Optimization of dosage regimen using a mechanism-based pharmacokinetic-pharmacodynamic modeling

降糖药的药代动力学（PK）和药效学（PD）可能受到给药时间的影响，因为血浆葡萄糖水平具有显著的昼夜节律。本研究的目的是评估昼夜变化对大鼠磺胺类药物甲苯丁胺（TB）的PK-PD关系的影响。在光照期结束时（约15:00 h），血浆葡萄糖的生理水平与顶相呈昼夜节律。采用双室模型描述静脉给药后6:00和18:00 h血浆游离结核浓度，PK参数未随给药时间变化。18:00 h静脉给药后的降糖效果大于6:00 h给药后的降糖效果。采用带昼夜节律的link和E_<max>模型构建PK-PD模型。然而，回归曲线不能同时很好地拟合两种剂量下观察到的降糖效果。对于TB的胰岛素分泌，血浆水平和离体胰岛释放量表明，在18:00 h时胰岛素的峰值分泌比6:00 h时增加，因此TB的剂量可以转化为胰岛素的表观剂量。为了评估TB剂量与降糖效果之间的关系，我们用血浆胰岛素水平代替TB构建了另一个模型。模型估计的效果解释了在两个给药时间同时观察到的数据。这些结果提示，当长时间使用苏类药物时，可能会由于某些生物昼夜节律而发生无法从血浆药物浓度预测的药理学变化。

Pharmacokinetics (PK) and pharmacodynamics (PD) of hypoglycemic drugs may be influenced by dosage time, because plasma glucose level shows a remarkable circadian rhythm. The purpose of the present investigation was to evaluate the effect of circadian changes on the PK-PD relationship of tolbutamide (TB), a SU agent, in rats. Physiological level in plasma glucose showed a circadian rhythm with acrophase in the end of the light phase (ca.15:00 h). Plasma concentrations of free TB after i.v. bolus administration at 6:00 and 18:00 h were described using a two-compartment model, and no change of the PK parameters by the administration time was observed. The hypoglycemic effect after i.v. administration of TB at 18:00 h was larger than the 6:00 h dosage. The PK-PD model was constructed using link and E_<max> models with a circadian rhythm. However, the regression curves did not fit simultaneously the observed hypoglycemic effects at both dosage times well. For the insulin secretion by TB, plasma level and release from isolated pancreatic islets indicated that the spiked secretion of insulin enhanced at 18:00 h comparing with 6:00 h. The dosage of TB was therefore translated into the apparent dosage of insulin. To evaluate the relationship between the dosage of TB and the hypoglycemic effect, another model was then constructed using plasma level of insulin instead of TB. The model-estimated effect explained the observed data at both administration times simultaneously. These results suggest that a pharmacological change, which cannot be predicted from plasma drug concentration, might occur because of some biological circadian rhythms when SU agents are continued for a long time.