Drug elimination during pregnancy may be affected by each of the following except

Protein Binding

Perhaps most confusing and least understood are the effects of protein binding on placental transfer. Amide local anesthetics are bound to AAG and to a lesser extent to albumin.19 The extent of protein binding varies among the local anesthetic agents (seeTable 13.1). For a given local anesthetic, the proportion of free drug increases as blood concentration increases because of the saturation of binding sites. Binding of local anesthetics in the fetal plasma is approximately half that in the mother.88,89

The fetal-to-maternal (F/M) blood concentration ratios of amide local anesthetic agents are listed inTable 13.1. The lower F/M blood concentration ratios of highly protein-bound drugs (e.g., bupivacaine) have been attributed to their more restricted placental transfer compared with less protein-bound drugs (e.g., lidocaine). Indeed, the rate of bupivacaine transfer across rabbit placenta perfusedin situ is lower than that of lidocaine transfer.156,157 Some investigators have suggested that protein binding in the maternal plasma should not affect the diffusion of drugs across the placenta because the dissociation from plasma proteins is essentially instantaneous.151,158 In subsequent studies, the relatively low umbilical vein–to–maternal vein blood concentration ratio for bupivacaine has been attributed to differences in protein binding between maternal plasma and fetal plasma (Fig. 13.5).88,89,159,160 Let us assume that the total concentration of lidocaine or bupivacaine in the maternal plasma is 2 mg/L. Lidocaine and bupivacaine are approximately 50% and 90% bound to maternal plasma proteins, respectively. Thus, the free concentrations of drug available for placental transfer are 1.0 and 0.2 mg/L, respectively. At equilibrium, the concentration of free drug is equal on the two sides of the placenta. In the fetus, however, lidocaine and bupivacaine are approximately 25% and 50% bound to fetal plasma proteins, respectively. Thus, the total lidocaine concentration in fetal plasma is 1.33 mg/L, resulting in an F/M ratio of 0.67; for bupivacaine, the corresponding values are 0.4 mg/L and 0.2.

In fact, accumulation of bupivacaine occurs in human fetuses whose mothers received the drug for epidural anesthesia.24 After delivery, measurable plasma and urine concentrations persisted for as long as 3 days.24In vitro studies using a perfused human placental model have found that the placental transfer of ropivacaine is similar to that of bupivacaine.161 Intravenous infusion of ropivacaine or bupivacaine to pregnant sheep results in steady-state maternal plasma concentrations of 1.5 to 1.6 µg/mL and fetal concentrations of approximately 0.28 µg/mL.117 Tissue concentrations of ropivacaine in fetal heart, brain, liver, lung, kidneys, and adrenal glands were similar to those of bupivacaine.117 Datta et al.30 noted that the free fraction of ropivacaine at delivery was approximately twice that of bupivacaine in neonates whose mothers received the drug for epidural anesthesia during labor or cesarean delivery.

Drug Protein Binding

Protein binding can involve plasma proteins, extracellular tissue proteins, or intracellular tissue proteins. Many drugs in circulation are bound to plasma proteins, and because bound drug is too large to pass through biologic membranes, only free drug is available for delivery to the tissues and to produce the desired pharmacologic action. Therefore the degree of protein binding can greatly affect the pharmacokinetics of drugs. Acidic drugs such as nonsteroidal antiinflammatory drugs (NSAIDs) tend to bind predominantly to albumin.5 Albumin is the most abundant plasma protein, and it is critical to maintaining the colloidal oncotic pressure in the vascular system. As a negative acute phase protein, albumin concentration decreases during inflammation. Hypoalbuminemia results from decreased production, seen with severe hepatic insufficiency, or by loss through increased rates of urinary excretion, such as in glomerulonephritis or with mucosal damage, as with protein-losing enteropathies. Basic drugs typically bind to α-1 acid glycoprotein, which is an acute phase protein, whose hepatic production increases significantly with inflammatory conditions.6 Other proteins, including corticosteroid binding globulin, are important for binding of some specific drugs but are less important in overall drug-protein binding.7 There is equilibrium between free and bound drug, however, just like the relationship of ionized and nonionized drug molecules. Protein binding is most clinically significant for antimicrobial therapy, where a high degree of protein binding serves as a drug “depot,” allowing for increased duration of the time the drug concentration remains above the bacterial minimum inhibitory concentration, adding to antimicrobial efficacy.8 For other drugs changes in plasma protein binding can influence individual pharmacokinetic parameters, but changes in plasma protein binding usually do not influence the clinical exposure of the patient to a drug. Changes in protein binding caused by drug interactions are assumed to instantaneously change free drug concentrations and have been frequently cited as a cause of adverse drug reactions. But the increase in free drug concentration is only transient, because drug distribution and drug elimination change to compensate. The often-cited example of the concurrent administration of phenylbutazone and warfarin leading to bleeding caused by increased free concentrations of warfarin is erroneous. The true interaction is from phenylbutazone-induced inhibition of the hepatic metabolism of warfarin, which results in increased plasma concentrations and increased anticoagulant effect.7 Therefore adjustments in dosing regimens because of hypoproteinemia or concurrent administration of highly bound drugs are not necessary except in the rare case of a drug with a high hepatic extraction ratio and narrow therapeutic index that is given parenterally (e.g., IV dosing of lidocaine).9

