Background Information

1. Model-informed dosing
2. Drug pharmacokinetics
3. Pharmacokinetic changes in pediatrics
4. Pharmacokinetic changes in pregnancy
5. Project MADAM
6. Glossary
7. Template for creating an MID report

1. Model-informed dosing

Drug dosing recommendations for special populations, such as pediatric and pregnant patients, are often missing from the drug label. An important reason for this is the limited inclusion of these populations in clinical trials (for a variety of reasons including ethical, regulatory and logistical challenges), resulting in a lack of pharmacokinetic (PK) data and off-label drugs use (1, 2). Various age and pregnancy-related physiological and anatomical changes may affect drug absorption, distribution, metabolism and excretion (together called ‘pharmacokinetics’, PK) in children and pregnant women. The complex interplay of these changes makes adequate dosing challenging (3, 4). More information on PK changes in these populations: Pharmacokinetic changes in pediatrics and Pharmacokinetic changes in pregnancy.

Physiologically-based pharmacokinetic (PBPK) modeling can be a useful computer tool to characterize drug behavior within the context of human physiology. It can be used to predict the effects of dynamic physiological changes on the pharmacology of a drug. A PBPK model consist of several compartments representing organs and tissues of children or pregnant women (Figure 1) and can hence be seen as a computerized pediatric or pregnant body. The effects of development and maturation and pregnancy on the compartments/organs are described using (gestational) age-dependent equations. This set-up allows the creation of a set of virtual subjects (‘virtual population’), with anatomical and physiological characteristics of pediatric subjects or pregnant women at a specific developmental stage. A PBPK model can describe drug PK: how a drug enters the body (e.g., as an oral dose), the amount of drug that reaches the systemic circulation (described as the fraction of the drug that is absorbed, ‘fa’ in Figure 1), how a drug moves between the compartments/organs (described by blood flow, ‘Q’ in Figure 1) and the rate of elimination from the body (e.g., via hepatic clearance, ‘CLhepatic’ in Figure 1). With a PBPK model, a virtual clinical trial can be conducted. The virtual environment allows the user to run multiple trials with different dosing strategies. Predicted PK can form the basis for dosing decisions (5). 


Figure 1. Compartments of a population model reflecting a human body. Population-specific anatomical and physiological parameters are defined in a population model. Time-based changes in drug concentrations in compartments (tissues and blood) are described with tissue-specific blood flow (Q) and clearance values (CL). The subscripts denote the following tissues: lu = lung, ad = adipose, bo = bone, br = brain, he = heart, ki = kidney, mu = muscle, sk = skin, re = rest of body, liv = liver, spl = spleen, ha = hepatic artery, gut = gut wall. Abbreviations: CL, clearance; fa, fraction absorbed; Q, blood flow. Adapted from Freriksen and colleagues (6).

Interest in PBPK modeling as a way to predict PK in pediatric and pregnant subjects has increased dramatically over the past few years (7, 8). Of all regulatory submissions involving PBPK modeling received by the US Food and Drug Administration (FDA) in 2018 - 2019, 9% were focused on the pediatric population (9). While PBPK modeling is generally conducted to inform clinical trial design in pediatric and pregnant populations, it can also be applied after market introduction to inform off-label drug dosing (5).

Model-informed dosing: a stepwise approach 
There are a several considerations that should be taken into account when determining a model-informed dose (MID) for implementation in clinical practice. A stepwise approach is displayed below (Figure 2). After formulating the context and rationale for modeling, a PBPK model can be developed. Evaluation of PBPK model performance is a critical step prior to conducting prospective simulations to explore the suitability of alternative dosing strategies. Then, it is critical to conduct a careful evaluation of the proposed model-informed dose by looking at aspects such as model extrapolation, model influence, decision consequence and practical considerations for clinical implementation of the dosing regimen(s). These steps are described in more detail in the following sections. 


Figure 2. Stepwise approach for establishing PBPK-informed dosing recommendations and subsequent implementation in real-life clinical care.

Developing a PBPK model
A PBPK model needs to be fueled with information on the drug of interest (e.g., molecular weight and lipophilicity) and physiological data of the population of interest (e.g., weight and organ blood flows). This information is generally collected in so-called ‘drug models’ and ‘population models’ (Figure 3). Next to pediatric and pregnant population models, many other population models are yet available. Examples are models mimicking a population with liver cirrhosis or specific ethnicities (10, 11).

In user-friendly PBPK software platforms, a population model can be coupled to a drug model of interest. Then, a virtual clinical trial can be designed. A ‘trial design’ describes the study population (e.g., age range, proportion of females) and the dosing strategy (e.g., dose, dosing frequency, route of administration) of interest. A PBPK model consists of a population model, a drug model, and a trial design (Figure 3). With this model, a clinical trial can be conducted in a virtual environment. Various dosing strategies can be applied and drug plasma concentration-time profiles and PK parameters (such as the volume of distribution and clearance) can be predicted (7). These model predictions can drive dosing decisions (see section: ‘Prospective PBPK model simulations: Dose finding’).


Figure 3. Components of a PBPK model. Population-specific anatomical and physiological parameters are defined in the virtual population model and drug-specific physicochemical properties are defined in the drug model. The trial design includes, amongst others, a description of the dosing strategy and the age and gender of the virtual subjects and can be adjusted to reflect a clinical study design of interest (indicated with the dashed border). PBPK: physiologically-based pharmacokinetic.

The availability of many population and drug models in commercial modelling platform can be leveraged to expedite model-informed dosing, also by people with no or limited modeling skills. In theory, it is possible to simulate the PK of a specific drug in any defined virtual population by coupling a compound model to any population model of interest, without changing default input parameters. Existing models can be used pragmatically and hence specific expertise on how to code and parameterize such models is not needed. A tutorial on this pragmatic approach to model-informed dosing has been published by Van der Heijden and colleagues (12).

General-purpose modeling tools such as R, MATLAB and Berkeley Madonna can be used for PBPK modeling, but the Simcyp® Population-based Simulator, GastroPlus, and PK‐Sim are the most frequently used PBPK modeling platforms. The latter platforms have a graphical user-friendly interface and are relatively easy to use by people with no or limited modeling experience (13). Population and drug models can be retrieved from PBPK modeling software platforms, from dedicated repositories or from the scientific literature (which can often be obtained upon request from the corresponding authors). It should be noted that the quality of the model may vary. Drug models can also be built from scratch, by modelers themselves, based on model input parameters reported in the scientific literature (12).

