Age-Induced Differences

Drug Disposition in the Geriatric Animal

This discussion focuses on some of the clinically important changes that are likely to alter response to drug therapy in geriatric patients (Table 2-1) and on actions that can be implemented to compensate or reduce the sequelae of these changes. Age is associated with changes both in pharmacokinetics (gerontokinetics) and pharmacodynamics, with changes generally compared with those of the young adult.^1-3 Data in animals are limited, but in general, disposition pattern changes described for humans appear to extrapolate adequately to dogs and cats.⁴ Geriatric animals are at a greater risk for therapeutic failure because of normal aging changes in physiology, an increased incidence of disease, and the likelihood of polypharmacy in response to disease; moreover, a decrease in normal organ protective mechanisms may increase the risk of adverse events. The age at which body functions shift from a period of growth to a period of decay (16 to 18 years in humans) has not been established in dogs and cats. The age probably differs among canine breeds. Aging is accompanied by permanent loss of up to 30% of body cells, with a parallel loss in oxygen consumption. Body composition changes, as do regional blood flow rates. Physiologic functions generally decline steadily with increasing age. In humans basal metabolic rate decreases by 0.4% per year²; according to the National Research Council,⁵ energy requirements of dogs decrease as they age. Changes among the body systems can influence all four drug movements.

Table 2-1 The Impact of Age-Related Differences in Physiology on Xenobiotic Disposition

KEY POINT 2-1

As the geriatric animal ages, organ mass reduces in size and function by approximately 25%, resulting in corresponding changes in drug disposition.

Cardiovascular

As animals age, cardiac output decreases and circulation transit time increases. In humans cardiac output decreases by about 1% per year for a total decline of 30% to 40% in the aged. Regional and organ blood flow rates similarly decrease.² The net effect of these changes depends on the state of disease but can influence each drug movement. Absorption, metabolism, and excretion are likely to decrease, whereas distribution may increase or decrease depending on the state of vascular responses or fluid retention.^2,6,7 As cardiac function decreases, secondary compensatory responses can lead to further risks of adverse reaction.⁷ Blood flow is preferably redistributed to the brain and heart, increasing the risk of toxicity of drugs toxic to these tissues.

Central and peripheral nervous systems

As the geriatric patient ages, brain weight and peripheral fiber numbers decrease. Connective tissue infiltrates peripherally.² Oxygen consumption and cerebral blood flow decrease. In addition, decreased amounts of selected neurotransmitters have been documented. Morita and coworkers⁸ reported alterations in the blood–brain barrier in elderly dogs.

Respiratory

In human geriatric patients, residual lung volume decreases by 50% with accompanying decreases in vital capacity, arterial oxygen pressure (PO₂), and maximum oxygen uptake. In addition, the central response to hypoxia and hypercapnia, such as that induced by opioid analgesics, decreases.² Anesthetic or other sedating agents must be used more cautiously.

Gastrointestinal

As animals age, deglutition decreases as a result of decreased salivation and pharyngeal and esophageal motility. Gastric function is characterized by atrophy of the mucosa with a reduction of hydrochloric acid secretion and a subsequent increase in gastric pH. Gastrointestinal motility is generally reduced. The intestinal macrovilli and microvilli also atrophy, increasing the risk of bacterial overgrowth. These sequelae tend to reduce the absorption and thus the PDC of orally administered drugs. Changes in gastrointestinal function (including reduction of gastroprotective effects) may also predispose the geriatric patient to adverse effects induced by toxic drugs such as chemotherapeutic agents and nonsteroidal antiinflammatory drug (NSAID) analgesics. The latter should be used with caution; the clinician should anticipate and be prepared to treat toxicities. Digestibility of key nutrients, such as fat, tends to decrease in elderly cats.⁹ As with humans, the intestinal microbiota population shifts in elderly dogs and cats, with lactobacilli decreasing and clostridia increasing in dogs¹⁰ and bifidobacteria decreasing in cats.¹¹ In addition to response to antimicrobials, these changes might influence enterohepatic circulation of bile acids and thus, potentially, some drugs.¹²

