Room: Track 5
Purpose: Radiation oncologists (ROs) use surgical clips (SCs) for tumor bed (TB) delineation in boost and partial breast irradiation (PBI). However, SCs may not provide a reliable radiographic surrogate of TB locations post-oncoplastic breast surgery (OBS). The accuracy of RO-contoured TBs (TB_RO) post-OBS was investigated.
Methods: Six representative OBS techniques were performed on realistic breast phantoms. Eight SCs were used to mark the TB following tumor excision. Additional radio-opaque markers were used to track TB deformation post-OBS closure (TB_True). Each phantom was CT imaged at different surgery phases. Pre- and post-OBS closure SC displacements were recorded. Two ROs were asked to delineate TBs on CTs with homogenous breast parenchyma density and SCs. Inter- and intra-RO contour agreements were calculated using the dice similarity coefficient (DSC) and Hausdorff distance (HD). Clinical-target-volume (CTV) was defined by applying expansions from 5-15 mm to TB_RO contours. The DSC, HD, undercontoured volumes (UCV) and overcontoured volumes (OCV) were measured for RO contours compared to TB_True. Results were reported as an average for all cases.
Results: The SC displacement was 9.2 ± 8.7 mm. DSC and HD were 0.79 ± 0.06 and 12.7 ± 3.6 mm (intra-RO), and 0.77 ± 0.05 and 12.4 ± 2.8 mm (inter-RO). ROsAvg and TB_RO+15mm DSC and HD were 0.54 ± 0.12 and 25.2 ± 9.2 mm, 0.57 ± 0.08 and 23.8 ± 6.5 mm, respectively. ROsAvg and TB_RO+15mm UCVs and OCVs were 28.0 ± 16.7 cc and 35.8 ± 28.1 cc, 5.4 ± 7.6 cc and 154.4 ± 82.0 cc, respectively.
Conclusion: Intra- and inter-RO contours agreed but were systematically off from the True_TB. Defining CTV margins did not improve contour agreement and may cause significant over-irradiation of normal tissue. Following OBS-procedures, the accuracy of TB delineation by ROs is challenging and makes indication of dose boost or PBI questionable.