To assess the effect of BMI on patient setup reproducibility in hypofractionation of breast cancer during the COVID-19, to establish an imaging protocol

Abstract

Abstract Objective To analyze the setup accuracy among patients treated for breast cancer with hypofractionation radiation therapy (HFRT) regimen (five fractions instead of 15-16 fractions in standard regimen) and predict the necessity of performing the setup imaging in the 4th and 5th fractions as a function of setup accuracy in the first three fractions. Method This retrospective study reviewed setup displacements in each direction (lateral, longitudinal, and vertical) for 51 women with breast cancer treated with HFRT at the Radiation Therapy Unit between September 2020 and May 2022. Besides the five fractions (#1– # 5), the mean setup error was computed for the first three fractions (AVG-III) for each direction. Accuracy rates were computed for each direction and fraction as the percentage of fractions with setup error ≤ 0.5 cm. The correlations of #1, #2, #3, and AVG-III setup errors and their value in indicating #4 and #5 setup error and accuracy were analyzed using Pearson's coefficient and Receiver Operating Characteristics (ROC) curve, respectively. Furthermore, the effect of body mass index (BMI) on setup reproducibility was analyzed using logistic regression. Result The mean (SD) age of the participants was 54.41 (11.46) years. There was a high percentage of overweight (25.5%) and obese (53.0%). The mean setup error was <0.5cm for all five fractions and three directions, and accuracy rates were remarkably high ranging between 80.4%–90.2%, 84.3%–94.1%, and 94.1%–100.0% in the lateral, longitudinal, and vertical directions, respectively. The bivariate correlations analysis showed no significant correlations of fraction #4 (Pearson’s coefficient r = -0.057–0.269; p>0.05) and #5 (r = -0.128–0.254; p>0.05) within any of the first three fractions or AVG-III, in any of the directions. In the ROC curve, only #5 accuracy was indicated by #3 in the longitudinal direction (AUC=0.89, p=0.025). BMI was only associated with inaccurate setup for fraction #3 in the lateral direction, in a positive relationship (OR=1.15, 95% CI = 1.01–1.30; p=0.031). Conclusion Setup accuracy in the first fractions of HFRT does not predict accuracy in the two last fractions nor is predicted by the patient’s BMI. Consequently, women with breast cancer treated with HFRT require daily imaging for optimal setup before each radiotherapy fraction.