Anatomical and physiological alterations of pregnancy

Abstract

The extensive metabolic demands of pregnancy require specific physiological and anatomical changes. These changes affect almost all organ systems, including the cardiovascular, respiratory, renal, gastrointestinal, and hematologic system. The placenta adds another layer of complexity. These changes make it challenging for clinicians to understand presenting signs and symptoms, or to interpret laboratory and radiological tests. Furthermore, these physiological alterations can affect the pharmacokinetics and pharmacodynamics of drugs. Postpartum physiology includes the return to pre-pregnancy physiology, and lactation with drug safety concerns, commonly only supported by limited evidence. In addition, the teratogenic effects of medications are often extrapolated from animals, which further adds uncertainties. Unfortunately, pregnant women are only rarely included in clinical drug trials, while doses, regimens, and side effects are often extrapolated from studies conducted in non-pregnant populations.

In this comprehensive review, we present the changes occurring in each system with its effects on the pharmacokinetic variables. Understanding these physiological changes throughout normal pregnancy helps clinicians to optimize the health of pregnant women and their fetuses. Furthermore, the information on pregnancy-related physiology is also critical to guide study design in this vulnerable ‘orphan’ population, and provides a framework to explore pregnancy-related pathophysiology such as pre-eclampsia.

Introduction

The amount of drug prescriptions and herbal medicine use in pregnancy has substantially increased during recent decades. This change has been more evident during the first trimester, where the average drug usage has increased by 62.5% between 1976 and 2008 [1–4].

The risks of pharmacotherapy in pregnancy are two-fold: for the pregnant woman and her fetus; altered pharmacokinetics and pharmacodynamics may lead to inadequate therapeutic response or maternal toxicity, and maternal-fetal transmission may impair fetal organ formation or cause fetal demise. Studies on the pharmacokinetics and safety of drugs during pregnancy are scarce [4, 5]. Therefore, practitioners need to balance the risks and benefits of each drug before deciding to prescribe.

Pharmacokinetics (PK) of a drug includes the processes of drug absorption, distribution, metabolism, and excretion (ADME), whereas pharmacodynamics (PD) describes the mechanism of action and the pharmacological effect(s) of a drug [6]. Several physiological and anatomical changes that occur during pregnancy may alter some or all of these ADME processes (Table 1). Clinicians often lack the knowledge on how and why drug disposition is altered in pregnancy, and for that reason, determining the “appropriate” doses to result in appropriate target exposure is difficult [7].

Table 1.

The different pharmacokinetic variables and the physiologic changes that affect these variables during pregnancy.

Pharmacokinetic Variable	Physiologic change that affects it	Direction of Change [38]
Absorption	Nausea and vomiting	Decreased
	Gastric motility	Delayed absorption
Distribution	Increase in body weight	Increased
	Increase in fat deposition	Increased for lipophilic drugs
	Increase of plasma volume	Increased
	Decrease of plasma albumin	Increased for unbound drugs
Metabolism	Change in hepatic drug metabolizing enzymes	Variable; depends on specific enzyme
Excretion	Increase of hepatic blood flow	Increased clearance for high excretion ratio drugs
	Increase of renal blood flow	Increased clearance for unchanged drugs
	Increase of glomerular filtration rate	Increased clearance of unchanged drugs
Transport	Placental drug metabolism	Variable
	Placental transport
	Accumulation in amniotic fluid

The increase in plasma volume, weight gain, and change in fat composition, as well as other effects on the renal, cardiovascular, and gastrointestinal systems play a pivotal role in determining the PK variables of a drug. Randomized controlled trials for drugs rarely include pregnant women as clinical investigators encounter barriers that span across medical, logistical, legal, and ethical arenas [5, 8]. The doses of most drugs currently used in pregnancy are extrapolated from studies conducted in men and non-pregnant women. Also, post-marketing surveillance and results from retrospective studies are used to determine the safety profile of drugs [9, 10]. The fact that studies determining teratogenicity of drugs were historically based on animal-model studies, adds complexity to extrapolate the findings to the human species [11]. A recent review showed that less than 0.5% of all registered clinical trials target the use of pharmaceutical drugs among pregnant women. Most of these clinical trials were conducted in the therapeutic areas of anesthesia and analgesia, preterm birth and tocolysis, pre-eclampsia and pregnancy-induced hypertension in descending order [12].

The placenta and potential fetal drug exposure add additional complexity to this situation. Various methods can be used to study whether and to which extent a drug crosses the placenta, such as cell cultures, organ-on-a-chip techniques, ex vivo experiments and in silico models [13]. Ideally, the information obtained from such experiments is integrated in pharmacometric models, such as physiologically based pharmacokinetic (PBPK) models, to improve the understanding of a drug’s PK and PD profile in pregnant women. Yet, although PBPK models have retrospectively demonstrated a good predictive performance, in vivo concentration data from clinical trials are still needed for confirmation and increasing the confidence in these models which requires interdisciplinary collaboration between clinicians and pharmacometricians [14]. In vivo data on fetal exposure are generally limited to single umbilical cord blood sampling at delivery, preferably with paired maternal sampling.

