Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Variability in drug efficacy and safety is multifactorial. Both the pharmacokinetics (how the body handles the drug) and the pharmacodynamics (how the body responds to the drug) play significant roles in drug efficacy and safety. This chapter will discuss the effects of pregnancy on medication pharmacokinetics.
The physiologic changes that occur during pregnancy result in marked changes in the pharmacokinetics for some medications. Whether or not the physiologic changes will result in clinically significant pharmacokinetic changes for an individual medication depends on many factors. The discussion of these factors will be the focus of this chapter. Generally speaking, pharmacokinetic changes are most important clinically for medications with narrow therapeutic ranges. The therapeutic range includes all the concentrations above the minimum effective concentration, but less than the maximum tolerated concentration ( Fig. 3.1A and B ). Medications such as cyclosporine, tacrolimus, lithium, warfarin, carbamazepine, valproic acid, phenytoin, digoxin, vancomycin, and the aminoglycosides are examples of narrow therapeutic range drugs. These are medications for which the concentrations needed for therapeutic benefit are very close to those that result in toxicity. For these agents, small changes in drug concentrations can lead to inefficacy if the concentrations decrease or intolerable toxicity if the concentrations increase. Typically, when drug interactions, disease states or conditions alter the concentration-time profile for a medication, if no changes have occurred in the pharmacodynamics, the patient's dosage is adjusted to keep the concentrations similar to those prior to the altered state or similar to those for the population in which the drug has been approved. This dosage adjustment is done to maintain concentrations within the therapeutic range. For narrow therapeutic range medications, even a 25% change in drug concentration can be considered clinically significant. In contrast, for most medications, which have wide therapeutic ranges, small changes in pharmacokinetics have little to no clinical effect. However, given the magnitude of some of the pharmacokinetic changes that occur during pregnancy in which there can be 2–6-fold changes in drug exposure ( Fig. 3.2A ), even medications that have wide therapeutic ranges can be clinically affected.
A change in pharmacokinetics for a medication can result in the need to change dosage. As described above, altered concentrations during pregnancy can result in the need for higher ( Fig. 3.2A ) or lower ( Fig. 3.2B ) drug dosage to maintain concentrations within the therapeutic range. The changes in medication pharmacokinetics during pregnancy in some cases are so great that altered medication selection should be considered. For example, oral metoprolol concentrations are 2–4-fold lower during pregnancy than in the nonpregnant state [ , ]. Given the magnitude and variability in metoprolol concentrations during pregnancy, for those patients that require a beta blocker (e.g., cardiac rate control), selecting another agent such as atenolol, which is renally eliminated, should be considered. Even with the changes in renal function that are expected during pregnancy, atenolol will give much more consistent and reliable drug concentrations in pregnant patients than metoprolol [ ]. Although there are fetal risks with the utilization of beta blockers during pregnancy, such as intrauterine growth restriction, if a beta blocker is required during pregnancy, selecting an agent that will consistently and reliably achieve the desirable therapeutic effect requires consideration of pharmacokinetic changes in medication selection.
The following sections will discuss the commonly estimated pharmacokinetic parameters, their application, and how they might be altered during pregnancy. The actual calculation of these parameters will not be discussed in this chapter. However, the reader is referred to the many publications that discuss in detail the mathematical equations used to determine the pharmacokinetic parameters [ , ].
Extraction ratio (ER) is the fraction of drug that is removed from the blood or plasma as it crosses the eliminating organ (e.g., liver or kidney). Knowing whether a drug has a high (ER > 0.7; e.g., morphine, metoprolol, verapamil), intermediate (ER 0.3–0.7; e.g., codeine, midazolam, nifedipine, metformin, cimetidine), or low (ER < 0.3; e.g., phenytoin, indomethacin, cyclosporine, amoxicillin, digoxin, atenolol) ER is important in predicting which factors such as intrinsic clearance, protein binding, and/or blood flow will alter the pharmacokinetic parameters for the drug.
The area under the concentration-time curve is a measure of the overall systemic drug exposure ( Fig. 3.1B ). Since we rarely can measure the drug concentration at the site of action (e.g., brain, lung, or heart); blood, plasma, or serum concentrations are typically used to determine systemic drug exposure. The area under the concentration-time curve (AUC) is dependent on the dose, clearance, and bioavailability of the drug. For some medications, AUC is the key determinant of medication efficacy and safety; while for other medications, either the maximum concentration and/or minimum concentration are better correlated with outcomes. For low ER drugs (both oral and intravenous administration), an increase in enzyme activity and/or a decrease in plasma protein binding will lead to a lower total drug AUC with changes in blood flow having no effect. For high hepatic ER, intravenously administered drugs, a decrease in blood flow will increase the total AUC; whereas, enzyme activity and protein binding have no effect on the total AUC. For high hepatic ER, orally administered drugs, the decrease in clearance caused by a decrease in blood flow is equal to the decrease in bioavailability such that changes in blood flow have no effect on oral AUC. However, increased enzyme activity or decreased plasma protein binding will decrease the total AUC through their effect on oral bioavailability.
