The Impact of Pregnancy on Antihypertensive Drug Metabolism and Pharmacokinetics: Current Status and Future Directions

Abstract

Introduction:

Hypertensive disorders of pregnancy (HDP) are rising in prevalence, and increase risk of adverse maternal and fetal outcomes. Physiologic changes occur during pregnancy that alter drug pharmacokinetics. However, antihypertensive drugs lack pregnancy-specific dosing recommendations due to critical knowledge gaps surrounding the extent of gestational changes in antihypertensive drug pharmacokinetics and underlying mechanisms.

Areas covered:

This review (1) summarizes currently recommended medications and dosing strategies for non-emergent HDP treatment, (2) reviews and synthesizes existing literature identified via a comprehensive Pubmed search evaluating gestational changes in the maternal pharmacokinetics of commonly prescribed HDP drugs (notably labetalol and nifedipine), and (3) offers insight into the metabolism and clearance mechanisms underlying altered HDP drug pharmacokinetics during pregnancy. Remaining knowledge gaps and future research directions are summarized.

Expert Opinion:

A series of small pharmacokinetic studies illustrate higher oral clearance of labetalol and nifedipine during pregnancy. Pharmacokinetic modeling and preclinical studies suggest these effects are likely due to pregnancy-associated increases in hepatic UGT1A1- and CYP3A4-mediated first-pass metabolism and lower bioavailability. Accordingly, higher and/or more frequent doses may be needed to lower blood pressure during pregnancy. Future research is needed to address various evidence gaps and inform the development of more precise antihypertensive drug dosing strategies.

1. Introduction

Although physiologic, biochemical, and molecular changes occur during pregnancy that can alter the pharmacokinetics (PK) and pharmacodynamics (PD) of medications, most drugs used in obstetric patients lack dosing recommendations specific to pregnancy. The traditional exclusion of pregnant and post-partum individuals from clinical trials has resulted in a dearth of evidence informing the safety, efficacy, and optimal dosing of most medications prescribed to obstetric patients [1,2]. Despite efforts to improve inclusion of pregnancy information in pharmaceutical labeling, only 5% of over 200 new pharmaceuticals approved from 2003 to 2012 provided some human data in the pregnancy section of the label [3]. The lack of available data in pregnant individuals often leads to off-label prescribing and trial-and-error dosing, which may increase the risk of adverse medication outcomes during pregnancy [4].

Hypertensive disorders of pregnancy (HDP) are among the most common medical conditions requiring treatment during pregnancy and illustrate the clinical importance of having pregnancy-specific prescribing recommendations [5,6]. HDP are rising in prevalence and currently affect over 10% of pregnancies in the U.S. [7], which is due in part to higher average maternal age coupled with the rising prevalence of pre-existing risk factors such as chronic hypertension, obesity, and diabetes [8,9]. Maternal hypertension is a major contributor to maternal morbidity and mortality and fetal complications of prematurity, and blood pressure (BP) lowering medications are frequently prescribed to reduce the risk of maternal complications such as heart failure and stroke and prolong the pregnancy [10,11]. However, there is a lack of consensus recommendations for appropriate selection and dosing of antihypertensive therapy [12].

Accumulating evidence derived from drug metabolism, clinical PK, and PK modeling studies has improved our understanding of the presence and mechanisms underlying alterations in drug disposition during pregnancy, which has created opportunities for more precise medication dosing in this special population [13,14]. However, substantial gaps in knowledge persist surrounding the extent and mechanisms underlying gestational changes in the metabolism and PK of antihypertensive medications commonly prescribed during pregnancy [13]. In this review, we 1) summarize currently recommended antihypertensive medications and dosing strategies for the treatment of HDP, 2) comprehensively review and synthesize existing literature evaluating the presence and extent of gestational changes in the maternal PK of commonly prescribed HDP drugs (most notably labetalol and nifedipine), and 3) offer insight into the metabolism and clearance mechanisms underlying altered HDP drug PK during pregnancy. We also identify remaining knowledge gaps surrounding pregnancy-related changes in antihypertensive drug disposition and future research directions to inform more precise antihypertensive drug selection and dosing strategies during pregnancy.

The following databases were used for the literature search: PubMed, PubMed Central, and general internet search engines. We conducted the search from January to April 2021. Articles were identified without constraints on publication date. Additional articles were identified using the reference lists in the identified articles.

2. Hypertensive Disorders of Pregnancy

According to the American College of Obstetricians and Gynecologists (ACOG), HDP are classified into four distinct categories: gestational hypertension, preeclampsia, pre-existing chronic hypertension, and chronic hypertension with superimposed preeclampsia [9]. In all categories, hypertension is defined as a BP ≥140/90 mmHg. However, diagnosis of HDP is complex and criteria vary across organizations and countries. Diagnostic criteria according to ACOG are summarized in Table 1 [9,15].

Table 1 –

Summary of Diagnostic Criteria for Distinct Categories of Hypertensive Disorders of Pregnancy

Hypertensive Disorder of Pregnancy	American College of Obstetricians and Gynecologists (ACOG) Diagnostic Criteria	References
Gestational Hypertension	SBP ≥140 mmHg or DBP ≥90 mmHg, or both, on ≥2 occasions at least 4 hours apart occurring after 20 weeks of gestation in a previously normotensive individual (without proteinuria or severe features)	[15]
Preeclampsia	SBP ≥140 mmHg or DBP ≥90 mmHg, or both, on ≥2 occasions at least 4 hours apart occurring after 20 weeks of gestation in a previously normotensive individual AND Proteinuria (defined as ≥300 mg/24-hour urine collection, protein/creatinine ratio of ≥0.3, or dipstick reading of 2+ if other quantitative methods are not available) OR In the absence of proteinuria, new-onset hypertension accompanied by any of the following severe features: severe-range hypertension (SBP ≥160 mmHg or DBP ≥110 mmHg, or both), thrombocytopenia, acute renal insufficiency without pre-existing renal disease, impaired liver function, pulmonary edema, new-onset headache refractory to treatment and not attributable to other causes	[15]
Chronic Hypertension	SBP ≥140 mmHg or DBP ≥90 mmHg, or both, on ≥2 occasions at least 4 hours apart that precedes pregnancy or is diagnosed before 20 weeks of gestation (without proteinuria or severe features), or hypertension that is first diagnosed during pregnancy without resolution during the post-partum period	[9]
Chronic Hypertension with Superimposed Preeclampsia	Preeclampsia (see above) in an individual with a diagnosis of hypertension occurring pre-pregnancy or prior to 20 weeks of gestation	[9]

Abbreviations: SBP=systolic blood pressure; DBP=diastolic blood pressure; mmHg=millimeters of mercury; mg=milligrams

Gestational hypertension is the most common HDP in the U.S., affecting between 6–14% of all pregnancies depending on the population studied and diagnostic threshold applied [16,17]. ACOG defines gestational hypertension as hypertension without proteinuria or severe features that develops after 20 weeks of gestation in an individual with a previously normal BP and resolves in the postpartum period [15]. Although not completely understood, the underlying pathophysiology of gestational hypertension may be related to reduced uteroplacental perfusion leading to increases in vascular responsiveness to angiotensin II, coupled with decreases in vasodilator biosynthesis [18]. Approximately 25% of patients with gestational hypertension will progress to preeclampsia, a complex medical disorder defined as hypertension (>140/90 mmHg) after 20 weeks of gestation accompanied by proteinuria or evidence of end-organ dysfunction [15,19]. The pathophysiologic transition from gestational hypertension to preeclampsia, and the superposition of preeclampsia on chronic hypertension, also remain poorly understood; however, the pathogenesis appears to be linked to sustained reductions in placental perfusion and resultant maternal vascular endothelial dysfunction [20]. Individuals that develop gestational hypertension or preeclampsia are at higher risk of developing chronic hypertension and cardiovascular disease later in life [15,21]. Preeclampsia can include severe features such as BP ≥160/110 mmHg, thrombocytopenia, impaired liver function, renal insufficiency, pulmonary edema, and/or visual disturbances, which are associated with higher risk of adverse outcomes such as preterm delivery, infants born small for their gestational age (SGA), and placental abruption [15,22]. Mortality due to preeclampsia is less common in high-income compared to low-income countries; however, preeclampsia prevalence continues to rise in the U.S. [15].

Chronic hypertension during pregnancy, which is defined as hypertension that precedes pregnancy or is diagnosed before 20 weeks of gestation, is prevalent in approximately 4–14% of pregnancies depending on the population studied and diagnostic threshold applied [17]. The underlying etiology of chronic hypertension during pregnancy is multifactorial, and is attributed to primary (or essential) hypertension in approximately 85–90% of cases and secondary to underlying renal, endocrine, or vascular conditions in approximately 10–15% of cases [9]. Although absolute risks for morbidity and mortality remain low overall, chronic hypertension is associated with 5–6-fold higher risk of cerebrovascular events, pulmonary edema and kidney failure, 10–20-fold higher risk of placental abruption, and 2–4-fold higher risk of fetal mortality compared with normotensive pregnancies [9,23]. Furthermore, the risk of adverse fetal outcomes such as fetal growth restriction and preterm birth has been correlated with the severity of maternal hypertension.

