Pharmacometrics in pregnancy: An unmet need.

PA54CH04-Unadkat

ARI

ANNUAL REVIEWS

3 December 2013

15:14

Further

Annu. Rev. Pharmacol. Toxicol. 2014.54:53-69. Downloaded from www.annualreviews.org by University of Central Florida on 01/27/14. For personal use only.

Click here for quick links to Annual Reviews content online, including: • Other articles in this volume • Top cited articles • Top downloaded articles • Our comprehensive search

Pharmacometrics in Pregnancy: An Unmet Need Alice Ban Ke,1 Amin Rostami-Hodjegan,2,3 Ping Zhao,4 and Jashvant D. Unadkat5 1 Drug Disposition, Lilly Research Laboratories, Indianapolis, Indiana 46285; email: [email protected] 2 School of Pharmacy and Pharmaceutical Sciences, University of Manchester, Manchester M13 9PT, United Kingdom; email: [email protected] 3

Simcyp Limited (now part of Certara), Sheffield S2 4SU, United Kingdom

4

Office of Clinical Pharmacology, Office of Translational Sciences, Center for Drug Evaluation and Research, United States Food and Drug Administration, Silver Spring, Maryland 20993-0002; email: [email protected] 5 Department of Pharmaceutics, University of Washington, Seattle, Washington 98195; email: [email protected]

Annu. Rev. Pharmacol. Toxicol. 2014. 54:53–69

Keywords

The Annual Review of Pharmacology and Toxicology is online at pharmtox.annualreviews.org

modeling, simulation, PBPK, CYP, POP-PK

This article’s doi: 10.1146/annurev-pharmtox-011613-140009

Abstract

c 2014 by Annual Reviews. Copyright All rights reserved

Pregnant women and their fetuses are orphan populations with respect to the safety and efficacy of drugs. Physiological and absorption, distribution, metabolism, and excretion (ADME) changes during pregnancy can significantly affect drug pharmacokinetics (PK) and may necessitate dose adjustment. Here, the specific aspects related to the design, execution, and analysis of clinical studies in pregnant women are discussed, underlining the unmet need for top-down pharmacometrics analyses and bottom-up modeling approaches. The modeling tools that support data analysis for the pregnancy population are reviewed, with a focus on physiologically based pharmacokinetics (PBPK) and population pharmacokinetics (POP-PK). By integrating physiological data, preclinical data, and clinical data (e.g., via POP-PK) to quantify anticipated changes in the PK of drugs during pregnancy, the PBPK approach allows extrapolation beyond the previously studied model drugs to other drugs with well-characterized ADME characteristics. Such a systems pharmacology approach can identify drugs whose PK may be altered during pregnancy, guide rational PK study design, and support dose adjustment for pregnant women.

53

PA54CH04-Unadkat

ARI

3 December 2013

15:14

CURRENT STATUS AND NEED FOR CLINICAL STUDIES IN PREGNANT WOMEN PK: pharmacokinetic(s)


POP-PK: population pharmacokinetic(s)

54

Pregnant women and their fetuses are orphan populations with respect to the safety and efficacy of drugs. Recent statistics indicate that 64% of pregnant women ingest at least one medication for the treatment of a variety of clinical conditions, including viral (e.g., HIV), fungal, or bacterial infections; smoking cessation; epilepsy; and pregnancy-induced conditions such as hypertension, depression, and gestational diabetes (1, 2). The use of over-the-counter medications during pregnancy is also common. For example, a study conducted in pregnant women in rural West Virginia found that 95.8% of participants used prescription medications, 92.6% took over-the-counter medications, and 45.2% used herbal medicines (1). Approximately 5–10% of pregnant women receive US Food and Drug Administration (FDA) category D or X drugs, which are potential teratogens, and the frequency of drug use is higher in early versus late pregnancy (2, 3). Despite these concerns, cessation of drug therapy during pregnancy for the treatment of serious clinical conditions is not possible as it may be detrimental to the health and well-being of the pregnant woman and therefore her fetus. Ethical, legal, and practical considerations often preclude enrollment of pregnant and lactating women in clinical trials. Despite the call to include pregnant women in drug trials in recent years, there appears to be little movement in that direction. A survey of registered clinical trials at R (hydroxyprogesClinicalTrials.gov indicates that in the past five years, only one drug—Makena terone caproate)—has been approved by the FDA for pregnancy indications (4). Drugs used in pregnancy are often developed informally; that is, they are used for maternal disease, fetal disease (e.g., fetal tachycardia), and placental dysfunction (e.g., malaria) after they are tested in a general population devoid of pregnant participants (5). As a result, although drugs are routinely prescribed during pregnancy, they are almost always prescribed off-label, i.e., without necessary clinical data about the dose, pharmacokinetics (PK), safety, or efficacy in pregnant women. PK studies, when conducted in pregnant women, can be generally categorized into two types: opportunistic PK studies of drugs administered for therapeutic intervention purposes, and dedicated PK studies of drugs (e.g., probe/model drugs to phenotype activity of a certain enzyme or transporter) for mechanistic understanding of PK changes. Opportunistic PK studies typically enroll pregnant patients who chronically receive a variety of dosing regimens. Because these studies do not impose additional risks to the study subjects as the subjects receive the drugs as part of clinical care, obtaining institutional review board approval has been relatively easy. Dedicated PK studies of (probe) drugs are usually single-dose, intensive-sampling PK studies performed in a small number of healthy pregnant women during pregnancy and in the postpartum period, during which the drugs are not used for a therapeutic purpose. Enrollment can be challenging as subjects are more reluctant to participate owing to fear of fetal toxicity, especially during the first two trimesters (T1 , T2 ). Therefore, although difficult, such studies are mostly possible during the third trimester (T3 ). When sampling is sparse (as is usually the case in opportunistic studies), the PK data need to be analyzed using a population pharmacokinetic (POP-PK) modeling approach. When intensive sampling is conducted, the PK data, owing to the small sample size, are analyzed by classical noncompartmental or classical compartmental analysis. Ideally, the study should include assessment of the PK of the drugs during each trimester and postpartum (6). This paired comparison (during pregnancy versus postpartum) increases the power of the study and therefore minimizes the number of subjects studied. Considerable data demonstrate that the PK of drugs can be significantly affected by pregnancy (see Table 1 for examples). Altered drug disposition in this special population may lead to either underdosing or overdosing of medication when the standard adult dose is prescribed, resulting in possible changes in safety and efficacy (7). For example, the systemic exposure to the HIV-protease

Ke et al.

PA54CH04-Unadkat

ARI

3 December 2013

15:14

Table 1 Summary of pregnancy-induced changes in the pharmacokinetics of clinically used drugs Effect on CL/F (%)a Metabolizingenzyme activity Drug/probe


Caffeine∗

Indication

T1

T2

T3

changes

Reference

CNS stimulant

↓ 33

↓ 48

↓ 65

Theophylline

Asthma

↔

↔

↓ 34

Nicotine

Smoking cessation

NA

↑ 54

↑ 54

↑ CYP2A6

50

Phenytoin∗,b

Epilepsy

↑ 43

↑ 51

↑ 61

↑ CYP2C9

51

↓ CYP2C19

52

↑ CYP2D6

53

Proguanil

Malaria

NA

↓ 60

↓ 60

Metoprolol∗

Hypertension

NA

NA

↑ 459

Dextromethorphanb

Cough

↑ 26

↑ 35

↑ 48

Midazolam∗

Sedation

NA

NA

↑ 99

Indinavir

HIV infection

NA

NA

↑ 277

Glyburide

Diabetes

NA

NA

↑ 106

Methadone

Addiction

NA

↑ 101

Labetalol

Hypertension

NA

Lamotrigine

Epilepsy

↑ 200

Zidovudinec

HIV infection

NA

NA

48

↓ CYP1A2

49

48 14 ↑ CYP3A4

8

↑ 65

↑ CYP2B6

54

↑ 30

↑ 30

↑ UGT1A1

55

↑ 200

↑ 300

↑ UGT1A4

19

↔

↔ UGT2B7

56

↑ Renal CL

20

Amoxicillin

Bacterial infection

NA

↑ 23

↑ 20

Metformin∗

Diabetes

↑ 22

↑ 28

↑ 11

Digoxin∗

Cardiac diseases

NA

NA

↑ 19

9

22 14

Asterisk indicates that the probe drug is commonly used to measure specific hepatic enzyme activity or transporter activity in vivo. ↓ indicates decrease; ↑ indicates increase; ↔ indicates no effect; NA indicates that the data are not available. a Mean percentage change relative to postpartum value. b Phenytoin data are based on plasma trough concentration (total). Dextromethorphan data are based on urinary metabolic ratio (dextromethorphan/dextrorphan). c Morphine intravenous clearance (mediated by UGT2B7), determined at time of delivery, was found to increase 59% compared with nonpregnant control data (57), suggesting increased UGT2B7 activity and/or increased hepatic blood flow. Abbreviations: CL, clearance; CL/F, apparent oral clearance, where F represents bioavailability; CNS, central nervous system; CYP, cytochrome P450; T1/2/3 , first/second/third trimester; UGT, uridine diphosphate glucuronosyltransferase.