Protein Binding

The degree of protein binding will also determine its removal. Hemofiltration and hemodialysis can remove only unbound poison because the poison–protein complex size exceeds the pore size of the hemofilter or dialyzer. Diffusion (IHD) and convection (HDF, CRRT) can remove poisons with protein binding up to 80%, with a few exceptions. Hemoperfusion, however, may be more effective in poisons with protein binding up to 90% to 95%, because binding to the adsorbent (activated carbon or, less commonly, a resin) competes with binding to plasma proteins. The clearance therefore depends on the affinity of the poison for the adsorbent.

The degree of protein binding can be influenced by acute alterations in poison or protein concentration and the presence of different pathologic states.27 For example, in the context of hypoalbuminemia, there is less protein available to bind poison. As such, the concentration that is free (unbound) is higher with hypoalbuminemia, which will result in increased removal by ECTRs. Similarly, accumulation of organic acids in uremia leads to a reduction in binding sites for some xenobiotics (e.g., salicylates, warfarin, phenytoin), which also increases the unbound concentration and favors removal by ECTR. Furthermore, in toxic concentrations, there may be saturation of the protein binding sites (e.g., valproic acid, salicylate, 4-chloro-2-methylphenoxyacetic acid [MCPA]), increasing the fraction of poison that is unbound in relation to its total concentration, which also increases the amount that is amenable to removal by ECTR.28

Restrictively Eliminated Drugs

Because protein binding reduces the hepatic clearance of restrictively eliminated drugs, the hypoalbuminemia that results from severe liver disease will decrease the protein binding of these drugs and increase their hepatic clearance. The hepatic clearance of these drugs will also be increased by the decreased protein binding that accompanies impaired renal function and some drug interactions. Although this will result in a decreased total drug concentration, steady state free concentrations will not be increased unless intrinsic clearance is also reduced. Because pharmacologic effects are related to free rather than total drug concentrations, reductions in protein binding will be of little consequence unless the decrease in total drug concentration prompts an inappropriate dose increase, as sometimes occurs with phenytoin.

Evaluation of drug intrinsic clearance is complicated by the fact that different metabolic pathways differ in the extent to which they are impaired as liver function deteriorates. This is demonstrated in Figure 13-14, in which probe drugs for several metabolic pathways were administered to normal subjects and to patients with liver disease. However, even when the metabolic pathway for a given drug is known, prediction of hepatic drug clearance in individual patients would be complicated by the effects of pharmacogenetic variation, variability in extent of enzyme induction, and drug interactions.

11.1 Introduction

Plasma and tissue protein binding of drugs is a major factor that affects both pharmacokinetics and pharmacodynamics of the drug. It is usually the free (unbound) form of the drug that can exert pharmacological activity, while the bound form of the drug is usually pharmacologically inactive (Ascenzi et al., 2014). Many drugs can bind to plasma proteins to form a drug–protein complex, the binding is usually reversible, and the unbound (free) form of the drug exists in equilibrium with the bound form (Li et al., 2015). Drugs bind mainly with plasma proteins such as albumin, alpha-1-acid glycoprotein, lipoproteins, and other biological moieties, e.g., red blood cells (RBCs) (Pellegatti et al., 2011).

The reversible binding of drugs to proteins has a significant impact on many pharmacokinetic parameters such as volume of distribution and clearance of the drug (Berezhkovskiy, 2010). Since the drug–protein complex has a large size, this will limit its ability to leave the vascular space and enter into cells thus restricting its distribution, while the unbound (free) drug can readily diffuse into cells. Also, the drug–protein complex is usually too large to be filtered by the glomeruli, and only the unbound drug can be filtered and excreted by the kidney. Thus, plasma protein binding also affects clearance of the drug by the kidney, and sometimes if the drug has a higher affinity for the plasma proteins than the liver enzymes, the drug will not be available for metabolism and clearance by the liver. Hence, only the unbound drug will be metabolized (Han et al., 2010).