Chronic hypertension with superimposed preeclampsia is chronic hypertension during pregnancy that is subsequently complicated by features of preeclampsia, including new-onset proteinuria or other end-organ dysfunction. It is estimated that 50% of all pregnant women with chronic hypertension will develop superimposed preeclampsia and risk increases in patients with a DBP of 100–110 mmHg, those with greater than a 4-year pre-pregnancy history of hypertension, or preexisting end-organ disease [9,17,24]. Superimposed pre-eclampsia is associated with higher rates of adverse fetal outcomes, including approximately 10-fold higher rates of fetal growth restriction compared with the general population [25,26] and a 3.6-fold higher perinatal mortality risk compared to pregnant individuals with chronic hypertension alone [27].

2.1. Indication for Antihypertensive Drug Treatment during Pregnancy

Antihypertensive drugs are clinically indicated during pregnancy for the treatment of severe acute elevations in BP (SBP ≥160 or DBP ≥110), and for the treatment of chronic hypertension [9,21,28,29]. In pregnant individuals with severe acute-onset hypertension, guidelines recommend urgent treatment with intravenous (IV) labetalol, hydralazine, or immediate-release nifedipine within 30–60 minutes of diagnosis to lower BP and prevent maternal morbidity [19,21,29]. However, the PK and dosing of acutely administered IV antihypertensive medications are beyond the scope of this review.

Regarding the treatment of chronic hypertension during pregnancy, there remain limited data surrounding specific BP treatment goals and clinical guideline recommendations vary across organizations and countries [12,28]. Recent ACOG practice bulletins summarize treatment recommendations for chronic hypertension [9] as well as gestational hypertension and preeclampsia [15]. Initiation of pharmacologic treatment is recommended for severe hypertension (SBP ≥160 or DBP ≥110), and continuing previously prescribed oral antihypertensives in preexisting chronic hypertension patients who become pregnant is reasonable, provided use of the agent is not contraindicated during pregnancy [9,15]. Although treatment may be indicated at lower BP thresholds among HDP patients at high risk to develop cardiovascular complications (e.g. diabetes, renal dysfunction), further research into the benefits and risks are needed before recommending treatment at non-pregnant adult BP thresholds [9,30]. Moreover, there is limited evidence to support initiation of oral antihypertensive therapy in patients who become hypertensive before 20 weeks gestation; therefore, treatment decisions in such cases warrant clinical judgement on a per-patient basis [9,30].

The ACOG guidelines recommend management of gestational hypertension and preeclampsia without severe features similarly to the management of chronic hypertension in pregnancy [15]. However, there is no evidence to demonstrate that lowering BP with antihypertensive medications prevents progression to preeclampsia. A 2018 Cochrane review suggested that, with moderate-certainty, antihypertensive use in pregnant women with mild-to-moderate hypertension (defined as SBP 140–169 mmHg and/or DBP 90–109 mmHg) may reduce the risk of developing severe hypertension by as much as 50% without any concurrent reduction in risk for development of preeclampsia, preterm birth, SGA, or fetal death [31].

Use of antihypertensive medications to target more precise BP treatment goals remains controversial and limited evidence supports the clinical utility of more intensive treatment strategies. Although maternal benefits may be derived from stricter BP control, this could be offset by fetal risk such as placental hypoperfusion and/or adverse effects of increased fetal drug exposure [6]. Consequently, the ACOG recommends treatment goals of SBP 120–159 mmHg and DBP 80–109 mmHg [9,15]. The Control of Hypertension in Pregnancy Study (CHIPS) was an international randomized control trial that compared a “less tight control” (target DBP <100 mmHg) strategy to a “tight control” (DBP <85 mmHg) strategy in 987 individuals at 14–33 weeks gestation with chronic or gestational (75%) or non-proteinuric hypertension (25%) [32]. Although the tight control strategy significantly decreased the progression to severe hypertension compared to less tight control (27.5% versus 40.6%), no difference in the primary outcome of pregnancy loss or the need for high-level neonatal care for ≥48 hours was observed (30.7% versus 31.4%). In the absence of clear evidence demonstrating the clinical utility of using antihypertensive medications to achieve lower BP treatment goals, the decision to initiate antihypertensive medications during pregnancy will likely continue to occur per provider discretion.

2.2. Selection and Dosing of Oral Antihypertensive Drugs during Pregnancy

For the chronic treatment of moderately elevated BP during pregnancy (BP 140/90 to 160/110 mmHg), if treatment is indicated, both ACOG and ACC/AHA recommend oral labetalol (a combined alpha-1 and non-selective beta-adrenergic receptor antagonist), slow-release nifedipine (a dihydropyridine calcium channel blocker), or methyldopa (an alpha-2 adrenergic receptor agonist) as preferred initial agents [9,21]. Due to the lack of clinical trial evidence comparing the efficacy and safety of antihypertensive drugs during pregnancy, specific recommendations and algorithms to guide the selection and dosing of particular medications are not available. A summary of first-line and second- or third-line agents recommended by ACOG is provided in Table 2 [9,33–41]. Of note, angiotensin-converting enzyme inhibitors and angiotensin II receptor blockers are contraindicated during pregnancy due to their association with fetal renal failure, intrauterine growth restriction, congenital malformations, and mortality [12,34].

Table 2 –

Summary of Oral Antihypertensive Agents Prescribed during Pregnancy

Place in Therapy	Drug	Mechanism of Action	Suggested Dosing*	Old FDA Pregnancy Category	New FDA Pregnancy Category	References
First Line Agents *	Labetalol	Combined alpha-1 antagonist and beta blocker	Initial dosing: 100–300mg twice daily Maintenance dosing: 200–2,400 mg daily in two to three divided doses	C	No well-controlled studies in pregnant women. Use during pregnancy should only occur if the potential benefit justifies the potential risk.	[9,33,34]
	Nifedipine (slow release)	Dihydropyridine calcium channel blocker	Initial dosing: 30–60 mg daily Maintenance dosing: 30–120 mg daily	C		[9,34,35]
	Methyldopa	Central alpha-2 agonist	Initial dosing: 250 mg twice-three times daily Maintenance dosing: 500–3,000 mg daily in two to four divided doses	B		[9,34,36]
Second or Third Line Agents	Hydrochlorothiazide	Thiazide diuretic	Maintenance dosing: 12.5–50 mg daily Note: Limited dosing data available in pregnancy	B		[34,37]
	Hydralazine	Vasodilator	Initial dosing: 10 mg three to four times daily Maximum dose: 300 mg daily Note: Limited dosing data available in pregnancy	C		[34,38]
	Clonidine	Central alpha-2 agonist	Maintenance dosing: 0.15–0.30 mg daily in divided doses Note: Limited dosing data available in pregnancy	C		[34,39]
	Amlodipine	Dihydropyridine calcium channel blocker	Note: Limited dosing data available in pregnancy	C		[34,40]
	Atenolol	Beta blocker	Use is no longer advised in pregnancy	D	Can cause fetal harm; associated with fetal growth restriction and bradycardia. Administration in second trimester associated with infants born small for their gestational age. Use during pregnancy should only occur if the potential benefit justifies the potential risk.	[34,41]

^*Suggested dosing adapted from ACOG Practice Bulletin #203 [9] and Kaye et al. [34]

A recent feasibility clinical trial in chronic hypertension in pregnancy demonstrated that oral labetalol and nifedipine exhibited similar BP lowering effects and were well-tolerated [42]. Although methyldopa is still commonly prescribed, it is generally considered to be less effective and have a less favorable safety profile [9]. These findings are consistent with a 2019 open-label randomized trial comparing the safety and efficacy of oral labetalol, nifedipine, and methyldopa for the management of severe hypertension in pregnancy, which found that nifedipine achieved BP control (defined as SBP 120–150 mmHg and DBP 70–100 mmHg within six hours with no adverse outcomes) significantly more often compared to methyldopa; no significant differences were observed between nifedipine versus labetalol or labetalol versus methyldopa [43]. A Cochrane review of 22 trials demonstrated that methyldopa was less effective in preventing severe hypertension compared to labetalol or calcium channel blockers (CCBs); labetalol and CCBs also decreased the risk of developing preeclampsia compared with methyldopa [31].