inhibitor indinavir is significantly lower during pregnancy than in the postpartum period (or in nonpregnant women or men): 3.8-fold-lower plasma AUC (area-under-the-concentration-time curve), 2.8-fold-lower Cmax (peak concentration), and 3.8-fold-lower Cmin (trough concentration) (8). The 3.8-fold induction of oral clearance resulted in subtherapeutic indinavir plasma concentrations that could result in breakthrough of resistant virus, leading to progression of disease in the mother and perhaps a higher incidence of maternal-fetal HIV transmission and transmission of resistant virus to the baby. The systemic exposure to glyburide is also significantly lower in pregnant women with gestational diabetes mellitus than in nonpregnant women with type 2 diabetes mellitus: 2.1-fold-lower plasma AUC, 2.2-fold-lower Cmax , and 3.2-fold-lower Cmin (9). Determining the magnitude of change in PK caused by pregnancy is important in the design of rational dosing regimens of drugs for pregnant women. However, it is neither feasible nor desirable to perform extensive PK studies in pregnant women (including fetal exposure) of all the drugs consumed by this common, at-risk population. Therefore, alternate approaches are required. One such approach is to conduct focused studies that shed light on the mechanisms by which pregnancy alters maternal drug disposition. Probe drugs (indicated by asterisks in www.annualreviews.org • Modeling and Simulation of PK/PD in Pregnancy

AUC: area-under-theconcentration-time curve Cmax : peak concentration Cmin : trough concentration

55

PA54CH04-Unadkat

ARI

3 December 2013

ADME: absorption, distribution, metabolism, and excretion PBPK: physiologically based pharmacokinetic(s)


Tmax : time at which the peak concentration occurs CYP: cytochrome P450

15:14

Table 1) are drugs commonly used to probe in vivo enzyme or transporter activity, as they are selective substrates of a single enzyme or transporter. Therefore, compared with nonprobe-drug studies, probe-drug studies are more informative in delineating hepatic enzyme activity during pregnancy, and the information generated can be used to predict the change in systemic exposure to nonprobe therapeutic drugs during pregnancy. Such a prediction is possible only if the absorption, distribution, metabolism, and excretion (ADME) characteristics of the nonprobe drugs, such as the fractional contribution of the enzyme or transporter toward the total clearance of the drug, are well characterized and if information on physiological changes that could affect the ADME of the drug is considered. Because the latter are multidimensional (e.g., simultaneous and time-dependent changes in volume of distribution, plasma protein binding, cardiac output, and renal function; see Review of Physiological Changes During Pregnancy, below), an efficient methodology for incorporating such prior knowledge to predict appropriate dosing regimens of drugs during pregnancy is physiologically based pharmacokinetic (PBPK) modeling (see PBPK, below). The coupled maternal-fetal physiology imposes additional concerns about both efficacy and safety of treatment. Recommendation of dose adjustment in this population should also be carefully evaluated in terms of assessing the risk of fetal exposure to drugs administered to the mother (10). Because the teratogenic or fetal toxicity potential of a drug is usually not known, treatments should also attempt to minimize fetal exposure to drugs, except in cases of fetal conditions or HIV infection, where fetal exposure to drugs is important for therapeutic reasons (11). In general, for most drugs, exposure targets of pharmacotherapy in pregnant women are not established. For drugs with an established relationship between systemic exposures and efficacy/safety in nonpregnant women or men, a reasonable goal for a dosing regimen in pregnancy is to achieve an unbound drug exposure equivalent to that obtained in nonpregnant adults who receive the standard dosing (12). Efficacy and/or fetal outcome data, whenever available, will greatly enhance the confidence in making a safe recommendation for dose adjustments during pregnancy.

REVIEW OF PHYSIOLOGICAL CHANGES DURING PREGNANCY Pregnancy is associated with a multitude of temporal physiological and metabolic changes. The mean change in physiological parameters during each trimester, based on a comprehensive metaanalysis of data obtained from adult healthy women (13), is summarized in Table 2. The causative mechanism of these changes is poorly understood, and most of them are believed to be regulated by pregnancy-related hormones. A number of these changes in maternal physiology can have a direct effect on drug ADME as discussed in greater detail below.

Absorption Systematic review of data reported in the literature does not appear to support a significant reduction in gastric motility or an increase in gastric pH (13). The rate of drug absorption does not appear to be altered by a significant extent as a result of any of these changes in pregnant women, as suggested by similar antepartum versus postpartum/nonpregnant Tmax (time at which the peak concentration occurs) values and none-to-modest changes in the half-lives of drugs (8, 14, 15). However, the extent of presystemic elimination (by enzymes or transporters) may be greater or lesser, depending on the pathway(s) contributing to the elimination of the drug [e.g., cytochrome P450 (CYP) enzymes or transporters]; the result is a lower or higher Cmax in this population. 56

Ke et al.

PA54CH04-Unadkat

ARI

3 December 2013

15:14

Table 2 Summary of pregnancy-induced changes in maternal physiology T1 a


Parameter

T2 a

T3 a

Total body weight (kg)

↑ 6%

↑ 16%

↑ 23%

Total fat mass (kg)

↑ 11%

↑ 16%

↑ 32%

Total body water (L)

↑ 11%

↑ 27%

↑ 41%

Cardiac output (L)

↑ 18%

↑ 28%

↑ 33%

Plasma volume (L)

↑ 7%

↑ 42%

↑ 50%

Red blood cell volume (L)

↑ 4%

↑ 20%

↑ 28%

Hematocrit (%)

↓ 3%

↓ 8%

↓ 14%

Albumin (g/L)

↓ 5%

↓ 16%

↓ 31%

α1-AGP (g/L)

↓ 1%

↓ 22%

↓ 19%

Glomerular filtration rate (mL/min)b

↑ 19%

↑ 37%

↑ 40%

Effective renal plasma flow (L/h)

↑ 38%

↑ 48%

↑ 31%

Creatinine clearance (mL/min)

↑ 28%

↑ 58%

↑ 26%

Uterine blood flow (L/h)

↑ 923%

↑ 1,567%

↑ 2,771%

Hepatic blood flow (L/h)c

↔

↔

↔

↓ indicates decrease; ↑ indicates increase; ↔ indicates no effect. a Mean percentage change (%) relative to prepregnancy level. b Glomerular filtration rate measurement is based on inulin clearance. c Literature data on hepatic blood flow are contradictory; hence, no effect is assumed. Data sources are provided in Reference 13. Abbreviations: α1-AGP, alpha 1 acid glycoprotein; T1/2/3 , first/second/third trimester.

Distribution Various pregnancy-related hemodynamic changes lead to an increase in plasma volume (up to 50%) and a decrease in plasma protein binding, which can alter the apparent volume of distribution (Vd ) of drugs (16). Serum albumin and alpha 1 acid glycoprotein (α1-AGP) concentrations decrease up to 31% and 19% during late pregnancy, respectively (13). Consequently, the change of unbound plasma fraction of highly bound drugs (e.g., phenytoin) during pregnancy can be as much as ∼30% (17). In the literature, the reported change in apparent Vd (Vd /F, where F represents bioavailability) of oral drugs during pregnancy may result from the physiological changes during pregnancy that affect Vd or F. The two effects cannot be easily dissociated on the basis of oral PK data only, and the data should be interpreted with caution. Through changes in Vd and clearance, pregnancy can cause an increase or a decrease in the terminal elimination half-lives of drugs.

