Session Item

The trend towards partial breast irradiation (PBI) and increased hypofractionation stresses the importance of the geographic accuracy of radiation delivery. As there are fewer fractions to compensate for a geographical miss in one fraction, the risk of local recurrence might increase. Furthermore, the increased use of full thickness closure leads to a seroma that is often not clearly visible on daily imaging. Fiducials are used as surrogates for patient positioning. To assess the accuracy of treatment delivery in this situation, we investigated the motion of fiducials relative to the tumor bed, assuming patients would be treated with single fraction PBI. Possible geographical miss and compensation for this error are evaluated by calculating CTV to PTV margins.

Sixty-eight patients treated in a Cyberknife PBI cohort study were included. All patients had a minimum of 3 titanium surgical clips placed in the tumor bed during lumpectomy. They also had 3 interstitial gold markers inserted postoperatively around the tumor bed under US guidance. A planning CT and a diagnostic-quality in-room CT at the first fraction were made with identical scanning parameters with a median interval of 15 days. The two CT scans were co-registered on the tumor bed area using MIM 6.9.3 software. The residual displacement was calculated for each fiducial individually and for the center of mass (CoM) of all interstitial markers and all surgical clips. These errors were used to calculate the uniform three-dimensional CTV to PTV margin required for the motion of the fiducials relative to the tumor bed. We calculated margins varying the percentage of patients with a completely covered CTV during treatment. We used the 2-sample Kolmogorov Smirnov test to compare the error distributions for the two fiducial types.

The distributions of the motion of the CoM per fiducial type relative to the tumor bed are shown in figure 1. The distributions were significantly different for the two fiducial types (p < 0.001). In figure 2, the margin is plotted against the percentage of patients with 100% CTV coverage. The uniform margin required to compensate for fiducial motion in 90% of patients was 2.3 mm for the interstitial markers and only 1.7 mm for the surgical clips. Applying this margin of 2.3 mm to positioning based on surgical clips would result in 95% of patients with 100% CTV coverage.

There is substantial fiducial CoM motion relative to the tumor bed between simulation and PBI treatment. Its magnitude was influenced by the type of fiducial. The margin required to cover the CTV of 90% of patients was 2.3 mm for positioning based on interstitial markers. This margin would cover the CTV of 95% of patients if surgical clips were used, potentially decreasing the risk of local recurrences. Our method of calculating the percentage of patients with 100% coverage for each margin allows for an informed decision on the trade-off between irradiated volume and risk of geographical miss.