Organ-specific equivalent uniform dose constraints and radiobiological parameters for radiation treatment planning of abdominal tumors.

BACKGROUND: Radiation therapy planning (RTP) for abdominal tumors often requires multiple dose-volume constraints for each organ at risk (OAR), which vary with different fractionation schemes. This variability can complicate biologically optimized treatment planning. PURPOSE: This study aims to extract fractionation-independent dose constraints based on the concept of equivalent uniform dose (EUD), utilizing published clinical data for a range of fractionation regimens. The goal is to derive organ-specific radiobiological parameters as well as the EUD for each OAR that can be used in biologically optimized treatment planning for abdominal tumors, independent of the specific fractionation scheme applied. METHODS: Clinical dose-volume constraints for duodenum, stomach, and small bowel were compiled from available literature sources. These dose constraints were obtained for conventionally fractionated radiotherapy (CFRT), hypofractionated RT, and stereotactic body radiotherapy (SBRT) and were associated with less than a 10% risk of grade 3 toxicity as categorized by CTCAE v.3 or v.4, RTOG, and EORTC. For each OAR, an iso-EUD fitting with EUD calculated based on the linear-quadratic (LQ) model or linear-quadratic-linear (LQ-L) model and a dose volume histogram generated from these dose volume constraints for each fractionation was applied to extract model parameters such as α/β ratio and 'n' (the volume effect of the OAR). Based on the obtained parameters, the dose constraint in EUD and the equivalent physical dose in 2Gy fraction (EQD2) were calculated. RESULTS: The EUD constraints for LQ-L (LQ) models are 52.62 (55.10) Gy, 48.22 (48.60) Gy, and 46.10 (45.06) Gy, α/β values are 5.42 (5.59) Gy, 7.67 (6.76) Gy, and 12.15 (8.20) Gy, and 'n' values are 0.06 (0.01), 0.03 (0.02), and 0.06 (0.08) for duodenum, small bowel, and stomach, respectively. Additional two parameters for the three OARs in LQ-L are 5.15 Gy, 8.65 Gy, and 5.20 Gy for dt and 5.40, 4.72, and 3.33 for γ/α. The LQ-L model fits the clinical data better than the LQ model. CONCLUSIONS: The obtained α/β values are comparable with those published previously. The extracted EUD constraints together with the volume effect parameter 'n' can be used for plan optimization and evaluation.