Even if coping well with the logistics of conducting a randomized controlled trial including pregnant women to study mechanisms governing a drug’s PK and PD, the effects of the physiological changes in pregnancy will need to be taken into consideration in the study design, data collection and interpretation. In this review we provide an overview of the systems that are affected in pregnancy and how each can contribute to changing certain variables related to the pharmacokinetics and pharmacodynamics of frequently used drugs during pregnancy.

Cardiovascular system

The cardiovascular system undergoes several significant anatomical and physiological changes during pregnancy. The heart is displaced more laterally and to the upper left of the chest by the effect of the progressively elevated diaphragm. Moreover, the left ventricle muscle wall is enlarged to accommodate the increase in blood volume, and as a result, the ejection fraction remains unchanged, maintaining compliance [15]. In terms of hemodynamic variables, the end-systolic volume and end-diastolic volume are increased, whereas the end-systolic and end-diastolic pressures remain the same [15]. On cardiac exam, there is evidence of changes in physiology, including mild tachycardia, peripheral edema, jugular venous distention, and lateral displacement of cardiac apex [16]. As for heart sounds, a louder first sound with the exaggerated splitting of the second sound and a third heart sound can be heard [16]. Also, systolic murmurs can be heard at the left lateral sternal border in 90% of pregnant women, and this is considered a flow murmur, reflecting the increased blood volume passing through the aortic and pulmonary valve [17].

Cardiac output denotes the amount of work performed by the heart and is the product of stroke volume and heart rate. It is affected by the blood volume, heart rate, and autonomic regulation of the heart and vessels, yet the exact contribution of each remains controversial [18, 19]. Cardiac output increases in a non-linear fashion at the beginning of the pregnancy, reaches its peak by early third trimester which can be up to 45% in a singleton pregnancy, then decreases slightly towards term (Table 2) [20, 21]. That early increase is more pronounced in multiple gestations [22]. During the first trimester, there is a reduction in the systemic vascular resistance (afterload) stimulating the sympathetic nervous system and leading to an increase in heart rate. Thus, a reduction in afterload is responsible for the increase in cardiac output during the first trimester as opposed to the previous attribution to change in stroke volume [20, 23].

Table 2.

Summary of key physiological changes during pregnancy.

System	Parameter	Direction of change (quantitative) [30, 183, 184]
Cardiovascular	Cardiac output	Increased (20%–45%)
	Heart rate	Increased (10 to 20 bpm)
	Stroke volume	Increased (20% to 30%)
	Plasma volume	Increased (30% to 50%)
	Blood Pressure
	Systolic	No change
	Diastolic	Decreased (10 mmHg)
Respiratory	Respiratory rate	No change
	Residual volume	Decreased
	Tidal volume	Increased (up to 40%)
	Minute ventilation	Increased (up to 50%)
	Functional Residual Capacity	Decreased (20%)
	Oxygen Consumption	Increased (20%)
	FEV1	Unchanged
Gastrointestinal	Serum Albumin	Decreased (20% to 40%)
	Gastric Emptying	Prolonged
	Hepatic artery blood flow	No change
Renal	Glomerular filtration rate	Increased (up to 50%)
	Renal Plasma Flow	Increased (up to 50%)
	Total Body Water	Increased (up to 20%, 75% in ECF and 25% in blood)

FEV1=Forced Expiratory Volume at end of 1 second; ECF= Extracellular fluid.

During the third trimester, the enlarging size of the gravid uterus compresses the vena cava [19, 24]. In addition, the blood flow to the utero-placental circulation is also at its peak in the third trimester [25]. Both factors contribute to decreasing the preload and might lead to a decrease in cardiac output during the third trimester. However, this is offset by the increase in the resting heart rate. Moreover, there is preferential blood flow to the uterus and placenta (up to 10-fold increase in uterine arterial blood flow) near term [26–29]. During labor, the cardiac output increases again as a result of increased blood volume accompanying contractions, and the phenomenon of “auto-transfusion ” as blood from the utero-placental unit is shunted back to the maternal circulation immediately after delivery [30].

