OVERVIEW 349 The goal of therapeutics is to achieve a definite beneficial effect with minimal adverse effect, and the approach to this has been through combining the principles of pharmacoki netics and pharmacodynamics. Pharmacokinetics (PK) is the branch of pharmacology that deals with the absorption, distribution, and elimination of drugs, while pharmaco dynamics (PD) deals with the actions of drugs on the organ ism. PK governs the relationship between the drug dose and its concentration whereas PD relates the concentration to the effect; together PK/PD clarify a drug’s dose–effect rela tionship. Since the 1950s there have been major advances in understanding the genetic basis of the interindividual variability observed in the pharmacokinetics, efficacy, and toxicity of various drugs. Recently, the European Medicines Agency (EMA) published guidance on the use of pharmaco genetics to investigate PK properties of new medicines [1] and PK pharmacogenetics feature in an increasing number of drug labels approved by the FDA [2]. Many pharmaceuti cal companies investigate the associations between PK gene variants and the observed interindividual pharmacokinetic and pharmacodynamic variability of drugs in early clinical development (phases I and II) [3]. In this chapter, we outline
the basic principles of pharmacokinetics and summarize the progress in the pharmacogenomics of key pharmacokinetic systems. 17.2 PRINCIPLES OF PHARMACOKINETICS PK studies are usually based on measuring the drug levels in blood and urine at different times following drug admin istration in order to understand how rapidly and for how long the drug appears in the target organ [4]. This involves mathematical modeling of data obtained from these stud ies around four domains: drug absorption (A), distribution (D), metabolism (M), and excretion (E). Together these are referred as ADME [4,5]. 17.2.1 Absorption Drugs are normally administered through various routes: orally, subcutaneously, intramuscularly, intravenously, rectally, or sublingually. Once a drug is administered, it is absorbed and distributed to different cells and organs. Drug absorption is generally defined as the rate and extent to
which the drug moves from its site of administration to its intended target (site) of action [6,7]. Absorption is a criti cal component in drug PKs, and several barriers must be overcome for a drug to reach its effect site. The rate and extent of absorption, as well as the time required to notice an observable effect, dictate the dose that should be admin istered [8]. Bioavailability indicates the proportion of the drug that passes into systemic circulation after administration, taking into account both absorption and local metabolic degradation [7]. Intravenous delivery provides 100% bio availability, while an orally ingested drug may be incom pletely absorbed by the GI tract before it reaches systemic circulation (around 75% of an orally administered drug is absorbed in 1–3 hours). In addition, the gastrointestinal lin ing expresses several cytochrome P450 enzymes and drug efflux transporters (e.g., ATPbinding cassette, or ABC) that can decrease bioavailability [6,9]. Bioavailability depends on several factors, including the drug’s physicochemical properties and formulation, firstpass metabolism, concomitant drug therapies, compli ance, and disease state. The area under the curve (AUC) calculated from time zero to infinity following drug admin istration is an important PK parameter that measures the patient’s “exposure” to a drug and depends on dose, bio availability, and clearance [6]. The patient’s plasma drug concentration time profile can be developed by measuring the plasma drug concentration at many time points and then estimating AUC, which is proportional to the absorbed frac tion only when clearance is constant and the concentration is standardized. Following linear kinetics, AUC is directly proportional to drug dose and inversely proportional to drug clearance. Therefore, the higher the clearance, the less time the drug spends in the systemic circulation and the earlier the decline in the concentration. As a result, the body’s exposure to the drug and the AUC are smaller. Clearance is slightly depen dent on the shape of the concentration time profile and is calculated by dividing the dose absorbed by the AUC [10]. Orally administered drugs are absorbed from the GI sys tem and transferred to the liver via the portal vein before entering systemic circulation. The drugmetabolizing systems in the liver can thus exert a substantial effect on bioavailability; this is called firstpass metabolism or pre systemic elimination. The larger the firstpass metabolism, the smaller the bioavailability of an orally administered drug. The liver drug metabolizing enzymes can be com pletely avoided by sublingual or buccal administrations or partially avoided by rectal administration [6]. Drugs that undergo extensive firstpass metabolism and cannot be administered orally (e.g., nitroglycerin) require alterna tive routes of administration, such as sublingual or intra venous. Drugs with extensive firstpass metabolisms can still be administered orally with higher doses. For example, a typical IV dose of verapamil is 1–5 mg, compared to the usual single oral dose of 40–120 mg. Firstpass metabolism is a major reason for recogniz able differences in drug bioavailability among individuals, as even healthy people show considerable metabolizing differences in liver capacity. Moreover, in patients with severe liver disease, firstpass metabolism may be sharply decreased, leading to greater bioavailability [6]. Marked interindividual variations in terms of firstpass metabolism lead to unpredictable consequences. Examples of drugs that undergo significant “firstpass effect” are aspirin, morphine, levodopa, verapamil, salbutamol, and lidocaine [7,11]. 17.2.2 Distribution Once a drug has reached the bloodstream, it is distributed into the interstitial and intracellular compartments. As a general rule, the movement of lipophilic drugs is faster than that of hydrophilic drugs, and small lipophilic mol ecules distribute across cell membranes more easily than do large polar molecules [12–14]. Lipophobic drugs are primarily confined to plasma and interstitial fluids and gen erally do not go through the brain tissue after acute dos ing. Hydrophilic drugs, such as the aminoglycosides, are mostly distributed into extracellular fluid, and their volume is affected by fluid retention or dehydration, both of which can happen in a number of renal diseases. Passive diffusion, facilitated diffusion, and active trans port are the three mechanisms involved in drug distribution across cell membranes [15]. Solute carrier (SLC) and ATP binding cassette (ABC) transporters play an important role in active drug transport [16]. Cardiac output, regional blood flow, capillary permeability, lipid solubility, and plasma protein binding determine the rate and amount of drug dis tributed into tissue [12]. Sites with high blood flow primarily receive greater amounts of a drug compared to sites with low or disturbed blood flow. Consequently, drug concentrations increase faster in organs such as the brain, heart, and kidneys as compared with skin, muscle, and bone. The structure and permeability of capillaries varies depending on the organ, and this affects how the drug is distributed. For example, capillaries in the kidney and liver sinusoids show more permeability as compared with the tight junctions between endothelial cells that line the brain capillaries, creating a relatively impermeable blood–brain barrier (BBB). The BBB allows only selective transport of lipophilic molecules and prevents passive entry of lipophobic/ionized molecules into the brain [12,13]. 17.2.2.1 Plasma-Protein Binding In blood, many drugs are bound to plasma protein. This binding is reversible with bound and unbound drug frac tions in dynamic equilibrium. Any change in unbounddrug
concentration is directly followed by a change in bound drug concentration. Only free (unbound) fractions are able to cross membranes or interact with drug targets and are considered pharmacologically active. In general, the amount of a drug that is bound to protein depends on three main factors: freedrug concentration, binding site affinity, and protein concentration [17,18]. The most significant plasmaprotein binding is to albumin, which is the major carrier of acidic drugs (e.g., warfarin, sulfonamides). Other plasma proteins, such as α1acid glycoprotein, bind basic drugs. In addition, certain drugs may bind to proteins that function as specific hor mone carrier proteins, such as the binding of estrogen or testosterone to sex hormone–binding globulin or the bind ing of thyroid hormone to thyroxinbinding globulin. For many drugs, more than 90% in the plasma is bound to a protein (e.g., warfarin, diazepam), whereas other drugs may have less extensive protein binding (digoxin and gentami cin). Because binding of drugs to plasma proteins such as albumin is nonselective, and because the number of bind ing sites is relatively large (highcapacity), many drugs with similar physicochemical characteristics can compete with each other and with endogenous substances for these bind ing sites, resulting in displacement of one drug by another and leading to an increase in its pharmacological activity or toxicity. Changes in plasmaprotein concentrations can also lead to variability in pharmacological effects of drugs, which are also affected by diseaserelated factors; for example, hypoalbuminemia secondary to nephrotic syndrome results in decreased drug binding, which increases the drug’s unbound fraction of the drug. In addition, acutephase reac tion responses (e.g., cancer, arthritis) can lead to high levels of α1acid glycoprotein and can enhance basic drug bind ing. Misinterpretation of plasma drug concentration mea surements is a common problem that results from drugs competing for plasmaprotein binding sites, as most assays do not differentiate unbound from bound drugs. A number of drugs concentrate in body tissues at higher levels than those in extracellular fluids and blood; for exam ple, the concentration of an antimalarial agent (quinacrine) in the liver can be many thousand times higher than that in the blood if the drug is administered for long periods. Reversible tissue binding of drugs regularly happens with proteins and phospholipids. Fractions of a drug in the body may be bound in this way and can act as a reservoir that prolongs drug action in the same tissue or at a distant site. Lipophilic drugs can accumulate in adipose tissue [12]. Halothane can concentrate in fat during long operations, and its slow release can cause postoperative prolonged cen tral nervous system (CNS) depression. Tissue binding and drug accumulation may lead to local toxicity, as in the case of aminoglycoside antibiotic (gentamicin) accumulation in the kidneys and vestibular system [12]. Volume of distribution (Vd), represents the apparent vol ume into which the drug is distributed to provide the same concentration as it currently is in blood plasma. It is calcu lated by the amount of the drug in the body divided by the plasma concentration [19]. Thus, Vd reflects the extent to which the drug is present in extravascular tissues but not in plasma. Lipid solubility can affect Vd, as highly lipidsolu ble drugs have good cell penetration, resulting in high Vd. Plasmaprotein binding, particularly to albumin, reduces the Vd, while tissue binding increases it [17]. Vd can be used to determine the size of a loading dose in order to quickly reach the required therapeutic plasma concentration, assuming that successful therapy is directly linked to the plasma concentration and that there are no adverse effects if a quite large dose is rapidly administered. In addition, Vd is helpful in predicting the initial maximum concentration for an IV bolus and can be used to predict the effectiveness of dialysis in treating drug intoxication during an emergency [18]. 17.2.3 Metabolism Once drugs have been distributed throughout the body, they are metabolized into more polar, inactive metabo lites in order to eliminate them from the body. The lipo philic characteristics of drugs that promote their passage through biological membranes and subsequent access to their site of action also serve to hinder their excretion from the body. Metabolism includes two main processes: one, the molecule is made more lipophobic in order to reduce the possibility of reabsorption in the renal tubules; two, the molecule is conjugated to reduce its effect and increase its excretion. While for some drugs the metabo lites have the actions of the parent drug (e.g., diazepam and its metabolite, nordiazepam), for others, the metab olite may result in toxicity (e.g., paracetamol). Also, a number of drugs are “prodrugs” (i.e., inactive) but trans form into an active drug in the body. Prodrugs are often designed to improve oral bioavailability in cases where the intended drug is poorly absorbed through the gastro intestinal tract; for example, enalapril is hydrolyzed to the active compound enalaprilat [12]. Metabolizing enzymes are distributed in many tissues in the human body, with the highest levels found in GI tract tissues (e.g., the liver and the small and large intes tines). The liver is the major metabolizing organ for both drugs and endogenous chemicals (e.g., cholesterol, fatty acids, and proteins). The small intestine also plays a vital role in drug metabolism, as most orally administered drugs become inactivated metabolically in either the intestinal epithelium or the liver before they reach systemic circula tion [20]. Drug metabolism reactions are classified as either phase I or phase II.
