A case of drug-induced hematotoxicity: from in vivo to in vitro assessment

As part of the early preclinical development of a new antipsychotic compound, Wistar rats of both sexes were dosed orally for upto 7 days. At high doses, expected changes in appearance and behavior, decreases in bodyweight gain and feed intake but also a fluid and pale bone marrow (BM) were observed. Blood cell counts were normal as were clinical chemical values. BM sections showed a red cell hypoplasia. Circulating reticulocytes and erythroblasts on BM smears were decreased suggesting that the compound might have a selective toxicity for the erythroid lineage. In a mechanistic experiment, rats were dosed for 9 days and phlebotomized after 7 days of exposure to stimulate erythroid regeneration. Red-blood cell mass, reticulocytes and erythropoietin (EPO) levels were monitored before and upto 48 h after bleeding. Results showed that an EPO-mediated pathogenesis could be excluded. The effect of the drug on the formation of Colony-forming units (CFU)-E and CFU-GM was then quantitatively measured in vitro after direct exposure to the compound. In two successive assays, rat or human BM cells were incubated with the drug dissolved in the collection medium at final concentrations of 0.3×10^–7 –3×10^–5 M. In the presence of adequate growth factors, CFU-E and CFU-GM were cultured and cell proliferation was compared between treated and control groups. Our results showed an expected inhibition by the drug of the growth of erythroid progenitors associated to a similar effect on myeloid progenitors. The CFU-E and CFU-GM of both human and rat sources were totally inhibited from the concentration of 3×10^–5 M. The IC₅₀ values were consistent with rat peak plasma levels reached in vivo by the drug. Therefore, the short-term cloning assays performed on rat BM cells were sensitive indicators of the hematotoxicity of the compound and were considered as predictive for human toxicity.