Clearance via Metabolism and Excretion Intravenous (IV) administration of a drug provides the best measure of the clearance and Vd of a drug. Changes in IV clearance of drugs may result from changes in hepatic blood flow, protein binding, or hepatic intrinsic clearance. On the basis of the well-stirred model of hepatic elimination, an increase in hepatic blood flow would result in increased clearance after IV administration of a drug with a high extraction ratio (18). Despite a marked 33% increase of cardiac output during pregnancy, there are contradictions within the existing data—gathered via various techniques such as indocyanine green clearance and Doppler ultrasonography—on the change in hepatic blood flow during pregnancy (13). Owing to this inconsistency, we have assumed that pregnancy has no effect on hepatic blood flow, as shown in Table 2. Because most drugs administered to www.annualreviews.org • Modeling and Simulation of PK/PD in Pregnancy

Vd : volume of distribution Vd /F: apparent volume of distribution, where F represents bioavailability

57

PA54CH04-Unadkat

ARI

3 December 2013

CL: clearance CL/F: apparent oral clearance, where F represents bioavailability


NCA: noncompartmental analysis SMPK: semimechanistic pharmacokinetic(s)

15:14

pregnant women are given orally, changes in the drugs’ oral clearance are likely due to altered hepatic enzyme/transporter activity and/or decreased plasma protein binding. The majority of PK studies in pregnant women have focused on CYP activity (see Table 1). The change in maternal hepatic enzyme activity is CYP-isoform specific. Several groups have utilized model (probe) drugs that measure CYP enzyme activities to delineate the magnitude of changes in the activity of major CYP enzymes. These studies have been carried out mostly during T3 , often using caffeine for CYP1A2, midazolam for CYP3A, metoprolol for CYP2D6, and phenytoin for CYP2C9 (16) (see Table 1). For example, the metabolism of drugs catalyzed by selective CYP isoenzymes (e.g., CYP3A4, CYP2D6, and CYP2C9) and uridine diphosphate glucuronosyltransferase (UGT) isoenzymes (e.g., UGT1A4 and UGT1A1) is increased during pregnancy. In contrast, CYP1A2 and CYP2C19 activity is decreased during pregnancy (16). The magnitude of pregnancy-related induction in drug-metabolizing enzyme activities can be as high as 300%, as suggested by a study of lamotrigine PK (19). The renal excretion of unchanged drugs is increased during pregnancy owing to an increase in glomerular filtration rate and also possibly to an increase in renal secretion via transporters (16). During T3 , renal secretion clearance (CLsecretion ) mediated by organic cation transporters and P-glycoprotein is increased by 50% and 120%, respectively, according to studies of metformin CLsecretion (20) and digoxin CLsecretion (14). However, despite these significant increases in the renal secretory clearance of drugs, they will not translate into a significant increase in apparent oral clearance (IV clearance/bioavailability, or CL/F) unless secretion is the predominant route of elimination of the drug (21). For example, despite the large increase in metformin and digoxin CLsecretion , the reported increase in oral clearance of metformin and digoxin throughout pregnancy is less than 30% each (Table 1).

MODELING AND SIMULATION METHODOLOGY In consideration of the ethical and logistical barriers to including pregnant women in clinical trials, using modeling and simulation to study the disposition of xenobiotics in the maternal-fetal unit has emerged as a promising approach. It can guide dose adjustment in pregnancy to ensure adequate efficacy and to prevent undesirable toxicity (10, 22, 23), and it can ultimately help predict untested dosage adjustments. The pharmacometric tools that support PK analysis and that are most often applied to the special population of pregnant women include noncompartmental analysis (NCA) and compartmental analysis, the former of which includes the use of POP-PK. To evaluate and confirm specific underlying mechanisms that might be responsible for potential alterations in drug disposition during pregnancy and to predict the PK of a drug that has no pregnancy PK data, researchers can employ mechanistic modeling. Such models include (a) semimechanistic pharmacokinetic (SMPK) modeling, which often extends existing compartmental models of a particular drug by integrating essential mechanistic (often physiologically related) components within the model structure, and (b) PBPK. The role of each of these tools in study design and analysis—as well as the pros, cons, and impact of each methodology—is described below using a range of case examples.

Data-Driven Methods This section describes two data-driven methods for obtaining PK parameters/exposure information in the pregnant population: These include the aforementioned NCA and POP-PK analysis. NCA. NCA is a data-driven method of analysis that requires densely sampled PK data to empirically calculate integral measures (moments) such as AUC. These can then be used to obtain 58

Ke et al.

PA54CH04-Unadkat

ARI

3 December 2013

15:14


estimates for valuable PK parameters such as CL or CL/F, Vd or Vd /F, and half-life. Cmax , Cmin , and Tmax values are usually determined by direct inspection of data. NCA methods work mainly when carrying out a dedicated PK study in pregnant women such that the PK data are dense enough to reliably estimate the above parameters for each individual. Moreover, when compared with the compartmental analysis modeling approach, the NCA approach requires fewer model-based assumptions in the calculation of PK parameters. The majority of PK studies conducted in pregnant women to date are dedicated PK studies with intensive sampling, and the data are analyzed by the NCA approach. For example, studies on the disposition of antiretroviral drugs for the prevention of HIV transmission from mother to child during pregnancy have shown decreased exposure of these drugs (atazanavir, fosamprenavir, indinavir, lopinavir, nelfinavir, ritonavir, and saquinavir) during pregnancy compared with nonpregnant adults when standard doses of these drugs are administered (24). POP-PK analysis. When a dedicated PK study in pregnant women is logistically difficult to conduct, investigators can use a POP-PK approach, which offers a less restricted study design (such as sparse sampling and less homogeneous demographics of the study population). The advantage of a sparse-sampling strategy is that a missed clinic visit or limitations on the number of samples or the duration of follow-up have less impact on study integrity. A POP-PK model consists of (a) a structural model that is often a compartmental PK model and (b) an error model. POP-PK analyses allow simultaneous estimation of typical (or mean) population PK parameters, as well as between-subject variability (e.g., through covariate analysis) and residual, unexplained variability (e.g., assay error) (25). Once the structural (base) model is identified, covariates can be incorporated to elucidate the influence on PK parameters of various demographic and clinical factors, such as body size measures (e.g., total or lean body weight, body mass index), categorical variables (e.g., pregnancy state, concurrent medication, disease state), and continuous variables (e.g., gestational age, creatinine clearance, serum albumin, dose) (26). Specifically, in covariate analysis, inclusion of pregnancy state as a categorical variable, or gestational age as a continuous variable, in the covariate model for CL/F and V/F allows the effect of pregnancy on drug PK to be quantitatively assessed. Model validation is usually achieved by bootstrap analysis and visual predictive checks (27). POP-PK analysis can be employed prospectively or retrospectively to analyze PK samples collected via a sparse-sampling strategy from pregnant women, postpartum women, and/or nonpregnant women receiving the same drug treatment. The POP-PK approach has been widely used for antimalaria drugs (see below) and sometimes for antiviral drugs (28). The studies of antimalarials are usually conducted in remote areas, where study conditions are challenging. Interestingly, artesunate, artemether, dihydroartemisinin, sulfadoxine, atovaquone, proguanil, cycloguanil, pyrimethamine, and lumefantrine concentrations and cure rates are lower in pregnant women than in nonpregnant adults of childbearing age (7). A meta-analysis by McGready et al. (29) showed that 9 out of 12 PK studies recommended dose optimization in pregnant women with malaria. One interesting case is azithromycin (AZ), an azalide antibiotic with antimalarial activity that is considered safe in pregnancy. AZ is also among 15 drugs most frequently prescribed to pregnant women and is commonly administered for community-acquired respiratory, skin, and gynecological infections (2). AZ has incomplete oral absorption (34%), extensively distributes into tissues, and is eliminated by hepatobiliary excretion mediated in part by P-glycoprotein and multidrug resistance-associated protein 2 (26). The understanding of the effect of pregnancy on these hepatic transporters is limited. In one study (23), two 2-g doses of AZ were given 24 h apart to 31 pregnant and 29 age-matched nonpregnant Papua New Guinean women. The only significant relationship between the PK of AZ and a range of potential covariates, including malarial parasitemia, was www.annualreviews.org • Modeling and Simulation of PK/PD in Pregnancy

59

PA54CH04-Unadkat

ARI

3 December 2013


Vc /F: apparent volume of distribution of the central compartment, where F represents bioavailability

