Abschätzung der Radiosensibilität des NSCLC für die Photonenbestrahlung und darauf basierende Berechnung der Relativen Biologischen Wirksamkeit für die Bestrahlung mit Kohlenstoffionen

Objectives: This study estimates the radiobiological parameters alpha, beta, the alpha/beta value and the threshold dose dt for photon radiotherapy of non small cell lung cancer (NSCLC). Based on these data the relative biological effectiveness (RBE) for carbon ion radiation was calculated. Methods: A mathematical tumor control probability (TCP) model was established, that allows for the estimation of radiobiological parameters based on the local tumor control after radiotherapy with photons. This model includes fractionation schemes and clinical follow-up data on local tumor control. The clinical data was obtained by a selective literature review. The estimation of radiobiological parameters based on the clinical data was performed by regression analysis. The analysis was performed time-dependent and time-independent, considering overall treatment duration. Normo-, hyper-, and hypo-fractionation were analyzed separately. For the calculation of RBE for carbon ions the local-effect-model (LEM) was used, based on the radiobiological parameters alpha, beta, the alpha/beta-ratio and the threshold dose dt for the photon radiotherapy of NSCLC. Results: 16 studies were available for analysis. The clinical data showed a clear dose-response relation, justifying the application of a mathematical TCP model. Evaluation of hyper- and normofractionated radiation revealed alpha/beta values of 18,4 +/- 11; 5 Gy without consideration of treatment duration and 23,6 +/- 21,7 Gy with inclusion of treatment duration. After hypofractionated irradiation a linear regression of the dose-response curve for local tumor control could be observed. By comparison of normo- and hyperfractionated data to hypofractionated data, a threshold dose dt of 9,3 +/- 2,5 Gy and 7,3 +/- 3,1 Gy (without and with inclusion of treatment time) was calculated. Based on these pooled clinical parameters a calculation of RBE was performed. By varying input parameters of the LEM their influence on the obtained RBE was characterized. Discussion: The presented method allows a clinical based alpha/beta value calculation for the estimation of radiosensitivity in NSCLC treated with normo- or hyperfractionated photon radiotherapy. The obtained radiobiological parameters confirm propagated alpha/beta values around 10 Gy for photon irradiation of NSCLC. The demonstration of a linear dose-response relation for hypofractionated radiotherapy endorsed the limitations of the TCP model and could be quantifyed by the calculation of the threshold dose dt. Error estimation identified the low number of clinical studies and radiobiological assumptions of tumor cell numbers and growth kinetics as error-prone influencing factors. The determined RBE is consistent with radiobiological findings and could be translated for clinical application by the LEM. The by this means generated RBE showed a specific dependence on assumed modeling-parameters.