The systolic blood pressure remains the same, but diastolic blood pressures decrease during pregnancy, reaching a nadir at 24–26 weeks and then rises again [23]. Blood pressure variations are common, and it is dependent on the position of the mother and on the systemic vascular resistance which decreases due to the interplay of progesterone, nitric oxide, endothelin, and estrogen [31]. The mean arterial pressure is determined by the systemic vascular resistance and cardiac output and serves as an indicator of organ perfusion. The autoregulation compensates the decrease systemic vascular resistance by increasing the cardiac output. [32]. In the case of pulmonary pressures, the pulmonary vascular resistance is decreased; however, the pulmonary capillary wedge pressure remains the same. This is advantageous because it allows to accommodate the increased cardiac output. In case of a pulmonary vascular disease during pregnancy, the increased resistance would lead to pulmonary artery hypertension which will not tolerate the increased cardiac output and would result in right sided heart failure [33]. Supine hypotension is a known physiological change that occurs in 10% of women during pregnancy as a result of compression of great vessels by the gravid uterus [34] and must not be mistaken for true hypotension.

There is an increase in the plasma volume during pregnancy to accommodate the increased needs of the placenta and the fetus [35, 36]. The plasma volume increase is evident as early as 6–8 weeks, and continues to increase throughout the pregnancy, peaking at 32 weeks [37]. This also causes an increase in the apparent volume of distribution of drugs, which implies that hydrophilic drugs would require a higher initial dose and maintenance dose to obtain plasma concentration levels that are comparable to those in non-pregnant women [38]. In addition, there is a 13% decrease in albumin concentration from non-pregnant levels by 32 weeks [39], which may cause an increase in the unbound drug concentration; this is very important clinically as drugs which are highly bound would have much higher concentrations of their free form during pregnancy (i.e. digoxin and phenytoin) [28, 30]. In addition, the α1-glycoprotein, which is a binding protein for many basic compounds, slightly decreases during healthy pregnancy [40].

Gastrointestinal System

Heartburn is common during pregnancy. This can be attributed to the decrease in gastric secretion pH, increase in the amount of secretions along with a decrease in the lower esophageal sphincter tone [41]. Lower esophageal sphincter tone decreases as a result of progesterone action on smooth muscle cells and is responsible not only for the heartburn symptom but also contributes to nausea and vomiting [42]. Nausea and vomiting of pregnancy is also common in around 80% of pregnant patients but with variable severity and presentation [43, 44]. It can present as early as the 2^nd week and last up until the 2^nd trimester, and in some cases until 37 weeks’ gestation or full term [45–47]. This is important to keep in mind as clinicians need to control the symptoms before the absorption of medication and bioavailability of oral drugs is altered. For that reason, it is recommended that patients take their medication at a time when nausea is minimal to avoid decreased bioavailability of the drug due to lower concentrations at the absorption site [38].

There is a controversy as to whether gastric emptying and motility are delayed during pregnancy and if that impacts the peak serum concentration of oral medication. Some studies showed that the delay is a result of increased progesterone and estrogen levels, causing the relaxation of smooth muscle cells in the alimentary tract [48, 49]. This decrease can often lead to constipation and bloating and cause discomfort to patients in addition to its role in augmenting nausea and vomiting during pregnancy resulting theoretically in a delay in the peak concentration of oral drugs in the system. Other studies have noted delayed gastric emptying specifically early in pregnancy [50, 51], compared to those showing no delay in motility throughout pregnancy [50, 52]. Absorption of certain medications, such as acetaminophen, can be affected by delayed gastric emptying during pregnancy. Furthermore, concomitant use of some medications may alter absorption profiles for both pregnant and non-pregnant patients [53]. For example, due to the fact iron supplements and proton pump inhibitors may decrease levothyroxine absorption, patients are instructed not to take them concurrently [54].

The physiological effect of pregnancy on liver enzymes involved in drug metabolism varies [55, 56]. Several phase I enzymatic activities are either enhanced or decreased. For example, activity of the CYP2D6 enzyme is increased resulting in increased clearance of fluoxetine, one of its substrates, which might require dose adjustment to reach a pharmacological effect [57]. Other phase I enzymes, such as CYP2A6, 2C9 and 3A4 appear to be induced as well resulting in decreased plasma concentrations of drugs metabolized via these enzymes, such as cotinine, phenytoin and midazolam [58–60]. Of note, as illustrated for phenytoin[59],a highly protein-bound anticonvulsant, the increase in total drug clearance can be further enhanced by increases in the free fraction which stems itself from reduced serum albumin levels. On the other hand, a decreased activity of drug metabolizing enzyme CYP1A2 will result in a decrease in the clearance of caffeine, olanzapine, and clozapine, and thus dose adjustment is needed to avoid toxic concentrations [57]. The activity of essential enzyme in the phase II metabolism pathway of the liver, including many members of the uridine diphosphate glucuronosyltransferase (UGT) 1 family [61] is also altered. For example, UGT1A1 appears to be induced by increased progesterone concentrations during pregnancy,, while others, such as UGT2B7 appear to be unchanged [62]. Labetalol, an α−1 and β-adrenoreceptor blocker and a substrate of UGT1A1, is an important drug for treating elevated blood pressure in pregnancy (Table 3) [63]. It is believed to have a lower availability due to increased hepatic clearance and first pass excretion, with a bioavailability averaging 20–40%. Its clearance is dependent on phase II metabolism that is enhanced during the second and third trimester compared to the post-partum period, which is similar to metabolism of acetaminophen during pregnancy [62, 64]. As a consequence, labetalol is often dosed using twice or three times daily regimens.

