Noncompartmental Pharmacokinetics

ALAN TALEVI

The ADME Encyclopedia. A Comprehensive Guide on Biopharmacy and Pharmacokinetics

Springer Nature

Lugar: Basignstoke; Año: 2022; p. 670 - 677

Compartmental treatment of experimental pharmacokinetics (PK) data is the most traditional approach in pharmacokinetic studies (see entries on One-compartment pharmacokinetic model and Two-compartment pharmacokinetic model). However, in the last decades analysis of PK data has been increasingly performed using a non-compartmental approach, after early contributions in the 1970s like the ones from Oppenheimer et al. and Yamaoka and Uno [1, 2]. Non-compartmental PK analysis is often also referred as model-independent PK, in the sense that it does not rely on assumptions about body compartments (i.e., there is no underlying structure model describing the organism). In the context of classic compartmental PK, there is often ambiguity concerning which model (among a series of options of different complexity) should be assigned to one or more sets of data [3, 4]. Furthermore, the reductionist perspective of classic compartmental PK models, where the organisms is unrealistically depicted as a few kinetically homogenous compartments, results in the fact that kinetic micro-constants and compartments themselves are mere abstract, mathematical artifacts with no immediate anatomo-physiological interpretation (this limitation is also addressed and overcome by physiologically-based PK modelling, a particular case of compartmental PK where the compartments are linked to anatomo-physiological entities, e.g., organs, and the exchange of matter between these compartments is also expressed in terms of physiological parameters, e.g., blood flow. See the chapter on Physiologically based pharmacokinetic modeling - Definition and History, for more details).Noncompartmental methods are straightforward and more readily automated, minimizing user?s decision-making [4] (and thus less prone to data manipulation!). It is more reliably applied when frequent sampling is performed from the subject(s) during the PK study, although, as discussed later, there are options to implement noncompartmental analysis when dealing with sparse data, provided that certain conditions are met. Noncompartmental PK are typically favored for characterizing PK within a single study, e.g., they are used in interim analyses to make dose escalation decisions, and also to obtain summary parameters used in bioequivalence studies (such as total area under the plasma drug concentration versus time profile, AUC0-∞ or the peak plasma concentration Cmax or the time to the peak concentration). Compartmental models are not universally accepted for these purposes by the main regulatory agencies, though the US Food and Drug Administration (FDA) is increasingly evaluating model-based bioequivalence as a way to solve particular situations (e.g., sparse design pharmacokinetic studies). Additionally, model-independent approaches are commonly used for establishing the initial exposure characteristics of a drug in nonclinical PK and toxicology studies. Many of the relationships used for noncompartmental PK were originally derived in the context of compartmental analysis for supplemental use in such models.In the framework of non-compartmental PK analysis, the theory of statistical models is used describe the characteristics of the time courses of plasma levels (area under the curve, mean residence time (MRT), variance of residence time) and of the urinary elimination rate [2]. The standard noncompartmental analysis is applied to analyze drugs with linear PK behavior, though it is also frequently used to distinguish if a drug´s PK are nonlinear when several dose strengths are administered; applicability of advanced noncompartmental analysis methods to nonlinear PK has also been investigated [4-8], but these techniques are more seldomly applied. Linear drug disposition implies that the systemic drug levels follow the superposition principle with respect to systemic drug input [7]. In a linear system, if Q1(t) is the response to the input P1(t) and Q2(t) is the response to the input P2(t), the system is considered linear if the response to the input P1(t) + P2(t) corresponds to the sum of Q1(t) + Q2(t) [9]. For instance, let Cp,1(t) be the plasma concentration (response) of a drug after the administration of an intravenous dose D1, and let Cp,2(t) be the plasma concentration of a drug following the administration of an intravenous dose D2. For a linear system, the administration of a dose D1 + D2 will result in a plasma concentration Cp,1(t) + Cp,2(t). While living systems are not actually linear, the use of a linear model is subject to the condition that under the concentrations found the processes linked to drug transference and elimination in the organism follow apparent first-order kinetics.