When conducting a virtual clinical trial using dedicated PBPK modeling software, several default settings are usually applied. Several features require attention:
1)   Biological intersubject variability is generally incorporated in the model to reflect the heterogeneity in vivopopulation adequately. In Simcyp®, this variation is implemented by applying a certain coefficient of variation (CV%) around distinct physiological parameters. In this way, virtual subjects are established with different weights, organ volumes, etc. When conducting 10 virtual trials with 10 subjects each, one obtains a notion of the expected variability in PK in a population.
2)   The effect of age on renal function is included in the model as well as ontogeny profiles for several drug transporters and metabolizing enzymes.
3)   It should be noted that the physiology of pediatric subjects can be redefined over time, meaning that physiological and biochemical growth that occurs during the dosing period. This option is most relevant when conducting virtual trials with children below 1 year of age since significant growth and maturation takes place during this period.

PBPK model verification
Credible MID recommendations can only be established with a PBPK model of sufficiently high quality. Therefore, the credibility of the established model should be assessed and its predictive performance should be evaluated. Key points are:
1) Is the PBPK model (i.e. the drug and population model) adequately parameterized?
2) Are there any assumptions made regarding model input parameters?
3) Is the PBPK model verified using PK data from a comparable population using the same drug for a similar indication?
4) Is model performance considered adequate?

The required level of model verification depends on the context of use and the clinical impact of the model, meaning that more trust in the model is needed when the clinical impact of the established model-informed dose is bigger. That is, for instance, the case when modeling results are being used to establish a new dosing recommendation (i.e. for a drug and/or age group for which no dosing recommendation is yet available) rather than to support an existing dosing recommendation. Also, the weight of the model-informed dose in the totality of available evidence (i.e. is their other data supporting the dose?), the potential consequences of over- or underdosing and the vulnerability of the population intended to use the model-informed dose are aspects that need to be considered.

Model verification describes the process of assessing the accuracy of the model in predicting observed data. To this end, clinical PK data are collected from the literature (PK data availability depends on the drug and population). Using the model, ‘virtual clinical trials’ are conducted in which the study design of the virtual trial mimics the characteristics of the PK studies that have been conducted (e.g. the same age range, proportion of females and dosing strategy). Predicted plasma concentration-time profiles and PK parameter values are subsequently compared to what has been reported in the corresponding PK studies.

Model performance should first be evaluated for (non-pregnant) adults, before verifying model accuracy for pediatric or pregnant subjects. If modelers do not comply with this rule and model performance is first assessed for pediatric or pregnant subjects, it is impossible to state whether inadequate model performance is due to inaccurate parameterization of the basic model (i.e.. healthy non-pregnant adults; the population for which the accuracy of the model is known to be best) or whether it is caused by inaccurate input for pediatric or pregnant-specific model parameters (e.g., wrong liver volume or plasma protein concentration). If PK simulations have to be conducted for oral drug administration, the accuracy of the model in predicting PK should first be evaluated for intravenous administration. This is needed to visualize distribution and clearance processes in the model first, independent from drug absorption and first-pass metabolism. If model performance is considered adequate for predicting PK upon intravenous administration, model verification can be conducted for oral administration. This stepwise approach is depicted in Figure 4.

Figure 4. Workflow of pragmatic pediatric PBPK modeling. After establishing a PBPK model for adults and assessing PBPK model performance by means of verification, the PBPK model can be verified for its use to predict PK in pediatric subjects. Abbreviations: IV, intravenous; PK, pharmacokinetic. Figure from: Van der Heijden and colleagues (12). 

Predictive performance of PBPK models is generally evaluated by 1) a visual predictive check of the agreement of the observed versus predicted plasma concentration-time profile (i.e. does the predicted curve visually match the observed curve?) and 2) calculating the ratio of predicted-to-observed PK parameter values. There is no general consensus on criteria for acceptable model performance. The most frequently applied criterion is the two-fold acceptance range (i.e., the predicted-to-observed PK parameter ratios should be within 0.5 and 2) (14). An example of a verification plot and a figure showing PK parameter ratios are depicted in Figure 5. 

Clinical PK data in pediatrics and pregnancy are generally less abundantly available compared with adult data. Hence, it is not always possible to verify a PBPK model for the specific pediatric age group or pregnant women. Different scenarios, representing a certain level of PK data availability for pediatric PBPK modeling, are outlined by Van der Heijden and colleagues (12). In general, it can be stated that PBPK model credibility decreases with decreasing PK data availability. If there is a paucity of data for the group of interest, PK data from another population or disease state can be exploited and the model can be verified for this specific population or disease state. Modeling results are still useful if there is adequate understanding of the effect of age/pregnancy, indication or disease on PK.


Figure 5. Example of PBPK model verification results. A. Visual predictive check of the predicted and observed midazolam plasma concentration-time curve, upon intravenous administration in adults. The solid line is the predicted mean and the shaded area represents the 5th and 95th percentile interval around the predicted plasma concentration in the virtual population. Observed data is presented in circles. Insets depict log-transformed plasma concentration–time data. B. Predicted-to-observed PK parameter ratios. Clearance of midazolam in adults is depicted as an example. Data from Van der Heijden and colleagues (15).

Prospective PBPK model simulations: dose-finding
With a verified PBPK model, an infinite number of dosing scenarios can be simulated to find the most optimal dose. The drug plasma concentration has been widely used as a surrogate for the drug concentration at the site of action. Also, PK drives the effect of a drug on the body (i.e. therapeutic response and toxicity, called ‘pharmacodynamics’) and therefore dosing can be guided based on plasma exposure. Generally, the aim is to reach a certain PK target that has previously been established based on laboratory and/or clinical studies. Depending on the drug, the aim is to reach a plasma concentration within a specified therapeutic window (range of plasma concentrations that provide therapeutic response without significant adverse effects), or to have a plasma concentration that remains above a defined level for a minimum amount of time (in case of antibiotics, for example). If a PK target is lacking, an MID can be established using exposure matching. In this scenario, one aims for drug exposure in the population of interest (e.g. children or pregnant women) to be equal to the exposure that is accepted as effective and safe in a population for which more pharmacological data and/or clinical experience are available (e.g. adults) (12). With a PBPK model, virtual trials with various dosing strategies can be conducted in a relatively timely manner. By exploring predicted PK, a model-informed dose can be established (7).

It should be noted that MID recommendations are generally established based on PK simulations conducted with a set of virtual healthy subjects. It is hence expected that drug plasma levels will be optimal when the particular MID is given to healthy subjects in real-life clinical care. Disease, comorbidities or treatment of real-world patients may, however, affect PK hence requiring an adjusted dose. The potential effect of any condition on drug PK should be evaluated on a case-by-case basis by the clinical teamSchermafbeelding-2024-03-19-om-10-50-20

Figure 6. Dose-finding methods for establishing a PBPK-informed dosing recommendation. Abbreviations: AUC, area under the curve; Ctrough, trough plasma concentration; MIC, minimum inhibitory concentration; PK, pharmacokinetic; PK, pharmacodynamic.