Table 3.

Pharmacokinetic variables of selected clinical drugs commonly used pregnancy.

Drug	Trimester	Comparison group	Overall observed PK changes			References
			VD	CL	t1/2
Ampicillin	3rd	Post-partum	↑38%‡	↑50%‡	↓12%‡	[185]
Cefazolin	2nd	Post-partum	NR	↑31‡	↓35%‡	[186]
Enoxaparin	All	Non-pregnant	↑48%‡	↑49%‡	↑13%	[187]
Fluoxetine	3rd	Historical controls	NR	NR	NR	[188]
Lamotrigine	3rd	Non-pregnant	NR	↑140‡	NR	[189]
Levothyroxine	All	No grouping	NR	no change in dose	NR	[190]
Labetalol	1st /term	Post-partum	↑90%	↑40%‡&60%‡ respectively	NR	[191]
Nifedipine	3rd	Non-pregnant	↑ (magnitude unspecified)	↑308%	↓(magnitude unspecified)	[192]
Phenytoin	3rd	Post-partum	NR	↑256%	NR	[59]

^‡statistically significant

PK=Pharmacokinetics, NR= not reported, VD= Volume of Distribution, CL=clearance, t_1/2=Half-life, IV=intravenous, IM=Intramuscular.

Renal system

During pregnancy, the kidneys are in a state of glomerular hyperfiltration since there is increased plasma volume, increased effective renal plasma flow (eRPF), and decreased renal plasma oncotic pressure. The eRPF increases by about 80% at the end of the first trimester as a result of massive vasodilation from relaxin-mediated nitric oxide release [65, 66]. This is clinically manifested by a decrease in creatinine levels by an average of 0.4 mg/dL reflecting an increased glomerular filtration rate (GFR), which can be up to 50% from baseline [67]. At the beginning of pregnancy, the GFR and eRPF both increase, however, the eRPF declines during late pregnancy while the GFR plateaus from the second trimester till birth and returns to normal levels in the postpartum period [67, 68]. Frequency and urgency, which are common symptoms seen even early in pregnancy, are believed to be due to these changes in GFR and eRPF [69]. Later in pregnancy frequency and urgency are also secondary to the physical compression of the bladder by the growing uterus and fetus. Serum creatinine is a rapid evaluation of glomerular filtration, though a 24-hour urine collection sample is the more accurate assessment of creatinine clearance [70].

Furthermore, this physiological change can be deemed cumbersome when dealing with renally cleared drugs. Lithium, which is prescribed as a mood stabilizer for bipolar patients, is one of the renally cleared drugs of which its clearance is doubled during the third trimester of pregnancy, with a fast return to pre-pregnancy clearance in the first 2 weeks after delivery [71, 72]. Other drugs, among the many that can be affected, include cephalosporins, piperacillin, atenolol, and digoxin [71]. Although the change in GFR is uniform as the pregnancy progresses, it is noted that the renally-excreted drugs are not affected equally, and that is because of the role of renal transporters responsible for secretion and reabsorption in the tubules and changes in the plasma protein binding of some drugs [73].

Urinalysis in pregnancy may show glucosuria resulting from increased GFR and impaired resorption at the proximal tubule level. However, even though it might be physiologic, the finding of glucosuria always warrants investigation for pregnancy related diabetes mellitus [74]. For the same reason, proteinuria is commonly found on urinalysis during pregnancy with values below the threshold of 300 mg/dL, and any value above that would require investigational studies to rule out pregnancy-specific abnormalities. This cut-off is double that of the non-pregnant states [75, 76].

Water and sodium regulation is changed during pregnancy. Despite the rise in all components of the renin-angiotensin system, resistance to pressor effects contribute to a more substantial increase in extracellular volume along with water and sodium retention [77]. As a result, plasma osmolarity is slightly decreased to an average of 280 mOsm/kg with an average value about 270 mOsm/kg and sodium to 136 meq/dL [78]. The body retains sodium despite the increased filtration by the kidneys, through upregulation of mechanisms responsible for reabsorption in the distal tubules.

In addition to the physiological changes, there are several anatomical modifications to be noted in pregnancy. The size of the kidney increases by 1 cm at the end of the pregnancy, as a result of the increase in vasculature and plasma volume [79]. Moreover, the renal calyces and ureters are dilated occurring more than 80% of pregnant women by mid-gestation and more commonly on the right side. Clinically, this is reflected by an increased risk of ascending urinary tract infection from urine stasis and sonographic difficulties in differentiating pathology from normal changes like hydronephrosis [80–82].