15:14

with pregnancy, which accounted for an 86% increase in the apparent volume of distribution of the central compartment (Vc /F). The plasma AUC0–∞ of pregnant subjects was similar to that of nonpregnant subjects. On the basis of this study, no dose adjustment was proposed for pregnant women (23). In a more recent study conducted in an urban US population (26), the POP-PK analysis included 53 pregnant and 25 nonpregnant women. Lean body weight, pregnancy, ethnicity, and the coadministration of oral contraceptives were identified as covariates that significantly influenced the oral clearance of AZ. Compared with nonpregnant women who were not receiving oral contraceptives, a 21% to 42% higher dose-adjusted AZ AUC occurred in non–African American women who were pregnant or receiving oral contraceptives. Fischer et al. (26) proposed that although higher levels of maternal and fetal AZ exposure suggest that lower doses be administered to non–African American women during pregnancy, consideration of AZ pharmacodynamics during pregnancy should guide any dose adjustments. Potential immunological changes in pregnant women (which may alter the bacterial responsiveness to AZ) and a limited understanding of the safety of fetal AZ exposure warrant further studies to determine the clinical implications of the observed PK differences. These two examples illustrate the utility of the POP-PK approach in providing appropriate estimates of PK parameters as well as in identifying important covariates that contribute to PK variability in women of childbearing age. As is the case for any POP-PK study, owing to the less controlled nature of the study when compared with dedicated PK studies, lack of adherence to the prescribed regimen as well as lack of accurate reporting of the timing of the dose and blood sampling may compromise interpretation of the data. Also, because PK samples are usually collected during clinic visits, they may cluster around certain time windows (such as in the morning) with insufficient PK samples in other time windows to estimate the PK parameters with confidence. Another limitation of POP-PK analysis is that the F of the drug is usually fixed in the structural model, so the impact of enzyme activity alterations on F, i.e., an inverse relationship between CL and F, is ignored. As a result, covariate effects related to the change of CL will not be captured properly in POP-PK analysis.

Predictive Methods Because NCA and POP-PK approaches do not readily incorporate the multidimensional changes in physiology produced by pregnancy that can affect the ADME of a drug (e.g., Vd , renal function), they cannot accurately predict an appropriate dosing regimen to use during pregnancy. Therefore, in order to predict dosing regimens of drugs during pregnancy, while incorporating the multidimensional changes in the physiology of pregnancy and the ADME of drugs, it is necessary to use a mechanistic model. PBPK. Under the overarching umbrella of systems pharmacology, PBPK modeling has the advantage of incorporating both (a) physiological parameters (e.g., tissue blood flow, tissue composition, and glomerular filtration rate) that are important for ADME processes and (b) drug-specific parameters (e.g., physicochemical and drug disposition characteristics) (30, 31). Review of methodology and available PBPK models. PBPK models are multicompartmental models that aim to describe major human tissues and organs in a mechanistic manner, and they have been used in drug development and regulatory review (32). The separation of drug- and systemdependent variables (31) offers many advantages compared with other modeling and simulation platforms. For example, the structural model and system-specific parameters are “generic”—i.e., they remain the same for a given population—whereas drug-specific parameters are overlaid onto the system model (30). Thus,“extrapolation” from one drug to another drug is possible; the 60

Ke et al.


PA54CH04-Unadkat

ARI

3 December 2013

15:14

structural model is mechanistic and integrative, and users can incorporate the effects of multiple “extrinsic” or “intrinsic” patient factors on drug disposition (32). Such incorporation makes this approach particularly useful in the case of evaluating/predicting drug exposure changes in the pregnant population when a multitude of pregnancy-related physiological changes in the ADME of drugs are expected. Because of the complexity of PBPK models, they require rich experimental data and often rely on in silico prediction tools to obtain the required physicochemical or ADME parameters. Because drug-specific parameters may not always be available, assumptions are often made on the basis of general knowledge. Thus, it is important to validate model predictions with available data. For example, validation of the drug-specific parameters using IV and oral data obtained in the nonpregnant population is necessary to demonstrate the adequacy of the model in nonpregnant adults. Also, assessing the sensitivity of model predictions to changes in the system model’s key parameters is necessary. As a general practice, only when the model has been verified for the nonpregnant subjects can it be modified and used to study the effect of pregnancy on drug PK. Many of the existing PBPK models that were developed to model maternal-fetal disposition have typically focused on toxicokinetics and risk assessment of environmental chemicals. Lu et al. (33) provided a systematic review of the existing human or animal PBPK models. These models considered the dynamic changes in physiological parameters including tissue volume, blood flow rate in the maternal body, and the growth of the placental-fetal unit. However, few models accounted for pregnancy-induced changes in the ADME of drugs (33). As discussed above, pregnancy effects on the ADME processes of drugs, particularly metabolism and elimination, can be profound and therefore cannot be neglected. For example, the midazolam model developed by Andrew et al. (34) sought to account for time-dependent physiological changes due to gestation. The developed model consists of 20 maternal compartments and 16 fetal compartments. Owing to the lack of data on fetal physiology, many assumptions were made to incorporate a detailed fetal structure. By assigning intrinsic clearance calculated from reported midazolam IV clearance in nonpregnant and postpartum subjects, the authors demonstrated the feasibility of their PBPK model in assessment of midazolam disposition during pregnancy. Development and application of a pregnancy PBPK model for drugs metabolized by CYP enzymes. Recently, a meta-analysis of literature data on pregnancy-induced physiological and metabolic changes in healthy pregnant women was carried out (13). On the basis of this work, Lu et al. (35) proposed to extend the perfusion-limited form of a 13-compartment PBPK model used for the nonpregnant population to the pregnancy population by applying known maternal physiological changes to all model components (Figure 1). These included changes in gestational weight gain, plasma protein and lipid concentration, individual organ/tissue volumes and blood flows, glomerular filtration rates, and hepatic enzyme activity (i.e., activity of CYP1A2, CYP2D6, and CYP3A). A lumped placental-fetal component was added to the model to represent the placenta, fetal organs, and amniotic fluid. Such simplification reduces the uncertainty that stems from fetal physiological parameters. Because the majority of drugs administered to pregnant women are metabolized by multiple CYP enzymes, it is important to determine if single-CYP activity data can be synthesized to predict the disposition of these drugs. Of the various CYP enzymes, CYP3A, CYP1A2, CYP2D6, CYP2C9, CYP2C19, and CYP2B6 are likely more important than other CYP enzymes in terms of metabolism of drugs that are administered to pregnant women. Therefore, we populated the PBPK model with probe-drug data (for CYP2B6 in vitro data) delineating hepatic CYP3A, CYP1A2, CYP2D6, CYP2C9, CYP2B6, and CYP2C19 enzyme activity during pregnancy, and we validated the model performance for well-characterized nonprobe www.annualreviews.org • Modeling and Simulation of PK/PD in Pregnancy

61

PA54CH04-Unadkat

ARI

3 December 2013

15:14

Lung

Adipose

Bone

Brain


Heart

Kidney Venous blood

Arterial blood Muscle

Liver

Spleen

Gut

Placentalfetal unit Figure 1 A schematic representation of the pregnancy physiologically based pharmacokinetic (PBPK) model. The PBPK model is an extension of Simcyp Ltd.’s 13-compartment full-PBPK model; it includes a lumped compartment to represent placental-fetal organs including the fetus, placenta, and amniotic fluid. Reproduced from Reference 36 with permission.

drugs that were cleared primarily by the same CYP enzyme that cleared the probe drug. Specifically, the refined PBPK model, assuming 99% induction of hepatic CYP3A during T3 (on the basis of midazolam data), quantitatively predicted T3 -induced changes in the disposition of two other CYP3A-metabolized drugs, nifedipine and indinavir (i.e., predicted/observed mean AUC and Cmax fell within the 0.80–1.25 range) (36). The PBPK model, assuming 65% suppression of hepatic CYP1A2 during T3 (on the basis of caffeine data), successfully predicted theophylline disposition during T3 (37). We defined the range of CYP2D6 induction during T3 to be 100% to 200% through the modeling of metoprolol, paroxetine, dextromethorphan, and clonidine disposition during pregnancy. Owing to the lack of a probe-drug study for CYP2B6, we incorporated CYP2B6 induction by estradiol based on in vitro–to–in vivo extrapolation. The expanded PBPK model assumed (a) hepatic CYP2B6 induction of 40% and 90% during T2 and T3 , respectively; (b) hepatic CYP2C9 (on the basis of phenytoin data) induction of 50% and 60% during T2 and T3 , respectively; and (c) hepatic CYP2C19 (on the basis of proguanil data) suppression of 62% and 68% during T2 and T3 , respectively. Using the expanded model, the disposition of methadone 62

Ke et al.