Implementation of a model-informed dose
Before a PBPK model-informed dose can be implemented in clinical care, a thorough evaluation of the benefits as well as the potential risks is needed. The benefit of the model-informed dose is its clinical value: more effective treatment, reduced toxicity, having a dosing recommendation for a drug and/or age group for which a dosing recommendation was lacking, etc. Potential risks of implementing a model-informed dose are under- or overdosing and the associated consequence. A careful analysis of risks and benefits requires collecting and appraising other available evidence to assess the ‘model influence’ (i.e. the weight of the model-informed dose in the totality of other data). In addition, a good understanding of model assumptions and limitations together with an analysis of the probability of therapeutic failure and adverse effects is crucial. This is called the ‘decision consequence’ (16).
This process involves multiple stakeholders (e.g., modeling experts, pharmacists and clinicians prescribing the drug, parents, patients), and is a drug-, indication-, and patient-specific effort (17).

Questions that can help guide this process:
1)     To what extent does the virtual population match the real-life patients that will receive the MID in clinical care? Are there differences in PK or pharmacodynamics expected which may require adjustment of the model-informed dose (which was established based on healthy physiology)?
2)     Does the model-informed dose recommendation deviate from the current dosing strategy?
3)     What is the level of certainty in the PK target used to establish the model-informed dose?
4)     What are the consequences and probability of over- or underdosing and are the associated risks deemed acceptable? 
5)     Is it possible to monitor plasma levels in real-life patients (e.g., with therapeutic drug monitoring or a clinical measurable effect) and to intervene by changing the dose?
6)     Is it practical and feasible to administer the model-informed dose, taking into account the drug formulation, difficulty of dose calculation, and excipient’s safety?

Writing a report on model-informed dosing
The model-informed dosing (MID) drug reports, that are included in the MELINDA database and available via the website, are created using the following template: Template for creating an MID report. This file can be used as a guidance on how to write an MID drug report and facilitates inclusion of emerging modeling data in the MELINDA database.

1.       Schrier L, Hadjipanayis A, Stiris T, Ross-Russell RI, Valiulis A, Turner MA, et al. Off-label use of medicines in neonates, infants, children, and adolescents: a joint policy statement by the European Academy of Paediatrics and the European society for Developmental Perinatal and Pediatric Pharmacology. Eur J Pediatr. 2020;179(5):839-47.
2.       Wisner KL, Stika CS, Watson K. Pregnant women are still therapeutic orphans. World Psychiatry. 2020;19(3):329-30.
3.       Kearns GL, Abdel-Rahman SM, Alander SW, Blowey DL, Leeder JS, Kauffman RE. Developmental pharmacology--drug disposition, action, and therapy in infants and children. N Engl J Med. 2003;349(12):1157-67.
4.       Koren G, Pariente G. Pregnancy- Associated Changes in Pharmacokinetics and their Clinical Implications. Pharm Res. 2018;35(3):61.
5.       Freriksen JJM, van der Heijden JEM, de Hoop-Sommen MA, Greupink R, de Wildt SN. Physiologically Based Pharmacokinetic (PBPK) Model-Informed Dosing Guidelines for Pediatric Clinical Care: A Pragmatic Approach for a Special Population. Paediatr Drugs. 2023;25(1):5-11.
6.       Freriksen JJM, Schalkwijk S, Colbers AP, Abduljalil K, Russel FGM, Burger DM, et al. Assessment of Maternal and Fetal Dolutegravir Exposure by Integrating Ex Vivo Placental Perfusion Data and Physiologically-Based Pharmacokinetic Modeling. Clin Pharmacol Ther. 2020;107(6):1352-61.
7.       Verscheijden LFM, Koenderink JB, Johnson TN, de Wildt SN, Russel FGM. Physiologically-based pharmacokinetic models for children: Starting to reach maturation? Pharmacol Ther. 2020;211:107541.
8.       Chaphekar N, Dodeja P, Shaik IH, Caritis S, Venkataramanan R. Maternal-Fetal Pharmacology of Drugs: A Review of Current Status of the Application of Physiologically Based Pharmacokinetic Models. Front Pediatr. 2021;9:733823.
9.       Zhang X, Yang Y, Grimstein M, Fan J, Grillo JA, Huang SM, et al. Application of PBPK Modeling and Simulation for Regulatory Decision Making and Its Impact on US Prescribing Information: An Update on the 2018-2019 Submissions to the US FDA's Office of Clinical Pharmacology. J Clin Pharmacol. 2020;60 Suppl 1:S160-S78.
10.     Johnson TN, Boussery K, Rowland-Yeo K, Tucker GT, Rostami-Hodjegan A. A semi-mechanistic model to predict the effects of liver cirrhosis on drug clearance. Clin Pharmacokinet. 2010;49(3):189-206.
11.     Barter ZE, Tucker GT, Rowland-Yeo K. Differences in cytochrome p450-mediated pharmacokinetics between chinese and caucasian populations predicted by mechanistic physiologically based pharmacokinetic modelling. Clin Pharmacokinet. 2013;52(12):1085-100.
12.     van der Heijden JEM, Freriksen JJM, de Hoop-Sommen MA, Greupink R, de Wildt SN. Physiologically-Based Pharmacokinetic Modeling for Drug Dosing in Pediatric Patients: A Tutorial for a Pragmatic Approach in Clinical Care. Clin Pharmacol Ther. 2023;114(5):960-71.
13.     El-Khateeb E, Burkhill S, Murby S, Amirat H, Rostami-Hodjegan A, Ahmad A. Physiological-based pharmacokinetic modeling trends in pharmaceutical drug development over the last 20-years; in-depth analysis of applications, organizations, and platforms. Biopharm Drug Dispos. 2021;42(4):107-17.
14.     Sager JE, Yu J, Ragueneau-Majlessi I, Isoherranen N. Physiologically Based Pharmacokinetic (PBPK) Modeling and Simulation Approaches: A Systematic Review of Published Models, Applications, and Model Verification. Drug Metab Dispos. 2015;43(11):1823-37.
15.     van der Heijden JEM, Freriksen JJM, de Hoop-Sommen MA, van Bussel LPM, Driessen SHP, Orlebeke AEM, et al. Feasibility of a Pragmatic PBPK Modeling Approach: Towards Model-Informed Dosing in Pediatric Clinical Care. Clin Pharmacokinet. 2022;61(12):1705-17.
16.     Kuemmel C, Yang Y, Zhang X, Florian J, Zhu H, Tegenge M, et al. Consideration of a Credibility Assessment Framework in Model-Informed Drug Development: Potential Application to Physiologically-Based Pharmacokinetic Modeling and Simulation. CPT Pharmacometrics Syst Pharmacol. 2020;9(1):21-8.
17.     van der Zanden TM, Mooij MG, Vet NJ, Neubert A, Rascher W, Lagler FB, et al. Benefit-Risk Assessment of Off-Label Drug Use in Children: The Bravo Framework. Clin Pharmacol Ther. 2021;110(4):952-65.