Bioavailability is the fraction of the dose administered that reaches the systemic circulation unchanged. Sometimes, the bioavailability term is used to encompass both the rate and extent of absorption from the site of administration to the systemic circulation. For orally administered drugs, the bioavailability is affected by the amount of drug that is absorbed from the gut as well as first pass metabolism as the drug crosses the intestine and liver on it's way to the systemic circulation. An increase or decrease in bioavailability directly impacts the oral AUC or total drug exposure. For low hepatic ER drugs, bioavailability is not affected by enzyme activity, hepatic blood flow, or protein binding. In contrast, for high hepatic ER drugs, bioavailability is decreased by an increase in enzyme activity, decreased hepatic blood flow, and/or a decrease in plasma protein binding. In addition to the above-described changes in enzyme activity, protein binding and blood flow which can alter medication pharmacokinetics, other physiologic changes that occur during pregnancy which might influence the bioavailability of drugs include: gastric acidity, gastrointestinal transit time, and hypertrophy of duodenal villi, which can alter drug absorption [ ].
Clearance is a parameter used to describe how well the body can metabolize or eliminate drugs. The clearance directly affects total drug exposure as well as average steady state drug concentrations and is utilized to determine maintenance dosage. There are three major determinants of hepatic drug clearance: hepatic blood flow, protein binding, and the intrinsic activity of hepatic drug metabolizing enzymes. Hepatic blood flow plays an important role in determining the hepatic clearance of drugs, particularly those with high ERs. Physiologic, pathologic, and drug-induced changes in hepatic blood flow can alter the systemic clearance and oral bioavailability of many important therapeutic agents, resulting in changes in patient response. For high hepatic ER drugs, clearance is directly affected by hepatic blood flow such that an increase in blood flow will increase clearance. The rate-limiting step for metabolism of high hepatic ER drugs is the delivery of the drug to the liver. Visualizing this process in which everything that is delivered to the eliminating organ, such as the liver, will be cleared from the body can be helpful. This process will proceed so that the faster the drug is delivered to the eliminating organ, the faster the drug is eliminated from the body.
In contrast, for low ER drugs, the rate-limiting step is not blood flow; therefore, a change in organ blood flow does not alter clearance. Instead, clearance is affected by the enzyme activity and protein binding, such that an increase in enzyme activity or a decrease in protein binding will increase the drugs total clearance. For intermediate ER drugs, clearance will be dependent on changes in enzyme activity, protein binding, and organ blood flow.
As described above, plasma protein binding can affect the pharmacokinetics of medications. There are multiple issues to consider with regards to protein binding of medications. Some of the plasma proteins are known to be altered both in normal pregnancy as well as pathologic conditions [ ]. In normal pregnancy, albumin concentrations decrease by approximately 1% at 8 weeks, 10% at 20 weeks, and 13% at 32 weeks [ ]. In pregnant patients with pathologic conditions, albumin concentrations can be substantially lower. Changes in albumin concentrations are important for many medications (e.g., phenytoin, valproic acid, carbamazepine). Other plasma proteins such as α-1-acid glycoprotein are involved in binding of drugs like betamethasone, bupivacaine, lopinavir, and lidocaine. Plasma α-1-acid glycoprotein has been reported to be 52% lower in late pregnancy (30–36 weeks gestation) than postpartum (2–13 weeks) [ ]. In addition, some agents (e.g., cyclosporine, tacrolimus) concentrate within the red blood cells. For these agents, binding might be altered as a result of anemia during pregnancy. Hematocrits are known to fall during normal pregnancy by 2% at 8 weeks and 4% at 20–32 weeks [ ]. For some medications, disease states or conditions during pregnancy can lead to severe anemia, which would be expected to have a much greater effect on binding of these medications. In contrast, there are plasma proteins which increase during pregnancy. For example, corticosteroid-binding globulin (CBG or transcortin) binds glucocorticoids such as cortisol and progesterone in the plasma/serum [ ] and is increased by 2–3-fold during pregnancy [ ]. The increase in CBG impacts the binding of drugs such as prednisone and prednisolone.
