Session Item

In routine IGRT of patients with breast cancer, positioning is often based on a tangential MV and an orthogonal kV image (kV-MV setup) prior to treatment delivery. It is convincing to see the target (breast) in the beams-eye view with the MV image, but it can be challenging to interpret setup difficulties in the form of e.g. rotations or arm position based on a kV-MV setup. In the present study a setup based on an anterior-posterior (AP) kV image and a lateral kV image (kV-kV setup) is evaluated using MV images acquired during treatment delivery  for patients treated in Free-Breathing (FB) or Deep Inspiration Breath-Hold (DIBH). In addition, the yaw rotation setup error is quantified. 

The analysis was based on 84 fractions from 11 patients treated with 3D conformal radiotherapy with tangential fields after breast conserving surgery, with and without lymph node involvement. The DIBH technique was used for 7 patients. To evaluate the setup deviations in the AP direction (vertical) between the kV-kV setup and MV images, MV images were acquired during treatment delivery for the two open tangential fields (n = 168). The setup deviations between the two unpaired groups treated with FB or DIBH was compared using a Wilcoxon rank sum test. Furthermore, the yaw setup correction from the initial patient position, based on in-room lasers and tatoo marks, were retrospecitvly collected based on the kV-kV setup. Yaw setup corrections above 3 degrees requires repositioning in the clinical setting in our institution.

Absolute vertical setup deviations were ≤ 4 mm in 98 % and 97 % of the acquired MV images in DIBH and FB respectively (Figure 1). The setup deviations was found to be statistically significantly larger (p = 0.01) for patients treated in FB (median 2 mm) compared to DIBH (median 1 mm). The patient position correction based on the kV-kV setup led to yaw setup corrections within 1° and 2° in 56 % and 85 % of the treatment fractions, respectively (Figure 2). In 4 % of the treatment fractions the yaw setup corrections was above 3°, which would require patient repositioning.

Setup of breast cancer patients based on planar kV-kV images was observed to be in good agreement with tangential MV images in both FB and DIBH. The setup deviation was statistical significantly larger for patients treated with FB compared to DIBH, however the difference in median was small (1 mm). With a kV-kV based setup it is possible to correct for yaw rotations, which based on the present study was above 2° in 15 % of the treatment fractions. Other advanatages using kV-kV for breast setup are visualization of surgical clips and heart.