Drugs metabolized by phase I reactions usually lose their pharmacological activity, although there are some examples of enhanced or altered pharmacological activity. If not elim inated rapidly through urine, phase I reaction products can react with endogenous compounds to form a highly water soluble conjugate [12]. Phase I reactions “functionalize” the drug for phase II, as they introduce a functional group—such as –OH, –COOH, –SH, –O–, or NH—which increases the polarity of the drug molecule and provides a site for phase II reactions. Phase I metabolic reactions include oxidation, reduction, and hydro lysis [20] (see Table 17.1). Oxidations are the most common type of reaction and are catalyzed by the cytochrome P450 (CYP) system, which is located in the smooth endoplasmic reticu lum. Metabolizing enzymes responsible for oxida tion are mainly CYPs, alcohol dehydrogenase (ADH), aldehyde dehydrogenase (ALDH), dihydropyrimidine dehydrogenase (DPD), monoamine oxidase (MAO),
and flavincontaining monooxygenase (FMO) [20,21]. Oxidation via phase I enzymes adds or exposes a func tional group, allowing phase I metabolites to act as substrates for the phase II conjugating enzyme [20]. The oxidation of a drug by the CYP system requires a P450 enzyme, molecular oxygen, nicotinamide adenine dinucleotide phosphate (NADPH), and a flavoprotein (NADPHP450 reductase). Hydrolysis is not limited to the liver but also takes place in various tissues. The main metabolic enzymes involved in drug hydrolysis are epoxide hydrolases (EH), ester ases, and amidases [21]. Inactive prodrugs are converted rapidly to active metabolites mostly through an ester or amide linkage hydrolysis; for example, enalapril, the angiotensinconverting enzyme inhibitor, is quite inactive until it is transformed by esterase into its diacid metabo lite (enalaprilat) [12]. Examples of drugs that undergo hydrolysis include carbamazepine, aspirin, meperidine, lidocaine, procainamide, and indomethacin [20]. Reduction reactions are much less common than oxi dation reactions. The main metabolic enzyme involved in reduction is NADPHCYP reductase [21]. A few drugs undergo reduction reactions, and some reductive reactions are important; for example, warfarin is inacti vated by conversion of a ketone to a hydroxyl group by the CYP2A6 enzyme. CYPs, FMO, and EH, in addition to some phase II conjugating enzymes, particularly Uridine 5′diphospho glucuronosyltransferase (UDPglucuronosyltransferase, or UGT), are located mainly in the endoplasmic reticulum, while phase II enzymes are mostly cytosolic [20]. 17.2.3.2 Phase II Reactions Whereas various phase I reactions inactivate a drug bio logically, phase II enzymes enable drug elimination by significantly increasing water solubility, abolishing phar macological activity, and increasing molecular weight. Phase II reactions are generally biosynthetic (conjugation) reactions, such as glucuronidation, sulphation, acetylation, methylation, and glutathione conjugation [20]. These highly polar conjugates in general are inactive and are eliminated rapidly through urine and feces. Phase II enzymes comprise a number of conjugat ing enzyme superfamilies, including the glutathione Stransferases (GST), UGT, sulfotransferases (SULT), Nacetyltransferases (NAT), and methyltransferases (MT). Phase II reactions as a whole depend on catalytic reactions for cofactors such as UDPglucuronic acid (UDPGA) for UGT and 3phosphoadenosine5phosphosulfate (PAPS) for SULT, which respond to the substrate functional groups. These functional groups are mainly formed via phase I CYPs. In cholestasis, toxic bile acids accumulate in hepatic cells, leading to their damage and functional impairment. Glucuronidation, which is catalyzed by UGT enzymes, is considered an essential metabolic pathway for hepatic bile acids [20,22]. 17.2.3.3 Non-CYP450 Enzymes NonCYP450 enzymes play a vital role in both the metab olism and the elimination of several drugs; for example, UGT, SULT, thiopurine Smethyltransferase (TPMT), DPD, NAT, and GST have been indentified for their clinical significance [23]. UGTs are membranebound enzymes found in the hepatic endoplasmic reticulum and various other extra hepatic tissues [21]. They involve a superfamily of vital proteins that catalyze the glucuronidation reaction on a wide range of structurally different endogenous and exogenous compounds. UGTs catalyze the transfer of the glucuronic acid group of uridine diphosphoglucuronic acid to the func tional group (e.g., hydroxyl, carboxyl, amino, sulfur) of a specific substrate [24]. SULTs are localized in the cytosol and catalyze the transfer of the sulfonyl group from the cofactor PAPS to the nucleophilic sites of a range of substrates, including hor mones and drugs. In humans, 11 SULT isoforms have been recognized and have been classified into four main groups based on evolutionary projections (SULT 1, 2, 3, and 4) [20,21]. They play a significant role in human homeostasis. For instance, SULT1B1, which is mostly expressed in the skin and brain, is responsible for cholesterol and thyroid hormone catalysis [20]. TPMT is recognized for its important role in the metab olism of thiopurine drugs (e.g., 6mercaptopurine (6MP), azathiopurine (AZA), and 6thioguanine) and is catalyzed through the smethylation of aromatic and heterocyclic sulfhydryl [21]. AZA and 6MP are used to treat inflamma tory bowel disease and autoimmune disorders such as sys temic lupus erythematosus and rheumatoid arthritis. 6MP is also used for the treatment of childhood acute lympho blastic leukemia (ALL). 6thioguanine is used to treat acute myeloid leukemia (AML), and since TPMT is responsible for the detoxification of 6MP, any deficiency in it can result in severe toxicities in patients taking those drugs [20,25]. DPD is the ratelimiting enzyme in pyrimidine and 5fluo rouracil (5FU) degradation. 5FU is commonly prescribed to treat GI malignancies, and DPD deficiency can increase concentrations of bioavailable 5FU anabolic products, leading to 5FU–related toxicity [26]. GST catalyzes glutathione conjugation to a broad range of endogenous metabolites and drugs. Human GSTs are classified into three major families: cytosolic/nuclear, mito chondrial, and microsomal. They also have nonenzymatic functions, as they work as regulators of cell signaling and the posttranslational modification pathway in response to stress, growth factors, DNA damage, cell proliferation, cell
nd other processes that eventually lead to tumor growth and drug resistance. Because of their functionalities, GSTs are seen as significant determinants of cancer suscep tibility, therapeutic response, and prognosis [20,21,27]. NATs are responsible for the metabolism of both drugs and environmental compounds that contain an aromatic amine or hydrazine group. NATs catalyze the transfer of an acetyl group from acetyl coenzyme A to arylamines, arylhy droxylamines, and arylhydrazines [20,21]. 17.2.3.4 Enzyme Induction The pharmacological effect of a drug is dependent on its concentration at its site of action, which is partly dependent on its metabolism rate. Any changes in this enzyme activity can affect drug action. Enzyme induction can be defined as the increased synthesis (higher amount) or decreased degra dation (increased activity) of enzymes that occurs as a result of the presence of an exogenous substance [20,28]. It is usu ally associated with a reduction in drug efficacy but may also change the toxicity of particular substances. CYP3A4 plays a role in the metabolism of about 50% of the drugs that are currently prescribed. Induction of CYP3A4 by rifampicin can increase oestrogen metabolism, thus reduc ing the effectiveness of birth control pills. Certain drugs can increase the rate of synthesis of CYP450 enzymes; consequently, this enzyme induction can enhance the clearance of other drugs. Generally, such induction needs exposure to the inducing agent for more than a week before effects can be observed [28]. Clinical studies have shown that carbamazepine, phenytoin, and rifampicin are the most potent enzyme inducers in clinical use as they produce many clinically significant drug inter actions mainly associated with increases in the metabo lism of CYP2C9, CYP2C19, and CYP3A4 substrates (see Table 17.2). However, enzyme induction is not limited to the administration of prescription drugs. St. John’s Wort, an herbal medicine, can also induce metabolizing enzymes, as can tobacco, which induces a CYP1A2 substrate, such as theophylline [29]. 17.2.3.5 Enzyme Inhibition Enzyme inhibition refers to a decrease in enzymerelated processes, enzyme production, or enzyme activity. A num ber of clinically important interactions between drugs result from CYP450 inhibition. CYP450 inhibitors are different in their selectivity toward enzymes and are classified by their mechanisms of action. Some drugs are potent com petitive inhibitors and compete for the active site, but they are not a substrate for the enzyme (e.g., quinidine and CYP2D6), while other drugs are noncompetitive inhibitors (e.g., ketoconazole and CYP3A4). Enzyme inhibition can cause many adverse drug interactions that tend to happen more rapidly (within a couple of days) than those seen with enzyme induction, as they occur once the concentration of the inhibiting drug becomes high enough to compete with the affected drug. Examples of enzymeinhibiting agents are cimetidine, erythromycin, ciprofloxacin, and isoniazid. In certain cases, enzyme inhibition can cause potentially serious adverse events; for example, ketoconazole reduces the metabolism of the CYP3A4 substrate (terfenadine), resulting in a prolonged QT interval and torsades de pointes. As with enzyme induction, enzyme inhibition is not lim ited to drug interactions [30]; for example, grapefruit juice (a CYP3A4 inhibitor) can cause clinically significant inter actions with many drugs, such as midazolam, simvastatin, and terfenadine. Resulting in much higher plasma concentra tions of the inhibited drug than intended, enzyme inhibition can be a major safety issue, such as in coadministration of ketoconazole or ritonavir with midazolam, which increases midazolam plasma exposure (AUC) by 15–20 times—a condition that should be avoided [8]. 17.2.4 Excretion Drugs are eliminated from the body through two main mechanisms—liver metabolism and renal excretion. Some drugs are excreted in insignificant amounts via other routes, such as sweat, saliva, and tears, and elimination through these routes depends mostly on nonionized lipophilic diffu sion through the epithelial cells of the glands and on urine pH [12]. 17.2.4.1 Renal Excretion The kidney is the principal organ of excretion for a drug and its metabolites. Drugs that are water soluble are mostly excreted unchanged through the kidneys. Lipidsoluble drugs are not eliminated efficiently by the kidneys and first need to be metabolized to more polar products. Renal dis ease influences the excretion of particular drugs. The extent to which excretion is impaired can be deduced by measur ing creatinine clearance. Drug and metabolite excretion in urine includes three different processes: glomerular filtra tion, active tubular secretion, and passive tubular reabsorp tion; changes in kidney function in general affect all three processes to a similar degree [12]. Glomerular filtration is the most common route of renal elimination. The amount of drug entering the tubular lumen through filtration depends on the glomerular filtration rate (GFR) and the level of plasma binding of the drug. Only unbound drugs are cleared by filtration; proteinbound drugs remain in circulation, where some of them dissociate to restore equilibrium. The size of the molecule is the only limiting factor at this step. In relation to filtered drugs, it is expected that great declines in renal function, as reflected by a decreased GFR, lead to drug accumulation, and there fore renal patients may need dose adjustments. However, for actively secreted drugs, GFR is less important than renal plasma flow [31].