Given that antihypertensive prescribing practices during pregnancy vary per clinician and are patient-specific, data from clinical trials and registries provide insight into the most commonly prescribed agents in the maintenance treatment of HDP. In the CHIPS trial, participants could remain on any previously prescribed agent for the duration of the study (other than atenolol); however, labetalol was recommended as the preferred initial agent following randomization unless contraindicated. In both the less tight control and tight control groups, labetalol or methyldopa monotherapy were used most frequently at baseline (labetalol: 20% vs. 23%; methyldopa: 25% vs. 24%) and following randomization (labetalol: 25% vs. 33%; methyldopa: 14% vs. 20%). Following randomization, nifedipine was occasionally combined with labetalol (10% vs. 13%) and with methyldopa (3% vs. 4%) [44]. Dosing was per provider discretion and not reported. These clinical trial results are consistent with HDP registries [45–47], including a retrospective cohort analysis of 2000–2007 Medicaid utilization claims that examined outpatient antihypertensive medication use patterns throughout pregnancy [45]. During the third trimester (T3), central alpha-2 receptor agonists (e.g., methyldopa; 28%), combined alpha- and beta-receptor antagonists (e.g., labetalol; 12%), beta-blockers (e.g., atenolol; 11%), and dihydropyridine CCBs (e.g, nifedipine; 51%) were most commonly prescribed. Notably, most patients prescribed a CCB (86%) were diagnosed with preterm labor, likely indicating use as a tocolytic rather than a BP lowering agent [45]. Although these studies provide important insight into patterns of antihypertensive drug selection during pregnancy, the range of doses prescribed per-provider throughout gestation was not reported. Thus, it remains unknown whether dose requirements needed to control BP are altered during pregnancy.

3. Gestational Changes in Drug Disposition and Pharmacokinetics

Extensive physiologic, biochemical, and molecular changes occur during pregnancy that can alter maternal drug disposition and PK through various mechanisms, which have been reviewed and presented in detail previously [48–51]. These changes may significantly increase or decrease systemic exposure of certain drugs during pregnancy via alterations in bioavailability (F), volume of distribution (Vd), and/or end-organ mediated clearance (CL).

Pregnancy-associated reductions in gastrointestinal motility can delay gastric emptying [52] and diminish gastric secretions, and increased gastric pH can alter drug absorption processes [49,51,53]. Pregnancy-related changes in plasma volume and blood flow, body weight and fat composition, and plasma protein concentrations can alter drug Vd [50,52]. For instance, expansion in total body water can increase the apparent Vd of hydrophilic drugs, while elevations in body fat can increase Vd of lipophilic drugs [54]. Increases in Vd can lower or delay peak plasma concentrations [49,52], which could necessitate higher doses in order to maintain therapeutic concentrations [54]. Albumin and α₁-acid glycoprotein concentrations decrease by roughly one-third compared to non-pregnant individuals, which begins in the second trimester (T2), continues through the end of pregnancy, and can impact the fraction unbound (pharmacologically active drug) relative to total drug concentrations [53] as well as Vd and CL [50]. Highly protein bound drugs are most likely to exhibit changes in fraction unbound during pregnancy, which could potentially influence maternal efficacy, maternal and fetal toxicities, and dosing requirements [54].

Cardiac output increases during the first trimester (T1) and peaks late in T2, which leads to increased glomerular filtration rate (GFR) and renal plasma and blood flow [55]. Collectively, these changes augment renal CL, which can lead to subtherapeutic concentrations of drugs that are primarily eliminated via renal excretion [50,52,54]. Moreover, pregnancy-related increases in renal drug transporter activity, specifically P-glycoprotein (P-gp), organic cation transporter (OCT)2 and organic anion transporter (OAT)1, have been reported [13,50], and can increase renal CL through transporter-mediated renal secretion processes.

Pregnancy-related changes in hepatic drug metabolism and CL can also impact the systemic PK and effects of drugs that undergo hepatic elimination. Studies have shown varying effects of pregnancy on maternal hepatic blood flow, which may increase or remain unchanged throughout gestation [56,57]. The hepatic CL of orally administered drugs with a high hepatic extraction ratio (ER) is predominantly dependent on hepatic blood flow; however, data describing the impact of hepatic blood flow changes during pregnancy on drug PK remain limited and variable [55]. In contrast, the hepatic CL of low-to-moderate hepatic ER drugs is principally dependent on protein binding and hepatic intrinsic CL. Clinical PK studies have described significant gestational changes in the CL of certain drugs that extend beyond changes in hepatic blood flow and plasma protein binding, and involve marked alterations in maternal hepatic intrinsic CL [50,53,55,58,59]. Accumulating evidence demonstrates significant gestational changes in hepatic drug metabolizing enzyme (DME) activity occur that can alter drug disposition during pregnancy [48,50]. Notably, induction of metabolic activity by cytochrome P450 (CYP) isoforms CYP2C9, CYP2D6, CYP2E1, CYP3A4, and uridine diphosphate-glucuronosyltransferase (UGT)1A1 and UGT1A4, and suppression of CYP1A2 metabolic activity is known to occur [50]. Pregnancy-associated changes in CYP- and UGT-mediated hepatic metabolism and CL can alter systemic drug exposure via changes in oral bioavailability through first-pass metabolism and/or changes in elimination half-life [52,60].

Due to the challenges of conducting PK studies in pregnancy, the need for a better quantitative understanding of PK changes during pregnancy has prompted a growing body of physiologically-based pharmacokinetic (PBPK) modeling studies [52,61]. PBPK models can leverage and incorporate in vitro, in vivo and clinical PK data with known physiologic parameters (e.g., blood flow, protein binding) and pharmacologic parameters (e.g., fractional CL by distinct metabolism pathways) in order to make predictions regarding drug disposition changes in special populations with sparse PK data, including pregnancy [13]. For instance, this modeling approach has been used to quantify and provide mechanistic insight into pregnancy-associated changes in hepatic CYP mediated drug metabolism and CL of drugs eliminated by distinct pathways across a range of gestational ages, demonstrating potential utility of model-informed prediction of drug exposure and dose optimization for CYP3A4 and other CYP substrates during pregnancy [59,62].

3.1. Alteration of Hepatic Drug Metabolism by Pregnancy Related Hormones

It is well-established that steroid hormones regulate hepatic DME expression and function [63], and numerous hormones including 17β-estradiol, progesterone, and cortisol increase substantially during pregnancy [64]. Cultured primary human hepatocytes are widely used as an in vitro model system to investigate the mechanisms underlying altered hepatic drug disposition [65]. Accumulating evidence demonstrates that pregnancy related hormones (PRH) significantly alter the hepatic expression and metabolic activity of various CYP and UGT enzymes, suggesting that these effects mediate gestational changes in hepatic DME expression observed in pregnant rodents and altered hepatic CL of various drugs during pregnancy in vivo [13,61,66,67]. PRH alter DME expression through activation of nuclear receptors, including pregnane X receptor (PXR), constitutive androstane receptor (CAR), estrogen receptor (ER), and glucocorticoid receptor (GR) [68–70].

Early work by several groups demonstrated that PRH significantly alter mRNA levels of certain CYP isoforms in cultured human hepatocytes and the metabolism of prototypical probe substrates [67,69,71–74]. CYP3A4 is primarily responsible for the metabolism of over 25% of FDA-approved drugs, including nifedipine [75]. Multiple studies have demonstrated that estradiol, progesterone, and cortisol increase CYP3A4 mRNA levels and the metabolism of the CYP3A probe substrate midazolam in primary and immortalized human hepatocytes [71,73,74]. These effects are consistent with observed increases in CYP3A-mediated midazolam CL during pregnancy at gestational age 28–32 weeks compared to 6–10 weeks post-partum in 13 pregnant volunteers [58]. Studies have also shown that estradiol increases CYP2B6 mRNA levels and CYP2B6-mediated metabolism of the probe substrate S-mephenytoin and antidepressant bupropion in human hepatocytes via activation of CAR and ERα dependent transcription [69,71,72]. More recently, Khatri et al. [76] demonstrated that PRH increased CYP3A4 and CYP2B6 protein expression to a greater degree in human hepatocytes compared to other CYP isoforms, and the PRH-induced increase in CYP3A4 expression yielded a significant increase in nifedipine metabolism.

The UGTs are an important family of phase II DMEs responsible for the glucuronidation and CL of drugs commonly prescribed during pregnancy, including labetalol and the anti-seizure drug lamotrigine. Liver expression of UGT1A1, UGT1A4, and other UGT1A isoforms are increased in mice during pregnancy via activation of PXR- and CAR-dependent transcription [77,78]. These effects appear to be mediated by PRH, since estradiol and progesterone induced UGT1A1 mRNA levels in hepatocytes isolated from humanized UGT1 mice, progesterone increased UGT1A1 promoter activation in immortalized HepG2 cells, and estradiol increased UGT1A4 mRNA levels and promoter activation and UGT1A4-mediated lamotrigine glucuronidation in HepG2 cells [68,77,79]. In contrast, neither estradiol nor progesterone altered UGT2B7 promoter activation in HepG2 cells [68]. Consistent with these effects, Khatri et al. [80] recently reported that PRH induced UGT1A1 and UGT1A4 protein expression in primary human hepatocytes, did not change UGT2B7 expression, and increased UGT1A1-catalyzed glucuronidation of labetalol.