2. Drug pharmacokinetics

After administration of a drug, several processes take place before the drug reaches its site of action and can exert its pharmacological effect. First, a drug needs to be absorbed in case of oral administration. Once a drug is within the systemic circulation, it can be distributed in the body. A drug can exert its effect within the systemic circulation or in a tissue or organ. A drug can undergo metabolism and will eventually be excreted. The processes contributing to the journey of a drug through the body are together called ‘pharmacokinetics’ (Figure 7). Pharmacokinetics is also referred to as ‘what a body does to a drug’. Drug absorption, distribution, metabolism and excretion (‘ADME’) are explained in further detail below.Schermafbeelding-2024-03-19-om-10-50-27

Figure 7. Pharmaockinetic processes.

The extent of drug absorption into the systemic circulation depends on:
̶         The drug formulation and way of administration:
            o   tablet, capsule and liquid (drink) for oral administration
            o   suppository for rectal administration
            o   intravenous fluid for direct administration into the systemic circulation
            o   fluid for injection under the skin or into the muscles
            o   local application formulations, such as ointments, nose drops and inhalers
̶         The acidity of the gastrointestinal tract. Drug dissolution, solubility and stability as well as intestinal permeability (dependent on the drug’s acid-dissociation constant, ‘pKa’) are influenced by the pH of the gastrointestinal tract.
̶         The rate at which the stomach is emptied. Slow gastric emptying causes delayed absorption into the systemic circulation and thus a lower plasma concentration.
̶         The rate at which the drug passes through the intestine. In case of slow passage, there is more time for absorption into the systemic circulation. 
̶         Intestinal surface area. The rate of drug absorption is influenced by the intestinal surface area available for absorption. Villi increase the surface area of the wall of the small intestine.

Bioavailability (abbreviation: F) is a measure of the amount of a drug from a formulation that reaches the systemic circulation. It is represented as a fraction or percentage of the total administered dose. The time it takes for a drug to reach the maximum plasma concentration is abbreviated as tmax.

The drug concentration in the systemic circulation (also referred to as ‘plasma concentration’) ultimately determines the concentration at the site of action and hence the effect of the drug. A typical plasma concentration-time curve upon oral administration is depicted in Figure 8. There is a high risk of toxicity (i.e., adverse effects) if the plasma concentration is too high. A plasma concentration that is too low concentrations can result in therapeutic failure. The plasma concentration at which sufficient efficacy is achieved without any signification toxic effects is called the therapeutic window. In case of a narrow therapeutic window, the margin between under- and overdosing is very small and small changes in drug dosing may hence cause therapeutic failure or toxic effects. In such cases, drug dosing should be strictly controlled.

. Schermafbeelding-2024-03-19-om-10-50-34

Figure 8. Plasma concentration-time curve upon oral administration. 

Once the drug is within the systemic circulation, it spreads through the different tissues and organs of the body. There are three major body compartments a drug can go to:
̶ Extracellular compartment, which includes:
            o interstitial fluid (between blood vessels and cells)
            o intravascular fluid (systemic circulation)
̶ Intracellular compartment (inside cells)
Drug distribution is influenced by:
̶ The amount of fat and water in the body.
̶ The extent to which the drug dissolves in fat or in water. Water soluble drugs (hydrophilic) tend to stay in the extracellular compartment whereas fat soluble drugs (lipophilic) can be transported over the cell membrane into cells. The latter are therefore characterized by a higher volume of distribution.
̶ Ionization degree of the drug. The drug’s pKa (acid-base dissociation constant, a drug-specific physicochemical property) and the pH of the environment together determine whether most molecules are in their ionized or unionized state. Only the unionized form can cross cell membranes and exert an effect. Example: if the pKa of a drug is higher than the pH of mother milk, a drug will ionize in milk and will not be able to diffuse back to the plasma. This phenomena is called ‘ion trapping’ and can cause dangerously high drug levels in specific compartments.
̶ Binding of the drug to plasma proteins. Human serum albumin is the main protein of human plasma. Only the amount of drug that is not bound to proteins (the free fraction), is effective as this can be transported over the cell membrane into cells.
̶ Hemodynamic factors such as cardiac output and resistance of peripheral blood vessels.
̶ Permeability of blood-tissue barriers like the blood-brain barrier and the placental barrier.
The degree to which a drug is distributed in body tissue is expressed as the volume of distribution (abbreviation: Vd). The volume of distribution is the ratio (i.e., proportionality factor) of the amount of drug in the body (the dose) to the plasma concentration of the drug. The smaller the volume of distribution, the more likely that the drug is confined to the systemic circulation. A high volume of distribution (it can go up to 10,000 L!) indicates that the drug is found in tissues of the body.

The body gets rid of many drugs by converting it into molecules that can be excreted more easily. Though, some drugs can also be excreted in their unchanged form. Drug conversion is called ‘metabolism’ or ‘biotransformation’ and the products are called metabolites. Commonly, metabolites are pharmacologically inactive, but they can also be active and produce significant effects in the body. Cytochrome P450 (CYP450) enzymes are responsible for the metabolism of the majority of drugs, but also other enzyme systems can play a role. CYP450 enzymes are mainly located in the gastrointestinal wall and in the liver.

Most drugs are eliminated from the body via the kidneys. The kidneys are able to excrete water-soluble drugs in their unchanged form or as metabolites via the urine. Renal drug excretion is affected by:
̶         Blood flow to the kidneys.
̶         Glomerular filtration. Only drugs that are small enough and not bound to plasma proteins can reach the pre-urine via glomerular filtration.
̶         Tubular secretion. Drugs that are not easily filtered can undergo tubular secretion from the blood to the pre-urine. This is an active process since it does require energy to transport drugs against a concentration gradient. Limited transport capacity may result in competition when two drugs must be excreted simultaneously.
̶         Tubular reabsorption. Once excreted into the pre-urine, a drug and/or its metabolite(s) can be reabsorbed back into the systemic circulation. This happens continuously with useful substances such as water and solutes. Tubular reabsorption occurs primarily in the proximal tubules and is a passive process since it does not require energy. 
Drug clearance (abbreviation: CL) is a measure of the efficiency of drug excretion. It is defined as the volume of blood cleared of a drug per unit of time. The elimination half-life (abbreviation: t1/2) is the time it takes to halve the amount of the drug in the blood. A long half-life usually means that the drug has a high volume of distribution and/or that the drug's metabolism or excretion is impaired, for instance due to strong binding to plasma proteins or the presence of enzyme inhibitors.