Endocrinology system

The several metabolic changes occurring during pregnancy are essential for meeting the demands of the growing fetus and placenta. By the third trimester, the basal metabolic rate is increased by 20% [83]. There exists a state of hyperinsulinism from hyperplasia of islet cells in the pancreas along with a state of peripheral insulin resistance. Human placental lactogen and the human placental growth hormone are thought to contribute to insulin resistance [84, 85]. The former acts as an insulin-like and anti-insulin molecule causing pancreatic cells to secrete insulin, and at the same time inducing peripheral resistance to insulin [84]. This ensures glucose availability for the fetus by keeping the postprandial glucose levels elevated [86, 87]. Also, augmentation of hepatic gluconeogenesis in the third trimester contributes to sustaining the postprandial glucose levels [88]. Patients with pre-existing insulin resistance or diabetes will need to adjust their insulin medication to higher doses as the pregnancy progresses [89].

Maternal lipid levels are increased under the effect of insulin resistance and estrogen [90]. During the first part of pregnancy, lipid synthesis is favored, and stored fat is increased. However, during the third trimester, the fat stored will be used by the mother for energy production [91]. This is important since fatty acids and glycerol serve as the nutritive fuel for the mother, whereas glucose and amino acids are preferentially utilized by the fetus [86]. In addition to the metabolic status, several hormone-secreting organs like the pituitary and thyroid undergo drastic changes. Estrogen induces hyperplasia and hypertrophy of lactotrophs resulting in an enlarged pituitary gland with prolactin levels, subsequently increasing as well [92, 93]. Also, there is a decrease in gonadotrophs from increased progesterone and estrogen. However, the corticotrophs and thyrotrophs remain the same [94]. The placenta also contributes to the hormonal milieu, by secreting its form of the growth hormone synthesized by the syncytiotrophoblast. The placental growth hormone is increased in mid-pregnancy, stimulating the release of insulin-like growth factors which contribute to the acromegalic features of some women during pregnancy [95].

Pregnancy is a state of hypercortisolism. Corticotropin-releasing hormone is synthesized by the placenta and increased until term [96]. This corticotropin-releasing hormone stimulates the release of both placental and pituitary adrenocorticotropic hormone, and characteristically, the former is insuppressible by the low-dose dexamethasone test [97, 98]. Moreover, the hepatic production of cortisol-biding globulins is upregulated under the effect of estrogen, ensuring an increased half-life and decreased hepatic clearance [99]. Consequently, cortisol levels are increased, and mainly, a cortisol surge occurs at labor [100].

The total triiodothyronine and tetraiodothyronine levels are increased during pregnancy as a result of an estrogen-induced increase in thyroid-binding globulin, decreased hepatic clearance and an increase in production of hormones by the thyroid gland [101, 102]. These hormones play a pivotal role in fetal neurogenesis. In particular, early pregnancy levels of thyroid hormones are positively correlated with cognitive and behavioral development in infancy and childhood [103]. Historically, there has been some controversy about treating subclinical maternal hypothyroidism and improvement of neurodevelopmental outcomes. A recent clinical trial definitively showed that neurodevelopmental outcome through 5 years of age is not significantly improved in the treatment arm [104]. Therefore, the American Congress of Obstetrics and Gynecology along with the American Thyroid Association do not recommend universal screening of thyroid dysfunction during pregnancy [105, 106]

The thyroid gland is increased in size during pregnancy. In addition, an iodine deficiency state ensues because of the active transport of iodine across the placenta to the fetus, increased renal excretion and increased consumption of iodine by the mother’s thyroid [107]. Despite the increase in total hormones, the free-from remains stable, maintaining a euthyroid status in the pregnant woman [108]. In the case of hypothyroidism, the dose of levothyroxine, if required, should be increased by 30% in early pregnancy to meet the maternal demands, and then usually returns to pre-pregnancy doses postpartum to avoid excessive exogenous thyroid hormone exposure [109].

Hematologic changes

As discussed earlier, the plasma volume increases in pregnancy nonlinearly and non-monotonically. It increases by 15% in the first trimester, and more rapidly during the second trimester, continuing at a slower rate during the third trimester till it plateaus in the few weeks before delivery [110, 111]. As a result, there is an average gain of 30 to 50% or more than one liter of blood volume compared to non-pregnant levels [112]. This is advantageous for mother and fetus in response to the increased metabolic demands of pregnancy, and more importantly, it serves as a protective buffer to minimize consequences of large volume blood loss encountered during delivery.

