EVIDENCE LIBRARY
Rifapentine Modeling Used Midazolam Interaction Data as a Check Before Predicting Another Drug Pair
A pharmacokinetic model used published midazolam interaction observations to assess a CYP3A-induction component. Its main predictions concerned rilpivirine, not a new midazolam clinical trial.
A pharmacokinetic model used published midazolam interaction observations to assess a CYP3A-induction component. Its main predictions concerned rilpivirine, not a new midazolam clinical trial.
The model’s evidence chain
The investigators built a physiologically based model for rifapentine and rilpivirine and compared components against reported clinical datasets. Those checks included daily rifapentine with midazolam and weekly rifapentine with another antiretroviral.
The final simulations predicted substantial reductions in rilpivirine exposure during rifapentine treatment and persistence of induction after interruption.
Calibration does not transfer every result
A probe-drug dataset can test part of a model without making every predicted percentage applicable to the probe. The rilpivirine exposure reductions should not be restated as measured midazolam reductions from this study.
Predictions also depend on enzyme, transport, and physiologic assumptions and the available input data. The paper supports explaining how interaction models combine evidence. It cannot establish an individual benzodiazepine concentration, impairment level, or universal time to recovery of drug clearance after stopping an inducer.
Source paper
Original analysis of the supplied journal article, identified as PMC9797969.