4. Pharmacokinetics of HDP Drugs during Pregnancy

4.1. Labetalol

Labetalol is a combined alpha-1 and non-selective beta-adrenergic receptor antagonist that lowers heart rate and BP, and has arterial vasodilatory effects that become more prominent at higher doses [51,81]. Labetalol is available in oral and IV formulations for both chronic and acute treatment of HDP, has a fetal:maternal placental transfer of 0.1:1.0, and has not been associated with teratogenicity [48,51].

The PK parameters of labetalol in non-pregnant individuals are summarized in Table 3. Following oral administration, labetalol is readily absorbed with peak plasma concentrations within 60–90 minutes. As an intermediate hepatic ER drug, labetalol undergoes extensive first-pass metabolism resulting in an average bioavailability of 20–35%. Labetalol is moderately protein-bound (50–60%), and the elimination half-life is approximately 6–8 hours [51,81]. Negligible amounts of unchanged labetalol are excreted in the urine; accordingly, labetalol PK is not altered in patients with impaired renal function [81]. Labetalol is extensively metabolized in the liver by UGT1A1 and UGT2B7 to glucuronide metabolites [68]. However, due to the lack of analytical standards and methods that can quantify absolute concentrations of distinct labetalol glucuronides, the fractional contribution of UGT1A1- and UGT2B7-mediated metabolism to labetalol CL in vivo remains unknown. Patients with hepatic impairment exhibit elevated labetalol steady-state plasma concentrations, without changes in elimination half-life, due to diminished first-pass hepatic metabolism and increased bioavailability [33].

Table 3 –

Pharmacokinetic Properties of Labetalol, Nifedipine, and Methyldopa in Non-Pregnant Individuals

Drug	%F	% Protein Binding	Hepatic Extraction Ratio	Vd (L/kg)	T ½ (hr)	CLoral	Renal Elimination %	Metabolism %	Metabolic Enzymes	Notes	References^
Labetalol	20–35	50–60	Intermediate	3–16	6–8	25 mL/min/kg	<5	>90	UGT1A1, UGT2B7	Bioavailability increases with age and food intake, and exhibits significant inter-patient individual variability that can range from 11–86%. In patients with hepatic or renal impairment, elimination half-life is not altered; however, hepatic impairment increases bioavailability through diminished first-pass metabolism.	[33,48,51,81]
Nifedipine	30–60	92–98	Intermediate	0.8–2	SR tablets: 6–11 † IR capsules: 2.5–3.4 †	35–38 L/hr	Trace	>90	CYP3A4, CYP3A5	In patients with hepatic impairment, elimination half-life and AUC increase by 4- and 2-fold, respectively. Coadministration with grapefruit juice (intestinal CYP3A inhibitor) results in 2-fold increase in AUC and Cmax.	[35,48,51,91]
Methyldopa	25–50	0	Not reported	0.6	1–2	Not reported	50	50	SULT1A1/SULT1A3*, COMT**	Clearance is decreased in patients with renal failure leading to accumulation of active drug and metabolites and potentially prolonged hypotensive effects.	[48,107,108]

Abbreviations: %F=oral bioavailability; Vd=volume of distribution; T ½=elimination half-life; CL_oral=oral clearance; CYP=cytochrome P450; SULT=sulfotransferase; UGT=uridine diphosphate glucuronosyl transferase; SR=sustained release; IR=immediate release; AUC=area under the curve; C_max= maximum plasma concentration

^{^}Data summarized in these references are derived from pharmacokinetic studies comprised of both non-pregnant female and male individuals

^†Nifedipine half-life dependent on formulation;

^*SULT1A1 is also known as phenol sulfotransferase,

^**Catechol-O-methyltransferase.

Rubin et al. [82] evaluated the PK of IV labetalol longitudinally in 10 hypertensive women age 18–37 years at 33–38 weeks gestation; 7 of the participants had a follow-up visit 3–4 months post-partum. In parallel, 10 normotensive, non-pregnant female controls (18–35 years old) were also studied. Labetalol (50 mg bolus) was administered and serial blood samples (15 over 480 minutes) were collected following the first dose. When comparing hypertensive T3 pregnant subjects, post-partum longitudinal controls, and non-pregnant parallel controls, respectively, no significant difference in median [range] total CL (24.8 [19.7–38.8] vs. 34.8 [27.6–51.4] vs. 33.8 [20.4–41.6] mL/min/kg), Vd (3.7 [2.4–6.8] vs. 6.3 [3.3–9.3] vs. 3.9 [3.0–6.5] L/kg), or elimination half-life (2.5 [1.9–2.8] vs. 2.5 [1.9–3.3] vs. 2.1 [1.6–2.6] hrs) was observed. Due to the IV route of administration, the impact of pregnancy on first-pass metabolism and bioavailability was not evaluated.

The PK of oral labetalol during pregnancy has been evaluated in a limited series of studies, which are summarized in Table 4 [83–85]. Given the challenges of conducting PK studies in pregnant individuals, the studies had relatively small samples sizes (ranging from 7 to 9 patients) which limited precision of PK parameter estimation. Rogers et al. [83] evaluated the PK of oral labetalol (100 mg every 8 hours) in 8 hypertensive patients during T3. Participants were 19±4.5 (mean±standard deviation) years of age at 29.7±3.8 weeks gestational age with persistent elevations in SBP (>140–160 mmHg) or DBP (>90–110 mmHg) 24 hours after hospitalization. Serial blood samples were collected at steady-state over 8 hours after the last dose. The observed apparent CL/F (21.8±6.8 mL/min/kg) and elimination half-life (1.7±0.27 hours) was comparable to pregnant patients receiving IV labetalol in Rubin et al. [82] (24.8 mL/min/kg and 2.5 hrs, respectively). Saotome et al. [84] evaluated the PK of oral labetalol (150–450 mg twice daily) in 7 hypertensive women during T3. Patients were age 33.0±2.6 years at 33.3±2.0 weeks gestational age. Labetalol was initiated at 150–200 mg twice daily, and then escalated over 3–5 days until BP was adequately controlled. Serial blood samples were collected at steady-state over the 12-hour dosing interval. In this study, the observed CL/F (43.7±15.0 mL/min/kg) was higher, on average, when compared to PK data from historical non-pregnant female and male literature controls (Table 3). The observed elimination half-life (5.8±0.9 hrs) during pregnancy was similar to previously reported values of 6–8 hours in non-pregnant individuals. These data suggest a potential decrease in systemic exposure following oral labetalol administration during pregnancy due to increased first-pass metabolism and lower bioavailability. However, a key limitation of these studies was lack of either longitudinal evaluation during pregnancy and the post-partum period, or a parallel group of non-pregnant female controls for direct comparison. Thus, interpretation of pregnancy PK parameters by the authors were limited to qualitative comparisons relative to historical literature data derived from non-pregnant female and male subjects. Although sex differences in labetalol PK have not been confirmed in large studies, lower labetalol CL/F and higher plasma concentrations have been reported in females versus males [86], which further challenges interpretation of pregnant versus non-pregnant PK comparisons.