Passive reabsorption in the distal tubule occurs only with unionized (lipidsoluble) drugs. Urine PH determines whether weak acids and bases are reabsorbed, and this in turn determines the degree of ionization. If renal function is impaired, for instance, by disease, the clearance of drugs that usually undergo renal excretion is decreased. Tubular reabsorption follows the role of passive nonionic diffusion and depends on both urinary pH and the drug’s pKa. A pH that prefers the nonionized state of a drug increases both its lipophilicity and passive reabsorption across the tubular membrane. The extent of passive reabsorption also depends on urinary volume. The clinical relevance of renal tubular reabsorption relates generally to the disease effect on the process [31,32]. Active secretion in the proximal tubule is predominant in that the proximal convoluted tubule can actively secrete some compounds—mostly weak acids and bases—into the lumen of the nephron. Actively secreted drugs accumu late in patients with reduced renal function, as renal blood flow and GFR generally also decrease. The SLC transport ers, which include organic cation transporters (OCTs) and organic anion transporters (OATs), are involved in active transport of many endogenous substances and drugs [32]. Actively secreted drugs are subject to competition for trans port, so there are some concerns with particular clinical conditions in which coadministered drugs inhibit transport, which can expose patients to high drug concentrations [31]. As about 80% of a drug delivered to the kidney is subjected to the carrier, tubular secretion is probably the most efficient mechanism of renal drug elimination. A number of organic acids are actively secreted (e.g., penicillins, probenecid), as well as quite a few diuretics (e.g., furosemide, thiazides) and different conjugates (e.g., glucuronic acid, sulphate). Some organic cations are also secreted, such as morphine and amiloride. 17.2.4.2 Biliary Excretion Liver cells transfer different substances, and bile excre tion is a main route of excretion for a small number of drugs (e.g., cromoglycate) and a larger number of drug metabolites (e.g., morphine glucuronide). Drug conju gates are excreted into the bile and later released into the intestines, where they are reabsorbed into the body (enterohepatic circulation). Enterohepatic circulation has important clinical consequences, as it prolongs the drug effect and increases its plasma halflife (t1/2). While some drugs are excreted mostly unchanged in bile (e.g., vecuronium), other drugs are excreted in bile after deacet ylation, which retains their biological activity and stops their reabsorption. Drugs pass from plasma to bile via transport systems similar to renal tubules, such as OCTs, OATs, and Pglycoproteins. 17.2.5 Half-Life The halflife (t1/2) is one of the simplest PK parameters to calculate. It is the time taken for a drug concentration to fall to half of its original value, measured in hours. However, not only does it determine the time needed for drug concentra tions to fall to immeasurably low levels following a single bolus; it is also the main determinant of the time required to achieve steadystate plasma concentrations (Css) after any changes in dose. When a drug is administered chroni cally and its amount administered per unit time equals the amount eliminated per unit time, this situation represents a steady state. While Css is stable in a continuous IV infusion, it varies during the dosing interval in chronic oral adminis tration although the profile of timeconcentration between dosing intervals is still stable. The t1/2 is related to elimination rate constant, clearance, and Vd. The “initial” concentration achieved after a single IV bolus, then, decreases by a constant amount per unit time as the drug is eliminated from the body. The parameter that describes this rate of decline is the elimination rate constant (k), and it depends on both clearance and volume of distri bution. Clearance measures the ability of the body to elim inate a drug; therefore, as clearance decreases (e.g., with disease), t1/2 is expected to increase. However, this kind of relationship is applicable only when the disease does not affect the Vd; for example, diazepam t1/2 increases with patient age, but it is the volume of distribution that changes, not the clearance [12,18]. The t1/2 represents the time taken to eliminate a drug and the time taken for the drug to accu mulate to a steady state with multiple dosing or during a constantrate infusion. This takes about 4–5 t1/2 (when start ing from zero). Drug concentration data are generally plot ted against time. A drug achieves 50% of its Css after one t1/2, 75% after two t1/2, 88% after three t1/2, 94% after four t1/2, and 97% after five t1/2. t1/2 is an important determinant of dose frequency and time required to achieve Css; however, it provides little information about differences in dose requirements asso ciated with clinical conditions or diseases. For instance, a change in t1/2 might reflect a change in Vd or clearance or both. In the same way, if both clearance and volume change in proportion, the t1/2 may be unaltered although average steadystate concentrations may change. Clearance and vol ume should therefore be calculated individually whenever possible [18,33]. In a number of diseases, clearance and distribution vol ume can be affected by changes in protein binding, result ing in unpredictable t1/2 changes. For example, in patients with acute viral hepatitis, tolbutamide t1/2 exhibits enhanced clearance which is unexpected. This occurs because the dis ease changes both plasma and tissues protein binding, does not affect volume of distribution, but increases clearance
because free drugs are present in the bloodstream in higher concentrations. Although it can be a poor indicator of drug elimination from the body, t1/2 does give a good indication of the time required to reach a steady state whenever a drug regimen is initiated or changed, the time needed for a drug to be cleared from the body, and a means to predict the appropriate dosing interval [12,17]. 17.2.6 Clearance Clearance (CL) is the most important concept to consider when designing a rational regimen for longterm drug administration. CL (expressed as volume/time) describes the volume of fluid that is completely cleared of drug per unit time, mainly through hepatic metabolism and renal excretion [12]. While elimination t1/2 determines the time required to achieve Css, the level of that steady state is determined through CL and dose alone. If the drug is given for long enough, the amount of drug eliminated from the body during one dosage interval is equal to the amount of drug that enters the systemic circulation during each dosage interval. This is known as a steady state. CL depends essen tially on liver and/or kidney efficiency to eliminate a drug, and it varies in certain diseases that affect those organs or the blood supply to them. In a stable clinical situation, CL remains constant and is directly proportional to dose rate. The clinical importance of this is that, for most drugs, if the dose rate is doubled, the Cssaverage doubles, and, if the dose rate is halved, the Cssaverage is halved. The concept of CL is very useful in clinical PKs because its value for certain drugs is generally constant over the clinically encountered concentration ranges. Drug elimination systems, such as metabolizing enzymes and transporters, are usually not saturated; therefore, the abso lute elimination rate is basically a linear function of the drug’s concentration in plasma. To be precise, the elimi nation of most drugs follows firstorder kinetics, where a constant fraction of drug in the body is eliminated per unit of time. If the elimination mechanisms of a particular drug become saturated, the kinetics move toward zeroorder kinetics, in which a constant amount of drug is eliminated per unit of time. In such situations, CL changes with drug concentration [12]. It is vital to understand that CL does not indicate how much of a drug is being cleared; rather, it indicates the vol ume of blood or plasma from which the drug would have to be totally removed to account for the CL. The CL by differ ent organs (GI tract, kidney, liver, and other organs) is addi tive. Diseases, drug interactions, or even genetic variants that decrease the activity of drugmetabolizing enzymes or mechanisms of execration may decrease CL; consequently, a dose adjustment to avoid drug toxicity is required. On the other hand, a number of drug interactions and genetic vari ants increase CYP expression, and therefore increased drug dosage might be required to maintain a therapeutic effect [12]. 17.2.7 Zero- and First-Order Kinetics The hallmark of firstorder (linear) kinetics is the pro portionality between dose rate and Css. Most drugs show firstorder kinetics, where a constant fraction of drug in the body is eliminated per unit time. Linear kinetics explains that the decrease in drug levels in the body is dependent on the plasma concentration (a concentration dependent process). The higher the drug concentration, the larger the absolute amount of drug eliminated per unit time. Therefore, the rate of elimination is proportional to the amount of drug in the body, while CL remains con stant. Generally, for drugs with firstorder kinetics, Vd, t1/2, k, and CL are all interrelated [12]. However, if the elimination mechanisms of a particular drug become saturated, the kinetics move toward zeroorder (nonlinear) kinetics, in which a constant amount of drug is eliminated per unit of time. Various drugs follow zero order kinetics at high or toxic concentrations. Metabolism in general, which involves particular enzymes, is one of the most important elements contributing to a drug undergo ing zeroorder kinetics. When metabolism enzymes reach a point of saturation, the rate of elimination does not increase in response to a concentration increase but rather becomes constant (a concentrationdependent process). Common examples of drugs that follow nonlinear kinet ics are aspirin, phenytoin, alcohol, heparin, and ethanol. The t1/2 is not constant for zeroorder reactions, but rather depends on concentration. The higher the concentration, the longer the t1/2. The clinical significance of nonlinear kinetics is that any small increase in dose can cause a large increase in concentration. This is mainly important when toxic side effects and concentration are strongly related, which is the situation with, for example, phenytoin [12]. 17.3 ADME: PHARMACOGENOMICS As discussed previously, there are many sources of varia tion in enzyme activity; age, enzyme induction or inhibi tion, and diseases (especially of the liver) are among them. Variation in the DNA sequence of genes encoding enzymes can abolish, reduce, or increase the expression and activity of an enzyme, and this can manifest as the “metabolizer” phenotype in an individual. Individuals who are homozy gous for the two alleles coding for “normal” enzyme func tion are termed extensive metabolizers (homozygous EM
or “wildtype”); those who are homozygous with two vari ant alleles resulting in inactive or absent enzymes are “poor metabolizers” (PM); those who are heterozygous manifest an intermediate metabolizer (IM) phenotype with reduced function (heterozygous EM). Intermediate and extensive metabolizers are often collectively referred as extensive metabolizers, especially in studies in which metabolizer status is assigned using phenotype. Gene duplication or multiplication, as, for example, seen in CYP2D6, can result in “ultrarapid metabolizer” (UM) phenotype. Standard drug doses achieve normal concentra tions and effect in homozygous EMs (which usually make up the largest proportion of the population), but they may be toxic in PMs (possibly in heterozygous EMs or IMs) and ineffective in UMs, who may require a higher dose to achieve therapeutic effective drug concentrations [9,21,34]. In general, the most important factor affecting drug action is the AUC at the site of action, reflecting the “metabolizer” phenotype; AUC is the best pharmacokinetic end point for the assessment of pharmacogenetic effects. However, even if there is a clear association between con centration and drug effect/toxicity, there are multiple fac tors that determine the clinical relevance of a functional polymorphism in a drug metabolizing gene [34]. These are summarized here: l Clear association between concentration and drug effect/toxicity. Genetic variation that affects the func tioning of the drug metabolizing gene can affect the AUC and hence can predict drug effects. lThe clinical significance of a genetic variant depends on the drug’s therapeutic index (the ratio between dose effectiveness and safety). A high therapeutic index indi cating that a drug is safe over a wide range of concentra tions and variations due to genetics may be irrelevant, whereas a low therapeutic index indicates that minor variations in drug concentration, such as from polymor phisms, may be important. lIf the drug effect is mediated by multiple active moi eties, the relevance of genetic variation is diminished. For example, if the product of the drug–enzyme reac tion is an active metabolite with activity similar to that of the parent molecule, an altered ratio of parent drug to metabolite (through polymorphism) may have little clinical effect. lIf a drug is metabolized or eliminated by multiple path ways, total ablation of one pathway (as in a PM) may result in minimal alteration of overall drug concentra tions and hence have minimal effect. lSometimes enantiomers have different pharmacological activity and pathways of elimination. This is called ste reoselective metabolism. lPhenocopying is the conversion of a patient from a phe notypic normal metabolizer into a slow metabolizer as a result of inhibition of the enzyme by another drug or by itself (“autophenocopying”). This can result in a reduc tion in a trait’s variation in a population. For example, all patients chronically taking a drug are poor metabo lizers either because of their genotype or because of autophenocopying. Phenocopying sometimes explains why differences between PMs and EMs are less at steady state than after single doses. 