A simulation study to evaluate limited sampling strategies to estimate area under the curve of drug concentration versus time following repetitive oral dosing: limited sampling model versus naive trapezoidal method.

The area under the curve (AUC) can be associated with the therapeutic or toxic effect of a drug. The limited sampling model (LSM) is an approach that is gaining popularity due to its simplicity for the estimation of AUC using 1-3 blood samples. The aim of the present simulation study was to compare the performance of LSM for various hypothetical drugs with that of the naive trapezoidal method (Trap). The 3-point (trough, peak, and downhill) sampling design following repetitive oral dosing was assumed for LSM (LSM3) and Trap (Trap3). The 2-point (trough and peak) sampling design was also assumed for LSM (LSM2) and Trap (Trap2). In addition, trough-sampling and peak-sampling designs for LSM were designated as LSM1 and LSM1', respectively. As a result, the rank order of precision of the AUC estimation designs/methods was summarized as follows: LSM3 asymptotically equal to Trap3> or =LSM2> or =Trap2 asymptotically equal to LSM1>LSM1'. The finding suggested that LSM can not always improve the estimation performance of AUC in the 3-point sampling design, and that LSM1' is insufficient to estimate the performance of AUC for the hypothetical drugs evaluated in the present study. Accordingly, LSM2 and LSM1 may be an efficient approach for estimating AUC following repetitive oral dosing. In addition, Trap3 and Trap2 may be promising alternatives, because Trap does not require a high investment to recruit a full-sampling model-development group.