PA54CH04-Unadkat

ARI

3 December 2013

15:14

(cleared by CYP3A, CYP2B6, and CYP2C19) and glyburide (cleared by CYP3A, CYP2C9, and CYP2C19) during T2 and/or T3 was successfully predicted (38). Moreover, our PBPK approach to modeling drug disposition during pregnancy allowed us to bridge knowledge gaps that are not currently addressed or are difficult to test through clinical studies in this special population, including discerning the site (hepatic, intestinal, or both) of CYP3A induction in pregnancy. For drugs cleared predominantly by CYP3A, the site of CYP3A induction during pregnancy is expected to have a differential impact on pregnancy-induced changes in AUC. Our sensitivity analysis suggests that the observed change in systemic exposure to midazolam, nifedipine, and indinavir was driven mostly by the induction of hepatic CYP3A activity, with modest to little contribution from intestinal CYP3A induction. Based on this finding and our observation that hepatic, but not intestinal, luciferase activity is increased by pregnancy in CYP3A4-promoter-luciferase transgenic mice, we propose that human pregnancy induces hepatic, but not intestinal, CYP3A activity. Furthermore, this research helps shed light on knowledge gaps that warrant further studies (see Conclusions, below). SMPK modeling. SMPK models are usually based on standard one- or two-compartment models, while incorporating both hepatic and/or intestinal metabolism. This approach uses compartmental PK parameters available from the literature or derived from clinical study data (39). Metabolic enzyme activity can be estimated from in vitro rate constants or from clinical data. In essence, the SMPK model is a minimal PBPK model that incorporates the key organs and systems involved in the ADME of drugs: intestine, liver, portal circulation, and systemic circulation. Other physiologic compartments are usually collapsed into the central and peripheral compartments. Such models are easier to implement using standard pharmacometric platforms compared with PBPK models. This approach has been successfully used in drug-drug interaction predictions of CYP3A substrates (39) and has also been adopted to incorporate certain physiological changes associated with pregnancy, such as changes in CYP3A enzyme activity, to predict the disposition of CYP3A substrates (40). The validation process for SMPK models is similar to that for PBPK models, i.e., demonstrating the adequacy of the model in nonpregnant adults and assessing the sensitivity to key model parameters of model predictions for multiple drugs for the pregnant population.

Clinical Trial Simulation Clinical trial simulation (CTS) uses Monte Carlo sampling methods to identify trial designs that are most likely to achieve trial goals (41). A CTS may consist of an input-output model and a trial execution model. The input-output model involves the stochastic simulation of PK profiles based on a prior or assumed PK model, along with a dosing schedule. Using a CTS, researchers can explore different trial designs to determine which is most likely to meet the study objectives. These design elements include placement of PK sampling times (when using a sparsesampling regimen) based on D-optimal design and determination of the required sample size according to defined or assumed between- and within-subject variance. A case study (7) used the example of enoxaparin, a low-molecular-weight heparin used for the prevention and treatment of thromboembolic conditions, to explore whether PK sampling would need to be modified during pregnancy. Because enoxaparin is eliminated primarily via the kidneys, physiological variables such as total body water and creatinine clearance were incorporated in a SMPK model. The authors concluded that the need for a changing sampling design over the course of pregnancy was not necessary for enoxaparin, presumably owing to the modest magnitude of PK change (60% increase in CL/F) observed during pregnancy (although this change could be substantial from a clinical perspective). www.annualreviews.org • Modeling and Simulation of PK/PD in Pregnancy

63

PA54CH04-Unadkat

ARI

3 December 2013

15:14

PHARMACODYNAMIC CONSIDERATIONS Pharmacodynamic (PD) modeling has been applied in the assessment of the commonly used oral hypoglycemic agent glyburide in the treatment of women with gestational diabetes mellitus (GDM). Normal pregnancy is characterized by insulin resistance with compensatory augmentation of insulin production, particularly in T3 . GDM is characterized by more severe insulin resistance and impaired β-cell compensation. Hebert et al. (9) observed a significant gestational augmentation of glyburide oral clearance compared with nonpregnant patients with type 2 diabetes mellitus. In addition, insulin sensitivity was estimated from glucose and insulin concentrations using the minimal model of glucose kinetics following a mixed-meal tolerance test. The authors found that although insulin secretion was significantly enhanced with glyburide treatment in GDM subjects, the effects of glyburide were still not enough to compensate for the degree of insulin resistance exhibited by the women with GDM, as demonstrated by the significantly lower β-cell response to glucose concentration corrected for the insulin sensitivity in these subjects. These results suggest that patients with inadequate glucose control might benefit from increased glyburide dosage. Other PD alterations including known changes in hemodynamics, cardiac output, and peripheral vascular resistance during pregnancy may have an impact on the PD response to antihypertensive drugs such as clonidine (42). Immunosuppression during pregnancy may impact the PD response to antibiotics (26), and disease activity changes in pregnancy may also occur. For example, improvement in rheumatoid arthritis, i.e., reduced disease activity, is noted in pregnancy (43). Because the efficacy of drugs will be influenced by alteration in the PK and the PD of drugs, the latter, when readily quantified, should be used as another component to guide dose adjustment in pregnancy.


PD: pharmacodynamic(s)

KEY KNOWLEDGE GAPS IDENTIFIED FOR PBPK MODELING OF MATERNAL-FETAL DRUG DISPOSITION To fully populate a PBPK model, further probe-drug studies to delineate the longitudinal changes of other CYP enzymes and UGT enzymes are needed. For example, owing to limitations of published CYP2D6 probe-drug studies, a PK study that uses dextromethorphan, the most sensitive CYP2D6 probe, and measures its plasma concentration-time profile should be conducted during various stages of pregnancy and in the postpartum period. Future studies with the CYP2B6 probedrug bupropion during various stages of pregnancy and postpartum are highly desirable to further refine the in vivo fold induction of CYP2B6 activity that has been predicted from in vitro data (38). In addition, no studies have evaluated the effect of pregnancy on CYP2C8 and CYP2E1 activity in vivo. PK studies of repaglinide (an antidiabetic) and chlorzoxazone (a muscle relaxant) could be conducted to assess the pregnancy effect. For the majority of the probe-drug studies conducted in pregnant women, data are obtained only in T3 . This issue stems from the limitation that some probe drugs cannot be safely administered to pregnant women during early gestation if they are not used for therapeutic purposes. Opportunistic PK studies in pregnant women are also likely restricted to the phase of pregnancy during which the patients are afflicted with the disease. In any case, extrapolating the magnitude of change in CYP enzyme activity from one trimester to other trimesters is difficult. An alternative approach would be to predict the magnitude of enzyme induction or suppression in vivo at gradually rising concentrations of specific hormones in plasma during each trimester, utilizing in vitro experiments in human hepatocytes (44). This approach has successfully predicted CYP3A induction in T3 , and, once validated, it can be expanded to study other CYP isoforms. Finally, the effect of pregnancy on certain physiological parameters, such as hepatic blood flow, should be further evaluated given the inconsistency in the data as described above. Currently, there is some knowledge of pregnancy effect on renal transporter organic cation transporter 2 and intestinal/renal transporter P-glycoprotein, based on metformin and digoxin 64

Ke et al.


