calculated and compared.

RESULTS: The median number of catheters amounted to ten (range: 6 to 13) and the average volume of planning target volume to 63.4 cm(3) (range: 17.7-122 cm(3)). The mean target coverage was 70%, 61%, 57% and 87%; the D90 72%, 64%, 60% and 94%; the DNR 0.35, 0.25, 0.25 and 0.55; the EI 0.62, 0.54, 0.08 and 0.15; the COIN 0.40, 0.34, 0.50 and 0.74 for the CONV, PDS, GOS and CONF systems, respectively.

CONCLUSION: With CT-based optimized dose planning the target coverage can be significantly increased compared to the conventional dosimetry systems, but the target dose distribution will be more inhomogeneous. To improve the quality of brachytherapy implants, the image-based three-dimensional information should be used not only for dose plan evaluation, but also previously, for planning the geometry of the catheter positions and performing the insertions.