The effect of a drug
When the drug reaches its site of action, the drug binds to specific receptors. Receptors are specialized proteins that are located inside the cell and on the cell membrane. This binding initiates a signal cascade within the cell that ultimately generates the effect. The effect can be a desired therapeutic effect or an adverse effect. ‘Pharmacodynamics’ is the study of drug effects and is also referred to as ‘what a drug does to a body’.

3. Pharmacokinetic changes in pediatrics

General differences in drug pharmacokinetics between pediatric subjects versus adults and consequences thereof are discussed in the table below (Table 1).

Table 1. Pharmacokinetic differences between pediatric versus adult subjects





Route of administration

Oral: a drink is often considered distasteful and administration can therefore be difficult.

Tablets and capsules can be administered to children of approximately 5 years of age and older.

Variable absorption.

Intravenous: obtaining venous access can be difficult.

Drug does not reach the systemic circulation. 

Intramuscular: low muscle mass.

Variable absorption.

Rectal: a suppository is often expelled.

Incomplete absorption, loss of the drug via feces.

Transdermal: a developing stratum corneum, intense hydration of the skin and a high surface area to body volume ratio enables efficient permeation of drug molecules across the skin.

Relatively high drug concentrations in the systemic circulation.

Inhalation: technique is difficult, especially for children below 4 years of age.

Less drug disposition in the lungs.

Acidity of the gastrointestinal tract

Milk products can increase gastric pH as they can buffer stomach acids.

Reduced bioavailability of drugs that require a pH of less than 2.5 for absorption.

Intestinal surface area

Larger intestinal surface area in children younger than 1 year of age and the intestinal wall is more permeable for larger molecules as compared with adults.

Absorption rate is delayed in neonates and young children. The time it takes for a drug to reach the maximum plasma concentration (tmax) is prolonged. The large area available for absorption compensated the delayed gastric emptying time and unregular peristalsis. Children between 4 and 13 years of age resorb strong lipophilic drugs fast, which is in contrast to neonates.

Gastric emptying time

Slower gastric emptying in neonates, reaching the adult rate at 6 to 8 months of age.


Unregular intestinal motility. Intestinal transit time is longer in neonates and shorter in children between 2 and 6 years of as compared with adults.








Body water content

Premature neonates: 85%.

Neonates: 75%

Child 5 months of age: 60%

Adults: 50%

Volume of distribution of hydrophilic drugs is larger in neonates and children of approximately 1 year of age because of the high volume of extracellular water. With a similar weight-based (mg/kg) dose, a lower plasma concentration is reached. Neonates therefore often need a higher weight-based dose as compared with older children and adults. Because of the low body fat percentage in neonates, lipophilic drugs tend to have a lower volume of distribution (so a higher concentration within plasma) compared with older children. 

Body fat content

Lower body fat percentage in neonates (15%), with a slow increase in the first year of life and subsequently a decrease starting at 2 years of age.

Extracellular water content

High extracellular water volume in neonates (45% of the total body) as compared with children of approximately 1 year of age (28%) and adults (15%).

Plasma protein binding

The amount of plasma proteins is relatively low, binding affinity is diminished and there is a high concentration of endogenous competing substrates, together resulting in relatively low drug binding to plasma proteins in neonates and infants. Binding of drugs to the most important plasma protein (albumin) reaches adult values at approximately 10 to 12 months of age.

A lower plasma protein binding results in a higher drug plasma concentration. Though, after 1 month of age, the increase in drug plasma concentration is not of clinical significance anymore.

Despite the fact that lower plasma protein binding suggests a higher drug concentration in the extracellular compartment, the effect on the plasma concentration appears to be negligible in most cases. This can be explained by the fact that low plasma protein binding also leads to a larger volume of distribution (outside plasma) and often an increased clearance of the drug.

Permeability of the blood-brain barrier

Pharmacodynamic observations reveal that drug passage seems to be more extensive in young children. The blood-brain barrier matures at approximately 4 months of age.

Potentially altered brain drug exposure and extrapyramidal effects.


Liver enzyme activity

The development of enzyme systems in neonates is incomplete (e.g., lower activity of CYP540 enzymes) but rapid maturation takes place in the first few weeks after birth. The metabolic capacity of some enzymes is bigger in children between 1 to 6 years of age as compared with adults, resulting in an increased clearance of several drugs in this age group.



Lower hepatic drug clearance and a prolonged elimination half-life. Because of the relatively quick changes in enzyme activity after birth, drug dose adaptations are often needed to prevent under- or overdosing.



Kidney blood flow

Blood flow through the kidneys is lower at birth and reaches adult values at 5 months of age.

Consequences are not known but lower renal drug clearance is likely.

Glomerular filtration rate (GFR)

The GFR is strongly reduced in term neonates (10-20 mL/min/1.73m2) and it doubles by 1 week of age. In premature neonates, the GFR is <1 mL/min/1.73m2. Adult values are reached by 3 to 5 months of age. 

In children between 1 to 3 years of age, the GFR is higher than in older children and adults (i.e., an ‘overshoot’ in this specific period) and returns to adult values in adolescence.

Prolonged elimination half-life. Because of relatively quick changes in GFR in the first few months after birth, drug dose adaptations are often needed to prevent under- or overdosing.

Children between 1 and 3 years of age may have a  lower plasma concentration of drugs that are excreted via glomerular filtration. 

Tubular secretion  

Tubular secretion matures less rapidly than glomerular filtration. At birth, tubular secretion is 20 - 30 % of adult values and it takes approximately 15 months before adult levels are reached.

Low activity of tubular drug secretion in young children, especially in neonates.


Tubular reabsorption

Maturation of tubular reabsorption is slower than maturation of glomerular filtration and tubular secretion. Adult levels are reached at approximately 2 years of age.

Low activity of tubular drug reabsorption children, especially in neonates.

Other aspects


It is presumable that there is interindividual variability in response (as it is in adults) and the response is drug- and population-specific.

A drug can be less effective or there is a higher risk of adverse effects.

Therapy adherence 

Largely depending on the formulation and route of administration (see ‘Absorption’). Adherence can be increased by improving the taste of a drug and by administering the drug in child friendly manner.

A lower drug plasma concentration (or even ineffective drug plasma levels).

4. Pharmacokinetic changes in pregnancy

Throughout pregnancy, a woman’s body undergoes changes to support the growth and development of her unborn child(ren). These physiological changes, along with the presence of a placenta and a growing fetus, affect how drugs are being processed. In medical terms, this can be described as altered drug pharmacokinetics in pregnancy. These changes may affect the dosing needs of pregnant women and their unborn children.

The influence of pregnancy on pharmacokinetics
Pregnancy affects all phases of pharmacokinetics. Examples of important changes in pharmacokinetics as part of pregnancy are summarized in Figure 9 and further explained below.  


