Why do we care about Drug Exposure?

Quantifying drug exposure is the foundational first step in Pharmacometric Analysis. Typically we are examining the concentration of the drug in the easiest tissue to which we have access: the blood. Interestingly, drug exposure is of very little interest on its own—it is only in relation to a biological or clinical effect that drug exposure finds meaning. Drug exposure can be considered in a hierarchy of complexity, depending on how rapidly and persistently the biology changes following a change in exposure. For example, a rapidly changing clinical effect, such as blood glucose level in a patient with Type 2 diabetes, suggests an acute relationship between circulating drug concentration and its impact. Alternatively, a more slowly moving effect, e.g., change in bone mineral density in osteoporosis, suggests that long-term exposure to the drug might be more relevant.

The answers to each of the following questions are helpful measures of drug exposure in different contexts:

1. Interrogation questions:
   1. Did the patient get the drug?
   2. What dose did they receive?
   3. How often did they receive it?
2. Questions requiring measurement but minimal analysis beyond the characterization of concentration:
   1. Concentration at the current sampling time?
   2. Maximum observed concentration (C_max)?
   3. Minimum observed concentration (C_min)?
   4. Time of maximum or minimum concentration (T_max or T_min)?
   5. An aggregate of the concentration measurements over a dosing interval (AUC)?
3. Questions requiring mathematical prediction between or beyond observations:
   1. Concentration at an unobserved timepoint?
   2. Exposure at an unstudied dose or schedule?
   3. Exposure in different groups of patients?
   4. Exposure under different routes of administration?
   5. Effects of formulation changes?

Types of Pharmacokinetic Analysis

As implied, questions in Category A above can be answered by simply asking the patient or clinician about their experience. A current and urgent example of this type of characterization is exposure to an effective COVID-19 vaccine as a predictor of susceptibility to future infection with SARS-CoV2.

The questions under Category B are what non-practitioners sometimes consider to encompass PK analysis: these are addressed through Non-Compartmental PK Analysis (NCA). NCA is a simple summary of concentration measurements over time after a single administration event (either as the first and only dose or after a series of doses). This involves a relatively simple summary of concentration measurements, including C_max, C_min, T_max, T_min and AUC, largely free of assumptions about the biology. Some limited, but still informative, conclusions about rates of elimination, variability between study subjects, and dose proportionality can be drawn from NCA. However, NCA is best viewed as a post-hoc characterization of observed data and is not particularly useful to predict exposure under alternative doses, dosing schedules, or target populations.

Compartmental PK analysis to address Category C questions involves a mathematical description of the change in drug concentration over time in a PK model. In the population setting, this includes quantifying how drug exposure changes in a “typical” study subject (whether that be a patient or healthy volunteer). These models also characterize variability between patients, evaluating the effect of factors such as patient sex, body size, ethnicity, and health of the organs of elimination (typically liver and kidney). These are all examples of “intrinsic” factors, i.e., those out of the patient’s or clinician’s immediate control. “Extrinsic” factors that may also be relevant include effects of the dose and or formulation, administration schedules, dosing under fed or fasted conditions, etc. Finally, random effects quantify the remaining, unexplained differences between individual patients or subjects and variation caused by differences within a particular subject. Nominally, the parameters produced in a compartmental analysis can have biological relevance, and importantly, they can be used to predict exposure under conditions which were not used to build the model (obviously, care must be exercised in evaluating the reliability of predictions that stray widely from observed experimental conditions). These predictions are most valuable when coupled to models of biological and clinical effects, i.e., in PK/PD and or exposure-response models.

Cognigen and PK Analysis

As one of the most experienced providers of Pharmacometric Analysis services (we have been in business for almost 30 years!), Cognigen has deep experience in helping sponsors characterize drug exposure. We literally wrote a book on population PK, PK/PD, and exposure-response analysis.[1] We have produced hundreds of NCA, population PK, and exposure-response analysis reports to help with strategic decisions in drug development and in support of regulatory interactions with global authorities.

We would love to hear more about your development program and brainstorm about how we might be able to apply our data analytical skills to help move your research forward.

[1] Joel S. Owen, Jill Fiedler-Kelly. Introduction to Population Pharmacokinetic / Pharmacodynamic Analysis with Nonlinear Mixed Effects Models. Wiley, 2014; DOI:10.1002/9781118784860.