Table 4 –

Summary of Pharmacokinetic Studies of Oral Labetalol during Pregnancy

First Author (Year)	Pregnant Population	Gestational age range (weeks)	Dosing	PK Samples Collected	Exposure		Elimination		Key Findings Relative to Non-Pregnant (Literature) Controls	Notes	Ref.
					Conc. (ng/mL)	AUC (ng × hr/mL)	CL/F (mL/min/kg)	T1/2 (hr)
Rogers (1990)	N=8 hypertensive patients	24–32	100 mg TID	9 samples over 480 minutes	Cmax: 881 ± 219	AUC0–8: 1083 ± 342	21.8 ± 6.8	1.7 ± 0.3	↔ CL/F ↓ t 1/2	Steady-state Fetal cord/maternal serum: 0.50±0.15 AF/MP: 0.16±0.13	[83]
Saotome (1993)	N=7 hypertensive patients	30–36	150–450 mg BID	7 samples over 720 minutes	Css: 82.0 ± 22.7^	AUC0–12: 984.2 ± 272^	43.7 ± 15.0	5.8 ± 0.9	↑ CL/F ↔ t 1/2	Steady-state Variable dosing per clinical team; titrated to desired antihypertensive effect Pharmacodynamic analysis conducted: EC50 (58–180 ng/mL for SBP; 34–133 ng/mL for DBP) Variability in blood pressure lowering effects of labetalol correlated with inter-individual variability in CL/F and EC50	[84]
Carvalho† (2011)	N=9 hypertensive patients	37.7 (33.3–38.9)	Single dose 100 mg	16 samples over 720 minutes (with a stereoselective analytical method)	Cmax: Active (RR): 14.4 (7.5–19.9) (SR): 16.9 (11.9–31.8) Less active (SS): 20.8 (14.5–35.7) (RS): 16.2 (9.8–27.7)	AUC0-∞: Active (RR): 45.6 (40.3–74.4) (SR): 84.2 (63.8–119.3) Less active (SS): 89.4 (79.7–140.1) (RS): 78.3 (50.2–85.3)	Active (RR): 73.3 (60–123.3) (SR): 48.3 (33.3–81.7) Less active (SS): 35 (31.7–61.7) (RS): 61.7 (50–101.7)	Active (RR): 5.5 (4.7–6.0) (SR): 5.5 (4.0–8.9) Less active (SS): 5.9 (4.4–10.7) (RS): 4.7 (3.5–6.6)	Comparison to historical CL/F data for individual isomers not available	Unique analytical method that quantifies concentrations of labetalol stereoisomers Stereoisomer differences in PK parameters were observed following oral administration, but not following IV administration Following oral administration, higher CL and lower AUC was observed for pharmacologically active (RR)-isomer compared with the less active (SS)-isomer, suggesting stereoselectivity in first pass metabolism	[85]

Abbreviations: PK=pharmacokinetic; Ref.=references; E_max=maximum BP-lowering effect attributable to labetalol; EC₅₀=drug concentration providing half-maximal response; BID=twice daily; TID=three times daily; Conc. = concentration; C_max: maximum plasma concentration; C_ss: steady-state concentration; AUC: area under the time concentration curve; CL/F = apparent oral clearance; T ½ = half-life; AF/MP = amniotic fluid/maternal plasma concentration ratio; ↔ = no change.

Data reported as mean ± SD/median (95% confidence interval) unless otherwise indicated;

^{^}Css and AUC_0–12 normalized for 150 mg dose equivalent;

^†Data reported as median (95%CI)

Carvalho et al. [85] evaluated the influence of the route of administration (IV versus oral) and the effect of gestational diabetes on the stereoselectivity of labetalol PK during T3 in 16 patients (9 non-diabetic, 7 diabetic). The median maternal age and gestational age was 28.0 years and 37.7 weeks, respectively. Serial blood samples were obtained following a single 100 mg oral dose, and plasma concentrations of (RR)-labetalol (beta-blocker activity), (SR)-labetalol (alpha-blocker activity), and (RS)- and (SS)-labetalol (minimal activity) were quantified. The PK characteristics following oral administration in the non-diabetic group are reported in Table 4. Stereoselective PK was observed following oral administration, such that the median [95% confidence interval] CL/F values for the active (RR)-labetalol were higher than the less active (SS)-labetalol (73.3 [60–123.3] vs. 35 [31.7–61.7] mL/min/kg). In contrast, stereoselective PK was not observed following IV administration, suggesting that higher labetalol first-pass metabolism and lower bioavailability during pregnancy following oral administration may be driven by the active (RR)-labetalol stereoisomer. The CL/F and AUC of (RR)-labetalol was not different in pregnant patients with and without diabetes. However, diabetic patients had higher AUC and lower CL/F for the alpha-1 blocking (SR)-isomer compared to non-diabetics, which was attributed to a stereoselective reduction in (SR)-labetalol glucuronidation in patients with gestational diabetes. The impact of pregnancy, compared to post-partum or non-pregnant controls, on the stereoselective PK and glucuronidation of labetalol warrants further study.

A recent prospective, open-label longitudinal population PK study investigated the influence of gestational age on oral labetalol PK in 57 hypertensive patients from 12 weeks gestation through 12 weeks post-partum [87]. Median maternal age and gestational age was 30 (range 18–41) years and 20 (11–39) weeks, respectively. Labetalol dosing occurred per clinical care. An opportunistic longitudinal sampling method was implemented, which occurred across T2, T3, and post-partum, coincided with clinic visits, and enabled individual patients to serve as their own controls. Population PK modeling was used to estimate PK parameters following collection of 649 plasma samples at various time points. The model estimates for apparent CL/F and steady-state Vd (V_ss/F) in a typical non-pregnant individual were 188 L/hr (per 50 kg lean body weight) and 691 L (per 70 kg total body weight). At 12 weeks and 40 weeks gestation, respectively, the model predicted that CL/F increased 1.4-fold and 1.6-fold on average compared to post-partum estimates, which exhibited substantial inter-patient variability (approximately 6-fold). The increase in labetalol CL/F, which resulted in lower labetalol plasma concentrations during pregnancy, returned to non-pregnant values by 12 weeks post-partum and was most significantly influenced by gestational age and lean body weight. The model predicted a 1.9-fold increase in V_ss/F during pregnancy regardless of gestational age. Differences in the fraction of labetalol bound to plasma proteins during T2, T3 and postpartum were not apparent. Due to opportunistic sampling and population PK design, elimination half-life was not estimated. The investigators concluded that the increases in labetalol CL/F during T2 and T3 were most likely mediated by an increase in hepatic intrinsic CL and resulting decrease in oral bioavailability due to a pregnancy-associated increase in hepatic first-pass metabolism.

Recent in vitro data demonstrating that PRH significantly increase labetalol glucuronidation in primary human hepatocytes by inducing UGT1A1 protein expression [80] provide mechanistic insight into these PK data. These effects are consistent with prior studies demonstrating induction of hepatic UGT1A4 expression and lamotrigine glucuronidation (an anti-seizure drug and UGT1A4 substrate) by PRH in vitro [79] and increases in lamotrigine glucuronide metabolite formation, CL/F, and dose requirements in pregnant individuals [88–90]. Collectively, the available evidence suggests that increased hepatic UGT1A1-mediated metabolism may be largely responsible for the observed and PK model-predicted increases in labetalol CL/F and decreases in labetalol plasma concentrations that occur during pregnancy.

4.2. Nifedipine

Nifedipine is a dihydropyridine CCB that causes arterial vasodilation and lowers BP [48]. Nifedipine is available in immediate-release (IR) capsule and sustained-release (SR) tablet formulations for treatment of HDP, has a fetal:maternal placental transfer of 0.1:1.0, and has not been associated with teratogenicity [48,51]. In addition, nifedipine induces uterine relaxation and is frequently prescribed off-label as a tocolytic agent to prevent premature delivery [51].

The PK properties of nifedipine in non-pregnant individuals are summarized in Table 3. Nifedipine undergoes rapid absorption after oral administration, is very highly protein bound (92–98%), and has a variable elimination half-life that is dependent on formulation [48,51,91]. Nifedipine undergoes substantial hepatic and intestinal first-pass metabolism that results in an oral bioavailability of approximately 50%, and is an intermediate hepatic ER drug that is primarily cleared via hepatic metabolism by CYP3A4 and to a lesser degree CYP3A5 [48]. Variability in systemic exposure is linked to interindividual differences in first-pass metabolism and patients with cirrhosis exhibit significantly increased AUC and elimination half-life, demonstrating the important contribution of hepatic metabolism to nifedipine systemic PK [51]. Although the presence of sex differences in nifedipine PK remain unclear, a population PK study reported higher nifedipine CL/F in females versus males [92], which is consistent with some studies suggesting higher hepatic CYP3A4 metabolism in females [93].

The PK of oral nifedipine during pregnancy have been evaluated in a series of studies conducted in patients receiving treatment for HDP or preterm labor, which are summarized in Table 5 [94–99]. Given the challenges of conducting PK studies in pregnant individuals, the studies had relatively small samples sizes (ranging from 7 to 15 patients). A key limitation of each study was the lack of a parallel group of non-pregnant controls for direct comparison. Moreover, the studies did not longitudinally evaluate PK changes throughout gestation or draw direct comparisons to the post-partum period. Thus, the evaluation of differences in PK parameters during pregnancy were limited to qualitative comparisons with historical literature data in non-pregnant individuals.

