What is Empirical PK/PD?
Empirical PK/PD modeling is a data-driven approach that relates drug concentration (pharmacokinetics) and drug effect (pharmacodynamics) based on observed data. The selection of mathematical expressions is driven by the observed data. There are a few commonly used models, such as Emax, logistic regression, direct response and indirect response models. Empirical PK/PD can also be referred to as traditional PK/PD or classical PK/PD.
When to use Empirical PK/PD
Empirical PK/PD modeling can be applied at both preclinical and clinical stages once PK and PD data is available. Here are some examples of when they are applied:
- Characterize exposure vs. efficacy relationship in mice
- Characterize exposure vs. toxicity relationship from toxicity studies
- Characterize exposure vs. PD relationship in animal or human studies
- Help understand the translation of PK/PD relationship from preclinical to clinical
- Predict PK and PD at new dose levels or dosing regimen
Empirical PK/PD models focus on fitting the data with the simplest model needed. It often takes less time than mechanistic models and may provide a better fit to the data. However, it will not provide mechanistic understanding when comparing across different compounds, different experiments, or different patient populations, except determining if a difference exists. It is also best used when interpolating between doses instead of extrapolating to higher dose levels that have never been tested. Given the limitations, it can still be a very useful tool when quick characterization of the data is needed.