A Comparison of Two Set-up Techniques for Tangential Breast Radiotherapy Treatments Using Portal Imaging and In-vivo Transit Dosimetry Data

Abstract

Background Accurate patient positioning for tangential breast treatments is crucial as this can influence the dose the patient receives. Two different set-up techniques have been examined for tangential breast radiotherapy treatments to determine which results in more accurate patient positioning and ultimately gives a better dose distribution across the target volume. Method This single centre service evaluation used a quasi-experimental design consisting of two technique groups; '5mm deviation' and the 'set to medial point' group. Each group contained 30 patients. Patients were found using relevant filters in Aria Report Manager and a method of stratified random sampling was used to find tangential treatments between 6-10MV. Image match analysis was carried out and any discrepancy in magnitude and direction between the Digitally Reconstructed Radiograph (DRR) and the integrated image noted and the corresponding transit dosimetry results were recorded from the DosimetryCheck program (Math Resolutions, LLC, Columbia, MD). Results The image match results for the 'set to medial point' technique showed 40% had a match discrepancy of 1mm compared to 3.3% in the '5mm deviation' group, only 3.3% of the matches in the 'set to medial point' group were above the 5mm accepted tolerance for analysing images offline compared to 23.3% in the '5mm deviation' group. The transit results for the '5mm deviation' group showed that two patients were outside the accepted +/- 10% tolerance and an additional five readings were over +/- 9%. In the 'set to medial point' group no in-vivo readings were outside +/- 10% and only 9.9% had readings above +/- 5%. A Spearman's Rho test showed a statistically significant, weak, positive monotonic correlation for the image match and transit results for the '5mm deviation' technique (rs,=0.358, n=30, p=0.052), in contrast to a non-significant, very weak correlation in the 'set to medial point' group (rs,=-0.041 n=30, p=0.830). Conclusions These findings illustrate that the 'set to medial point' technique provides a more stable patient set-up and also an improved dose distribution in comparison to the '5mm deviation' technique. Additionally, the DosimetryCheck program has proven its usefulness as a tool capable of identifying an unacceptable dose distribution when the patient's alignment differs significantly from the planned position.