SU‐E‐T‐712: Study of the Linear‐Quadratic‐Linear and Universal‐Survival‐ Curve Models for Lung SBRT Using BED and EQD2 Concepts

Abstract

Purpose:Linear‐quadratic‐linear(LQ‐L) and Universal‐Survival‐Curve(USC) models are investigated for the changes in biologically effective dose(BED) and equivalent dose at 2Gy/fraction(EQD2) for SBRT.Method:Survival data were extracted from published reports for 5 small‐cell(SC) and 6 non‐small cell(NSC) lung cancer cell lines. The best‐fit regression and chi‐square fit used to compute α and β of LQ, D0, and n of multitarget(MT) and y of LQ‐L models, and used to calculate BED and EQD2 for lung SBRT.Results:The value of parameters α and β was determined by best‐fit regression for low dose survival data(R2≥0.97) and by interactive‐inspection and chi‐square best‐fit to initial curvature points for LQ model, respectively. D0 and n were calculated by best‐fit regression to final slope points. Mean value of α, α/β, D0 and n with 95%CI was 0.756±0.152Gy‐1, 16.31±7.78Gy, 1.014±0.308Gy and 1.688±0.696, respectively, for SC; and 0.18±0.081Gy‐1, 7.74±5.893Gy, 2.90±1.704Gy and 3.69±3.36, for NSC, respectively. Mean value of α, α/β, D0 and n for lung tumors, averaging parameter values for all cell lines, was 0.442±0.194Gy‐1, 11.63±5.241Gy, 1.28±1.07Gy and 2.78±1.88, respectively. BED and EQD2 for LQ, LQ‐L and USC models calculated for dose levels I‐IX of RTOG0813 protocol were 67.52–121.59Gy and 57.61–104.02Gy, 78.25–132.55Gy and 66.77–113.10Gy, 59.14–94.49Gy and 50.46–80.62Gy, respectively. USC curves closely overlap to survival data. For transition dose Dt=2α/β (α/β>6Gy) LQ‐L curves have significant difference at higher doses than USC and survival data. When Dt(LQ‐L)‐Dt(USC), LQ‐L curves exceptionally close to USC and experimental data. In SBRT doses, BED and EQD2 calculated by LQ and LQ‐L(α/β>6Gy) are higher than that of USC model.Conclusion:USC curves closely match with experimental data. BED and EQD2 for dose levels of RTOG0813 protocol for LQ and LQ‐L(α/β>6Gy) have significant difference than USC values. BED matching for designing SBRT treatment should consider comprehensive calculations to include BED of USC model.