Hepatic

Changes in hepatic function are important to the geriatric animal because of the liver’s role in the metabolism of drugs.¹³ Hepatocyte number and function decrease, as do hepatic and splanchnic blood flow, hepatic oxidation, and cytochrome P450 content (the primary drug-metabolizing enzyme). Both flow-limited and capacity-limited drugs are affected. For example, hepatic clearance of both opioid analgesics (which are characterized by first-pass metabolism; i.e., flow limited) and nonsteroidal analgesics (eliminated principally by hepatic metabolism; i.e., capacity limited) is decreased in geriatric patients. Increased response of human geriatric patients to opioid analgesics—they require 60% to 75% less drug than younger patients do—has been attributed to changes in drug elimination.^14,15 Changes in hepatic function, oxygenation, and nutrition may also predispose the liver to drug-induced hepatotoxicity. Because of reduced hepatic function, the geriatric patient may be less able to generate endogenous hepatoprotectant agents, increasing the risk of drug-induced hepatotoxicity (Chapter 4).

Urinary

As renal blood flow decreases, the glomerular filtration rate and active secretory capacity of the nephron unit progressively decrease with age. Both result in a similar decline in renal clearance. Renal excretion is the major route of elimination of many drugs. Changes in renal clearance tend to prolong the elimination and thus increase PDCs in the geriatric patient. Changes in renal function also render the geriatric patient more susceptible to adverse drug reactions such as those induced by aminoglycosides, angiotensin-converting enzyme inhibitors, and NSAID analgesics.

Body weight and composition

Changes in body composition may be among the most complex in the geriatric animal. Like humans, dogs tend to lose lean body mass and accumulate body fat as they age (Figure 2-1).¹⁶ In male humans, fat increases from approximately 18% in young adults up to 50% in the aged.² Increased proportion of body fat is accompanied by a decrease in total body water and cell mass. Although extracellular fluid does not change in total amount, the relative proportion of total body water that it makes up increases. Thus the proportion of intracellular to extracellular fluid decreases. The sequelae of these changes depend on the drug. As distribution of water-soluble drugs decreases with total body water, PDCs tend to increase. The distribution of lipid-soluble drugs increases as the proportion of body fat increases, however, which tends to decrease PDCs unless the patient is dosed on a mg/kg basis. The impact of aging on body composition in dogs varies with breeds. For example, in one study body fat increased with age in Great Danes but not Labrador Retrievers or Papillons,¹⁷ although this did occur within 1 year of death in elderly Labrador Retrievers in another study.¹⁶ Although middle-aged cats tend to be overweight, geriatric cats often become thin.⁹ Thus PDCs of water-soluble compounds might be expected to be higher in older animals compared with those of young adults, even if dosing on an mg/kg body weight basis, whereas for lipid soluble drugs, dosing on an mg/kg body weight basis might compensate for potential changes in plasma concentrations.

Figure 2-1 The geriatric dog and cats may represent two extremes in regard to the impact of aging on disposition. Body mass frequently increases in dogs (A) but decreases in geriatric cats (B). Mass loss in cat reflects lean tissue and fat. In contrast to the geriatric animal, adult cats, like dogs, may be characterized by marked differences in body composition, as is exemplified by the obese cat or dog. Water-soluble drugs generally do not distribute to fat, and dosing of water-soluble drug on the basis of total body weight may result in overdose. Dosing on lean body weight is more prudent. Distribution of a lipid-soluble drug to the fat compartment, on the other hand, may increase the volume of distribution, decreasing drug concentrations. For such animals dosing on total body weight (per kg) is more reasonable.

Serum albumin

Although total serum plasma protein content probably remains the same in the geriatric animal, the proportion represented by albumin decreases and that by gamma globulins increases. Changes in serum albumin can be clinically important to patients receiving highly protein-bound drugs, such as NSAIDs. Decreased albumin can result in a greater proportion of free drug: Most NSAIDs are close to 99% protein bound. A decrease of only 1% (i.e., 99% to 98% binding) doubles the concentration of a pharmacologically active drug. The sequelae of increased PDC may be offset by a compensatory increased clearance, because only unbound drugs are generally conducive to hepatic or renal clearance. However, this balance might be minimized if organs of clearance are negatively affected.