17.3.1 CYP450 Enzymes The major causes of interindividual and intraindividual variability in CYP activity are environmental factors (inducers and inhibitors), biological factors (gender, dis ease, and circadian rhythms), and genetic polymorphisms in CYP450 genes and their regulators. There are large varia tions between individual CYP450 isoforms in terms of their susceptibility to these mechanisms [35,36]. Overall, 57 CYP450 genes and 58 pseudogenes have been identified, 42 of which play a role in the metabolism of both exoge nous xenobiotics and endogenous substances (e.g., steroids and prostaglandins), and 15 of which are involved in the metabolism of drugs in humans [35,36]. CYP450 genes are highly polymorphic and can exhibit clinically significant genetic polymorphisms. In general, CYP3A4/5, CYP2D6, CYP2C9, CYP2C19, CYP2A6, CYP2B6, and CYP2C8 are the most important and most studied metabolic enzymes. The variations in CYP450 genes—deletions, missense mutations, deleterious mutations (creating splicing defects or premature stop codons), and duplications—can result in abolished, reduced, normal, or enhanced enzyme activity [21,35,37,38]. 17.3.1.1 CYP2D6 CYP2D6 is a member of the cytochrome P450 superfamily involved in metabolizing and eliminating many prescribed medications; it accounts for approximately 2% of total hepatic CYP450 content. The CYP2D6 gene is highly poly morphic, with at least 100 genetic variants and 120 alleles identified. Additionally, its variants are the best character ized among all CYP450 variants, and the distributions of these alleles exhibit notable interethnic differences. (See Chapter 16.) Approximately 7–10% of Caucasians and 1% of Chinese, Japanese, and Koreans are PMs of CYP2D6 [34,39,40]. In Caucasians, CYP2D6*3, *4, and *5 produce inactive enzymes or no protein products and are the variants most commonly implicated in the PM pheno type. CYP2D6*4 is the most common variant allele in Caucasians (allele frequency ∼21%), but it is virtually absent in Chinese, although overall CYP2D6 activity is lower in Chinese than in Caucasians as a result of the high allele frequency of CYP2D6*10 (∼50%), which is largely
absent in Caucasians. Gene duplication occurs in ∼1–7% of Caucasians and 29% of black Ethiopians; also, it is pre dictive of an UM phenotype [34]. This variant produces an unstable enzyme with reduced (but not absent) ability to metabolize substrate drugs. CYP2D6 is responsible for the metabolism of almost 20–30% of drugs and is the most widely studied enzyme in relation to polymorphisms. Examples include tricyclic antidepressants, selective serotonin reuptake inhibitors (SSRIs), antiarrhythmics, betablockers, neuroleptics, opi oid analgesics, antiemetics, and anticancer drugs. CYP2D6 is not inducible, and therefore variations in the enzyme expression and activity are largely attributable to genetic polymorphisms [21,41]. Powerful CYP2D6 inhibitors have been shown to significantly decrease EM metabolic capac ity; therefore, EM subjects may seem to be PMs during the coexisting administration of potent inhibitors (e.g., fluox etine and paroxetine) [39]. Recently, ADRs have been reported in UMs, mainly of a 10 to 30fold increase in drug metabolite concentrations. For example, codeine is converted through CYP2D6 to mor phine, which is more pharmacologically active. UMs receiv ing standard doses of codeine have been reported to display symptoms of narcotic overdose significantly related to higher morphine concentrations [39,42]. Perhexiline is an antiangi nal drug that is almost entirely metabolized by CYP2D6 to hydroxyperhexiline (inactive). PMs have trough concentra tions up to 6fold higher than EMs after a single dose, with evidence of saturable metabolism. The major toxicity of per hexiline is hepatotoxicity and peripheral neuropathy, which are concentrationdependent. Therapeutic drug monitoring has assisted dosing, with a suggested range of 0.15–0.6 mg/l, supported by both concentrationdependent efficacy and tox icity [43]. This supports a case for prospective genotyping. In most individuals, only a small fraction (∼10%) of codeine is metabolized to morphine via CYP2D6, with most of that fraction glucuronidated to codeine6glucuro nide and the remainder metabolized by CYP3A4 to nor codeine. The AUC of codeine is similar in PMs and EMs, whereas morphine is virtually undetectable in PMs as well as EMs taking quinidine (phenocopying). Clinical studies in volunteers generally support the lack of analgesia in PMs, which is consistent with the belief that morphine is the key metabolite responsible for codeine’s antinociceptive effects. Theoretically, UMs may convert codeine to morphine more rapidly, resulting in increased opioid effects for a given dose. However, studies to date have been with volunteers and not the target population. Overall there is a strong argument for a gene–concen tration effect resulting from the failure of prodrug conver sion to morphine in PMs and phenocopied EMs. As far as the gene–effect relationship is concerned, there seems to be a predictable failure of analgesia in healthy volunteers but a less clear relationship with adverse effects. Although nonresponse to codeine may be explained by PM status, it also may be due to other factors, including phenocopying by a CYP2D6 inhibitor such as paroxetine [34]. The beta1selective metoprolol appears to have both consistent gene–concentration and gene–effect relation ships in healthy volunteers, suggesting that dose reduc tion to ∼25% should occur in PMs or those phenocopied by other drugs [34]. Metoprolol is administered as a race mate, with Smetoprolol thought to produce most of the βblockade. The main metabolite, Odesmethylmetoprolol (essentially inactive), accounts for approximately two thirds of the metabolism and occurs via various path ways, including CYP2D6 (mainly Rmetoprolol); another pathway, to αhydroxymetoprolol, accounts for ∼10% of the dose in EMs and seems to be under CYP2D6 control because very little is produced in PMs. AUCs of metoprolol are 4 to 6fold higher in PMs than in EMs after one dose and 3 to 4fold higher after repeated dosing. UMs achieve metoprolol concentrations that are half those observed in EMs. The strong CYP2D6 inhibitor, paroxetine, increases the mean AUCs of S and Rmetoprolol by 5 and 8fold, respectively. Enhanced or prolonged βblockade has been observed in both PMs and EMs receiving CYP2D6 inhibitors [44,45]. Three recent prospective studies failed to show a relation ship between CYP2D6 and adverse effects with metoprolol [46–48]. The high therapeutic index of the drug and the fact that the effects of excessive βblockade are usually easy to detect clinically (e.g., bradycardia) lessen the need for genotype testing. The long history of metoprolol use with out genotyping or phenotyping suggests that these tests are unlikely to happen in practice. 17.3.1.2 CYP2C9 CYP2C9 is the most abundant of the CYP2C enzymes and accounts for about 30% of total hepatic CYPs. It plays an important role in the metabolism of about 10% of the drugs available on the market, including nonsteroidal antiinflam matories (e.g., ibuprofen), antiepileptics (e.g., phenytoin), oral anticoagulants (e.g., warfarin), and antihypertensives (e.g., losartan). The human CYP2C9 and CYP2C19 genes are highly homologous at the nucleotide level [39,49], and more than 30 variants of CYP2C9 have been identified. Generally, CYP2C9 polymorphisms encode proteins with a loss of catalytic function, with the extent this reduction often being substratedependent. Many variants have been associated with reduced enzyme activity (see http://www.cypalleles.ki.se), with CYP2C9*3 and, to a lesser extent, *2 having the most clin ical relevance. In vitro studies show that *3 is associated with a lower intrinsic clearance of substrate drugs than is *2. The effects of CYP2C9*2 are more substrate spe cific (e.g., warfarin and phenytoin), whereas CYP2C9*3 displays reduced catalytic activity toward the majority of
CYP2C9 substrates. The clinical importance of CYP2C9 polymorphisms is demonstrated by the dose adjustment of an oral anticoagulant, warfarin, based on the CYP2C9 genotype[4,39,49,50]. At least one CYP2C9*2 or *3 allele is carried by ∼20% and 12% of Caucasians, respectively, with ∼2.5% being homozygotes for *2 or *3, or for compound heterozygotes for both alleles. The remaining twothirds of the Caucasian population are wild types and have normal enzyme activity. The small proportion of individuals (∼0.4% of Caucasians) homozygous for CYP2C9*3 have the lowest ability to metab olize substrate drugs. The CYP2C9*2/*2 and *2/*3 geno types may also cause important reductions in the metabolism of some drugs (e.g., phenytoin). CYP2C9*2 and *3 are rare in African American and Asian populations, with the wild type making up more than 95% of these populations [51]. The three oral coumarin anticoagulants—warfarin, acenocoumarol, and phenprocoumon—exist as S and Renantiomers. The Senantiomers are CYP2C9 sub strates and are responsible for most of the effects of war farin and phenoprocoumon. In contrast, although S and Racenocoumarol have comparable activities, rapid elim ination of the Senantiomer means that Racenocoumarol produces most of the anticoagulant effect. All three oral anticoagulants have low therapeutic indices, and the dose required to produce a target prothrombin time is largely unpredictable [34] (see Chapter 24). CYP2C9 is impor tant for the metabolism of this class, with CYP2C9*3 (but not CYP2C9*2) being clearly implicated in impaired clearance of tolbutamide, glyburide, and glimepiride [52–54]. Losartan is metabolized by CYP2C9 via an aldehyde intermediate (E3179) to E3174, the predominant active moiety. E3174 is at least 10fold more potent than losar tan at the AT1 receptor. Furthermore, although only 14% of losartan is metabolized to E3174, the AUC of the latter is 4fold to 8fold higher than that of the parent and is thought to be responsible for most of the activity [55]. This find ing suggests that individuals with CYP2C9 variants might exhibit a reduced losartan response. Only one of the three studies in healthy volunteers that included blood pressure assessments reported a significant influence of CYP2C9; they reported reduced response among those with the *1/*3 genotype [56–58]. Studies on the impact of genotype on AUC have been inconclusive, and it seems unlikely that CYP2C9 variants significantly affect parent losartan con centrations since production of E3174 constitutes a quanti tatively minor route of elimination [56–59]. Phenytoin is a commonly used antiepileptic drug despite its complex nonlinear pharmacokinetics and low therapeutic index. It is primarily (80–90%) eliminated via 4′hydroxylation to 5(4phydroxyphenyl)5phenylhydan toin (HPPH), largely via CYP2C9, which preferentially produces the Senantiomer of HPPH. The presence of at least one CYP2C9*2 or *3 allele is associated with one third lower mean dose requirements (199 versus 314 mg/ day, respectively). Furthermore, a “genedose” effect seems to exist, with dose requirements of 314, 193, 202, 217, and 150 mg/day for the CYP2C9*1/*1, *1/*2, *1/*3, *2/*2, and *2/*3 genotypes, respectively [60]. In a singledose study in healthy volunteers 30% lower concentrations were seen in wildtype individuals com pared with carriers of CYP2C9*2 or *3 alleles. CYP2C9*2 and *3 were suggested as accounting for 31% of the varia tion in phenytoin concentrations taken 12 h after a single 300 mg dose [61]. As central nervous system toxicity (e.g., ataxia and nystagmus) is closely related to concentration, it is likely that individuals with CYP2C9 variants are predis posed to these effects. There is some evidence for an asso ciation between CYP2C9*1 and *3 genotype and cutaneous reactions [62]. A reasonable case for genotyping in relation to phenytoin to guide the initial choice of a maintenance dose can be made on the basis of a strong gene(dose) con centration effect, a moderate geneeffect relationship, a strong concentrationeffect relationship for both desired and adverse effects, and a low therapeutic index. Points against genotyping include the fact that an alternative pathway (CYP2C19) exists for phenytoin metabolism. Therapeutic drug monitoring continues to be necessary, and clinicians have long experience with phenytoin. 17.3.1.3 CYP2C19 CYP2C19 accounts for about 3% of total hepatic CYPs. The ability of individuals to metabolize Smephenytoin has enabled them to be classified as PMs or EMs. PM pheno types for CYP2C19 are common (20%) among Asians and rare (3–5%) in Europeanderived populations. CYP2C19 catalyzes the metabolism of several drugs, including pro ton pump inhibitors (PPIs) (e.g., omeprazole, lansoprazole, pantoprazole), antidepressants (e.g., citalopram and ami triptyline), antiplatelet drugs (e.g., clopidogrel), antifungals (e.g., voriconazole), and anticancer compounds (e.g., cyclo phosphamide). Seven variants (*2–*8) in the CYP2C19 gene have now been associated with reduced enzyme activ ity in vivo, largely due to production of inactive enzyme protein. A novel CYP2C19 variant (CYP2C19*17) that may pro duce an ultrarapid metabolizer phenotype was recently iden tified [63]. A splice site mutation, CYP2C19*2 (rs4244285, 19154G >A), and a premature stop codon, CYP2C19*3 (rs4986893, 17948G >A) represent the two most predomi nant null alleles [4,39,64]. Genotyping for CYP2C19*2 and *3 identifies most PMs in African American and Chinese populations, while genotyping for CYP2C19*2 identifies 70–85% of variant reducedactivity alleles and CYP2C19*2–*8 identifies more than 99% of PMs [65].
