Value of functional in-vivo endpoints in preclinical radiation research

Background: Cancer research faces the problem of high rates of clinical failure of new treatment approaches after positive preclinical data. We hypothesize that a major confounding factor to this problem in radiooncology is the choice of the preclinical endpoint. Methods: We present a comprehensive re-evaluation of large-scale preclinical in-vivo data on fractionated irradiation alone or simultaneously with Epidermal Growth Factor Receptor inhibition. Taking the permanent local tumour control assay as a gold standard, we evaluated different tumour volume dependent endpoints that are widely used for preclinical experiments. Results: The analysis showed the highest correlations between volume related and local tumour control endpoints after irradiation alone. For combined treatments, wide inter-tumoural variations were observed with reduced correlation between the endpoints. Evaluation of growth delay per Gray (GD/Gy) based on two or more dose levels showed closest correlation with local tumour control dose 50% (TCD50). Conclusions: GD/Gy with at least two dose groups correlates with TCD50, but cannot replace the latter as the goldstandard.