Intra‐ and inter‐laboratory reproducibility of the rat blood Pig‐a gene mutation assay

Abstract

The in vivo Pig‐a assay is being used in safety studies to evaluate the potential of chemicals to induce somatic cell gene mutations. Ongoing work is aimed at developing an Organization for Economic Cooperation and Development (OECD) test guideline to support routine use for regulatory purposes (OECD project number 4.93). Among the requirements for OECD approval are demonstrations of assay reliability, including reproducibility within and among laboratories. Experiments reported herein address the reproducibility of the rat blood Pig‐a assay using the reference mutagens chlorambucil and melphalan. These agents were evaluated for their ability to induce Pig‐a mutant erythrocytes in three separate studies conducted across two laboratories. Each of the studies utilized a common treatment schedule: 28 consecutive days of exposure via oral gavage. Whereas one laboratory studied Crl:CD(SD) rats, the other laboratory used Wistar Han rats. One or two days after cessation of treatment blood samples were collected for mutant reticulocyte and mutant erythrocyte measurements that were accomplished with the same analytical technique whereby samples were depleted of wildtype erythrocytes via immunomagnetic separation followed by flow cytometric enumeration of mutant phenotype cells (MutaFlow®). Dunnett's test results showed similar qualitative outcomes within and between laboratories, that is, each chemical and each study demonstrated statistically significant, dose‐related increases in mutant reticulocyte and erythrocyte frequencies. Benchmark dose analysis (PROAST software) provided a means to quantitatively analyze the results, and the relatively tight, overlapping benchmark dose confidence intervals observed for each of the two chemicals indicate that within and between laboratory reproducibility of the Pig‐a assay are high, adding further support for the development of an OECD test guideline.

Similarity of multiple dose-response curves in interlaboratory studies in regulatory toxicology

“…This simplification should be avoided if downturns at higher dose(s) are possible or if you are more interested in comparing the NOAEC instead of a trend itself. The simplification becomes clear looking on Table 2 for the above Ames assay example: only 7 contrasts, claiming for local equivalence for laboratories [1,2,3,4,5,6] and but not for global equivalence because laboratory 7 behaves borderline non-equivalent.…”

“…To characterize the Ames fluctuation test [19] used the lowest ineffective concentration as criterion, the fold change increase to compare standard miniaturized and Ames II and MPF Assay [4,21], where the confidence intervals for benchmark dose estimates were used for the blood Pig-a gene mutation assay [3] (requiring the same underlying nonlinear model for any condition).…”