Population pharmacokinetics, also referred to as population PK or popPK, seeks to understand the variability in drug concentrations among individuals in a group of interest (the “population”) receiving clinically relevant doses of a drug. Because patient characteristics (“covariates”), such as age, disease state, demographics, sex, concomitant medications, or presence of renal or hepatic impairment, can affect drug pharmacokinetics (PK) and pharmacodynamics (PD), understanding this variability can help establish a robust PK/PD profile and inform safe and effective dosing regimens.
Population PK analysis involves collecting sparse PK samples (i.e., drug concentration data in a relevant matrix, which is most often plasma) from many patients and often across multiple clinical studies, and then building mathematical models to describe those data. Using sparse PK sampling, only a limited number of samples are taken from any given patient. With appropriate sampling design and model selection, the resulting PK data can be pooled and analyzed to support conclusions about PK variability and the influence of covariates. The FDA guidance on Population Pharmacokinetics provides a framework for population PK modeling.
Advantages of Population PK Approaches
Population PK modeling approaches often provide advantages over standard noncompartmental PK analysis, including:
- Analysis of PK information from a patient pool that is reflective of the intended target population (i.e., not just healthy subjects or narrowly selected patients), including an assessment of inter-individual variability, which is often intentionally minimized in traditional clinical trial designs (and therefore missed by noncompartmental PK analysis)
- Identification, measurement, and explanation of variability, and the covariates that influence variability, in PK and/or PD
- Quantitative estimation of unexplained variability in patients
- Correlation of blood concentrations with PD response
- Use of sparse PK sampling. Sparse PK sampling is especially desirable in certain indications and populations.
- Potential to avoid full, standalone hepatic impairment, renal impairment, and/or thorough QT (TQT) studies by leveraging pooled, cross-study data.
When to Execute a Population Pharmacokinetic Analysis
Population PK analysis can be completed at different stages of development to fit your program needs:
- During early and mid-stage clinical development (Phases 1 and 2), population PK modeling and simulations are primarily used to evaluate PK variability, the impact of covariates, and the relationship of safety to exposure, with this information used to inform study design decisions (e.g., dose selection) in later stage studies. These results can be used to design more efficient and informative clinical trials and make Go/No-Go decisions about the drug development program.
- During later stages of development (Phases 2 and 3), population PK analysis is commonly used to provide additional information about PK and PD relationships, as well as to help characterize effects in special populations, such as geriatric patients or those with renal or hepatic impairment. These results are used to support marketing applications (i.e., New Drug Applications [NDAs] or Biologics License Applications [BLAs]) and product labeling claims.
- Following drug approval, postmarketing (Phase 4) studies often utilize sparse PK sampling approaches and are thus best suited for population PK modeling.
Population PK Services
Nuventra has an expert team of pharmacometricians led by our Senior Vice President of Modeling and Simulation, Mark Sale, MD. We take complex population PK analyses and distill them into the critical insights that drive common sense drug development. Our population PK services include:
- Expert advice on study design, including protocol and analysis plan development
- NONMEM datasets
- Population PK model generation, development, and refinement
- Identification & confirmation of predictive covariates
- Modeling & simulation to support Population PK/PD
- Parallel NONMEM to decrease run time for complex models
- Automated modeling selection using DARWIN
- Dose selection and justification
- Exposure-response analysis (please see our blog part 1, part 2)
- Concentration QT (cQT) modeling
- Submission-ready Population PK report writing