As for blood components, the upregulation of erythropoiesis, induced by erythropoietin hormone from the maternal kidneys, increases red blood cell (RBC) mass. However, this increase lags behind that of the plasma volume resulting in dilutional anemia, or more clinically known as “anemia of pregnancy” [113]. One advantage of this is that the blood viscosity decreases, with increasing carrying capacity of the RBCs, ensuring a more efficient exchange between mother and fetus [114]. The hemoglobin averages at 12.5 g/dl, and despite physiologic anemia, any value below 11 g/dL should be investigated [115]. Iron requirements increase by the latter half of the pregnancy, and it is imperative to start the patient on iron supplements early on in pregnancy especially in the subset of patients who are anemic at the start of their pregnancy to avoid maternal and neonatal morbidity, particularly low birth weight or severe hemorrhage [116–118].

Lymphocytes are higher in pregnancy, averaging around 15,000/mm³, and this value is even higher as we approach labor and delivery [119]. For that reason, diagnosing an infection, especially at the time of delivery like chorioamnionitis using the lab values, may be difficult and the clinician should always support the lab values with the clinical picture. In addition to looking at the number of band cells or immature cells which are not elevated in pregnancy [30]. Although studies have documented changes in platelet count, their level remains within the normal range [120]. It is believed that the drop in platelets is due to hemodilution along with an element of hypersplenism from splenic enlargement, reaching up to 50% during the first trimester, and resulting in platelets sequestration [121–123].

Pregnancy is a hypercoagulable state as evident by upregulation of procoagulation factors like all clotting factors, and especially thrombin (factor II) and fibrinogen (factor I) [124–126]. Moreover, regulatory proteins responsible for anticoagulation such as protein S and anti-thrombin are decreased during pregnancy, and there is increased resistance to activated protein C [127]. The result of the interplay between coagulation and anticoagulation pathway is evident, and this might play a role in decreasing blood loss during delivery and afterwards. However, pregnant women are at increased risk of venous thromboembolism in both pregnancy and postpartum period, and that is a leading cause of maternal death [128–131]. It is maybe that the thrombosis of pelvic veins offers an advantage for facilitating normal postpartum involution, as anti-thrombin falls and reaches a nadir at 12 hours postpartum before rising again and plateauing at 72 hours [132, 133]. Low-molecular-weight-heparin has become the traditional drug of choice of treating venous thromboembolism in pregnancy. Most women go on twice daily dosing for anticoagulation during pregnancy or postpartum, however, a population pharmacokinetic study was able to show efficacy of once-daily dosing, adjusted to weight, in treatment of venous thromboembolism [134]. Often women are continued for low-molecular-weight-heparin for a duration of at least 6 weeks depending on the indication. Further research is needed in the area of optimal anticoagulation dosing especially for women with active thromboembolism who have additional comorbidities such as morbid obesity.

Respiratory System

Increasing progesterone during pregnancy induces a change in the threshold of the respiratory centers of the brain, increasing the sensitivity to carbon dioxide, which is reflected by the slope of the ventilation curve in response to the changes in alveolar carbon dioxide [135, 136]. In addition to that, progesterone also mediates dilation of the respiratory airways and hyperemia and edema of the mucosal surfaces causing nasal congestion, resulting in rhinitis of pregnancy [137–139]. The effect of estrogen is by upregulating the progesterone receptors in the central nervous system, particularly in the medulla and hypothalamus, where the respiratory control center resides [140].

Anatomically, the upward displacement of the diaphragm as a result of the growing gravid uterus leads to a decrease in functional residual capacity [141]. Not only the mechanical effect causes this change, but it has been shown that the displacement occurs earlier in pregnancy when the uterus is not enlarged, and it is attributed to the effect of progesterone and relaxin on inducing the ligamentous attachments of the lower ribs to relax [142]. Other anatomical changes include an increase in the subcostal angle by 50% by the end of pregnancy from 68.5 to 103.5, and the transverse diameter of the rib cage increases by 2 cm [142]. Eventually, the chest wall compliance decreases by the third trimester due to increased abdominal content; however, lung compliance remains the same [143, 144]. Despite the changes in the anatomy, inspiratory, and expiratory maximum pressure values remain preserved throughout pregnancy [141]. Dyspnea is a common presentation during pregnancy as this physiological dyspnea occurs in 50–70% of patients by the third trimester [145].

As for the pulmonary function test, the tidal volume is increased by up to 50% (or 650 ml) of the non-pregnant values, but the respiratory rate remains the same. Consequently, the minute ventilation, which is the product the tidal volume and respiratory rate follow with the trend and increases to up to 50% [146]. The functional residual capacity (FRC) decreases by 20 to 30 percent during pregnancy, as a result of diaphragm elevation decreasing outward recoil of the chest. Pregnant women are at risk of hypoxemia during general anesthesia induction as a result of this decrease in FRC [147, 148]. The expiratory reserve volume and residual volume make up the FRC, and both decrease by about 20% [149]. On the other hand, the inspiratory capacity, which is defined as the maximum inhaled volume from FRC, is increased by 10% and the total lung capacity (a combination of Inspiratory Capacity and FRC) remains the same. These changes do not differ between singleton and twin pregnancy [150].

