A method to estimate composite doses for organs at risk in prostate cancer patients treated with EBRT in combination with HDR BT

<div><p></p><p><b>Background.</b> When evaluating late toxicity after combined external beam radiation therapy (EBRT) and high-dose rate brachytherapy (HDR BT) prostate cancer treatments, it is important that the composite dose distribution is taken into account. This can be challenging if organ-at-risk (OAR) dose data are incomplete, i.e. due to a limited ultrasound imaging field-of-view in the HDR BT procedure. This work proposes a method that provides estimates of composite OAR doses for such situations.</p><p><b>Material and methods.</b> Original EBRT, simulated HDR BT, and composite dose-volume histograms (DVHs) for 10 pelvic OARs in 30 prostate cancer cases were used for method implementation and evaluation (EBRT: 25 × 2.0 Gy + BT: 2 × 10.0 Gy). The proposed method used information from the EBRT DVH to estimate OAR BT doses (with or without fractionation correction). Coefficients of determination (R<sup>2</sup>) were calculated for linear relationships between several EBRT DVH parameters and a BT DVH parameter of interest. The largest R<sup>2</sup> value decided the relationship that best predicted the BT DVH parameter. The composite dose value was then calculated by adding the EBRT DVH and the estimated BT DVH parameter values and was compared to the reference composite value (in 1200 OAR/patient/parameter cases).</p><p><b>Results.</b> The linear relationships had an average R<sup>2</sup> of 0.68 (range 0.42–0.88). Only one ninth of the 1200 estimated composite DVH values differed more than 2 Gy from their reference values.</p><p><b>Conclusion.</b> Given a successful implementation, the proposed method only requires original or simulated BT plan data for a subset of patients to estimate composite doses for large study populations in a time-efficient manner. This can assist in evaluating radiation-induced late toxicity in multimodality treatments with limited OAR dose data.</p></div>