PA54CH04-Unadkat

ARI

3 December 2013

15:14

data (14, 20). Future studies should address other important renal transporters such as organic anion transporters. Understanding the effect of pregnancy on the expression/activity of the hepatic transporters OATP1B1 and OATP1B3 (organic anion-transporting polypeptides 1B1 and 1B3) is particularly important when the uptake of a drug into hepatocytes is the rate-limiting step of its clearance. To date, established probe drugs for OATPs are statins, but they are contraindicated in pregnancy. In the future, newly identified probe drugs for OATPs that are not contraindicated in pregnancy can be evaluated in pregnant women. It is important to recognize that changes in the activity of uptake or efflux transporters expressed at the target sites, such as hepatocytes and proximal kidney tubules, can result in significant alteration in tissue distribution. Such alterations, known as changes in the local PK of drugs (i.e., changes that may not be accompanied by changes in the systemic PK of drugs), further leads to variation in pharmacological or toxicological response. In this regard, understanding the pregnancy effect on transporter activity/expression will be important in interpreting potential drug PD differences observed in pregnant women. The coupled maternal-fetal physiology imposes additional concerns about both efficacy and safety of treatment. Recommendation of dose adjustments during pregnancy should also be carefully evaluated in terms of assessing the risk of fetal exposure to drugs administered to the mother (10). Although it is important to determine fetal exposure to a drug administered to the pregnant mother, logistical and ethical reasons prevent the determination of such exposure except at the time of birth, when it is feasible to obtain a single cord-blood sample. Fetal exposure to drugs depends not only on maternal PK but also on placental passage of drugs. An increased knowledge base of gestational-age-dependent changes in placental transporter activity of drugs would be especially helpful in improving the utility of PBPK models for further exploration of fetal exposure in silico. In this regard, some progress has been made with respect to assessing gestational-age-dependent placental P-glycoprotein activity in nonhuman primates using positron emission tomography imaging (45). Alternatively, quantification of the expression of placental transporters by liquid chromatography–mass spectrometry (46), quantification of transporter activity (Vmax , Km ) and expression in cell lines, and knowledge of gestational-age-dependent changes in fetal physiological parameters would help in developing a comprehensive fetal model and linking it to the maternal PBPK model to better predict fetal exposure to drugs.

CONCLUSIONS We have discussed the specific challenges related to the design, execution, and analysis of clinical studies in pregnant women, underlining the unmet need for pharmacometric analysis approaches. We believe that with the established base of knowledge of the profound physiological changes during pregnancy and the significant advances in PBPK approaches that have been applied in special populations such as pediatric populations (47), the PBPK approach during pregnancy offers added advantages compared with other approaches. By integrating physiological data, preclinical data, and clinical data to quantify anticipated changes in the PK of drugs during pregnancy, this approach allows extrapolation beyond the previously studied model drugs to other drugs with wellcharacterized ADME characteristics. Conducting trials in silico before executing them in vivo can also be helpful in optimizing the design of a first-in-pregnancy PK study, including prioritizing the study period (T1 , T2 , and T3 ), the sample size, and the dose selection. Furthermore, the coupled maternal-fetal physiology imposes additional concerns about both efficacy and safety of treatment. To assess fetal exposure to drugs administered to the mother, the PBPK approach is considered to be the most appropriate methodology, and the established maternal PBPK model has the potential to be expanded in order to provide quantitative predictions for these events. Using such a systems pharmacology approach can potentially allow us to identify drugs whose maternal-fetal PK, and www.annualreviews.org • Modeling and Simulation of PK/PD in Pregnancy

65

PA54CH04-Unadkat

ARI

3 December 2013

15:14

therefore efficacy and toxicity for the mother and/or the fetus, may be affected by pregnancy. Ultimately, the information that is generated will support the design of rational dosing regimens for pregnant women and their unborn children.

DISCLOSURE STATEMENT A.B.K. was supported by the Office of Women’s Health, US Food and Drug Administration; J.D.U. was a member of the scientific advisory board of the modeling and simulation company Simcyp (now part of Certara); A.R.-H. devotes 50% of his time to Simcyp and consults for various drug companies.


ACKNOWLEDGMENTS The authors acknowledge funding from the US Food and Drug Administration’s Office of Women’s Health and a visiting fellowship from Simcyp Ltd. (to A.B.K.). LITERATURE CITED 1. Glover DD, Amonkar M, Rybeck BF, Tracy TS. 2003. Prescription, over-the-counter, and herbal medicine use in a rural, obstetric population. Am. J. Obstet. Gynecol. 188:1039–45 2. Andrade SE, Gurwitz JH, Davis RL, Chan KA, Finkelstein JA, et al. 2004. Prescription drug use in pregnancy. Am. J. Obstet. Gynecol. 191:398–407 3. Andrade SE, Raebel MA, Morse AN, Davis RL, Chan KA, et al. 2006. Use of prescription medications with a potential for fetal harm among pregnant women. Pharmacoepidemiol. Drug Saf. 15:546–54 4. Endicott S, Haas DM. 2012. The current state of therapeutic drug trials in pregnancy. Clin. Pharmacol. Ther. 92:149–50 5. Anger GJ, Piquette-Miller M. 2008. Pharmacokinetic studies in pregnant women. Clin. Pharmacol. Ther. 83:184–87 6. US Dep. Health Hum. Serv., Food Drug Admin., Cent. Drug Eval. Res. (CDER). 2004. Guidance for industry: Pharmacokinetics in pregnancy—study design, data analysis, and impact on dosing and labeling. US Food Drug Admin., Rockville, Md. http://www.fda.gov/downloads/Drugs/ GuidanceComplianceRegulatoryInformation/Guidances/ucm072133.pdf 7. van Hasselt JG, Andrew MA, Hebert MF, Tarning J, Vicini P, Mattison DR. 2012. The status of pharmacometrics in pregnancy: highlights from the 3rd American conference on pharmacometrics. Br. J. Clin. Pharmacol. 74:932–39 8. Unadkat JD, Wara DW, Hughes MD, Mathias AA, Holland DT, et al. 2007. Pharmacokinetics and safety of indinavir in human immunodeficiency virus–infected pregnant women. Antimicrob. Agents Chemother. 51:783–86 9. Hebert MF, Ma X, Naraharisetti SB, Krudys KM, Umans JG, et al. 2009. Are we optimizing gestational diabetes treatment with glyburide? The pharmacologic basis for better clinical practice. Clin. Pharmacol. Ther. 85:607–14 10. Etwel F, Hutson J, Madadi P, Gareri J, Koren G. 2014. Fetal and perinatal exposure to drugs and chemicals: novel biomarkers of risk. Annu. Rev. Pharmacol. Toxicol. 54:295–315 11. Panel Antiretrovir. Guidel. Adults Adolesc. 2011. Guidelines for the use of antiretroviral agents in HIV-1infected adults and adolescents. AIDSinfo, US Dep. Health Hum. Serv., Rockville, Md. http://www.aidsinfo. nih.gov/contentfiles/AdultandAdolescentGL.pdf 12. Mirochnick M, Clarke D. 2011. Oseltamivir pharmacokinetics in pregnancy: a commentary. Am. J. Obstet. Gynecol. 204:S94–95 13. Abduljalil K, Furness P, Johnson TN, Rostami-Hodjegan A, Soltani H. 2012. Anatomical, physiological and metabolic changes with gestational age during normal pregnancy: a database for parameters required in physiologically based pharmacokinetic modelling. Clin. Pharmacokinet. 51:365–96 66

Ke et al.