CYP2C19 plays an important role in protonpump inhibitor (PPI) therapy for peptic ulcers and gastroesopha geal reflux disease. In EMs, it is responsible for >80% of the metabolism of omeprazole, lansoprazole, and pantopra zole [66], the remainder with CYP3A4 [39,64]. The metab olites produced are inactive. A fourth PPI, rabeprazole, may be less reliant on CYP2C19 as it undergoes nonen zymatic conversion to rabeprazole thioether. SOmeprazole (esomeprazole) was recently marketed as an individual entity to exploit its reduced variation in CYP2C19 geno type pharmacokinetics compared with the racemate or with Romeprazole [67]. The AUCs of both omeprazole and lansoprazole are 4 to 15fold higher in PMs than in homozygous EMs; with mul tiple dosing, the increase in AUC of omeprazole (but not of lansoprazole or pantoprazole) decreases to ∼2fold in EMs because of inhibition of its own metabolism by CYP2C19 [68]. This does not occur in PMs who lack a functioning CYP2C19 enzyme to inhibit. The AUCs of rabeprazole are also increased but less markedly in CYP2C19 deficiency. Individuals with CYP2C19 deficiency have superior acid suppression with conventional doses of omeprazole and lansoprazole [69], and increasing the lansoprazole dose from 30 mg once daily to four times daily in homozygous EMs leads to an increase in mean 24h intragastric pH from 4.5 to 7.0 [70]. Overall, omeprazole and lansoprazolebased regimens produce lower eradication rates in homozygous EMs than in heterozygous EMs or PMs [71]. If patients are confirmed as being PMs, dual therapy with PPI and amoxicillin may be appropriate, as the eradication rate is likely to be high (>90%) [72]. The PPIs are an exceptionally welltolerated class of drugs, and there seems to be no clear evidence of increased toxicity in PMs despite a markedly elevated AUC. However, individuals with a CYP2C19 deficiency may be predisposed to vitamin B12 deficiency during longterm use of this class [73]. 17.3.1.4 CYP2A6 CYP2A6 is a member of the cytochrome P450 superfamily and is expressed predominantly in hepatic cells, with some expression in specialized extrahepatic cell types. Compared with CYP2D6 and CYP3A4, relatively few clinically used drugs are metabolized to a significant point by CYP2A6. Its substrates include coumarin, halothane, methoxyflurane, valproic acid, disulfiram, and losigamone [74]. A number of significant variations in CYP2A6 have been identified, including CYP2A6*2 (rs1801272, 479T >A), CYP2A6*4 (gene deletion), CYP2A6*5 (rs5031017, 1436G >T), and CYP2A6*20 (rs28399444, frame shift). These polymor phisms are associated with abolished enzyme activity and have different allele frequencies among ethnic groups. The prevalence of CYP2A6 PMs in the Caucasian population is ≤1% [39]. A deletion of the CYP2A6 gene is common in the Asian population and accounts for significant differences in PMs compared with the Caucasian population. Since CYP2A6 is the highaffinity metabolizer of both nicotine and its oxi dized metabolite, cotinine, CYP2A6 variants have mainly been studied for treating tobacco abuse. Studies have revealed that the kinetics of nicotine metabolism are dif ferent in individuals carrying the variant CYP2A6 alleles. For example, in three studies, smokers who carried the CYP2A6 variants smoked fewer cigarettes and were more likely to quit smoking. These results reflect the possibil ity of increased nicotine concentrations and, subsequently, increased nicotine tolerance and ADRs from nicotine, espe cially in CYP2A6 PMs [39,74,75]. 17.3.1.5 CYP2B6 CYP2B6 accounts for 6–10% of the total CYP con tents in the liver, with a substantial (>100fold) varia tion in expression between individuals [76]; it is involved in the metabolism of an increasing number of clinically important drugs (∼8% of those on the market) [39,77], including bupropion, cyclophosphamide, efavirenz, and methadone. It also metabolizes some procarcinogens (e.g., 6aminochrysene) and drugs of abuse (e.g., Nmethyl 3,4methylenedioxyamphetamine, “ecstasy”). CYP2B6 is subject to inhibition and induction by drugs such as clopi dogrel and phenobarbital respectively [78,79]. A number of CYP2B6 variants have been identified. CYP2B6*6 (rs3745274, 516G >T and rs2279343, 785A >G) is the most common polymorphism. It commonly occurs in Caucasians and Asians, while CYP2B6*16 (rs2279343, 785A >G) and CYP2B6*18 (rs28399499, 21011T >C) are common in Africans [39]. CYP2B6 is the main catalyst of efavirenz metabolism (to its inactive 8OH metabolite); therefore, its polymor phisms may have major implications for the PKs and tox icity of this drug, which at present is recommended as an option in firstline combination therapy for HIV infections [77]. Individuals homozygous for the 516T variant or the CYP2B6*6 allele (516G >T and 785A >G) may have two to threefold higher efavirenz concentrations and may be pre disposed to side effects. Studies have shown that when the efavirenz dose is decreased, the plasma concentrations of the drug decreases proportionally. It is reported that patients who were *6/*6 homozygous and *6/*26 and *1/*26 het erozygous had lower plasma concentrations of efavirenz and less frequent CNS effects when the dose was decreased [39,80]. 17.3.1.6 CYP2C8 CYP2C8 accounts for approximately 7% of total hepatic content and plays a vital role in the metabolism of pio glitazone, amiodarone, paclitaxel, chloroquine, verapamil,
and ibuprofen. It also plays a secondary role in the metab olism of fluvastatin, amitriptyline, diclofenac, omepra zole, and carbamazepine. The metabolism of paclitaxel to 6αhydroxypaclitaxel, which is essentially inactive, has been used as an index of CYP2C8 activity in vitro [81]. The most common CYP2C8 variants are CYP2C8*2 (rs11572103, 11054A >T), CYP2C8*3 (rs11572080, 2130G >A and rs10509681, 30411A >G), and CYP2C8*4 (I264M substi tution); they lead to decreased enzyme activity [21,39,82]. CYP2C8 is primarily responsible for the hydroxylation of Ribuprofen; CYP2C9, for Sibuprofen. The mean AUC after a single dose of 400 mg of racemic ibuprofen was reported to be 2.2 and 8.7fold higher in individuals with the CYP2C8*1/*3 or *3/*3 versus *1/*1 genotype [83]. CYP2C9*2 has been associated with altered ibupro fen pharmacokinetics only when it is coinherited with CYP2C8*3 (linkage disequilibrium) [83]. Several studies suggest that reduced paclitaxel metabolism may occur with the CYP2C8*2 and *3 alleles, with the latter producing the greatest reduction [84]. In vitro data suggest that CYP2C8 and CYP3A4 are responsible for inactivating repaglinide, a nonsulfonylurea insulin secretagogue. CYP3A4 inhibi tors such as ketoconazole and clarithromycin increase repa glinide concentrations by small amounts (15% and 40%, respectively) compared with a 5.5 to 15.5fold increase with the CYP2C8 inhibitor gemfibrozil [85]. CYP2C8*2 is more common in Africans, while CYP2C8*3 and *4 are common in Caucasians. Both CYP2C8 and CYP2C9 play an important role in ibuprofen metabo lism, and it has been found that CYP2C8 plays a key role in the hydroxylation of the (R)enantiomer. For this reason, ibuprofen disposition has been used as a guide for CYP2C8 activity. When the effects of the different genotypes on ibu profen PKs were investigated, CYP2C8 polymorphisms were found to be common [82,86]. Studies have shown that the presence of the CYP2C8*3 allele causes a significant effect on the disposition of ibu profen, which suggests that a considerable proportion of Caucasian subjects may show alterations in the disposi tion of drugs that are CYP2C8 substrates. Moreover, in one study CYP2C8*3 carriers who received Ribuprofen had lower ibuprofen CLs; the CYP2C8*3 heterozygotes who were treated with the antidiabetic repaglinide had a higher metabolism of the drug when compared to CYP2C8*1 and *4 carriers [39,82,86]. 17.3.1.7 CYP3A4/5 The CYP3A locus comprises four genes that code for the functional enzymes: CYP3A4, CYP3A5, CYP3A7, and CYP3A43. Of these CYP3A7 is primarily fetal and CYP3A43 is minimally expressed and has low functional activity. CYP3A4/5 is responsible for metabolizing the larg est number of prescribed drugs (more than 50%), includ ing cyclosporin, indinavir, nifedipine and simvastatin. These genes are often considered collectively as “CYP3A” because of their promiscuous substrate specificity and the difficulty discerning the relative role of each isoform in drug metabolism [34]. The overall activity of CYP3A is unimodally distributed, exhibits wide interindividual vari ability (>10fold), and is highly susceptible to the effects of enzyme inhibitors and inducers [87]. More than 20 variants in the coding region of CYP3A4 have been described [39], and a number of them have altered enzyme activities, ranging from a modest to a significant loss in catalytic efficiency; however, their clinical signifi cance is uncertain. CYP3A4*2 (rs55785340, 664T >C) and CYP3A4*7 (rs56324128,167G >A) have higher frequencies in Caucasian populations, whereas CYP3A4*16 (rs12721627, 554C >G) and CYP3A4*18 (rs28371759, 878T >C) have higher frequencies in Asian populations [4]. A single point mutation (6986A >G) in intron 3 of CYP3A5 (designated CYP3A5*3) produces a truncated and nonfunctional protein. In contrast to the unimodal dis tribution of CYP3A taken as a whole, CYP3A5 exhibits a bimodal distribution that can be predicted by the pres ence or absence of this allele. Individuals homozygous for CYP3A5*3 produce little CYP3A5 enzyme (“low expres sors”), whereas the remainder have at least one wildtype (CYP3A5*1) allele and express a large amount of CYP3A5 (“high expressors”). The PKs of the immunosuppressive agent tacrolimus are mainly dependent on the CYP3A5 genotype, and in terms of the dose required to reach target trough blood concen trations (C0). Tacrolimus C0 and its dose requirements are related to CYP3A5 polymorphisms; for example, individu als who carry at least one CYP3A5*1 allele have a functional CYP3A5 and consequently need a higher dose of tacrolimus to reach the targeted blood concentration. CYP3A5 expres sors experience a delay in reaching target concentrations [4,88]. It is estimated that 10–20% of Caucasians, 40–50% of East Asians, 60–70% of Hispanics, and >80% of African Americans are high expressors of CYP3A5 [89]. 17.3.2 Non-CYP450 Enzymes Here we discuss the variations in the nonCYP450 enzymes: UGT, TPMT, NAT, and GST. 17.3.2.1 UDP Glucuronosyltransferases UDP Glucuronosyltransferases (UGTs) are expressed in a tissuespecific and frequently inducible way in most human tissues; their highest concentration is found in the GI tract and liver. UGTs contribute to about 35% of phase II drug metabolism and are involved in the glucuronidation of many endogenous compounds and xenobiotics [39]. Nineteen human genes are encoded by UGT proteins: 9 by UGT1 and 10 by UGT2. While both types are involved in the process
