Since cancer may be regarded as a disease of differentiation and sodium butyrate induces differentiation of malignant cells in vitro, a study of the clinical pharmacology of sodium butyrate was undertaken. Nine patients with acute myeloid (n = 1), acute monocytic (n = 1), acute myelomonocytic (n = 6) and acute undifferentiated (n = 1) leukemia were treated. Their median age was 52 (range, 27-78) years. Six of the nine patients were pretreated with cytostatic agents. Sodium butyrate was administered i.v. at a dosage of 500 mg/kg/day as continuous infusion over 10 days. A sensitive and reproducible high-performance liquid chromatographic separation was developed after derivatization of sodium butyrate with 2,4'-dibromoacetophenone employing crown ether catalysis. Plasma concentrations and urinary excretion of sodium butyrate were monitored during the 10 days of continuous infusion and for 2 days thereafter. During infusion, plasma concentrations increased 6-fold over the endogenous butyrate level and reached 39-59 microM. The area under the curve of the exogenous butyrate was 384 +/- 50 microM X day (mean +/- S.D.). After the end of infusion, concentrations declined rapidly with a half-life of 6.1 +/- 1.4 min, and reached pretreatment values within 1 hr. The total clearance rate was 83 +/- 12 ml/kg/min and the volume of distribution 738 +/- 245 ml/kg. The excreted amounts of butyrate in the urine were minimal as compared to the infused dose. Although excretion by other organs was not ruled out, it is suggested that the infused sodium butyrate was rapidly metabolized. A significant increase in peripheral blast cells was observed, whereas bone marrow cytologies before and after treatment did not reveal a significant change in blasts. Differential counts of peripheral white blood cells did not show significant changes. No toxicity was encountered. The apparent lack of clinical efficacy may be explained by the low plasma levels of sodium butyrate due to its short half-life in vivo. In comparison, concentrations reported for in vitro studies were at least 10 times higher.