Drug binding is important for many reasons. The first reason is that the unbound drug is in equilibrium with the site of action and is therefore considered the active moiety as well as being able to cross membranes including the placenta. Unbound drug not only will cause beneficial effects but also potentially toxic effects. For drugs that are highly bound to albumin such as phenytoin, changes in albumin concentrations during pregnancy can be associated with alterations in protein binding. Yerby et al. reported a significant increase in the percent of unbound phenytoin during the second and third trimesters of pregnancy as well as labor and delivery as compared to the prepregnancy state [ ]. This is particularly important clinically because phenytoin is a highly protein bound drug with a narrow therapeutic range. The second reason is that understanding protein binding is critical in the interpretation of total drug concentrations. For phenytoin, when interpreting total drug concentrations, knowing whether protein binding has been altered is important. Changes in phenytoin protein binding would require either measuring unbound phenytoin concentrations or accounting for protein binding changes in the interpretation of total phenytoin concentrations.
Fig. 3.3A illustrates the total concentration for the drug in plasma measured to be 10, and the unbound concentration measured to be 1. In contrast, in Fig. 3.3B , the total drug concentration is 5, but the unbound concentration is still 1. In this example, although the total concentration is reduced in half, since the unbound concentration is still the same, no dosage adjustment should be made clinically because the active form of the drug (unbound) is the same. This would be expected to occur if there was a change in protein binding and no change in enzyme activity, leading to a change in total clearance, but no change in unbound clearance.
For prednisone and prednisolone, both are highly bound to albumin and CBG. The protein binding of prednisone and prednisolone has been shown to increase with the increase in CBG during pregnancy [ ]. The total prednisolone clearance is lower during pregnancy compared to postpartum; however, unbound prednisolone clearance did not change [ ]. Since there is no change in unbound prednisolone concentrations, then no dosage adjustment is required based on pharmacokinetics alone. However, this does not take into account changes in immune response during pregnancy. Prednisone and prednisolone binding is a saturable process both during pregnancy and in the nonpregnant state resulting in increased fraction unbound at higher concentrations along with increased variability [ ].
An alternate situation could occur in which there is no change in total concentrations, but a change in protein binding, leading to an increase in unbound drug concentration and toxicity. It is essential that in the case of a highly bound drug with a narrow therapeutic range to either measure the unbound concentration or to mathematically account for the changes in protein binding when total concentrations are measured, such as in pregnant patients with low-albumin concentrations. Tacrolimus binds to red blood cells, albumin, and α1-acid glycoprotein, all of which significantly decrease during pregnancy [ , ]. Mean tacrolimus oral clearance was 39% higher during mid- and late-pregnancy compared to postpartum [ ]. Since tacrolimus has a narrow therapeutic index and dosage is adjusted based on whole blood trough concentrations, the increase in oral clearance might suggest a need for a dosage increase. However, the estimated unbound tacrolimus oral clearance did not change, and the unbound percent of tacrolimus in plasma was reported to be approximately two-fold higher in mid- and late-pregnancy compared to postpartum [ ]. If drug binding is not accounted for and the total drug concentration is measured in a patient with an increase in fraction unbound and no change in unbound clearance, the total concentration will be lower, but the dosage should not be adjusted. If the clinician does not account for the altered binding and increases the dosage, the patient might develop drug toxicity.
The physiologic changes that occur during pregnancy can translate into changes in multiple pharmacokinetic parameters that can alter the interpretation of drug concentrations. For example, you often have changes in both protein binding and unbound clearance during pregnancy, as is the case with phenytoin. These patients require consideration of both factors in interpreting total phenytoin concentrations. It is important to note that not all highly protein bound drugs have increased percent unbound during pregnancy. Some highly protein bound drugs such as midazolam and glyburide have little to no change in protein binding during pregnancy, but significant changes in their clearances [ , ].
Changes in hepatic and renal blood flows can alter drug clearance. As described above, changes in organ blood flow are particularly important for high ER drugs. During pregnancy, cardiac output is markedly increased. On average, during normal pregnancy, cardiac output has been reported to be 35% higher in the second trimester and 40% higher in the third trimester as compared to postpartum [ ]. In nonseptic critically ill patients, there is a good correlation (r = 0.92) between cardiac output and effective hepatic blood flow [ ]. In an animal model of reduced cardiac output, there was an associated decrease in portal venous flow [ ]. Unfortunately, there is limited information available evaluating the effects of pregnancy on hepatic blood flow. Nakai et al. [ ] studied the effects of pregnancy on hepatic arterial and portal venous blood flows during the first trimester of pregnancy (n = 13), second trimester (n = 25), third trimester (n = 29), and in nonpregnant women (n = 22). They found an increase in total liver blood flow (2.98 ± 1.13 L/min, P < .05) and portal vein blood flow (1.92 ± 0.83 L/min, P < .05) during the third trimester of pregnancy as compared to the nonpregnant women (1.82 ± 0.63 L/min and 1.25 ± 0.46 L/min respectively). Rudolf et al. [ ] reported indocyanine-green clearance in 16 women with hyperemesis gravidarum with all but one subject within the upper limit of normal. Robson et al. [ ] found no change in hepatic blood flow in 12 women 12–14 weeks, 24–26 weeks, and 36–38 weeks gestation as compared to 10–12 weeks after delivery. Probst et al. [ ] conducted a study in seven healthy pregnant women during labor and delivery and compared them to nonpregnant controls. They found that hepatic blood flow was decreased to 70% of the control value during labor. In a cross-sectional study of 210 pregnant women between 6 and 40 weeks gestation and 40 healthy nonpregnant women, Doppler resistance indices [hepatic artery pulsatility (PI) and resistive index (RI)] were measured to evaluate hepatic liver blood flow [ ]. Mandic-Markovic et al. [ ] found PI and RI in hepatic artery was decreased during the third trimester compared to nonpregnant state and the first two trimesters. This is likely a result of systemic arterial vasodilation during the third trimester of pregnancy [ ]. All of the studies were underpowered and in most cases did not have the pregnant women serve as their own control. At this point, it is unclear whether hepatic blood flow is increased or unchanged during pregnancy.