Table 5 –

Summary of Pharmacokinetic Studies of Oral Nifedipine during Pregnancy

Indication	First Author (Year)	Pregnant Population	GA range (weeks)	Dosing	PK Samples Collected	Distribution	Exposure		Elimination		Key Findings Relative to Non-Pregnant (Literature) Controls	Notes	Ref.
						Vd/F (L)	AUC (ng × hr/mL)	Conc. (ng/mL)	CL/F (L/hr/kg)	T ½ (hr)
Hypertensive Disorders of Pregnancy	Prevost (1992)	N=15 hypertensive patients	26–35	IR capsules; 10 mg QID	7 samples over 360 minutes	Not reported	AUC0–6 : 83.2 ± 42.6	Cmax : 38.6 ± 18	2.0 ± 0.8	1.3 ± 0.5	↓ Cmax ↑ CL/F ↓ t 1/2	Steady-state Fetal cord/maternal serum (0.93 ± 0.2) AF/MP (0.56 ± 0.15)	[94]
	Barton (1991)	N=8 Patients with preeclampsia	Immediate post-partum period	10 mg	9 samples over 360 minutes	Not reported	AUC0–6 : 40.8 ± 6.2	Cmax : 18.3 ± 1.9	3.3 ± 1.3	1.4 ± 0.3	↓ Cmax ↑ CL/F ↓ t 1/2	Single-dose study Formulation not specified	[95]
	Filgueira† (2015)	N=12 hypertensive patients	34.3–39.6	SR tablets; 20 mg BID	18 samples over 720 minutes	600.0 (490.9–709.1)	AUC0–12: 250.2 (190.2–310.3)	Cmax : 28.2 (22.7–33.6)	89.2 L/hr (70.7–107.7)	5.1 (3.5–6.7)	↑ Vd ↑ CL/F ↓ t 1/2	Steady-state AF/MP: 0.05 ± 0.03	[96]
Tocolysis	ter Laak (2014)	N=8 with threatened preterm labor	26.0–32.3	SR tablets; 20 mg QID	7 samples over 216 hours	6.2 ± 1.9 L/kg^	Not reported	Css : 18.7 ± 9.5	Not reported	3.2 (2–5)	↑ Vd ↓ t 1/2	Steady-state Only trough levels drawn	[97]
	Haas (2013)	N=14 treated for preterm labor	24–36	IR capsules; 10 to 20 mg q4-8h	9 samples over 8 hours	295 ± 262	AUC0–6 : 207 ± 139	Css : 41.2 ± 31.0	128 ± 80.4 L/hr	1.7 ± 1.6	↓ Css ↓ t 1/2	Steady-state single dosing interval study Variable dosing per clinical team CYP3A4/5 genotyping completed	[98]
	Juon (2008)	Group 1: N=7 treated for threatened preterm labor	30.4–41.4	Load: 10 mg IR × 4, then 60 mg SR per day × 5	23 samples over 5 days	Not reported	AUCss: 220.0 ± 160.1	Css : 32.0 ± 23.8	2.2 ± 1.3	Not reported	↓ Css	Prospective, open-label randomized study Steady-state Multiple formulations studied	[99]
		Group 2: N=7 treated for threatened preterm labor		Load: 10 mg IR × 4, then 90 mg GITS per day × 5	23 samples over 5 days		AUCss: 173.5 ± 26.2	Css : 14.5 ± 4.8	2.9 ± 0.5

Abbreviations: GA=gestational age; PK=pharmacokinetic; SR=slow release; IR=immediate release; BID=twice daily; QID=four times daily; GITS=gastrointestinal intestinal therapeutic system; Vd=volume of distribution; Vd/F=apparent oral volume of distribution; Conc.=concentration; C_max=maximum plasma concentration; C_ss=steady-state concentration; AUC=area under the time concentration curve; CL/F=apparent oral clearance; T ½=half-life; AF/MP=amniotic fluid/maternal plasma concentration ratio; SE=standard error; ND=not detectable.

Data are reported as mean ± SD or (range) unless otherwise indicated;

^{^}Calculated using a fixed bioavailability of 0.45 from literature due to lack of IV data in this population;

^†Filgueira et al. [96] PK data reported as mean (95% CI)

Prevost et al. [94] evaluated nifedipine PK in 15 hypertensive pregnant patients during T3. Participants were 24.6±5.9 years of age, at 32.1±2.7 weeks gestation, and had persistently elevated SBP (140–160 mmHg) and DBP (90–110 mmHg) 24 hours after hospitalization. Nifedipine IR capsules (10 mg) were administered every 6 hours for 8 doses, and serial blood samples were collected at steady-state following the final dose. Compared with a healthy non-pregnant female historical control population receiving the same dose [100], a lower maximum plasma concentration (C_max) (73.5±17.5 vs. 38.6±18 ng/mL), lower elimination half-life (3.4±10.6 vs. 1.3±0.43 hours), and higher CL/F (0.49±0.09 vs. 2.0±0.8 L/hr/kg) was observed in the pregnant population. Notably, 12 of 15 pregnant patients had undetectable plasma nifedipine trough concentrations at 6 hours, suggesting underdosing and/or too infrequent dosing.

Barton et al. [95] evaluated nifedipine PK following a single oral dose (10 mg) in 8 patients with preeclampsia during the immediate post-partum period. Consistent with the findings of Prevost et al. [94] during T3, patients in the immediate post-partum period also exhibited a lower C_max (18.3±1.9 ng/mL) and elimination half-life (1.4±0.3 hours), as well as a higher CL/F (3.3±1.3 L/hr/kg), compared to non-pregnant female historical controls [100]. Moreover, 6 of the 8 patients had undetectable nifedipine plasma concentrations at 6 hours.

Filgueira et al. [96] observed similar findings in a more recent study evaluating nifedipine PK in 12 hypertensive patients during T3. Participants were age 30.9±6.2 years and at 36.1±1.7 weeks gestation. Nifedipine SR tablets (20 mg) were administered twice daily for at least two weeks and serial blood samples were collected at steady-state following the final dose. The authors similarly observed lower C_max (mean [95% CI]: 28.2 [22.7–33.6] ng/mL), higher CL/F (89.2 [70.7–107.7] L/hr), and lower elimination half-life (5.1 [3.5–6.7] hours), as well as higher apparent volume of distribution (Vd/F) (600 [490.9–709.1] L) compared with non-pregnant female historical controls. Notably, due to use of the nifedipine SR formulation in the study, elimination half-life in this study was higher than reported by Prevost et al. [94] (1.3±0.43 hours). A subsequent study by the same investigators evaluated the effect of diabetes on nifedipine PK during T3 [101]. Diabetes did not significantly affect nifedipine PK during pregnancy, such that the median [25–75%] C_max (23.5 [21.2–27.9] ng/mL), CL/F (98.9 [85.1–107.8] L/hr), elimination half-life (5.0 [4.5–5.8] hours), and Vd/F (609.4 [357.3–742.1] L) were similar to non-diabetic pregnant individuals.

In addition to PK studies in the setting of HDP, multiple studies evaluating the PK of IR and SR nifedipine as a tocolytic agent have similarly demonstrated lower steady-state concentrations (C_ss) and elimination half-life, as well as higher CL/F and Vd/F [97–99] compared to non-pregnant individuals (Table 5). Additional studies over the past three decades have reported nifedipine C_max or C_ss at a single time-point in both normotensive pregnant women [102] and in women treated for preterm labor [103–106]. Despite the lack of serial sample collection and estimation of PK parameters, these studies have also collectively suggested lower average nifedipine plasma concentrations during pregnancy relative to non-pregnant individuals.

The challenges of conducting PK studies in pregnancy and need for a better quantitative understanding of pregnancy alterations in nifedipine PK has prompted a series of PBPK modeling studies, which aimed to predict the disposition of various nifedipine doses and formulations across a range of gestational ages [52,59,60,62]. Consistent with described nifedipine PK studies, these modeling and simulation analyses further illustrate higher CL/F during pregnancy compared to non-pregnant controls. Dallman et al. [52] used a PBPK approach to predict the longitudinal disposition of nifedipine throughout pregnancy. Because variation in the extent of CYP3A4 induction have been reported across different gestational ages in previous in vivo human studies, the investigators assumed a 1.6-fold induction of CYP3A4 activity during T1, T2 and T3 compared to non-pregnant individuals in their model. The model predicted a 2.25-fold average increase in nifedipine CL/F during pregnancy, which was similar to the predicted 2.17-fold increase in midazolam CL/F (a prototypical CYP3A4 substrate) and within 1.25-fold error of observed data. Notably, the model tended to underestimate C_max due to significant inter-subject variability; however, 93% and 54% of all predicted serum concentration values were within 2-fold and 1.25-fold error, respectively, relative to observed data. Abduljalil et al. [62] further modeled pregnancy-evoked increases in CYP3A4 activity across a range of gestational ages (1.3-, 1.8-, and 2.3-fold during T1, T2, and T3, respectively) and its impact on nifedipine PK. Model predictions for both the IR capsule and SR tablet nifedipine formulations were within 2-fold error compared with historical pregnancy data [94], and predicted higher CL/F in pregnant compared to non-pregnant individuals for both the IR (202±145 vs. 70±43 L/h) and SR (170±101 vs. 64±38 L/h) formulations.