of drug metabolism, UGT2 seems to have greater speci ficity for the glucuronidation of endogenous substances (e.g., steroids). UGT1A1 expression plays a vital role in drug metabolism because the glucuronidation of bilirubin by UGT1A1 is the ratelimiting step that assures effective bilirubin conjugation levels. This rate can be influenced by both genetic variation and drug treatments. If a drug under goes selective metabolism by UGT1A1, competition for its metabolism with bilirubin, glucuronidation occurs, result ing in marked hyperbilirubinemia in addition to decreased clearance of the metabolized drug [4,20,90]. UGT1A1 recently gained recognition as the first pharma cogenetic test to achieve FDA approval for use in conjunc tion with a specific drug (irinotecan). Irinotecan is a prodrug that is metabolized by carboxylesterases to the active topoi somerase inhibitor 7ethyl10hydroxycamptothecin (SN 38) and by CYP3A4 to inactive metabolites. Thereafter, SN38 is glucuronidated by UGT1A1, with the resultant SN38 glucuronide excreted into the intestine via bile [91]. The activity of UGT1A1 varies widely, with an in vitro study demonstrating a 17fold variation in SN38 gluc uronidation. UGT1A1*28 is the variant most frequently implicated in defective SN38 glucuronidation and involves an extra thymineadenine (TA) repeat in the TATA sec tion of the UGT1A1 promoter (i.e., (TA)7TAA instead of (TA)6TAA in the wild type). It is also the primary cause of Gilbert’s syndrome [92]. This variant occurs com monly, with the homozygous genotype found in 5–15% of Europeans, 10–25% of Africans and South Asians, and 1–5% of Southeast Asians and Pacific Islanders. There is an increased risk of severe neutropenia and diarrhea with irinotecan in homozygotes for UGT1A1*28 compared with homozygotes for UGT1A1*1 [93]. Gilbert’s syndrome can be caused by a number of genetic changes, and patients with it may experience adverse drug reactions (ADRs) because of reduced capacity to metabo lize drugs by UGT1A1. UGT1A1*37 has an 8TA insertion in the promoter and results in more decreased promoter activity than that of the UGT1A1*28 allele. UGT1A1*36 (rs8175347) has only a 5TA insertion in the promoter and is associated with increased enzyme activity and a decreased risk of neonatal hyperbilirubinemia [20,90]. 17.3.2.2 Thiopurine S-methyltransferase Thiopurine Smethyltransferase (TPMT) catalyzes the Smethylation of 6mercaptopurine (6MP), azathioprine (AZA), and thioguanine to inactivate these drugs, which are used for treating leukemia and autoimmune diseases. It is a cytosolic enzyme found in many tissues, with activity most commonly determined in red blood cells. In Caucasians, a trimodal distribution exists, with 0.3–0.6% having low or undetectable activity, 10% having intermediate activity, and the remaining 90% having high (normal) activity [94]. The TPMT gene exhibits significant genetic polymorphisms across all of the ethnic groups studied, and TMPT activ ity is highly variable, as well as polymorphic, in all large populations evaluated up to the present time. About 90% of Caucasians inherit high enzyme activity, 10% inherit intermediate activity (heterozygous), and 0.3% inherit low or no activity. A number of clinically significant SNPs have been identified for TMPT that can alter its activity, and since methylation is involved in both the activation and the metabolism of mercaptopurine, altered enzyme activity affects the concentration of both active and toxic metabo lites [4,21,39]. The genetic polymorphisms of TMPT have the strongest case for prospective pharmacogenetic testing. Generally, thiopurine agents have a narrow therapeutic index, with lifethreatening myelosuppression being a major concern. Patients with polymorphic TMPT frequently require a sig nificant dose reduction in order to prevent toxicity. More than 20 variants of the TPMT gene have been identified; however, the most studied are TPMT*2, *3A, *3B, and *3C, which are responsible for most cases of TPMT deficiency. These three variants account for 80–95% of intermediate and poor metabolizers [4,21,39]. *3A is the most common variant in Caucasians; *3C is the most common in Africans and Southeast Asians. Homozygotes for a variant allele (e.g., TPMT*3A/*3A) have negligible TPMT activity, whereas heterozygotes (e.g., TPMT*1/*3A—intermediate activ ity) have an average activity that is approximately half that observed in homozygotes for wildtype alleles (TPMT*1/*1, normal/high activity). Genotyping is reasonably accurate in predicting TPMT phenotype, which is defined as low, inter mediate, or normal/high TPMT activity [95]. The therapeutic niche of TPMT testing relates to its ability to identify prospectively the small proportion of patients (0.3–0.6%) with enzyme deficiency who will almost certainly develop lifethreatening myelosuppression if standard doses are used. Although it is clear that TPMT deficiency (particularly complete deficiency) predisposes to myelotoxicity, approximately 75% of cases of bone marrow depression occur in those without mutations (i.e., presumed normal enzyme activity) [96]. This indicates that TPMT metabolizer status is a useful adjunct to (but not replace ment for) regular bloodcount monitoring. The clinical use fulness of prospective determination of TPMT metabolizer status is supported by pharmacoeconomic studies [97]. 17.3.2.3 Dihydropyrimidine Dehydrogenase Dihydropyrimidine dehydrogenase (DPD), an enzyme encoded by the DPYD gene, metabolizes two endogenous pyrimidines—thymine and uracil—and facilitates the metabolism of the pyrimidine analog 5fluorouracil (5FU). DPD activity in peripheral blood mononuclear (PBM) cells has been used as a surrogate for total body DPD activity and varies up to
20fold. DPD activity is usually described
as normally distributed, perhaps reflecting the absence of cases of absolute deficiency in the studies [98]. DPD poly morphisms result in DPDdeficient phenotypes with a total frequency of about 3–5%. DPD variants, such as DPYD*13 (rs55886062, 1679T >G), DPYD*9A (rs1801265, 85T >C), DPYD*2A (rs3918290, IVS14+1G >A), and 2846A >T (rs67376798, D949V), are among the identified SNPs asso ciated with grade 3 and grade 4 toxicities in 5FU–treated patients. For example, DPYD*2A homozygous patients have a complete absence of normal DPD activity, whereas heterozygous carriers have a 50% absence of the enzyme activity, resulting in 5FU accumulation and significant life threatening 5FU–related toxicities [39]. Moreover, DPYD*9A is a common missense muta tion with a C29R substitution; however, its association with reduced DPD activity is still debatable. For exam ple, two different studies reported that cancer patients who were DPYD*9A heterozygous had severe 5FU tox icity in. Together, DPYD*2A and DPYD*9A polymor phisms have a high concordance with 5FU toxicity but a low concordance with enzyme activity. Determination of PBM cell DPD activity (i.e., phenotype) may identify up to 60% of patients who may develop severe toxicity, whereas screening solely for the DPYD*2A allele (geno type) identifies approximately 25% of these patients. Identification of a patient with absolute deficiency would allow selection of alternative chemotherapy, whereas those with partial deficiency can be treated with a lower dose of 5FU. The increasing availability of DPD inhibi tors (eniluracil) may make assessment of metabolizer status redundant. 17.3.2.4 N-acetyltransferases Human Nacetyltransferases (NATs) catalyze the acetyla tion of aromatic amines and hydrazines. There are two func tional NAT genes in humans, NAT-1 and NAT-2, which carry functional polymorphisms that influence enzyme activity [4]. More than 25 allelic variants of these functional genes have been identified. Based on the level of NAT activity, patients can be classified into two phenotypes: fast acety lator (FA—wildtype NAT acetylation activity) and slow acetylator (SA—reduced NAT enzyme activity). NAT1 activity is usually constant, whereas NAT2 polymorphisms result in individual differences in the rate at which drugs are acetylated. While several mutations have been identified in the NAT1 and NAT2 genes, the frequency of the slow acety lation patterns is attributed mostly to polymorphisms in the NAT2 gene [4,20,21]. Altogether, NAT2 polymorphisms result in more than 10 NAT2 alleles, and the variant alleles that account for the majority of SA phenotypes include NAT2*5 (rs1801280, 341T >C), which results in an I114T substitution; NAT2*6 (rs1799930, 590G >A), which causes a R197Q substitution; and NAT2*7 (rs1799931, 857G >A), which corresponds to a G186E substitution [39]. A number of compounds have been used as a probe for NAT2, and a good genotype–phe notype correlation is generally observed (>90%) with caf feine as a probe [99]. Slow acetylators are found in approximately 50% of European and Africanderived populations, but they are less common in Asians. Individuals phenotyped as slow acetylators are likely to have two slow activity alleles, whereas those phenotyped as fast acetylators can have one or two highactivity alleles (most likely NAT2*4, which is considered to be the wild type). In Caucasians, most of the fast acetylators are heterozygotes for slow and fast alleles, whereas individuals homozygous for fast alleles (e.g., NAT2*4/*4) comprise ∼5% of the population [99]. In contrast, ∼30% of Chinese individuals are homozy gous for rapid alleles, 45% are heterozygous, and 25% are homozygous for slow alleles [100]. NAT2 polymorphisms and their association with the slow acetylation of isonia zid (an antituberculosis drug) were one of the first fully characterized genotypes shown to influence drug metabo lism, which links the pharmacogenetic (PG) phenotype to a genetic polymorphism [20] (see Chapter 36). Since NAT1 and NAT2 catalyze the bioactivation (via Oacetylation) of aromatic and heterocyclic amine carcinogens, genetic variations in the NAT1/2 genes may alter the cancer risk associated with exposure to these carcinogens. NAT2 poly morphisms are related to individual susceptibility to par ticular cancers caused by industrial chemicals (e.g., α and βnaphthylamine); for example, individuals with poor acet ylator phenotypes have increased risks of lung, bladder, and gastric cancers if exposed to carcinogenic arylamines for a long period of time [4,21]. 17.3.2.5 Glutathione S-transferases The superfamily of human glutathione Stransferases (GSTs) facilitates the conjugation of glutathione (GSH) to different endogenous metabolites and xenobiotics. GSTs include 10 members, which are divided into 3 main groups: cytosolic/nuclear, mitochondrial, and microsomal. Cytosolic GSTs are also divided into 7 classes: alpha, mu, omega, pi, sigma, theta, and zeta, the most important of these being GSTM1, GSTT1, GSTP1, and GSTA1. In addi tion to their enzymatic function, GSTs also act as regulators of many physiological processes, including cell signaling and growth factors and DNA replication [21]. The PKs of many drugs can be affected by GST polymor phisms, and since they decrease the activity of metabolizing enzymes, biologically active parent drugs or metabolites can accumulate and reach toxic levels. A number of known GST variants are well characterized in their influence on drug disposition. Two polymorphisms of the GSTP1 gene have been identified: 1404A >G SNP (rs947894), with an I105V substitution, and 2294 C >T SNP (rs1799811), with an A114V substitution [39].