Figure 9. Pharmacokinetic changes in pregnancy

During pregnancy, stomach acidity decreases and the production of the hormone progesterone increases. Lower stomach acidity reduces the absorption of basic drugs (drugs with a high pH) while it increases the absorption of acidic drugs (drugs with a lower pH). Progesterone causes smooth muscle cells in the gastrointestinal tract to relax. This slows down gastric emptying and gut movements. Consequently, drugs are absorbed more slowly. On the other hand, the cardiac output (the amount of blood the heart pumps through the body per minute) increases during pregnancy. Because of this, more blood is pumped towards the intestines. This, in turn, enhances the absorption of drugs. In addition, many women experience nausea and vomiting in early pregnancy. This may decrease the amount of drug available for absorption, especially following oral intake.

Overall, the effect of pregnancy on absorption is drug-specific. However, taken together, the effects of pregnancy on drug absorption are likely to be minimal as pregnancy-induced changes in the gastrointestinal system that reduce drug absorption are compensated by the increase in cardiac output.

The amount of blood, body water and body fat significantly increases during pregnancy. This causes drugs to spread differently in the body and causes changes in the volume of distribution of a drug. The volume of distribution of a drug is a measure of how extensively it spreads across the body. The impact of pregnancy on the volume of distribution depends on the physical and chemical characteristics of the drug. Drugs can either be lipophilic (repelled by water and attracted to fat) or hydrophilic (favoring water and repelled by fat).
·       When the body has a higher water content (e.g. expanded plasma volume and water in tissues), hydrophilic drugs can become less concentrated in the bloodstream. This stems from the fact that there is more water to distribute across.. Drugs can then have a higher volume of distribution during pregnancy compared to the non-pregnant situation. This dilution can contribute to a lower peak plasma concentration of the drug, shortly after administration. Since drug concentration commonly drives the pharmacological effect, pregnancy-induced changes in bodily fluid can thus affect the effectiveness of drugs. Besides an effect on the peak plasma concentration, an increase in volume of distribution also contributes to extending a drug’s half-life.
·       As a woman gains more fat during pregnancy, this may also lead to an increased volume of distribution for lipophilic drugs that tend to spread to fat tissues. The same principles apply as were explained earlier for hydrophilic drugs.

Changes in plasma proteins during pregnancy can also impact drug distribution. Because of the increasing blood volume in pregnancy, the concentration of proteins circulating in the bloodstream decreases. An example of a plasma protein that shows a decrease in concentration during pregnancy is albumin. This, in turn, may affect the fraction of a drug bound to plasma proteins. These changes can impact the volume of distribution and potentially the (pharmacodynamic) effects of a drug. This is because, in principle, the effect of a drug is primarily determined by its unbound fraction in the bloodstream. Depending on how well the “excess” unbound drug distributes to the body tissues, and how fast it is cleared from the body via metabolism, an increased drug effect in pregnancy may occur.

Overall, due to a higher volume of fluids in the body, pregnancy often leads to lower peak concentrations of drugs in the bloodstream. This may lead to lower efficacy for certain drugs.   

The metabolism (that is, the breakdown) of drugs in the body is carried out by enzymes, primarily cytochrome CYP450 (CYP) enzymes. Breakdown of drugs facilitates their removal from the body through urine and feces. This process is known as drug metabolism. CYP enzymes are found throughout the body, with most of them residing in the liver. The breakdown of a given drug may rely on one CYP enzyme, a combination of multiple CYP enzymes, and/or other enzymes. The number and activity of CYP enzymes are altered by pregnancy hormones such as progesterone and estradiol. The direction of these changes depends on the type of CYP enzyme. Some enzymes become more abundant and/or active, while others become less abundant and/or active. This means that some drugs will be broken down faster, while others are broken down more slowly in pregnancy. This shift can lead to lower or higher plasma concentrations of the drug during pregnancy. Figure 10 illustrates this principle.

Pregnancy alters drug metabolism by increasing the liver blood flow and altering enzyme activity in different directions. For drugs whose breakdown is in part dependent on liver enzymes, the impact of pregnancy on their metabolism will depend on the particular enzyme(s) involved. Some drugs may be broken down more quickly, while others will be broken down more slowly. 

Alongside drug metabolism by liver enzymes, the kidneys play an important role in removing drugs from the body. Because of an increased cardiac output in pregnancy, there is an increased blood flow to the kidneys. This change starts in the second trimester and continues during the third trimester of pregnancy. In addition, due to the lower concentration of plasma proteins in the bloodstream, there is a higher concentration of free drug reaching the kidneys. As a result, drugs that are filtered by the kidneys are removed from the body more quickly during pregnancy. Faster renal filtration results in an increased drug clearance. On top of renal filtration, the kidneys contain small channels (‘tubules’) that can reabsorb drugs from urine as well as secrete drugs into urine. Renal absorption and secretion are also influenced by hormonal changes as part of pregnancy. This may impact the speed of drug removal from the body.

Overall, drugs tend to be removed faster from the body during pregnancy. This is mainly a result of increased renal filtration. These changes can result in shorter durations of therapeutic levels of the drug in the body during pregnancy. A higher dose may thus be needed for effective treatment in pregnancy.


Figure 10. The dose that was effective before pregnancy may not be as effective during pregnancy due to changes in the body.   

Placental transfer between mother and child
During pregnancy, alongside changes in a woman’s body, drug pharmacokinetics are affected by the growth of the placenta and the fetus. pregnancy, a connection is established between the mother’s bloodstream and the fetus’ bloodstream through the placenta. The placenta acts both as a gate and a barrier for the transport of substances between mother and fetus. Examples of substances that can cross the placenta are nutrients, oxygen, antibodies, and waste products. Most drugs can also cross the placenta, potentially reaching the fetus. The degree of placental transfer of a drug depends on drug characteristics such as molecular size and its lipophilicity (see definition in the paragraph ‘Distribution’).
The placental transfer of drugs is a key consideration in the context of pharmacotherapy during pregnancy for several reasons. First, some drugs are administered to a pregnant woman with the intention of treating the fetus. For instance, antibiotics can be used to treat infections of the uterus, thereby protecting the child. In these cases, placental transfer from the mother to the fetus is a prerequisite for the drug to have the desired effect. From a pharmacokinetic stance, placental transfer and metabolism, along with fetal metabolism (the breakdown of drugs in the fetus’s body) influence the amount of drug a mother and her fetus(es) are exposed to. This can affect the effectiveness of a drug. Lastly, some drugs may influence the development of the fetus., potentially causing adverse effects. The extent of placental drug transfer and the desired and unwanted effects of the drug on the fetus determine whether a drug can be used or not during pregnancy.