PA54CH04-Unadkat

ARI

3 December 2013

15:14

14. Hebert MF, Easterling TR, Kirby B, Carr DB, Buchanan ML, et al. 2008. Effects of pregnancy on CYP3A and P-glycoprotein activities as measured by disposition of midazolam and digoxin: a University of Washington specialized center of research study. Clin. Pharmacol. Ther. 84:248–53 15. Bryson YJ, Mirochnick M, Stek A, Mofenson LM, Connor J, et al. 2008. Pharmacokinetics and safety of nelfinavir when used in combination with zidovudine and lamivudine in HIV-infected pregnant women: Pediatric AIDS Clinical Trials Group (PACTG) Protocol 353. HIV Clin. Trials 9:115–25 16. Anderson GD. 2005. Pregnancy-induced changes in pharmacokinetics: a mechanistic-based approach. Clin. Pharmacokinet. 44:989–1008 17. Tomson T, Lindbom U, Ekqvist B, Sundqvist A. 1994. Epilepsy and pregnancy: a prospective study of seizure control in relation to free and total plasma concentrations of carbamazepine and phenytoin. Epilepsia 35:122–30 18. Wilkinson GR, Shand DG. 1975. A physiological approach to hepatic drug clearance. Clin. Pharmacol. Ther. 18:377–90 19. Fotopoulou C, Kretz R, Bauer S, Schefold JC, Schmitz B, et al. 2009. Prospectively assessed changes in lamotrigine-concentration in women with epilepsy during pregnancy, lactation and the neonatal period. Epilepsy Res. 85:60–64 20. Eyal S, Easterling TR, Carr D, Umans JG, Miodovnik M, et al. 2010. Pharmacokinetics of metformin during pregnancy. Drug Metab. Dispos. 38:833–40 21. Zamek-Gliszczynski MJ, Kalvass JC, Pollack GM, Brouwer KL. 2009. Relationship between drug/ metabolite exposure and impairment of excretory transport function. Drug Metab. Dispos. 37:386– 90 22. Andrew MA, Easterling TR, Carr DB, Shen D, Buchanan ML, et al. 2007. Amoxicillin pharmacokinetics in pregnant women: modeling and simulations of dosage strategies. Clin. Pharmacol. Ther. 81:547–56 23. Salman S, Rogerson SJ, Kose K, Griffin S, Gomorai S, et al. 2010. Pharmacokinetic properties of azithromycin in pregnancy. Antimicrob. Agents Chemother. 54:360–66 24. Mirochnick M, Best BM, Clarke DF. 2010. Antiretroviral pharmacology: special issues regarding pregnant women and neonates. Clin. Perinatol. 37:907–27, xi 25. Johnson TN, Kerbusch T, Jones B, Tucker GT, Rostami-Hodjegan A, Milligan PA. 2009. Assessing the efficiency of mixed effects modelling in quantifying metabolism based drug-drug interactions: using in vitro data as an aid to assess study power. Pharm. Stat. 8:186–202 26. Fischer JH, Sarto GE, Habibi M, Kilpatrick SJ, Tuomala RE, et al. 2012. Influence of body weight, ethnicity, oral contraceptives, and pregnancy on the pharmacokinetics of azithromycin in women of childbearing age. Antimicrob. Agents Chemother. 56:715–24 27. Mould DR, Upton RN. 2013. Basic concepts in population modeling, simulation, and model-based drug development—Part 2: Introduction to pharmacokinetic modeling methods. CPT Pharmacomet. Syst. Pharmacol. 2:e38 28. Benaboud S, Hirt D, Launay O, Pannier E, Firtion G, et al. 2012. Pregnancy-related effects on tenofovir pharmacokinetics: a population study with 186 women. Antimicrob. Agents Chemother. 56:857–62 29. McGready R, White NJ, Nosten F. 2011. Parasitological efficacy of antimalarials in the treatment and prevention of falciparum malaria in pregnancy 1998 to 2009: a systematic review. Br. J. Obstet. Gynaecol. 118:123–35 30. Rowland M, Peck C, Tucker G. 2011. Physiologically-based pharmacokinetics in drug development and regulatory science. Annu. Rev. Pharmacol. Toxicol. 51:45–73 31. Jamei M, Dickinson GL, Rostami-Hodjegan A. 2009. A framework for assessing inter-individual variability in pharmacokinetics using virtual human populations and integrating general knowledge of physical chemistry, biology, anatomy, physiology and genetics: a tale of ‘bottom-up’ vs. ‘top-down’ recognition of covariates. Drug Metab. Pharmacokinet. 24:53–75 32. Zhao P, Zhang L, Grillo JA, Liu Q, Bullock JM, et al. 2011. Applications of physiologically based pharmacokinetic (PBPK) modeling and simulation during regulatory review. Clin. Pharmacol. Ther. 89:259–67 33. Lu G, Abduljalil K, Jamei M, Johnson TN, Soltani H, Rostami-Hodjegan A. 2012. Physiologically-based pharmacokinetic (PBPK) models for assessing the kinetics of xenobiotics during pregnancy: achievements and shortcomings. Curr. Drug Metab. 13:695–720 www.annualreviews.org • Modeling and Simulation of PK/PD in Pregnancy

67

ARI

3 December 2013

15:14

34. Andrew MA, Hebert MF, Vicini P. 2008. Physiologically based pharmacokinetic model of midazolam disposition during pregnancy. Conf. Proc. IEEE Eng. Med. Biol. Soc. 2008:5454–57 35. Lu G, Abduljalil K, Jamei M, Johnson TN, Rostami-Hodjegan A. 2012. A pregnancy physiologically based pharmacokinetic (p-PBPK) model for disposition of drugs metabolized by CYP1A2, CYP2D6, and CYP3A4. Br. J. Clin. Pharmacol. 74:873–85 36. Ke AB, Nallani SC, Zhao P, Rostami-Hodjegan A, Unadkat JD. 2012. A PBPK model to predict disposition of CYP3A-metabolized drugs in pregnant women: verification and discerning the site of CYP3A induction. CPT Pharmacomet. Syst. Pharmacol. 1:e3. doi: 10.1038/psp.2012.2 37. Ke AB, Nallani S, Zhao P, Rostami-Hodjegan A, Isoherranen N, Unadkat JD. 2013. A PBPK model to predict disposition of P450 2D6 and P450 1A2 metabolized drugs in pregnant women. Drug Metab. Dispos. 41(4):801–13 38. Ke AB, Nallani SC, Zhao P, Rostami-Hodjegan A, Unadkat JD. 2013. Expansion of a PBPK model to predict disposition in pregnant women of drugs cleared via multiple CYP enzymes, including CYP2B6, CYP2C9 and CYP2C19. Br. J. Clin. Pharmacol. In press. doi: 10.1111/bcp.12207 39. Quinney SK, Zhang X, Lucksiri A, Gorski JC, Li L, Hall SD. 2010. Physiologically based pharmacokinetic model of mechanism-based inhibition of CYP3A by clarithromycin. Drug Metab. Dispos. 38:241–48 40. Quinney SK, Mohamed AN, MF H, Haas DM, Clark S, et al. 2012. A semi-mechanistic metabolism model of CYP3A substrates in pregnancy: predicting changes in midazolam and nifedipine pharmacokinetics. CPT Pharmacomet. Syst. Pharmacol. 1:e2. doi: 10.1038/psp.2012.5 41. Holford NHG, Kimko HC, Monteleone JPR, Peck CC. 2000. Simulation of clinical trials. Annu. Rev. Pharmacol. Toxicol. 40:209–34 42. Rothberger S, Carr D, Brateng D, Hebert M, Easterling TR. 2010. Pharmacodynamics of clonidine therapy in pregnancy: a heterogeneous maternal response impacts fetal growth. Am. J. Hypertens. 23: 1234–40 43. Østensen M, Villiger PM, Forger F. 2012. Interaction of pregnancy and autoimmune rheumatic disease. ¨ Autoimmun. Rev. 11:A437–46 44. Papageorgiou I, Grepper S, Unadkat JD. 2013. Induction of hepatic CYP3A enzymes by pregnancyrelated hormones: studies in human hepatocytes and hepatic cell lines. Drug Metab. Dispos. 41:281– 90 45. Ke AB, Eyal S, Chung FS, Link JM, Mankoff DA, et al. 2013. Modeling cyclosporine A inhibition of the distribution of a P-glycoprotein PET ligand, 11 C-verapamil, into the maternal brain and fetal liver of the pregnant nonhuman primate: impact of tissue blood flow and site of inhibition. J. Nucl. Med. 54:437– 46 46. Prasad B, Lai Y, Lin Y, Unadkat JD. 2013. Interindividual variability in the hepatic expression of the human breast cancer resistance protein (BCRP/ABCG2): effect of age, sex, and genotype. J. Pharm. Sci. 102:787–93 47. Leong R, Vieira MLT, Zhao P, Mulugeta Y, Lee CS, et al. 2012. Regulatory experience with physiologically based pharmacokinetic modeling for pediatric drug trials. Clin. Pharmacol. Ther. 91:926–31 48. Tracy TS, Venkataramanan R, Glover DD, Caritis SN. 2005. Temporal changes in drug metabolism (CYP1A2, CYP2D6 and CYP3A activity) during pregnancy. Am. J. Obstet. Gynecol. 192:633–39 49. Gardner MJ, Schatz M, Cousins L, Zeiger R, Middleton E, Jusko WJ. 1987. Longitudinal effects of pregnancy on the pharmacokinetics of theophylline. Eur. J. Clin. Pharmacol. 32:289–95 50. Dempsey D, Jacob P III, Benowitz NL. 2002. Accelerated metabolism of nicotine and cotinine in pregnant smokers. J. Pharmacol. Exp. Ther. 301:594–98 51. Tomson T, Lindbom U, Ekqvist B, Sundqvist A. 1994. Disposition of carbamazepine and phenytoin in pregnancy. Epilepsia 35:131–35 52. McGready R, Stepniewska K, Edstein MD, Cho T, Gilveray G, et al. 2003. The pharmacokinetics of atovaquone and proguanil in pregnant women with acute falciparum malaria. Eur. J. Clin. Pharmacol. 59:545–52 53. Hogstedt S, Lindberg B, Peng DR, Rega˚rdh CG, Rane A. 1985. Pregnancy-induced increase in ¨ metoprolol metabolism. Clin. Pharmacol. Ther. 37:688–92 54. Pond SM, Kreek MJ, Tong TG, Raghunath J, Benowitz NL. 1985. Altered methadone pharmacokinetics in methadone-maintained pregnant women. J. Pharmacol. Exp. Ther. 233:1–6


PA54CH04-Unadkat

68

Ke et al.