Receptor sensitivity and pharmacodynamics

Geriatric patients respond differently to some drugs, which suggests that tissue receptor sensitivity to the drugs is altered. Changes in receptor number or responsiveness have been implicated but not documented.^2,3 Physiologic changes such as altered neurotransmission or intracellular constituents have also been suggested. For example, geriatric patients are less likely to perceive, appreciate, or express pain. Thus the need for analgesic therapy is often not detected. In addition, geriatric patients are less able to respond to many analgesic drugs.

Disease

Aged animals are more likely to be suffering from diseases that affect not only drug disposition but also tissue receptivity to drugs and organ protection.¹⁵ The immune system of the geriatric patient is not as effective as that of the adult,¹⁸ leading to the use of bactericidal antimicrobials and minimizing the use of immunosuppressive drugs. In addition, the geriatric patient is more likely to be receiving multiple drugs, which increases the likelihood of drug interactions. Finally, diseases of selected organs may predispose these organs to drug-induced toxicity.

Drug Disposition in the Pediatric Animal

With regard to dogs and cats, pediatric generally refers to the first 12 weeks of life.¹⁹ Important developmental changes occurring within this time spectrum, however, justify further staging into neonatal (0 to 2 weeks), infant (>2 to 6 weeks), and pediatric (>6 to 12 weeks) periods of growth. Changes associated with each of these periods cause accompanying changes in drug disposition, thus rendering the pediatric patient more susceptible to drug-induced adverse reactions. All four determinants of drug disposition (i.e., absorption, distribution, metabolism, and excretion) undergo dramatic changes as the neonate matures (see Table 2-1).^20,21 However, the clinical significance of these sequelae varies.

KEY POINT 2-2

Drug clearance generally does not reach adult capacity until approximately 3 months of age.

Absorption

Young kittens have decreased energy, carbohydrate, and organic matter digestibility compared with kittens older than 19 weeks of age.²² However, because the surface area of the small intestine is large, even in neonates, the extent of drug absorption probably does not clinically differ between normal pediatric and adult animals. The rate of absorption tends to be slower in pediatric animals, however, probably because of decreased gastric emptying and irregular intestinal peristalsis. As a result, peak PDCs may be lower in pediatric patients. The decreased rate of absorption might actually protect against toxic drug concentrations.^23,24 However, protection may not be present before absorption of colostrum. During this period the permeability of the intestinal mucosa is increased, leading to increased rate and extent of drug absorption. Occasionally, drugs that normally are not absorbed from the gastrointestinal tract (e.g., aminoglycosides, carbenicillin, and other acid-sensitive beta-lactams and enteric sulfonamides) can reach systemic circulation. Intestinal permeability decreases rapidly after the ingestion of colostrum,^24,25 possibly because of the endogenous release of hydrocortisone or adrenocorticotropic hormone. Exogenous supplementation of either of these hormones by the 24-hour prepartum mother prevents increased permeability and colostrum absorption in the neonate.

A number of other factors may alter small intestinal drug absorption in pediatric patients. Gastric pH is neutral in the newborn; adult levels are not reached until sometime after birth, depending on the species.^23,25 Increased gastric pH (achlorhydria) may decrease the absorption of many drugs that require disintegration and dissolution or are ionized in a less acidic environment (e.g., weak acids such as penicillins). Milk diets can reduce drug absorption by either decreasing gastric emptying or directly interacting with drugs (e.g., milk can impair absorption of tetracyclines). The “unstirred water layer” adjacent to the surface area of the mucosal cells is thicker in the neonate than in the older pediatric patient and may limit the rate of absorption of some drugs. As biliary function matures, the absorption of fat-soluble drugs (e.g., griseofulvin and fat-soluble vitamins) increases. Microbial colonization of the gastrointestinal tract may alter response to antimicrobial drugs, extrahepatic metabolism, or enterohepatic circulation.^26,27