26.2.1 Factors influencing toxicokinetics-toxicodynamics correlations

Factors like administration route, protein binding, mode of toxicity, time of dosing (Choudhary et al., 2021; Vasdev et al., 2021), metabolic activity (Achanta et al., 2021), and so on should be addressed in developing TK-TD correlations. Further, the medication route in toxicity research should resemble that which is used in people. For some medicines, toxicity could be linked to a particular threshold drug level, closely related to drug solubility and bioavailability (Polaka et al., 2021b; Rajpoot et al., 2020f). In these kinds of situations, the number of dosages may also influence the TK-TD correlations. For instance, Powell et al. demonstrated that treatment of nephrotoxicity in dogs by continuous intravenous infusion of gentamicin as well as tobramycin is substantially more effective than the delivery of the medication every 4 h as an intravenous bolus. Treatment of dogs with 45 mg gentamicin/kg/day for 10 days resulted in a decrease in 86% of drug serum levels over to 4 h, decreasing to 29%. Likewise, treatment of 45 mg tobramycin/kg/day for 10 days to dogs decreased their serum levels by 40% whenever the medication was administered for 4 h decreased by 31%, and so by 12% while administered once daily. Their findings revealed that the toxicity is linked to the length of drug exposure above a specific threshold of plasma levels for such medicines, and not to the peak of drug levels (Powell et al., 1983).

On the other hand, embryotoxicity for valproic acid in rats was found significantly greater when such medication was given as intravenous bolus than that given via intravenous infusion. Furthermore, it showed 20% embryotoxicity after intravenous bolus injection (dose = 350 mg/kg/day) followed by continuous intravenous infusion (dose = 2700 mg/kg/day). With valproic acid, embryotoxicity was seemed to be linked to the peak plasma drug, which is significantly greater after the quick intravenous bolus delivery (Nau, 1991).

Protein binding also plays a crucial role in developing TK-TD correlations for firmly bound medicines. The free/unbound drug level also significantly correlates either with pharmacological or toxicological responses than complete medication does. A previous study revealed that the plasma level of total phenytoin in humans and mice was substantially different at the start of ataxia. Nevertheless, the free/unbound phenytoin levels appeared comparable when ataxia developed (i.e., ~6 mg/L) for both humans and rats (Ramzan, 1990). Thus species and concentration-related variations in protein binding should be addressed for every meaningful extrapolation across animals to humans. Whenever the plasma drug–drug levels are not in harmony with tissue drug levels, and so the drug levels will not correspond to pharmacological as well as toxicological response. Throughout the event of bumetanide (a diuretic), this PD impact corresponds well with an unaltered molecule in the kidney (Halladay et al., 1978).

In developing a TK-TD correlation using toxicology information, it is believed that the observed toxicological results on animals could be equally effective in humans. Nevertheless, this will not always be accurate. The antiinflammatory chemical ICI 54,450 is harmless in rats and dogs at plasma levels considerably above that seen in humans with therapeutic dosages, but humans experienced jaundice at therapeutic levels (Alcock, 1970). Therefore a key element in establishing TK-TD correlation is the selection of an easily observable TD endpoint. Since it is hard to define this as an endpoint for several medicines, creating TK-TD correlations often might be a little more challenging than establishing pharmacokinetics (PK)-PD correlations.

Before understanding the models in detail, let us first understand the basic terms used to frame the model, like TK and TD. The term TK is analogous to PK where the term kinetics itself explains a biological event related to kinetic events like ADME of compounds (Anup et al., 2021c; Polaka et al., 2021a; Tambe et al., 2021e). With the addition of the suffix, the term defines the kinetic parameters of a xenobiotic or a toxic compound, which can be evaluated and quantified. TK uses mathematical models for the quantification, which includes the course of time of a drug or a chemical compound to get absorbed and disposition in man and animals (Dixit et al., 2003; Tambe et al., 2021a; Anup et al., 2021b).

