RELATIONSHIP BETWEEN HYPOGLYCEMIC RESPONSE AND PLASMA-CONCENTRATIONS OF BTS-67582 IN HEALTHY-VOLUNTEERS

1 The relationships between blood glucose, plasma insulin and plasma BTS 67 582 concentrations were studied in a randomised, placebo-controlled, four-way crossover study involving 16 healthy male volunteers aged between 19 and 43 years. Single oral doses of 125, 250 and 500 mg BTS 67 582 were studied. 2 Fasting blood samples were taken pre-dose and half-hourly for 8 h post-dose. Mean peak plasma concentrations of BTS 67 582 were 518, 1076 and 2435 ng ml-l for doses of 125, 250 and 500 mg, respectively. Mean maximum reductions in blood glucose were 1.13, 1.59 and 1.78 mmol l(-1), and mean maximum increases in plasma insulin were 26, 14 and 21 mu u ml(-1) for the three doses, respectively Changes in incremental area under the curve (AUC) of blood glucose were correlated with changes in plasma BTS 67 582 AUC. The maximum reduction in blood glucose was correlated with the peak plasma BTS 67 582 concentration. No correlations between plasma insulin and plasma BTS 67 582 concentrations were observed. 3 Anticlockwise hysteresis was evident in concentration-effect curves, but less evident following subtraction of placebo data, and was mainly due to an underlying downward trend in fasted blood glucose levels with time evident under placebo treatment. This suggests that hypoglycaemic effects were related to systemic BTS 67 582 concentrations, suggesting that active metabolites of the drug do not make a major contribution to acute hypoglycaemic effects. 4 A log-linear model described the relationship between blood glucose and plasma BTS 67 582 concentrations for 14 of the 16 volunteers. Log-linear and E(max) models described this relationship when the data were pooled. 5 A single compartment pharmacokinetic model with no lag time was fitted to the plasma concentration data of BTS 67 582. The pharmacokinetics of the drug were generally linear over the three doses studied. 6 Concentration-effect and pharmacokinetic models were combined to produce a predictive model of the reduction in blood glucose with time following doses of BTS 67 582. A potential application of this model is for the prediction of reductions in glycosylated haemoglobin under longterm administration of BTS 67 582, using reported correlations with blood glucose.