In pregnancy there is a state of respiratory alkalosis with maternal pH ranging between 7.42–7.46 [151]. Arterial blood gases change as a result of the increase in minute ventilation. The PaO₂ is increased from 93 to 106 mmHg, and the PaCO₂ is decreased by 37 to 30 mmHg by the end of the third trimester [151–153]. The latter decrease provides an advantageous gradient between mother and fetus, allowing CO₂ of the fetus to be transferred to the mother. Fetal acidemia, among other factors, can affect local placental transportation of weakly basic compounds like the anesthetic drugs bupivacaine and meperidine, and can result in the entrapment of these compounds in their ionic form [154, 155]. The renal system compensates by increasing the excretion of bicarbonate, which its serum levels decrease from 23 to 18 mEq/L in late pregnancy [156]. Moreover, the chronic state of respiratory alkalosis in pregnancy would shift the oxyhemoglobin curve to the right, which facilitates the oxygen diffusion across the placenta [157].

Volatile anesthetic agents with minimum alveolar concentration (MAC) values more than 1.5 MAC can lead to dilation of uterine arteries and also reduce uterine tone, and pregnancy has been associated with increased sensitivity to these agents. Most induction agents, opioids, and neuromuscular blockade can be used in pregnancy. Thiopental is the most commonly used agent during rapid-sequence intubation in pregnancy with a lower dose than generally used [158]. More recently, propofol is also used as an alternative agent in early pregnancy after showing no teratogenicity in animal studies [159].

Placenta

The placenta was perceived initially as a protective barrier at the maternal-fetal interface. It was not until fetal malformations were attributed to thalidomide exposure in the early 1960s that physicians started discovering the transport of xenobiotics and other drugs across the placenta from mother to fetus and vice versa [160, 161]. It is well established that uncharged, lipophilic drugs with a relatively low molecular weight can pass readily, and the amniotic fluid surrounding the fetus can also be considered a site for drug distribution and accumulation [6]. However, newer studies are showing evidence of drug transporters in the placenta that can be responsible for active transmission of drugs from the mother to the fetus [162]. In fact, multiple drug transporters, such as P-glycoprotein and the breast cancer resistance protein, are expressed in the membrane of the fetal syncytiotrophoblasts and endothelium at varying levels during gestation [13].

Although the maternal and fetal blood never meet, many nutrients, waste products, and other molecules can and should pass through the brush border (apical) membrane of syncytiotrophoblast cells at the interface [163]. Three types of drug transport categories are recognized: the complete transfer (type 1 drugs), which indicates that a drug can readily cross the placenta, and its concentration will rapidly equilibrate between mother and fetus, or exceeding transfer (type 2 drugs) in which greater concentration is reached in the fetus compared to the mother, and the last is incomplete transfer where the drug is unable to pass through the placenta completely [164].

Like any other molecule, drugs depend on particular mechanisms to pass through the cell membrane. Simple diffusion, like in the cases of midazolam and acetaminophen, is the most common mechanism of drug transport across the placenta, especially for type 1 drugs. Another mechanism is facilitated diffusion like in the case of cephalosporins and corticosteroids, and active transport like the case of norepinephrine or dopamine [164]. However, despite the possibility of drug transport, some medications cannot cross the barrier and reach the fetus. An example of that is doxorubicin, a chemotherapeutic agent involved in several hematologic malignancy treatment protocols [165]. The transfer of molecules across the placenta is also dependent on size, where drugs with a molecular weight > 1000 Da (e.g. low-molecular-weight-heparin and insulin) usually do not cross the placenta [166, 167]. Monoclonal antibodies are a specific group of drugs used to treat oncological or a variety of inflammatory diseases, or to prevent infections in mother or infant (influenza, pertussis).The human placenta has the capacity of active transfer of antibodies from mother to fetus, and this may result in developmental toxicology [168, 169] However, the extent of transfer may also depend on biochemical modifications like PEGylation of the Fc fragment of the antibody. Furthermore, there is some controversy over their use in pregnant women with chronic conditions where the potential benefits outweigh the risks [170].

Summary of physiological changes and model applications

In addition to the previous changes described (Tables 1 & 2), trimester-specific changes are summarized in Figure 1. It is important to note that these percent changes are compared to pre-pregnancy state. These trimester-specific changes may not be valid in disease states like preeclampsia. Pathophysiological changes during preeclampsia result in an additional increase in body weight due to water retention [171], an additional decrease in albumin [172] and impaired GFR (about −15 %) when compared to the normal physiological increase in GFR [173].

