PK/PD modelling of high-dose diltiazem : absorption-rate dependency of the hysteresis loop

To investigate bioequivalence of 2 different sustained-release diltiazem formulations the preparations each containing 180 mg diltiazem-HCl were given to 20 healthy male volunteers in an open, randomized, 2-way crossover design. Blood samples were taken before drug administration and at 14 times until 30 hours post application. 12-lead ECGs were recorded at the same time points, and atrioventricular conduction time was monitored as a safety parameter. Plasma samples of 8 subjects were assayed by HPLC. Peak values of plasma concentrations and prolongation of the PQ interval were taken from the plasma concentration or ECG data directly, AUCs of pharmacokinetic and pharmacodynamic effects were calculated by the linear trapezoidal rule, MRTs were calculated as the first moment over AUC. Bioequivalence was tested according to Schuirman (ratios of parameters and shortest 90% confidence intervals) using pharmacokinetic and pharmacodynamic data sets. Relative bioavailability of the test preparation with respect to AUC 0-30 was 110% with the 90% confidence interval ranging from 100 to 130%. Bioavailability with respect to C max was significantly higher (190%) with a 90% confidence interval not even including 100%. Consequently, MRT was significantly lower with the test preparation (80%), again with a confidence interval not including unity. Relative bioavailability of the test product in terms of pharmacodynamic parameters was 160% in the extent of the effect (AUEC 0-10 ), 190% even with the rate of the effect (E max ) and 80% with the mean residence time (MRT E ). All parameters differ significantly between the products. Bioinequivalence was therefore concluded from these results. The functional relationship between pharmacokinetic and pharmacodynamic parameters could be described by hysteresis loops, however, with a clockwise rotation. This cannot be explained in the classical way by the time-lag between central and effect compartments. Two alternative conceivable explanations, namely formation of antagonistic metabolites or downregulation, were tested for plausibility. A comparison of the present results to literature data favors the model of downregulation/tolerance development. This model is additionally supported by the finding that the shape of the hysteresis is dependent on the absorption rate of diltiazem, calculated as mean input time according to MIT = MRT - 1/λ z . It is concluded that acute tolerance develops at least with the electrophysiological action of diltiazem after oral application and that the extent of tolerance development increases when decreasing its absorption rate. Bioequivalence assessment of diltiazem is possible using pharmacodynamic parameters, however, since PK/PD relationships are influenced by the absorption rate, extent parameters may be misinterpreted when rate parameters of the test formulations are different.