Effect of pharmacokinetic changes on drug doses in pregnancy
Pharmacokinetic changes in pregnancy can affect the dosing needs of pregnant women and their unborn children. An example of a drug for which some women may need a dose increase during pregnancy is sertraline. Sertraline is used for the treatment of depression and anxiety disorders, and can be prescribed during pregnancy (Sertraline MID report). Sertraline is primarily broken down by CYP enzymes. Altered CYP enzyme activity during pregnancy may result in increased sertraline metabolism, especially in the second and third trimesters. This increased breakdown, alongside other factors like changes in plasma protein concentration, can result in lower sertraline levels in the bloodstream of pregnant women. This can lead to increased depression or anxiety symptoms in certain women during pregnancy. In those cases, a dose increase may be needed to ensure adequate treatment. The proposed dose recommendation for sertraline in pregnancy (Sertraline MID report) has been chosen to maintain plasma levels deemed safe for the fetus.

1.     Westin AA, Reimers A, Spigset O. Should pregnant women receive lower or higher medication doses? Tidsskr Nor Laegeforen. 2018 Oct 30;138(17). English, Norwegian. doi: 10.4045/tidsskr.18.0065. PMID: 30378417.
2.     Kazma JM, van den Anker J, Allegaert K, Dallmann A, Ahmadzia HK. Anatomical and physiological alterations of pregnancy. J Pharmacokinet Pharmacodyn. 2020 Aug;47(4):271-285. doi: 10.1007/s10928-020-09677-1. Epub 2020 Feb 6. PMID: 32026239; PMCID: PMC7416543.
3.     Abduljalil K, Furness P, Johnson TN, Rostami-Hodjegan A, Soltani H. Anatomical, physiological and metabolic changes with gestational age during normal pregnancy: a database for parameters required in physiologically based pharmacokinetic modelling. Clin Pharmacokinet. 2012 Jun 1;51(6):365-96. doi: 10.2165/11597440-000000000-00000. PMID: 22515555.
4.     Pinheiro EA, Stika CS. Drugs in pregnancy: Pharmacologic and physiologic changes that affect clinical care. Semin Perinatol. 2020 Apr;44(3):151221. doi: 10.1016/j.semperi.2020.151221. Epub 2020 Jan 25. PMID: 32115202; PMCID: PMC8195457.
5.     van Donge T, Evers K, Koch G, van den Anker J, Pfister M. Clinical Pharmacology and Pharmacometrics to Better Understand Physiological Changes During Pregnancy and Neonatal Life. Handb Exp Pharmacol. 2020;261:325-337. doi: 10.1007/164_2019_210. PMID: 30968215.
6.     Eke AC. An update on the physiologic changes during pregnancy and their impact on drug pharmacokinetics and pharmacogenomics. J Basic Clin Physiol Pharmacol. 2021 Dec 8;33(5):581-598. doi: 10.1515/jbcpp-2021-0312. PMID: 34881531; PMCID: PMC9174343.
7.     Koren G, Pariente G. Pregnancy- Associated Changes in Pharmacokinetics and their Clinical Implications. Pharm Res. 2018 Feb 12;35(3):61. doi: 10.1007/s11095-018-2352-2. PMID: 29435666.
8.     Pariente G, Leibson T, Carls A, Adams-Webber T, Ito S, Koren G. Pregnancy-Associated Changes in Pharmacokinetics: A Systematic Review. PLoS Med. 2016 Nov 1;13(11):e1002160. doi: 10.1371/journal.pmed.1002160. PMID: 27802281; PMCID: PMC5089741.
9.     Pacifici GM, Nottoli R. Placental transfer of drugs administered to the mother. Clin Pharmacokinet. 1995;28(3):235–69. Available from:

5. Project MADAM

Background and aim of project MADAM
During pregnancy, changes in a woman’s body, influence how drugs move through and get processed in the body. In addition, some drugs reach the fetus through the placenta. In some cases, this may require altering drug doses for pregnant women and their unborn children. Recognizing the lack of well-researched doses in pregnancy, project MADAM aims to issue evidence-based dose recommendations for pregnant women and their unborn children. The goal is to issue dose recommendations for commonly used drugs in pregnancy.

Collaborators on project MADAM
Project MADAM is a collaboration between the Radboud and Maastricht University Medical Centers and Moeders van Morgen Lareb (the Dutch Teratology Information Service), funded by the Bill & Melinda Gates Foundation.  

Issuing dose recommendations for pregnancy
Issuing evidence-based dose recommendations for use in pregnancy relies on a structured assessment of the scientific literature. Information collected as part of project MADAM include data on the pharmacokinetics, safety, and efficacy of drugs during pregnancy in relationship to the dose. Alongside clinical data, some dose recommendations issued as part of project MADAM rely on evidence derived from pharmacokinetic models. Pharmacokinetic models are computer models used to predict adequate drug doses for specific groups of patients based on their (changing) characteristics, for example changes in a woman’s body and her fetus as part of pregnancy. Initial assessment of this data is conducted by researchers of project MADAM and results in a preliminary dose recommendation (Figure 11). This preliminary dose recommendation is also informed by the input of clinical experts (e.g., doctors prescribing the drug). Next, this preliminary dose recommendation and all underlying evidence are presented to a multidisciplinary working committee.

The working committee is a national body that comprises over 25 members with various types of experience and expertise). Members include various medical specialists that prescribe or give advice on drugs in pregnancy, pharmacists, pharmacologists, a midwife. alongside representatives of pregnant women and their partners, an ethicist and an expert modeler. The working committee assesses the proposed dose recommendation bringing in the diverse expertise and experience of its members. The aim is to issue a dose recommendation based on the latest available scientific knowledge as well as practical and ethical considerations for its use.
Drug doses that have been endorsed by the working committee are published on the website of Moeders van Morgen Lareb. The underlying evidence for a pregnancy-adjusted dose is shared on the drug pages of this website.