Pregnancy is associated with increased renal filtration, increased creatinine clearance, and increased renal clearance of drugs [ , , , ]. During normal pregnancy, effective renal plasma flow increases on average 50%–85%, with a corresponding 50% increase in glomerular filtration rate [ , ]. Because the estimated tubular ER for metformin is moderately high, the gestational changes in the metformin's renal clearance and net secretory clearance can in part be explained by enhanced renal plasma flow [ ].
The intrinsic clearance generally refers to the liver's inherent ability to metabolize drug. It is a term used to describe enzyme activity and is independent of protein binding and hepatic blood flow.
Drug metabolism is the conversion of one chemical structure to another. The formation of metabolites often occurs via drug metabolizing enzymes. There are many drug metabolizing enzymes involved in both phase I [e.g., CYP3A4, CYP3A5, CYP2D6, CYP1A1, CYP1A2, CYP2C8 CYP2C9, CYP2C19, CYP2E1, CYP2A6, CYP2B6, esterases, epoxide hydrolase, dihydropyrimidine dehydrogenase, alcohol dehydrogenase (ADH), carbonyl reductase 1 (CBR1)] and phase II (e.g., uridine diphosphate glucuronyltransferase (UGT), sulfotransferase, methyltransferase, N-acetyltransferase, catechol-O-methyl transferase, thiopurine S-methyltransferase, histamine methyltransferase, glutathione S-transferase) metabolism. Phase I metabolism usually precedes phase II metabolism, but not always. Phase I reactions typically include: oxidation, reduction, hydrolysis, cyclization, and decyclization reactions. Phase II reactions involve conjugation with glucuronic acid, sulfate, glutathione, or amino acids. Occasionally, there is back conversion of metabolites to the parent compound. For some medications that are administered as inactive compounds (prodrugs), metabolism is necessary to convert the drug to active compound.
As described above, many different drug metabolizing enzymes exist. The enzyme involved in the metabolism of a drug is dependent on the chemical structure of the agent. For some medications, only one enzyme is involved in the metabolism. For other drugs, multiple enzymes with differing affinities are involved in the formation of the metabolites. One method that has been used to evaluate the effects of pregnancy on drug metabolizing enzymes is to use probe substrates as markers for enzyme activity. A probe substrate is a drug that is primarily metabolized by a single enzyme. The drug is administered and a pharmacokinetic study is completed. From this, drug clearance, urinary excretion of metabolite, metabolite formation clearance, area under the concentration-time curve, or metabolite to parent concentration ratio are used as surrogate markers for enzyme activity. The discussion below will describe the effects of pregnancy on key drug metabolizing enzyme activities ( Table 3.1 ).
Pregnancy effect | Substrates | |
---|---|---|
Enzymes | ||
ADH | Unchanged (83) | Abacavir |
CBR1 | Decreased (87) | Doxorubicin |
CYP1A2 | Decreased (33, 56) | Caffeine, dacarbazine |
CYP2B6 | Unchanged (63), increased (62) | Efavirenz |
CYP2C9 | Increased (19, 60) | Glyburide, phenytoin |
CYP2C19 | Unchanged (58), decreased (54,57) | Proguanil, etravirine |
CYP2D6 | Increased (1, 2, 43) | Metoprolol, clonidine |
CYP2E1 | Increased (55, 64) | Acetaminophen, caffeine, theophylline |
CYP3A4 | Increased (10, 33–38) | Dextromethorphan, indinavir, nelfinavir, midazolam, rilpivirine |
UGT1A1 | Increased (75) | Labetalol |
UGT1A4 | Increased (71) | Lamotrigine |
UGT2B7 | Unchanged (78), increased (79) | Morphine, zidovudine |
Renal transporters | ||
OATP | Increased (10) | Digoxin |
OAT1 | Increased (97) | Amoxicillin |
OCT2 | Increased (32) | Metformin |
P-gp | Increased (10) | Digoxin |
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here