PA54CH04-Unadkat

ARI

3 December 2013

15:14


55. Hardman J, Endres L, Fischer P, Fischer J. 2005. Pharmacokinetics of labetalol in pregnancy. Pharmacotherapy 25:1493 R (zidovudine) Tablets, Capsules, and Syrup [package insert]. Research Triangle Park, 56. Retrovir NC: ViiV Healthcare; 2011. http://www.accessdata.fda.gov/drugsatfda_docs/label/2011/ 019655s052,019910s039,020518s022lbl.pdf 57. Gerdin E, Salmonson T, Lindberg B, Rane A. 1990. Maternal kinetics of morphine during labour. J. Perinat. Med. 18:479–87

www.annualreviews.org • Modeling and Simulation of PK/PD in Pregnancy

69

PA54-FrontMatter

ARI

12 December 2013

10:22

Annual Review of Pharmacology and Toxicology


Contents

Volume 54, 2014

Learning to Program the Liver Curtis D. Klaassen p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1 The Druggable Genome: Evaluation of Drug Targets in Clinical Trials Suggests Major Shifts in Molecular Class and Indication Mathias Rask-Andersen, Surendar Masuram, and Helgi B. Schiöth p p p p p p p p p p p p p p p p p p p p p p p 9 Engineered Botulinum Neurotoxins as New Therapeutics Geoffrey Masuyer, John A. Chaddock, Keith A. Foster, and K. Ravi Acharya p p p p p p p p p p p p27 Pharmacometrics in Pregnancy: An Unmet Need Alice Ban Ke, Amin Rostami-Hodjegan, Ping Zhao, and Jashvant D. Unadkat p p p p p p p p p53 Antiparasitic Chemotherapy: From Genomes to Mechanisms David Horn and Manoj T. Duraisingh p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p71 Targeting Multidrug Resistance Protein 1 (MRP1, ABCC1): Past, Present, and Future Susan P.C. Cole p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p95 Glutamate Receptor Antagonists as Fast-Acting Therapeutic Alternatives for the Treatment of Depression: Ketamine and Other Compounds Mark J. Niciu, Ioline D. Henter, David A. Luckenbaugh, Carlos A. Zarate Jr., and Dennis S. Charney p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 119 Environmental Toxins and Parkinson’s Disease Samuel M. Goldman p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 141 Drugs for Allosteric Sites on Receptors Cody J. Wenthur, Patrick R. Gentry, Thomas P. Mathews, and Craig W. Lindsley p p 165 microRNA Therapeutics in Cardiovascular Disease Models Seema Dangwal and Thomas Thum p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 185 Nanocarriers for Vascular Delivery of Anti-Inflammatory Agents Melissa D. Howard, Elizabeth D. Hood, Blaine Zern, Vladimir V. Shuvaev, Tilo Grosser, and Vladimir R. Muzykantov p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 205

v

PA54-FrontMatter

ARI

12 December 2013

10:22

G Protein–Coupled Receptors Revisited: Therapeutic Applications Inspired by Synthetic Biology Boon Chin Heng, Dominique Aubel, and Martin Fussenegger p p p p p p p p p p p p p p p p p p p p p p p p p p 227 Cause and Consequence of Cancer/Testis Antigen Activation in Cancer Angelique W. Whitehurst p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 251 Targeting PCSK9 for Hypercholesterolemia Giuseppe Danilo Norata, Gianpaolo Tibolla, and Alberico Luigi Catapano p p p p p p p p p p p p p 273


Fetal and Perinatal Exposure to Drugs and Chemicals: Novel Biomarkers of Risk Fatma Etwel, Janine R. Hutson, Parvaz Madadi, Joey Gareri, and Gideon Koren p p p p 295 Sodium Channels, Inherited Epilepsy, and Antiepileptic Drugs William A. Catterall p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 317 Chronopharmacology: New Insights and Therapeutic Implications Robert Dallmann, Steven A. Brown, and Frédéric Gachon p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 339 Small-Molecule Allosteric Activators of Sirtuins David A. Sinclair and Leonard Guarente p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 363 Emerging Therapeutics for Alzheimer’s Disease Karen Chiang and Edward H. Koo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 381 Free Fatty Acid (FFA) and Hydroxy Carboxylic Acid (HCA) Receptors Stefan Offermanns p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 407 Targeting Protein-Protein Interaction by Small Molecules Lingyan Jin, Weiru Wang, and Guowei Fang p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 435 Systems Approach to Neurodegenerative Disease Biomarker Discovery Christopher Lausted, Inyoul Lee, Yong Zhou, Shizhen Qin, Jaeyun Sung, Nathan D. Price, Leroy Hood, and Kai Wang p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 457 GABAA Receptor Subtypes: Therapeutic Potential in Down Syndrome, Affective Disorders, Schizophrenia, and Autism Uwe Rudolph and Hanns Möhler p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 483 Role of Hepatic Efflux Transporters in Regulating Systemic and Hepatocyte Exposure to Xenobiotics Nathan D. Pfeifer, Rhiannon N. Hardwick, and Kim L.R. Brouwer p p p p p p p p p p p p p p p p p p p 509 Turning Off AKT: PHLPP as a Drug Target Alexandra C. Newton and Lloyd C. Trotman p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 537 Understanding and Modulating Mammalian-Microbial Communication for Improved Human Health Sridhar Mani, Urs A. Boelsterli, and Matthew R. Redinbo p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 559

vi

Contents

PA54-FrontMatter

ARI

12 December 2013

10:22

Pharmaceutical and Toxicological Properties of Engineered Nanomaterials for Drug Delivery Matthew Palombo, Manjeet Deshmukh, Daniel Myers, Jieming Gao, Zoltan Szekely, and Patrick J. Sinko p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 581 Indexes Cumulative Index of Contributing Authors, Volumes 50–54 p p p p p p p p p p p p p p p p p p p p p p p p p p p 599


Cumulative Index of Article Titles, Volumes 50–54 p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 602 Errata An online log of corrections to Annual Review of Pharmacology and Toxicology articles may be found at http://www.annualreviews.org/errata/pharmtox

Contents

vii

IV iron in pregnancy: an unmet clinical need.

'Toxgnostics': an unmet need in cancer medicine.

Clinical perspective - cough: an unmet need.

Biomarkers and Microscopic Colitis: An Unmet Need in Clinical Practice.

Access to care--an unmet need in headache management?

Assessing cardiovascular risk in hepatitis C: An unmet need.

An unmet need: tailoring extended adjuvant endocrine therapy.

Unmet clinical need in autoimmune liver diseases.

Global Neurosurgery: The Unmet Need.

Discussing Sexual Health after Traumatic Brain Injury: an Unmet Need!

An unmet need: Feeding for critically ill vegans.

Unmet need, intention to use contraceptives and unwanted pregnancy in rural bangladesh.

Unmet need for contraception and its association with unintended pregnancy in Bangladesh.

Reducing unmet need by supporting women with met need.

Understanding unmet need: history, theory, and measurement.

Unmet need for contraception: issues and challenges.

An assessment of residents' and fellows' personal finance literacy: an unmet medical education need.

Assessing unmet medical need in India: A regulatory riddle?

[Unmet need in asthma management and future treatment options].

Sustained Remission: An Unmet Need in Patients with Giant-cell Arteritis.

Impact of venous thromboembolism on survival in patients with advanced cancer: an unmet clinical need.

Management of pulmonary exacerbations in cystic fibrosis: still an unmet medical need in clinical practice.

[Unmet need in asthma management and future treatment options].

Fall-risk prediction in older adults with cancer: an unmet need.