Abstract
Purpose
To examine the distortion of the linear quadratic (LQ) model of in vitro cytogenetic dose response over an extended range of γ-ray doses by analyzing the available literature data, and to establish the dose ranges, in which the LQ dose response curve (DRC) can be most accurately fitted for biological dosimetry.
Materials and methods
Data on yields of dicentrics (Dic) or dicentrics plus centric rings (Dic + CR) induced in vitro in human lymphocytes by acute γ-rays were extracted from 108 open sources. The overall dose response dataset in the dose range up to 50 Gy was fitted to a fractional-rational (FR) model, which included a ‘basic’ LQ function in the numerator, and a reduction factor dependent on the square of the dose in the denominator. Cytogenetic dose response data obtained at Grigoriev Institute for Medical Radiology, Kharkiv, Ukraine (GIMRO) in the range 0.1 − 20.3 Gy acute γ-rays were fitted to the LQ model with the progressive changing minimum or maximum radiation dose.
Results
The overall dose response, as expected, followed the LQ function in the dose range ≤5 Gy, but in the extended dose range appeared to be S-shaped, with intensive saturation and a plateau at doses ≥22 Gy. Coefficients of the ‘basic’ LQ equation in FR model were very close to many published DRCs; calculated asymptote was 17. Fitting of the GIMRO dataset to the LQ model with the shift of the dose range showed the increase in linear coefficient with the increment of either minimum or maximum radiation dose, while the decline of the quadratic coefficient was regulated mostly by the increase of the highest dose. The best goodness of fit, assessed by lower χ2 values, occurred for dose ranges 0.1 − 1.0 Gy; 0.5 − 5.9 Gy; 1.0 − 7.8 Gy; 2.0 − 9.6 Gy, 3.9 − 16.4 Gy and 5.9 − 20.3 Gy. The ‘see-saw’ effect in changes of LQ coefficients was confirmed by re-fitting datasets published by other laboratories.
Conclusions
The classical LQ model with fixed coefficients appears to have limited applicability for cytogenetic dosimetry at radiation doses >5 Gy due to the saturation of the dose response. Different response of the LQ coefficients to the changes of the dose range must be considered during the DRC construction. Proper selection of minimum and maximum dose in calibration experiments makes it possible to improve the goodness of fit of the LQ DRC.
Acknowledgments
This report makes used of data obtained, among other open sources, from the Human Radiation Cytogenetics Archives of the Radiation Biology Center, Kyoto University, and from the Dicentric Chromosome Assay Data Collection, version 1, of the Biodosimetry Group of the Department of Radiation Measurement and Dose Assessment of the National Institutes for Quantum Science and Technology, Japan. The author would like to thank Dr Elizabeth Ainsbury of the UK Health Security Agency, Harwell-Oxford, United Kingdom, for providing the DoseEstimate_v.5.1 software package for statistical treatment of data; Dr. Manuel Higueras of the Scientific Computation & Technological Innovation Center, Universidad de La Rioja, Logroño, Spain, and Dr. Eugenia Radzishevska of the S.P. Grigoriev Institute of Medical Radiology and Oncology of the National Academy of Science of Ukraine, Kharkiv, Ukraine, for their consultations and help in the technical aspects of statistical computations for the present study.
Disclosure statement
No potential conflict of interest was reported by the author.
Additional information
Funding
Notes on contributors
Volodymyr A. Vinnikov
Volodymyr A. Vinnikov, Ph.D., Radiobiologist, Leading Scientific Officer at the S.P. Grigoriev Institute for Medical Radiology and Oncology of the National Academy of Medical Science of Ukraine (GIMRO), Kharkiv, Ukraine; currently the Senior Researcher at the Department of Radiobiology, Cancer Research Institute, Biomedical Research Center of Slovak Academy of Science, Bratislava, Slovak Republic