Existing investigations about the precision of radioluminography (RLG) are restricted to descriptive analysis of the tissue samples. The aim of the present experiments was to obtain a general prospective statement about the precision that the RLG method can achieve. Several pharmaceutical companies in Europe participated in the experiments. Albino rats of various strains were dosed with various (14)C-labeled compounds. Whole-body sections were produced, and blood calibration scales were set up with standard radioactivity sources of dog or rat blood. Photostimulated luminescence was detected using Fuji imaging plate BAS-III. For each organ separately, variability was investigated on each of the levels: rat, section of rat, region within section, and residual, with the help of variance components. The producing company was seen as a fixed factor and adjusted for. A mixed linear model was fitted to the log-transformed data. The variance component (SD estimate) for the residual term gave the desired prospective statement about the achievable precision of the RLG method. Exponential back transformation from the logarithmic to the natural scale transformed the SD estimates to multiplication factors. In total, 29 organs were investigated. The RLG method was comparable in precision to the dissection/combustion method.
Copyright 2000 Academic Press.