A pharmacokinetic model can fit much of a concentration–time curve and still depart from measurements during a particular interval. A study modeling the interaction between itraconazole and midazolam examined that kind of discrepancy.
The researchers used baseline data from 19 genotyped adult men to build linked models for midazolam and two metabolites. They then simulated concentrations under inhibition by itraconazole and its metabolite. The predicted curves generally described the observed interaction data, but concentrations tended to be too high about six to twelve hours after midazolam.
The authors considered an assumption about inhibition of both CYP3A4 and CYP3A5 as a possible explanation. A useful model review compares predictions with measurements across the whole curve and within relevant subgroups. Agreement at one time or for total exposure can leave an error elsewhere unresolved.
Vossen et al., Theoretical Biology and Medical Modelling · Full text