Fig 1.

Percentage Change in Physiological Parameters across the Three Trimesters. Percent changes have been calculated at 12, 26 and 38 weeks based on the individual trend lines as described by Dallmann, A. et al. to reflect the first, second and third trimester [193]

HCT=hematocrit; AAG= α−1 acid glycoprotein; GFR= glomerular filtration rate.

During the postpartum period, many of these physiological changes return back to normal and for some this return may be gradual. For example, blood volume decreases rapidly in the first few days under effect of diuresis, and continues to decrease gradually over a period of eight weeks [174]. Labor increases the cardiac output, along with its components (heart rate and stroke volume), to an additional 80% from pre-labor values, and eventually cardiac output will return to normal within 6–8 weeks of pregnancy [175]. Postpartum GFR normalizes to pre-pregnancy estimates within 4 weeks after delivery [173]. Moreover, as uterine size decreases, the FRC also returns to normal within one to two weeks. In cases of chronic medical conditions, such as hypertension, women either resume their pre-pregnancy medications in the postpartum period or commonly stay on labetalol or nifedipine (Table 3) [176]. There is limited data for safety with breastfeeding and use of angiotensin-converting enzyme inhibitors and angiotensin II receptor blocker agents and often these are initiated by a primary care provider at some point beyond the 6 weeks after delivery. Caution is warranted for using thiazide drugs postpartum if they are breastfeeding due to a potential for decreasing breast milk production. The optimal pharmacologic management of diseases in this critical period of recovery from pregnancy is not well studied and it is a new front to discover, as breastfeeding and subsequent maturational changes in intestinal absorption of the infant is an additional factor in the complex model.

Application of mechanistic models for pregnancy

Several modeling tools may help to shed light on changes in pharmacokinetics and pharmacodynamics stemming mainly from the herein discussed physiologic alterations. Mechanistic models, such as PBPK models constitute a strong tool to leverage the accumulated knowledge about physiologic alterations in pregnant women within a biologically plausible framework. For example, previous studies have demonstrated that information obtained from ex vivo perfused placental cotyledons can be integrated in PBPK models to successfully predict and quantify fetal exposure to antiretroviral drugs [177, 178]. Other applications encompass the prediction of a maximum recommended dosage of lithium for pregnant women for treatment of bipolar disorder [179]; the prediction of a worst-case scenario for maternal exposure to the toxic metabolite N-acetyl-p-benzoquinone imine (NAPQI) following acetaminophen intake at various stages throughout pregnancy [180]; and the prediction of caffeine pharmacokinetics and pharmacodynamics in the fetus to support the identification of potential cut-offs for caffeine intake during pregnancy [181]. While PBPK models typically focus on pharmacokinetic and pharmacodynamics processes at the organ level, biologically based dose-response (BBDR) models shift the focus to a finer biological scale by incorporating data on biological processes at the cellular and molecular level. These models can be used to study the interactions between compounds and the biological system and to link exposure to a quantifiable biological response. For example, a BBDR model for the thyroid endocrine system was recently used to evaluate perturbations of maternal serum levels of free thyroxin following a range of dietary iodide intake. Through global sensitivity analysis, key parameters of the model were identified that critically drive maternal and fetal iodide kinetics, the thyroid function and their interactions thereby contributing to an improved mechanistic understanding of the thyroidal system during pregnancy [182].

These examples illustrate how modeling approaches can be used to integrate available data and generate new insights into the complex interplay of physiologic alterations in pregnant women and the functioning of the physiological system, and the pharmacokinetic and pharmacodynamics processes of a drug. In combination with clinical data for model development and evaluation, these tools can help to better understand and anticipate clinically relevant changes drug exposure and efficacy. Ultimately, this can lead to a more informed decision-making in the clinical setting and improve therapeutic outcome in pregnant women.

Conclusion

The myriad of changes occurring in pregnancy challenges clinicians to provide the optimal care when prescribing medication in pregnancy. Most notably, changes such that of the gastrointestinal and renal system, along with the increased plasma volume, decreased albumin plasma concentrations and altered enzymatic activity play a key role in determining the pharmacokinetic variables of medication during pregnancy. Understanding these changes is pivotal for the clinician when managing patients that might develop complications such as pre-eclampsia and hemorrhage or have pre-existing medical conditions. To date, pregnant women are still being excluded from many clinical trials studying pharmacokinetic and pharmacodynamic properties of medications. Extrapolating dosages and safety profiles of medications from the non-pregnant population can be misleading or inaccurate and inappropriate dosing foreshadows minor to profound adverse effects. Further pharmacometric and efficacy studies in pregnancy are needed to better understand optimal dosing, so that toxicities are minimized for the mother and the fetus.