In addition to predicting changes in systemic nifedipine PK during T3, PK modeling and in vitro studies have offered mechanistic insight into the most important site of CYP3A induction during pregnancy by studying varying extents of CYP3A metabolism in different organs. Ke et al. [60] assessed the relative contributions of intestinal and liver CYP3A4 metabolism changes to nifedipine PK changes during pregnancy, and concluded that increases in hepatic, rather than intestinal CYP3A metabolism, underlie the observed gestational increases in CYP3A-mediated nifedipine metabolism and CL/F during pregnancy. These findings corroborated results of a semi-mechanistic modeling study, which concluded that induction of maternal CYP3A-mediated CL/F of nifedipine and the prototypical CYP3A substrate midazolam during pregnancy is driven by increased hepatic CYP3A activity, as opposed to gestational alterations in liver blood flow, protein binding, and placental or fetal metabolism [59]. These results are further supported by recent in vitro data demonstrating that PRH significantly increase nifedipine metabolism in primary human hepatocytes by inducing CYP3A4 protein expression [76]. Collectively, these data provide mechanistic insight into the contribution of increased hepatic CYP3A4-mediated metabolism to the observed and PBPK model predicted increases in nifedipine CL/F and the lower nifedipine systemic exposure observed during pregnancy.

4.3. Methyldopa and Other Antihypertensives

Methyldopa is an alpha-2 adrenergic receptor agonist that centrally inhibits vasoconstriction by attenuating the release of catecholamines [48]. Methyldopa is not bound to plasma proteins and has an oral bioavailability of 25–50% [107]. Approximately 50% of the dose is excreted unchanged in the urine, and the remainder is metabolized via sulfation and O-methylation by sulfotransferases and catechol-O-methyltransferase, respectively [108]. Methyldopa’s half-life in non-pregnant females and males is approximately 1–2 hours (Table 3); however, accumulation of active drug and metabolites occurs in patients with renal failure, which can lead to prolonged hypotensive effects [36]. To date, there are no human or preclinical data evaluating the PK of methyldopa during pregnancy, although anecdotal experience suggests that pregnancy-associated PK alterations are either absent or not clinically significant [48].

Other second-line agents include hydralazine, hydrochlorothiazide, amlodipine, atenolol, and clonidine (Table 2). A recent study evaluated the steady-state PK of oral hydralazine (5–25 mg 4 times daily) in 12 hypertensive pregnant patients during T2 (n=5) and T3 (n=8) [109]. Although N-acetyltransferase (NAT2) genotype was significantly associated with hydralazine PK, the authors reported no differences in AUC, CL/F, or Vd/F in T2 versus T3 patients or when comparing hydralazine PK during pregnancy to literature data derived from non-pregnant females and males. In contrast, no data describing the PK of hydrochlorothiazide in pregnancy have been reported to date. A recent study evaluated the PK of furosemide, a loop diuretic and UGT1A1/9 substrate used to manage fluid overload during peripartum cardiomyopathy [110]. The authors reported approximately 6-fold higher Vd and 4-fold lower C_max in pregnant individuals compared to literature data from non-pregnant females and males, which was attributed to expanded plasma volume and total body water, and decreased plasma protein binding. Consistent with aforementioned increases in labetalol CL/F (a UGT1A1 substrate) during pregnancy, median [25–75%] furosemide CL/F in pregnant individuals (25.3 [13.8–31.4] L/hr) was approximately 2-fold higher compared to non-pregnant individuals (range 8.3–15.7 L/hr), a finding which the authors attributed to UGT1A induction during T3. In addition, a study evaluating steady-state PK of amlodipine, a dihydropyridine CCB and CYP3A4 substrate, reported a lower AUC (53.4±19.8 hr*ng/mL) and elimination half-life (13.7±4.9 hrs) in 11 hypertensive pregnant patients compared to literature data from non-pregnant females and males [111].

Hurst et al. [112] evaluated the PK of oral atenolol (50 mg twice daily), a selective beta-1 adrenergic receptor antagonist that is primarily eliminated by renal excretion and undergoes minimal hepatic metabolism, in a longitudinal study of hypertensive pregnant patients from T3 through week 6 post-partum. No differences in atenolol AUC (0.21±0.06 vs. 0.22±0.09 ng/ml/day), CL/F (0.11±0.04 vs. 0.11±0.04 L/kg/hr), Vd (0.72±0.14 vs. 0.64±0.15 L/kg), or elimination half-life (5.0±1.8 vs. 4.2±1.0 hrs), respectively, were observed. These findings were consistent with a previous single-dose PK study of oral atenolol (100 mg) in 13 pregnant women with severe pre-eclampsia during T3 [113], which concluded that there were no pregnancy-related alterations in atenolol PK.

The PK of oral clonidine, a centrally acting alpha-2 adrenergic receptor agonist that undergoes 50% renal excretion and 50% hepatic metabolism by CYP2D6 and to a lesser extent CYP3A4/5 and CYP1A2 [114], were evaluated in 17 hypertensive women during T2 or T3 [115]. Dosing occurred per clinical care (0.15 to 0.30 mg per day in divided doses) and steady-state PK revealed higher CL/F (440±168 mL/min) compared with literature control data derived from non-pregnant females and males (245±72 mL/min). Although creatinine CL (CrCl) is increased during pregnancy and approximately 50% of clonidine is renally eliminated, the observed renal CL of clonidine in pregnant participants was not different from historical data. Moreover, a relatively lower amount of unchanged drug was excreted in the urine in pregnant patients, suggesting increases in the non-renal CL of clonidine during pregnancy. Because clonidine is a low hepatic ER drug and is not highly protein bound (20–40%), the authors concluded that increases in CYP-mediated hepatic and/or intestinal clearance most likely underlie the observed increase in clonidine CL/F during pregnancy. A recent PBPK model predicted that a 2–4-fold induction of hepatic CYP2D6 activity during T3 accounted for the observed changes in clonidine PK [116].

5. Conclusion

Extensive physiologic and biochemical changes alter the PK and effects of certain medications during pregnancy. The diagnosis of HDP (gestational hypertension, preeclampsia, pre-existing chronic hypertension, and chronic hypertension with superimposed preeclampsia) and use of antihypertensive medications to lower BP and preserve maternal and fetal health has become increasingly common. Although there remains a paucity of data rigorously evaluating, comparing, and elucidating the mechanisms underlying gestational changes in the PK and dose requirements of antihypertensive drugs, a growing body of evidence has improved our understanding of antihypertensive drug PK changes during pregnancy.

When non-emergent treatment with an antihypertensive drug is indicated, oral labetalol, nifedipine, or methyldopa are recommended as first-line options and commonly prescribed. The majority of pregnancy antihypertensive drug PK studies have been completed with labetalol and nifedipine. In contrast, no human or preclinical studies evaluating methyldopa PK during pregnancy have been conducted to date. The existing literature is limited by the lack of direct comparisons of PK during T2 and T3 with postpartum or non-pregnant female controls, considerable inter-patient variability, and small sample sizes.

Collectively, for both labetalol and nifedipine, the CL following oral administration appears to be higher in pregnant compared to non-pregnant individuals. Associated decreases in plasma steady-state concentrations during pregnancy also have been observed. A series of PK modeling and simulation studies have provided a greater quantitative and mechanistic understanding of these pregnancy-related PK changes. Multiple PBPK modeling studies have predicted that nifedipine CL/F increases approximately 2.2–2.8-fold during pregnancy, and concluded that these effects are most likely mediated by increases in maternal hepatic CYP3A-mediated metabolism as opposed to gestational alterations in liver blood flow, protein binding, or placental or fetal metabolism [52,59,60,62]. Likewise, a population PK study predicted that labetalol CL/F increases approximately 1.4–1.6-fold during pregnancy, and concluded that these increases were most likely mediated by increases in maternal hepatic UGT1A1-mediated intrinsic CL [87]. Because pregnancy-related changes in labetalol total CL or elimination half-life have not been observed following IV administration [82], the observed changes in CL/F following oral dosing are likely due to decreased bioavailability secondary to pregnancy-associated increases in hepatic first-pass metabolism. Together, these findings suggest that higher labetalol and nifedipine doses and/or more frequent dosing may be needed during pregnancy to sustain desired concentrations and prevent therapeutic failure.