Figure 11. Our methods for issuing dose recommendations for use in pregnancy 

6. Glossary

  • Absorption: The process by which a drug enters the systemic circulation from its site of administration and becomes available for distribution within the body.                   
  • ADME: Absorption, distribution, metabolism and excretion of a drug.
  • Clearance: Rate of elimination of a drug from the body.                    
  • Decision consequence: The significance of an incorrect decision. In the context of modeling to inform drug dosing, this concerns the probability of therapeutic failure or adverse effects in case of an incorrect model-informed dose. 
  • Distribution: The transport of the drug from the systemic circulation to various tissues and organs.
  • Dose-exposure-response relationship: The presence of a causal relationship between the drug dose, the plasma concentration and the effect of the drug over time
  • Dose-finding method: The method applied to determine the optimal, model-informed, dose. For example, when the therapeutic window of the drug of interest is known, one can aim to reach plasma levels within this window. Another approach is to match the drug's exposure in the population of interest to exposure that is considered safe and effective in another population (‘exposure matching’).
  • Drug label: Any information provided with prescription drugs (i.e. package leaflet) under the regulation of the FDA in the United States or EMA in the European Union. Drug labels provide specific instructions or warnings for administration, storage and disposal. Drug labeling is also referred to as prescription labeling.
  • Drug model (in the context of physiologically-based pharmacokinetic modeling): Database containing drug-specific information such as its molecular weight and lipophilicity .This so-called ‘drug model’ (a file) can be selected in a modeling software tool, can be uploaded into it, or can be generated from scratch.
  • Excretion: Elimination of a drug from the body.
  • Exposure matching: Match the drug's plasma exposure in the population of interest to exposure that is considered safe and effective in another population.
  • In vivo: in a living organism (e.g., human body).
  • Metabolism: Describes the conversion of a drug into one or multiple metabolites. An important family of enzymes involved in this process are the cytochrome P450 (or CYP) enzymes. Of all tissues, the highest concentration of CYP enzymes is found in the liver. Some CYP enzymes are also located in other tissues, such as the lungs, intestines and placenta. Drug metabolism serves several purposes, including the facilitation of the drug elimination from the body (e.g., by making the molecule more water soluble, it can be excreted via the urine).
  • Model credibility: Measure of trustworthiness of a model. Model credibility is typically ascertained through different verification steps. Sufficient model credibility is crucial for applying a model for clinical dose optimization.
  • Model influence: The weight of the model in the totality of all available evidence. In the context of modeling to inform drug dosing, this concerns the availability of other data supporting the model-informed dose.
  • Model-informed dosing (MID): An approach to optimize drug dosing recommendations for a population of interest by conducting computer model simulations. 
  • Model performance: The accuracy of the model in predicting observed data.
  • Model robustness: The extent to which a model maintains its performance when faced with uncertainties or variations.
  • Model verification: The process of assessing the accuracy of the model in predicting observed data. Model verification is required to establish confidence in model predictions of drug pharmacokinetics (‘model credibility’).
  • Off-label drug use: There is often a need to administer drugs to a population for which the drug is not officially approved, for instance pediatric patients or pregnant women. This is called ‘off-label’ drug use and is legally permitted under certain conditions. Also if a drug is processed into a different formulation, is given in a different dosage, administered via a different route or used for another indication than stated in the label, its use is considered off-label.
  • Pharmacodynamics: What a drug does to the body. It refers to the pharmacological effect of a drug, which can be a desired therapeutic effect or side effect(s). A key mechanism is the drug’s ability to interact with its target receptors in the body, and the resulting cascade of biochemical reactions leading to the effect(s).                                              
  • Pharmacokinetics: What a body does to the drug. It refers to the journey of a drug through different organs and tissues, including the absorption, distribution, metabolism and excretion (ADME) of a drug from the body, and how these processes together govern the plasma concentration of a drug over time (or ‘exposure to a drug’).
  • Pharmacokinetic parameters: Parameters describing the absorption, distribution, metabolism and excretion (ADME) of a drug. Most important parameters are the Cmax (maximum plasma concentration), tmax (time to maximum plasma concentration), Vd (volume of distribution), CL (clearance) and t1/2 (elimination half-life).
  • Pharmacokinetic studies: Studies that analyse the drug plasma concentrations over time in a relatively small number of individuals who are administered the drug (as a fixed dose or different dosing regimens distributed over the study population). The measured plasma concentrations over time, or ‘plasma concentration-time profiles’ can then be used to determine pharmacokinetic parameters.
  • Physiologically-based pharmacokinetic (PBPK) models: Mathematical computer models that are based on our knowledge of anatomy and physiological processes (‘population model’) as well as drug-specific physicochemical properties (‘drug model’) that together govern drug pharmacokinetics in a population of interest. The modeler can design a virtual trial (e.g., age range of the population, duration of the study and dosing regimen) so that the plasma concentration over time and corresponding pharmacokinetic parameters of a certain drug in a given population can be predicted. Various dosing regimens can be explored and model-informed dosing recommendations can be established (amongst other applications).
  • Plasma concentration-time profile: Curve representing the drug plasma concentration (y-axis) versus time (x-axis). The area under the plasma concentration-time curve (AUC) reflects the actual total body exposure to a drug after administration.
  • Population model (in the context of physiologically-based pharmacokinetic modeling): Database containing population-specific information such as the populations’ weight distribution, organ blood flow and liver enzyme abundance. This so-called ‘population model’ (a file) can be selected in a modeling software tool, can be uploaded into it, or can be generated from scratch. Taking into account intersubject biological variability, a virtual population can be generated.
  • Population-based pharmacokinetic (Pop-PK) models: Mathematical computer models that rely on observed in vivopharmacokinetic data. These models seek to identify factors contributing to interindividual variability in drug pharmacokinetics (e.g., weight and renal function) and can be used to optimize dosing regimens in various populations.
  • Pragmatic physiologically-based pharmacokinetic (PBPK) modeling: Due to the mechanistic nature of PBPK modeling, it is possible to combine any ‘drug model’ with any defined virtual ‘population model’. Several PBPK software platforms provide a variety of population and drug models. Many models have been published and are available in databases or from scientific literature. Already established models can be used pragmatically for PBPK modeling. 
  • Predicted-to-observed PK parameter ratio: The ratio of the predicted to the observed PK parameter value is a measure of model performance. A predicted-to-observed ratio can be calculated for different PK parameters, depending on the availability of observed data. Generally, model performance is considered acceptable if the ratio falls within 2-fold (0.5 to 2).
  • Simulation (in the context of PK modeling): A virtual trial that is simulated with a computer PK model (PBPK or Pop-PK model) to explore drug pharmacokinetics in a given population.
  • Systemic circulation: The flow of blood in the circulatory system within the body (‘bloodstream’).
  • Therapeutic Drug Monitoring (TDM): Refers to the (routine) measurement of plasma concentrations of a drug in individual patients to monitor if the plasma level is adequate and to adjust the dose if needed.
  • Therapeutic window (therapeutic range): Refers to the plasma concentration range of a drug that is associated with optimal efficacy while minimizing the risk of toxicity. A drug’s therapeutic window is determined based on drug pharmacokinetic data (i.e. plasma exposure) and how it relates to pharmacodynamic properties of the drug (i.e. therapeutic efficacy and adverse effects). The therapeutic window of a drug may vary between populations (e.g., pregnant women versus non-pregnant adults). A narrow therapeutic window may ask for very accurate dosing and therapeutic drug monitoring while this is redundant in case of a wide therapeutic window.
  • Visual predictive check: Assessment of model performance by visually comparing the predicted plasma levels to observed in vivo plasma levels by overlaying the predicted plasma concentration-time curve with observed data points.

7. Template for creating an MID report

A model-informed dosing (MID) report can be created using this template.
This template can be used as a guidance document for creating an MID report. Similarity of these reports will facilitate inclusion of modeling and simulation data into the MELINDA database and thereby expedite public knowledge dissemination via the website. An MID report can be send to