Comparison of set-up errors by breast size on wing board by portal imaging.

To quantify and compare setup errors between small and large breast patients undergoing intact breast radiotherapy. 20 patients were inducted. 10 small/moderate size breast in arm I and 10 large breast in arm II. Two orthogonal and one lateral tangent portal images (PIs) were obtained and analyzed for systematic (Σ) and random (σ) errors. Effect of no action level (NAL) was also evaluated retrospectively. 142 PIs were analyzed. Σ(mm) was 3.2 versus 6.7 (p = 0.41) in the mediolateral (ML) direction, 2.1 versus 2.9 (p = 0.06) in the craniocaudal (CC) and 2.2 versus 3.6 (p = 0.08) in the anteroposterior (AP) direction in small and large breast, respectively. σ(mm) was 3.0, 3.3 and 3.3 for small breast and 4.1, 3.7 and 3.2 for large breast in the ML, CC and AP direction (p = 0.07, 0.86, 0.37), respectively. 3 D Σ(mm) was 2.7 versus 4.2 (p = 0.01) and σ(mm) was 2.5 versus 3.2 (p = 0.14) in arm I and II, respectively. The standard deviation (SD) of variations (mm) in breast contour depicted by central lung distance (CLD) was 5.9 versus 7.4 (p < 0.001), central flash distance (CFD) 6.6 versus 10.5 (p = 0.002), inferior central margin (ICM) 4 versus 4.9 (p < 0.001) in arm I and II, respectively. NAL showed a significant reduction of systematic error in large breast in the mediolateral direction only. Wing board can be used in a busy radiotherapy department for setting up breast patients with a margin of 1.1 cm, 0.76 cm and 0.71 cm for small breasts and 1.96 cm, 1.12 cm and 0.98 cm for large breast in the ML, AP and CC directions, respectively. The large PTV margin in the mediolateral direction in large breast can be reduced using NAL. Further research is needed to optimize positioning of large breasted women.