Development of a predictive DNA double strand break assay for the identification of individuals with high normal tissue radiosensitivity

A genetically determined high level of intrinsic normal tissue radiosensitivity may account for the 5% of patients who experience unexpectedly severe normal tissue side effects following radiotherapy. The pre-treatment identification of these individuals by a diagnostic test or “predictive assay “ may allow appropriate modification of treatment plans and improve the therapeutic index of radiotherapy. Results from studies of cell-based assays measuring the response of a single cell type taken from patients to in vitro irradiation have been inconsistent, leading to the opinion of many that they are of no value in the prediction of normal tissue radiosensitivity. A systematic review of the literature presented here, however, suggests that poor methodology of study design often with inadequate control for those factors other than normal tissue radiosensitivity which influence radiotherapy toxicity and lack of reporting of assay precision means that it is difficult to form any conclusions, positive or negative about the diagnostic accuracy of the cell-based assays studied so far. Analysis of individual patient data extracted from these studies suggests that at least some of these assays may possess some discriminatory value. This finding justified an attempt to develop a novel cell-based assay based on the kinetics of radiation-induced .H2AX in peripheral blood lymphocytes. Assay failure rate was high and intra- and inter-sample assay reproducibility was poor for quantification by microscopy but were better for flow cytometric analysis. A study of 8 volunteers, however, demonstrated that intra-individual variation was higher than inter-individual variation in assay results, strongly suggesting that poor assay reproducibility due to technical or biological factors may limit the assay’s potential to identify radiosensitive individuals. This suspicion needs to be confirmed in a clinical study of patients of known radiosensitivity. As blood sample storage conditions affect assay results these will need to be standardized to prevent confounding of results.