Organ at risk dose-volume metrics in a series of hypofractionated breast radiotherapy with integrated boost

Abstract

Dose and volume metrics to organs at risk are used for evaluation and optimization in radiotherapy planning. However, the numerous choices of metrics can be confusing. In a series of patients treated with hypofractionation and an integrated boost for breast cancer, we aim to determine if a parsimonious selection of representative metrics can be identified. The dosimetries of 42 patients receiving 42 Gy to the breast, with or without nodal irradiation, and 51 Gy integrated boost to tumor bed in 15 fractions were reviewed. For each organ-heart, lungs, and contralateral breast-cumulative dose-volume histograms were used to extract values for 3 basic metric classes: Two additional classes were considered: Pearson correlation coefficient R was calculated between pairs of values within each basic class and with the 2 additional classes for each organ. The interquartile ranges of correlations for D.yy, Vrel.xx, and Vabs.xx were as follows: The mean dose correlated with all basic classes for the heart and lungs, and with dose D.yy and volumes at Vrel.10-Vabs.10 for the contralateral breast. The standard deviation correlated with Vrel.xx and Vabs.xx for the heart and lungs (R ≥ 0.70). Among the D.yy, D.50 (median dose) correlated with the mean and standard deviation for all organs (R = 0.65-0.96). The mean, standard deviation, and median doses were the preeminent correlators. These statistics appear to be parsimonious representatives of doses to organs. Further studies with other radiotherapy series will be necessary to validate these observations.

Keywords: Correlations; Dispersion; Distribution summary measure; Dose-volume correlations; Standard deviation of the dose.