BP05 Presentation Time: 4:36 PM

Purpose

High dose-rate brachytherapy can be plagued by a multitude of user errors that could result in late toxicities for patients. A common one reported to NRC is wrong treatment length. This error will result in a shifted dose distribution between planned and delivered dose. To automate and verify delivery with correct treatment lengths, Varian released its Bravos remote afterloader which uses the device's dummy wire for accurate pre-treatment applicator measurements. Bravos measurements are compared with the planned lengths and if differences are small, treatment length is adjusted automatically. In this process, dwell positions are kept unchanged relative to the tip of the applicator. In breast interstitial implants, from our clinical experience, we noticed, through these automated measurements a significant catheter lengthening that occurs between the first and second fractions. The first fraction is typically delivered within 2h from implantation and the second fraction is more than 12h later. The catheters length does not seem to change much after that. We seek to quantify the dosimetric impact of such automated adjustments, identify cases and treatment parameters particularly sensitive to these effects and describe the best planning and delivery mitigation strategies to ensure accurate dose delivery.

Materials and Methods

A retrospective cohort of 14 successively treated breast cancer patients using multi-catheter implants were selected for analysis. Actual length measurements were used to create plans simulating the dosimetric impact of making or not making these adjustments. Daily QA for two years shows positional accuracy of 0.3±0.3mm and for the dummy and source wires. In these simulations, it was hypothesized that the target did not change in volume or position relative to breast. In planning these treatments, two targets are used, expansions of the lumpectomy cavity, a CTV1cm and CTV1.5cm for which we try to achieve V95>95% and V90>90% respectively. The max skin dose is limited to <100% Prescription dose regimens included 7.5Gyx3fx=22.5Gy(8/14 patients) and 4.3Gyx7fx=30.1Gy(6/14 patients). Dose inhomogeneity was evaluated using V150% and V200% for breast tissue. The skin contour encompassed an area 5mm from the surface of the breast and the maximum dose was evaluated using D0. 1cc.The change in catheter lengths, as measured by Bravos, were recorded for each fraction of a patient's treatment.

Results

The following results show single fraction changes between the original clinically delivered plan and the re-plan which does not consider catheter lengthening by the amount measured by Bravos prior to the second fraction) averaged over all 14 cases. The average reductions in the V95% of the cavity 1.0 and V90% of the cavity 1.5 structures were -0.49±0.99% and -2.93±2.85%, respectively. Negligible changes in normal tissue dose were observed with an average increase in the V200% and V150% for the body of 0.09±0.16% and 0.37±0.48%, respectively. The average change in the skin D0.1cc were 3.37±3.90Gy. The max increase in skin D0.1cc was11.43Gy corresponding to a case in which the lumpectomy cavity abutted the skin surface. In another case, the skin D0.1cc was reduced by 72.3cGy in which the cavity was located distally to the tips of the catheters. The average change in catheter lengths between the first and second fractions amongst all patients was 3.95±1.13mm with a max single catheter lengthening of 6mm.

Conclusions

Pre-treatment applicator length measurements are advised when the target is proximal to the skin surface to prevent late toxicities. However, such measurements may be redundant following the first fraction in cases in which the target is located medially or distally to the surface of the breast. Catheter lengthening following implantation generally peaks 24h after the implant. In general, the degree of lengthening increases with increasing lumpectomy cavity (breast) volume.