Although the mechanisms require further study, our review highlighted the central role that PRH play in the regulation of hepatic CYP and UGT expression and metabolic activity, which appears to be a primary contributor to increased CYP3A4-mediated nifedipine metabolism and UGT1A1-mediated labetalol metabolism in the liver during pregnancy [76,80]. Consistent with these effects, studies evaluating the PK of additional second- or third-line oral antihypertensive drugs have also demonstrated higher CL/F of several drugs that undergo significant CYP- or UGT-mediated hepatic metabolism during pregnancy, including clonidine (CYP2D6) [115], amlodipine (CYP3A4) [111], and furosemide (UGT1A1) [110]. In contrast, no apparent pregnancy-related differences in CL/F have been observed with either atenolol (renal elimination) [112] or hydralazine (hepatic NAT2 metabolism) [109]. Our review of available evidence suggests that oral antihypertensive drugs relying predominantly on hepatic CYP- or UGT-mediated metabolism for systemic CL may be more sensitive to pregnancy-related changes in PK. However, additional studies are needed to more rigorously and precisely quantify gestational changes in antihypertensive drug PK during pregnancy and elucidate the underlying mechanisms.

6. Expert Opinion

Most drugs used during pregnancy are prescribed off-label and lack dosing recommendations specific to this special population. “Precision dosing” offers enormous potential to individualize drug treatment based on patient factors that alter drug disposition and response. The need for such strategies is augmented in clinical scenarios where patients exhibit pronounced inter-patient variability in drug PK and response, have narrow therapeutic windows for efficacy and toxicity, and were underrepresented in or excluded from clinical trials [14]. We posit that treatment of hypertension during pregnancy provides a potential opportunity for the investigation and application of precision dosing.

Although the hepatic metabolism and CL/F of labetalol, nifedipine, and certain other antihypertensive drugs appear to increase during pregnancy, these studies are limited by small sample sizes and lack of internal non-pregnant female controls for direct comparison. Various barriers challenge the conduct of research in pregnant individuals, including logistical challenges and a lack of pharmaceutical industry interest in conducting pregnancy trials; however, the lack of rigorous PK and outcome evidence in this scientifically complex population may ultimately be harmful. Consequently, additional studies that rigorously and precisely quantify and compare gestational changes in antihypertensive drug PK are needed to inform pregnancy-specific dosing recommendations in drug labeling [6]. This includes studies evaluating longitudinal changes in PK parameters during T2, T3, and multiple post-partum timepoints as physiology returns to a pre-pregnancy state. Additionally, PK studies should quantify pregnancy effects on fractional CL via hepatic metabolism and renal elimination pathways to elucidate the mechanisms underlying pregnancy-associated changes in systemic exposure. Preclinical and clinical studies of pregnancy-related changes in hepatic and extra-hepatic drug metabolism and transport, endogenous biomarkers of key metabolism and transport pathways, and the molecular mechanisms mediating these effects should continue to be a major area of investigation.

A recent multicenter cohort study evaluating the frequency of seizures during pregnancy and post-partum demonstrated the potential clinical implications and provided a framework for investigation of pregnancy-related dose adjustment [90]. Among women with epilepsy on anti-seizure drug therapy, the percentage of those with a higher incidence of seizures during pregnancy was not significantly different compared to non-pregnant controls (odds ratio (OR) 0.93; 95% CI [0.54–1.60]). However, changes in drug dosing occurred more frequently in pregnant subjects (74% vs. 31%; OR 6.36; 95% CI [3.82–10.59]) and doses were more likely to be increased by the end of pregnancy compared to non-pregnant controls (70% vs. 24%; OR 7.49; 95% CI [4.37–12.84]). Notably, lamotrigine (a UGT1A4 substrate) was commonly prescribed. The authors concluded that diligent therapeutic drug monitoring and dose optimization is necessary to control seizure activity and minimize maternal and fetal adverse effects.

The reported increases in labetalol and nifedipine CL/F during pregnancy suggest a potential need for dose optimization strategies in HDP patients. Higher doses and/or more frequent dosing may be needed to sustain desired concentrations and prevent therapeutic failure during pregnancy. Some of this may be occurring already in clinical practice based on BP response; however, precision dosing in pregnancy might allow for more rapid achievement of BP goals and thereby reduce morbidity. Indeed, more tailored interventional strategies have exhibited considerable promise [117]. Although an increasing number of clinical trials and registries comparing the safety and effectiveness of different antihypertensives during pregnancy have been reported, these studies rarely report the range of doses used to control BP. Accordingly, it remains unknown whether antihypertensive dose requirements change throughout pregnancy, what patient factors are associated with these changes, and whether these effects are more prominent for certain drugs. Moreover, it is unknown whether insufficient dosing of oral antihypertensive drugs for non-emergent HDP confers a higher risk of developing severe HTN and various adverse maternal and fetal outcomes. Consequently, ongoing HDP treatment trials evaluating maternal and fetal outcomes may not be able to detect benefit if dosing is suboptimal.

Studies evaluating the clinical relevance of PK/PD relationships between pregnancy changes in antihypertensive drug exposure, dose requirements, and BP lowering effects are needed to guide development of more precise treatment regimens. To complete such analyses, there is a critical need for HDP clinical trials and registries to more systematically collect and report medication dosing data, ideally with accompanying biological sample collection (including maternal blood, cord blood post-delivery, and breast milk post-partum) for drug concentration and biomarker measurements. Use of modern observational study design and data analysis strategies can facilitate estimation of dynamic treatment regimens (DTRs), or sequences of decision rules, which offer enormous potential for the advancement of precision dosing [118]. DTRs can also be estimated in randomized controlled trials using Sequential Multiple Assignment Randomized Trial (SMART) designs, which can estimate the benefit of different sequences of treatment decisions. SMART designs mirror clinical care by systematically adjusting treatment dose and type in patients who do not respond well to their originally assigned therapy [119], and offer the potential to improve recruitment and retention of pregnant participants.

PK modeling and simulation will be instrumental to advancing knowledge of PK/PD changes during pregnancy, underlying mechanisms, and precision dosing strategies with enriched potential for improved outcomes. Several studies have demonstrated how PK modeling has enabled quantification of nifedipine CL changes during pregnancy and provided mechanistic insight into key drivers (most notably hepatic CYP3A4) of these changes [52,59,60,62]. Ongoing efforts should leverage emerging data to further improve precision of existing PBPK models and expand model development for additional drugs. These advancements will also allow investigators to quantify the relative impact of other factors, such as race/ethnicity, maternal age, genetic polymorphisms, plasma protein binding, comorbidities, and drug-drug interactions, on gestational changes in PK, PD, and dose requirements, which have not been well-studied to date. With increased understanding of the intrinsic and extrinsic factors that influence inter-patient variability in drug exposure and PD effects, more robust models can be created and validated to yield more precise dosing predictions that guide optimized treatment regimens for individual patients and can be evaluated in clinical trials.

Advancing knowledge surrounding the extent and mechanisms underlying antihypertensive PK changes and dose requirements during pregnancy is necessary to both inform the need for and develop strategies that deliver precision antihypertensive drug dosing during pregnancy. For example, it remains unknown whether pregnancy-associated drug metabolism or PK changes differ across the distinct diagnostic classifications of HDP due to underlying differences in the pathophysiology, hormonal changes, or other contributing factors. Future research is needed to address these evidence gaps and inform development of more precise HDP therapeutic strategies. While there is insufficient data to support integration of specific dosing recommendations in current clinical practice, clinicians should be advised that more frequent dosing and/or higher doses of labetalol and nifedipine may necessary to lower BP during pregnancy due to gestational increases in CL. At a minimum, clinicians should be vigilant to monitor BP as an indicator of effective treatment and have a low threshold to increase dosing as needed to achieve BP control. As more evidence is generated through PK studies and innovative data analysis strategies, it will be critical to carefully consider the appropriate selection and dosing of these and other therapeutic agents in HDP where PK alterations may lead to decreased therapeutic effect or increased maternal and fetal toxicity.

Article Highlights:

• Hypertensive disorders of pregnancy (HDP) increase risk for adverse maternal and fetal outcomes, and frequently require use of antihypertensive medications (most notably labetalol and nifedipine).

• Physiologic and molecular changes occur during pregnancy that can alter maternal drug disposition and pharmacokinetics through various mechanisms.

• Pregnancy related hormones increase UGT1A1 and CYP3A4 expression in human hepatocytes, which are primarily responsible for the hepatic metabolism and clearance of labetalol and nifedipine, respectively.

• Multiple studies have demonstrated higher oral clearance and lower plasma concentrations of labetalol and nifedipine following oral administration during pregnancy.

• These changes are most likely mediated by increased maternal hepatic UGT1A1- and CYP3A4-mediated metabolism, and decreased bioavailability secondary to enhanced first-pass metabolism.

• Pregnancy-related decreases in plasma labetalol and nifedipine concentrations following oral administration suggest the need for increased and/or more frequent dosing.

• A better understanding of the extent and mechanisms underlying antihypertensive pharmacokinetic changes during pregnancy will create opportunities for more precise medication selection and dosing in HDP patients.