Similarly, there are types of TD models that can be inculcated into the finalized TK-TD model enriching the researcher with the information of graded (gradual changes) to quantal (all or none response) responses where the lethal and sublethal effects are considerable (Fig. 26.2). Focusing on the graded responses of the TD model, it explicitly describes the relation between toxicant concentration and its site of action along with the individual’s reactivity or response. Graphically, they represent a sigmoidal-shaped dose–response curve determining the concentration of max and min responses of the toxicant along with its slope of the curve, which directly measures the responses generated due to the changes in the exposure of the toxic compound (Gehring and van der Merwe, 2014).

In one of the reported studies, the author has specifically focused the DEBtox TD model only for correlating the concentration of toxicant, giving stress to the organism’s growth. It is reported that the xenobiotic substance has a “no effect concentration” (NEC), indicating that below this concentration, none of the biological processes might face the generation of stressful conditions. The NEC has no relation with time and thus no relation with the exposure time too. It is assumed that any toxic compound, when it exceeds its NEC, creates a similar kind of stressful condition linearly, leading to the linear threshold relationship. The other parameter considered by the author in this model is the tolerance concentration (CT). NEC indicates stress exceeding zero, and an increase in the stress by 1 is denoted by CT (Martin et al., 2019).

Gergs et al. have contributed to this field by using the TK-TD model for assessing the lethal effects on Daphnia magna exposed to triphenyltin. Various scenarios were simulated using the individual-based model for comparing the predictions with the population data. The TK model identified the damage-causing factor, whereas the damage repair rate of the organism was used to represent the TD recovery. Two concepts of TD, namely, stochastic death (SD) and individual tolerance (IT), including under the general unified threshold model of survival framework, were used to link the dose metric for survival probability. In the SD model, authors concluded that the survival probability decreases when the dose exceeds the threshold value. The IT model shows a similar result to the SD model but follows a cumulative log-logistic frequency distribution within a population. Death of the organism occurs immediately once the dose metric exceeds the threshold of survival (Gergs et al., 2016).

Plasma protein binding of acidic drugs

Reidenberg and Drayer [46] have stated that protein binding in serum from uremic patients is decreased for every acidic drug that has been studied. Most acidic drugs bind to the bilirubin binding site on albumin, but there are also different binding sites that play a role. The reduced binding that occurs when kidney function is impaired has been variously attributed to reductions in serum albumin concentration, structural changes in the binding sites, or displacement of drugs from albumin binding sites by organic molecules that accumulate in uremia. As described in Chapter 3, reductions in the protein binding of acidic drugs result in increases in their distribution volume. In addition, the elimination clearance of restrictively eliminated drugs is increased. However, protein-binding changes do not affect distribution volume or clearance estimates when they are referenced to unbound drug concentrations. For restrictively eliminated drugs, the term intrinsic clearance is used to describe the clearance that would be observed in the absence of any protein-binding restrictions. As discussed in Chapter 7, CLH for restrictively eliminated drugs, when referenced to total drug concentrations, simply equals the product of the unbound fraction of drug (fu) and this intrinsic clearance (CLint):

(5.5)CLH=fu.CLint

Phenytoin is an acidic, restrictively eliminated drug that is classically used to illustrate some of the changes in drug distribution and elimination that occur in patients with impaired kidney function. In patients with normal kidney function, 92% of the phenytoin in plasma is protein bound. However, the percentage that is unbound or “free” rises from 8% in these individuals to 16% (or more) in hemodialysis-dependent patients. In a study comparing phenytoin pharmacokinetics in normal subjects and uremic patients, Odar-Cederlöf and Borgå [47] administered a single low dose of this drug so that first-order kinetics were approximated. The results presented in Table 5.3 can be inferred from their study.

Table 5.3. Effect of impaired kidney function on phenytoin kinetics.

Healthy subjects (n = 4)Uremic patients (n = 4)Percent unbound (fu)12%26%Distribution volume (Vd(area))0.64 L/kg1.40 L/kgHepatic clearance (CLH)2.46 L/h7.63 L/hIntrinsic clearance (CLint)20.3 L/h29.9 L/h

The uremic patients had an increase in distribution volume that was consistent with the observed decrease in phenytoin binding to plasma proteins. The threefold increase in hepatic clearance that was observed in these patients also was primarily the result of decreased phenytoin protein binding. Although CLint for this CYP2C9, CYP2C19, and P-gp substrate also appeared to be increased in the uremic patients, the difference did not reach statistical significance.

What affects drug elimination in pregnancy?

What are 4 factors that affect absorption of a drug?

What 4 systems are affected by pregnancy?

What properties of drugs are affected by pregnancy?