Prediction of area under the concentration-time curve for lopinavir from peak or trough lopinavir concentrations in patients receiving lopinavir-ritonavir therapy.

Abstract

A "single time point" strategy for predicting the area under the concentration-time curve (AUC) for lopinavir in patients receiving ritonavir-boosted lopinavir therapy was investigated. Linear regression equations describing the relationships of lopinavir peak and trough concentrations to lopinavir AUC values were established using pharmacokinetic data from published studies of patients or healthy subjects receiving lopinavir and ritonavir at standard dosages. The resulting "trough-AUC model" and "peak-AUC model" were used to predict lopinavir AUC values in the evaluated study populations (total n = 479); those values were then compared with reported AUC values. Lopinavir peak or trough concentrations were strongly correlated with lopinavir AUC values (r = 0.9947 and r = 0.9541, respectively). For about 94% of calculations using the peak-AUC model and 87% of calculations using the trough-AUC model, differences between predicted and observed AUC values were in the range of 0.76-1.5 fold; the associated r values were 0.9514 (p < 0.001) and 0.9345 (p < 0.001), respectively. The mean absolute predictive error was less than 6% with the use of either the peak-AUC model or the trough-AUC model, with corresponding values for root-mean-square error of 17.6% and 23.5%, respectively. Equations incorporating lopinavir peak and trough concentrations were found to satisfactorily predict lopinavir AUC values in data sets describing patients receiving lopinavir with ritonavir boosting. Variability in predictions was higher with use of the trough-AUC model.