17.3.3 Drug Transporters The movement of drugs across the cell membrane is through a combination of both passive diffusion and active transport, facilitated by certain drug uptake and efflux molecules. Drug transporters are membrane proteins that exist at vari ous endothelial and epithelial barriers, including the BBB, intestinal epithelial cells, hepatocytes, and renal tubular cells. These transporters can significantly affect drug dispo sition. For example, the influx of a drug from the blood to the liver, where it is subsequently metabolized and excreted, may increase the rate of elimination. These proteins and the genes that encode them are essential to drug uptake, bioavailability, targeting, efficacy, toxicity, and clearance. Genetic variation in the genes encoding these transporters can result in variable expression levels and transport effi ciency, which can have an impact on drug pharmacokinetics and response to treatment. More than 300 human genes code transporters or trans porterrelated proteins, most of which work on endogenous substrates, although some also transport xenobiotics. Drug transporters can generally be classified into two groups: the efflux adenosine triphosphatebinding cassette (ABC), known as the multidrug resistance (MDR) family of trans porters, and the uptake solute carrier singlelevel cell (SLC) family of transporters. SLCs mediate passive movement of solutes down their electrochemical gradient, while ABCs are active pumps fueled by ATP [39,101]. 17.3.3.1 ABC Transporters There are a total of 49 members of the human ABC trans porter family, and they are grouped into seven subclasses or families (ABCA through ABCG) [102]. These trans porters generally counteract uptake through the intestinal wall, efflux substrates out of tissues into systemic circula tion, and eventually mediate the clearance of drugs from the body. Of all ABC transporters, the best known are ABCB1 (Pglycoprotein, Pgp, or MDR1), ABCC1/2 (mul tidrug resistance–related proteins 1/2, MRP1/2), ABBC2 (multidrug resistance, MRP2) and ABCG2 (breast cancer resistance protein, BCR). ABCB1 and ABCG2 are the most characterized [103]. They are expressed in the enterocytes, colon, intestinal epithelium, canalicular plasma membrane of hepatocytes, proximal renal tubule, hematopoietic stem cells, blood–brain barrier, heart, nerves, testes, and pla centa. In all of these tissues, except the blood–brain barrier, they mediate efflux substrates out and into systemic circu lation. ABCB1 in the choroid plexus, transports molecules from circulation into cerebrospinal fluid. It is believed that the evolutionary role of ABC trans porters is to limit the penetration of toxic molecules into critical organs, thereby protecting blood–tissue barriers [104,105]. ABCC1 is expressed ubiquitously and is local ized to the basolateral, rather than apical, membranes of epithelial cells. Because of its basolateral localization, ABCC1 pumps drugs into the body rather than into the bile, urine, or intestine. For this reason, it is thought to serve mainly as a protective barrier in tissue epithelial cells rather than as a classic drug efflux pump. ABCC2 is functionally similar to ABCB1 and is expressed on the apical domain of epithelial cells in liver, intestine, and kidney [104,105]. ABCB1 ABCB1/MDR1 was the first recognized and most exten sively studied ABC transporter encoding a polypeptide (Pglycoprotein, or Pgp). Pgp has two halves, each con taining six hydrophobic transmembrane domains and an ATPbinding domain. The expression of ABCB1 on the apical surface of enterocytes and the canalicular part of the hepatocyte has been shown to influence intestinal drug absorption and limit oral bioavailability of a wide variety of structurally diverse drugs. Additionally, it facilitates hepa tobiliary excretion and renal excretion, and protects the brain and fetus from xenobiotics. ABCB1 overexpression in cancer cells is involved in multidrug resistance to chemo therapeutic agents [21,106,107]. ABCB1s are responsible for amphipathic lipidsoluble drugs. They transport a broad spectrum of structurally and functionally different drugs, such as anticancer agents (e.g., anthracyclines), antibiotics (e.g., erythromycin), immunosuppressants (e.g., cyclosporine), cardiac drugs (e.g., digoxin), calcium channel antagonists (CCB) (e.g., diltiazem), and HIV protease inhibitors (e.g., ritonavir). The most common SNPs are the synonymous 1236C >T (rs1128503) and 3435C >T (rs1045642), and the nonsynony mous 2677G >T (rs2032583) [21,106], all three of which have allele frequencies that vary in different ethnic popula tions, with a higher frequency among Caucasians and Asians. The 3435C >T SNP has strong linkage disequilibrium with other SNPs in the ABCB1 gene, creating common haplotypes consisting of 3435C >T combined with 2677G >T and/or 1236C >T. In general, SNP association studies on bioavail ability and efficacy are inconclusive [105]. 3435C >T SNP was first shown to be associated with reduced serum digoxin concentrations, whereas it was also associated with higher plasma digoxin levels. The association was stronger when the ABCB1 2677G >T/A and 3435C >T polymorphisms were evaluated together as a haplotype [108]. Many investigators have now found similar associations between these polymorphisms and plasma concentrations of several other drugs, although these observations have not been consistently confirmed. More recent evidence suggests that polymorphic ABCB1 expression influences not only plasma pharmacokinetics but also the degree to which drugs are able to penetrate into tissues that express ABCB1. In a study of chronic myeloid leukemia patients treated with imatinib, patients with the 1236C >T had higher imatinib plasma concentrations and also showed an improved therapy response, whereas the
presence of the wildtype 2677G variant worsened clinical response [109]. In a stable recombinant cell model, the anti cancer drugs doxorubicin HCl, daunorubicin HCl, vinblastine sulfate, vincristine sulfate, taxanes (paclitaxel), and epipodo phyllotoxin (etoposide, VP16) exhibited selectively reduced Pgpmediated resistance in 561A carriers [110]. In general, there is an overlap in substrate specificity and tissue distribution for ABCB1 and CYP3A4/5. Some drugs are dual substrates for both ABCB1 and OATP transport ers (e.g., digoxin, fexofenadine, atenolol). Cyclosporine is not only transported by ABCB1, but is also metabolized by CYP3A4; consequently, the possible ABCB1 effects can be influenced by the activity of OATP transporters or CYP3A4 [21,39]. These may explain the conflicting results obtained with ABCB1 studies. ABCC1 and ABCC2 ABCC1 and ABCC2 have overlapping substrate speci ficities, such as many anticancer agents (e.g., vincristine, doxorubicin), HIV protease inhibitors (e.g., ritonavir, saqui navir), and antibiotics (e.g., difloxacin, grepafloxacin). Both ABCC1 and ABCC2 need cotransport of reduced glutathione (GSH) to transport some of their substrates [12,21,111,112]. ABCC1 is ubiquitously expressed, whereas ABCC2 is mostly expressed in hepatocytes, renal proximal tubule cells, the intestine, and the brain. Genetic variants in ABCC1 are relatively rare, and around 16 SNPs in both of its non coding and coding regions have been determined to cause amino acid changes [21,112]. However, none of them sig nificantly influence the expression level, indicating that the amino acid exchanges do not substantially affect the syn thesis or stability of the protein. Gly671Val (rs45511401) in ABCC1 and a haplotype of ABCC2 were found to cause anthracyclineinduced cardiotoxicity, as a consequence of the higher concentration of the drug, among nonHodgkin lymphoma patients treated with doxorubicin [21]. In general, the influence of ABCC2medicated transport on the pharmacokinetics of drug substrates in humans is not fully understood. Tumor cells often show an inducible expression of ABCC2, which contributes to drug resistance. Because of the transport of bile acids and glutathione from the hepatocytes into the bile duct, ABCC2 plays a physio logically important role in forming bile flow and potentially in detoxification by delivering glutathione for conjugation of xenobiotics. The polymorphism −24C >T in the 5′UTR is in strong linkage disequilibrium with 3972C >T. Renal allograft transplant recipients harboring the −24T allele show a decreased oral clearance for the immunosuppressant mycophenolic acid (MPA), which is the active metabolite of mycophenolate mofetil. In consequence, these patients are more protected from a decrease in MPA exposure but with a higher association to mild liver dysfunction [113]. Highdose methotrexate treatment in pediatric ALL induced a twofold higher AUC and a ninefold higher risk of inten sification of folinate rescue in female patients carrying the −24 variant allele [114]. In nonsmallcell lung cancer patients, ABCC2−24TT and 3972TT genotypes were asso ciated with higher response rates and longer progression free survival, which was sustained in haplotype analysis. This suggests a more effective exposure to irinotecan [115]. A missense SNP 1249G >A (Val417Ile) is located in the substratebinding region of the first transmembrane domain and is associated with lower oral bioavailability and increased residual clearance after intravenous administra tion of the betablocker talinolol, which indicates higher intestinal transporter activity [116]. It has also been found to be correlated with renal proximal tubulopathy after treat ment with the HIV protease inhibitor tenovir disoproxil fumarate, probably because of a toxic concentration of the drug in renal tubular cells after being actively secreted into them from the blood by ABCC2 [117]. The silent polymor phism 1446C >G is associated with higher ABCC2 mRNA expression in the liver and with a decreased AUC and peak concentration (Cmax) of the cholesterollowering drug pravastatin [118]. ABCG2 ABCG2—also known as breast cancer resistance protein (BCRP) or mitoxantroneresistant protein (MXR) or pla centaspecific ATP binding cassette transporter (ABCP)—is an ABC halftransporter with six transmembrane domains and only one ATPbinding domain. Overall, it is expressed at various sites, including the liver, small intestine, colon, placenta, lung, kidney, adrenal and sweat glands, and cen tral nervous system vasculature. More than 80 polymorphisms in the ABCG2 gene have been identified in different ethnic populations, with a higher frequency among Asians and Caucasians. A number of ABCG2 SNPs have been found to influence the function and/or expression of the encoded protein. For example, ABCG2 variant alleles alter drug exposure by reducing the biliary excretion of diflomotecan and rosuvastatin, causing variations in drug effects [4,119]. The nonsynonymous variant, 421C >A SNP (rs2231142), affects the PK of many drugs, including irinotecan, rosuvas tatin, sulfasalazine, and topotecan. It has been reported that the 421A variant exhibits 1.3fold decrease in ATPase activ ity compared to the wild type, with elevated bioavailability of diflomotecan and topotecan. The AUC of rosuvastatin following a single oral administration has been shown to be greater in 421C >A heterozygotes and homozygotes, but the heterozygotes did not show any difference in the PK profile of irinotecan and its active metabolite, SN38 [119]. The AUC and peak concentration of rosuvastatin increased 2.4 fold in healthy individuals with the homozygous 421AA genotype compared to that in individuals with the 421CC
