SU-E-T-562: Do the Differences Between Photon Dose Distributions in the Breast Have Significant Impact On the Parameters of Radiobiological Models?

Abstract

Purpose: The comparison of advanced photon dose calculation algorithms in breast radiotherapy has focused on the physical dose distributions (DDs). Studies on their radiobiological impact are restricted to the normal tissue complication probability (NTCP) of the ipsilateral lung and the heart. The comparisons of the tumor control probability (TCP) and of the NTCP of moderate breast fibrosis are lacking. This study quantifies the radiobiological difference between four photon DDs in the breast. As the clinical outcome should be independent of the photon dose calculation algorithm, significant differences are eliminated by optimization of the model parameters found in literature. Methods: Thirty breast cancer patients, irradiated with two wedged tangential half-beams, were included in the study. Each DD, calculated with the Pencil Beam Convolution algorithm with modified Batho heterogeneity correction (PBC-MB) was recalculated using the Analytical Anisotropic Algorithm (AAA), AcurosXB (dose to medium) and AcurosXB (dose to water) with fixed monitors units (Eclipse, Varian Medical Systems, Inc.). The TCP and the NTCP of each DD were calculated with the Poisson model and the relative seriality model, respectively. The endpoint for the NTCP was moderate breast fibrosis. The differences were checked for significance with the paired t-test. Significant difference was eliminated by minimizing the difference between both the TCPs and the NTCPs. Results: AAA-TCP and AAA-NTCP are significantly lower than calculated for PBC-MB: 1.1% and 3.1%, respectively. AAA-TCP and AAA-NTCP are significantly higher than AcurosXB (dose to medium): 0.5% and 1.5%, respectively, and than AcurosXB (dose to water): 1.4% and 2.9%, respectively. New TCP and NTCP model parameters were determined for the radiobiological analysis of the PBC-MB and the AAA DD, respectively. Conclusion: This study shows that the differences between the photon DDs of advanced dose calculation algorithms have radiobiological significance which can be eliminated by determination of algorithm specific model parameters.