genotype, but did not affect the elimination of t1/2, which signifies that 421C >A affects the intestinal absorption of rosuvastatin [120]. 17.3.3.2 SLC Transporters The SLC family comprises 360 members, subdivided into 47 subfamilies, that encode membrane proteins identi fied as passive transporters, ioncoupled transporters, and exchangers. The betterknown SLC transporters are the organic anion–transporting polypeptides (OATPs) and the organic cation transporters (OCT). Organic Anion–Transporting Polypeptides Organic anion–transporting polypeptides (OATPs) have 12 transmembrane domains, with a large, highly conserved extracellular loop between the 9th and 10th transmembrane domains. Nglycosylation sites in extracellular loops 2 and 5 are consistent among the various members of the OATP family [121]. OATPs mediate the sodiumindependent transport of a wide range of amphipathic organic com pounds, including steroid conjugates, anionic oligopepties, thyroid hormones, bile salts, xenobiotics, and pharmaceuti cals. A total of 11 OATPs have been identified and classi fied into 6 families, and in humans OATP1A2, OATP1B1, OATP1B3 and OATP2B1 are the best characterized for their roles in drug PK [4,39,122]. OATP1A2s encoded by the SLCO1A2 gene are mainly expressed on the luminal membrane of small intestinal enterocytes and at the BBB, and play a role in the intes tinal absorption and brain penetration of their substrates. OATP1A2 substrates include rosuvastatin, methotrexate. and Dpenicillamine. A number of nonsynonymous poly morphisms have been identified in the SLCO1A2 gene, some of which exhibit diminished in vitro transport activity toward OATP1A2 substrates [21,122,123]. The SLCO1B1 gene encodes OATP1B1, also known as OATPC, and is mainly expressed on the basolateral mem brane of hepatocytes in the human liver. It plays a role in the hepatic uptake of substrate drugs, including statins, Dpenicillamine, and rifampin [21,39]. The primary role of OATP1B1 is believed to be removal of substrates from the blood into the liver for subsequent elimination. More than 40 nonsynonymous variants have been identified in the SLCO1B1 gene, some of them being associated with decreased transport activity and whose allele frequen cies vary markedly across different ethnic populations. Functionally impaired OATP1B1 SNPs may decrease the uptake of substrate drugs into the hepatocytes, resulting in decreased biliary excretion or hepatic metabolism, which is a consequence of increased systemic exposure [21,120,124]. OATP1B1 is one of the mechanisms underlying both drugdrug interactions, due to competition at the transporter, and pharmacokinetic variation, due to genetic polymorphisms in the gene encoding the OATP1B1 protein SLCO1B1. The most commonly studied OATP1B1 variants are the 521T >C SNP (rs4149056) and the 388A >G SNP (rs2306283), which are in linkage disequilibrium and exist together in several haplotypes. The variant allele 521T >C is more common among Caucasians (8–20%) and Asians (9–16%) than Africans [39,107]. It has consistently been found to cause a functional decrease in OATP1B1 activity based on altered in vitro transport of a number of substrates, including estrone3sulphate, estradiol17βDglucuronide, pravastatin, atorvastatin, cerivastatin, rifampicin, and SN38. Generally, there is a variantdependent change in pharmaco kinetics such that individuals homozygous for the 521T >C variant (CC) have the highest plasma concentrations, which is in line with in vitro data suggesting that this variant leads to a decrease in the function of OATP1B1 [111,125]. A genomewide association study (GWAS) of 85 patients with myopathy and 90 matched controls from Study of the Effectiveness of Additional Reductions in Cholesterol and Homocysteine (SEARCH) identified an association with rs4149056 in the SLCO1B1 gene [126]. Following the accumulation of evidence on the SLCO1B1 521T >C poly morphism, the FDA announced an update to its simvastatin productlabel recommendations in 2011. SLCO2B1 encodes OATP2B1, which is also known as OATPB, and shows substrate selectivity similar to that of OATP1B1. OATP2B1 is expressed in the lumi nal membrane of smallintestinal enterocytes, and have a role in drug absorption/disposition [39,122]. A num ber of OATP2B1 SNPs have been identified, including 1457C >T (rs2306168), 601G >A (rs35199625), 935G >A (rs12422149), and 43C >T (rs56837383). Patients who received montelukast, a leukotriene receptor antagonist, and who carry the 935A variant allele of the 935G >A SNP were reported to show lower plasma concentration of the drug as well as lower pharmacological response [39,122,123]. OATP1B3 (encoded by SLCO1B3) and OATP2B1 are mainly expressed on the sinusoidal membrane of hepato cytes, and mediate the hepatic uptake of their substrate drugs. OATP1B3 substrates include, digoxin, rifampin, methotrex ate, D penicillamine, and cyclosporine. According to cur rently published findings, OATP1B3 appears to be unique in transporting digoxin and possibly also the taxanes docetaxel and paclitaxel [127]. Polymorphisms in SLCO1B3 include 334T >G (rs4149117) and 699G >A (rs7311358), both of which occur at a high frequency in Caucasian popula tions. While OAT1B3 mediates the hepatic uptake of many drugs, the role of its polymorphisms in drug disposition and response is not fully understood [111]. Organic Cation Transporters Organic cation transporters (OCTs) belong to the solute carrier SLC22A family, which facilitates the intracellu lar uptake of a broad range of structurally diverse, small organic cations. Three OCT isoforms have been identified
in humans: OCT1, which is encoded by SLC22A1; OCT2, which is encoded by SLC22A2; and OCT3, which is encoded by SLC22A3. OCT1 is primarily expressed in the basolat eral membrane of hepatocytes, whereas OCT2 is expressed in the basolateral membrane of kidney proximal tubules. OCT1 and OCT2 substrates include metformin, cisplatin, imatinib, procainamide, citalopram, cimetidine, and quini dine. OCT3 is expressed in various tissues, including the placenta, heart, liver, and skeletal muscle. Additionally, OCT expression has been detected in several cancer cell lines and tumor tissue samples [21,39,107]. The 1393G >A polymorphism in SLC22A1 was found to reduce the localization of OCT1 to the surface of the baso lateral membrane of hepatocytes. Other variants, including 41C >T, 566C >T, 1201G >A, and 1256delATG (a deletion variant) have all been associated with decreased metformin uptake activity; increased AUC and Cmax; and the signifi cantly decreased ability of metformin to lower glucose lev els [111,128]. While several SNPs have been identified for the SLC22A2 gene, the most important is the 808G >T SNP (rs316019). Since metformin’s active renal secretion is through OCT2, any functional defect in OCT2 trans port results in decreasing the drug’s renal clearance. Homozygotes of low activity (808G >T SNP) have lower renal clearance and higher plasma concentrations of metfor min compared to homozygous carriers of wildtype 270A [39,129]. Additionally, homozygous and heterozygous car riers of various haplotypes of lowactivity alleles of some OCT1 variants—R61C (rs12208357), G401S (rs34130495), M420del (rs35191146), and G465R (rs34059508) among them—have shown significantly higher metformin renal clearance (20–30%). The clinical use of cisplatin is related to doselimit ing nephrotoxicity, which occurs in onethird of patients regardless of intensive prophylactic measures. OCT2 has been involved in the cellular uptake of cisplatin, but its role in cisplatininduced nephrotoxicity is not fully under stood. Moreover, the nonsynonymous SNP 808G >T has been associated with reduced cisplatininduced nephrotox icity in patients. These results indicate the critical impor tance of OCT2 in PK and related cisplatin renal toxicity [39,129,130]. Previous studies showed that homozygous carriers of the lowactivity OCT2 variant 270S have a significantly lower renal clearance and a higher plasma concentration of metformin than do homozygous carriers of the active vari ant 270A [21,129]. Additionally, homozygous and hetero zygous carriers of various haplotypes of lowactivity alleles of some OCT1 variants—including R61C (rs12208357), G401S (rs34130495), M420del (rs35191146), and G465R (rs34059508)—have shown significantly higher renal clear ance of metformin PK (20–30%). In addition, lowfunction OCT1 variants (R61C, G401S, M420del, and G465R) have been associated with a significantly decreased glucoselow ering metformin response in healthy volunteers, probably by reducing metformin metabolism in hepatocytes—the major target sites of metformin’s action. For OCT3, sev eral synonymous SNPs have been identified in SLC22A3; however, their functional consequence has not been fully clarified and needs further study [21,39]. Organic Anion Transporters Organic anion transporters (OATs) belong to the SLC22A family of solutecarriers, which facilitate the cellular uptake of a wide range of structurally diverse small hydrophilic organic anions. Several clinically important anionic drugs are OAT substrates, such as βlactam antibiotics, diuret ics, and nonsteroidal antiinflammatories (NSAIDs). There are at least six OAT members (OAT1–6) [21], and four SLC22A genes that have been identified for encod ing OATS: SLC22A6, which encodes OAT1; SLC22A7, which encodes OAT2; SLC22A8, which encodes OAT3; and SLC22A11, which encodes OAT4. In terms of tissue location, OAT1, OAT2, and OAT3 are expressed (mostly) in the basolateral membrane of the renal proximal tubules, mediating the uptake of drug substrates from the blood into the proximal tubule cells; OAT4, in contrast, is located at the apical side of the renal proximal tubule, functioning in the secretion of drug substrates into urine [21,107]. OAT1, OAT2, and OAT3 are responsible for the uptake of drugs into the tubular cells, and OAT4 mediates their secretion into the renal tubule [19,39]. Additionally, OAT1 is an example of transporterrelated toxicity. Different studies have shown that OAT1substrates, such as cephaloridine and ßlactam antibiotics, sometimes cause nephrotoxicity. OCT1 trans ports a number of drugs, such as metformin, which act partly through intracellular effects in hepatocytes [19]. Genetic variations in genes encoding OATs can contrib ute to interindividual variability in the renal clearance of many drug substrates. To date, a number of polymorphisms have been identified in the coding region and 5′ regulatory region of the SLC22A6, SLC22A7, SLC22A8, and SLC22A11 gene variants, and some of them have caused changes in the transport activity of their encoded proteins. While differ ent polymorphisms have been reported for genes encoding OAT—mainly SLC22A8 (encoding OAT3)—their allele frequency is ≤1%, and their functional significance has not been fully explained [21,39]. 17.4 CONCLUSIONS This chapter presented a broad outline of genetic varia tion affecting drug pharmacokinetics and consequently drug efficacy and toxicity that highlights the potential of pharmacogenomics to facilitate improved and more effec tive therapy. Important genetic associations that have been identified between variant genotypes and drug response phenotypes have prompted the FDA to revise labels for certain drugs to include relevant pharmacogenetic infor mation and recommendations. For successful clinical implementation, there must be parallel development in genomic discovery and collaborative partnerships among stakeholders.