Absorption from the rectal mucosa is rapid. Rectal administration of drugs or fluids can be used for pediatric patients when venous catheterization is difficult, to reduce complications associated with intravenous administration (e.g., sedation, anesthesia), or when oral administration is undesirable (e.g., antiemetics). Several pediatric drugs intended for systemic effects are available as rectal suppositories. Limited data from studies of human infants indicate that peak plasma concentrations after rectal administration may be higher than those obtained by other routes.²⁶

Absorption of drugs administered parenterally to pediatric animals also varies from that of adults. The rate of absorption after intramuscular administration changes with age as muscle mass and its accompanying blood flow increase and as vasomotor responses mature.²⁶ Because muscle mass is small, subcutaneous administration is frequently preferred for pediatric patients. Again, variability in subcutaneous absorption rates can be anticipated with age. Less fat but greater water may result in faster absorption compared with that in adults.²⁸ Environmental temperature probably influences subcutaneous absorption, particularly in newborns whose thermoregulatory mechanism functions poorly. Cold environments are likely to reduce subcutaneous drug absorption if the neonate is not kept warm. The same is true for patients in a state of hypothermia. Intraperitoneal administration can be a lifesaving route of blood and fluid administration, particularly for the newborn with inaccessible central veins. Isotonic fluids are rapidly absorbed, and up to 70% of red blood cells are absorbed in 48 to 72 hours.²⁹ Blood and fluids can also be administered into the medullary cavity of large bones.^30,31

Absorption of volatile anesthetics from the pediatric respiratory tract is rapid because minute ventilation is greater.¹⁹ Thus young animals are more sensitive to the effects of gas anesthetics. Although not a common route of drug administration, percutaneous absorption of drugs is likely to be greater in pediatric patients. Percutaneous absorption is directly related to skin hydration, which is greatest in neonates. Topical administration of potentially toxic lipid-soluble drugs (e.g., hexachlorophene and organophosphates) is not recommended.

Distribution

The most important factors contributing to differences in drug distribution in pediatric patients are differences in body fluid compartments and drug binding to serum proteins. Body fluid compartments undergo profound changes with the growth of the neonate. Both the percentage of total body water and the ratio of compartmental volumes change with maturation. The percentage of total body water decreases with age, but the decrease is more substantial in the extracellular versus the intracellular compartment (see Table 2-1; Figure 2-2).³² Daily fluid requirements are greater in neonatal and pediatric patients, in part because a larger proportion of their body weight is represented by body water (see Figure 2-2). The sequelae of these body compartment differences depend on the normal distribution of the drug. Most water-soluble drugs are distributed to extracellular fluids. In pediatric patients the volume to which these drugs is distributed is therefore higher than in adults; PDCs correspondingly decrease. Thus it may be necessary to increase doses to prevent therapeutic failure. A different pattern might be expected for lipid-soluble drugs because they tend to be distributed to total body water. Such drugs should be dosed according to body weight (e.g., mg/kg). Although decreased PDCs resulting from increased distribution may protect the pediatric patient from potentially toxic drug concentrations,³³ a poor therapeutic response may result from failure to generate therapeutic drug concentrations of water-soluble drugs. The disposition of several antimicrobials in neonatal animals has been reviewed by Baggott,³⁴ with differences in disposition generally found to be similar to those described for humans. Ampicillin is characterized by volumes that are larger in puppies and kittens (threefold to fourfold higher) compared with adults, resulting in clinically significant lower drug concentrations.^35,36 Distribution of enrofloxacin, a lipid-soluble drug, is somewhat unpredictable and age dependent in kittens, being smaller at 2 to 4 weeks but greater at 4 to 6 weeks.³⁷ Changes in the half-life of each drug parallel changes in distribution. Because many drugs are distributed to a larger volume in pediatric patients, a longer half-life should be anticipated, and it may be necessary to prolong the dosing interval. However, for enrofloxacin in kittens this effect was balanced by increased clearance at 6 weeks.³⁷

Figure 2-2 Differences in body composition can be appreciated between the two age extremes by following changes as the animal ages. The extracellular compartment and total body water of dogs and cats is greatest in the neonate and gradually decreases toward adult proportions as the animal ages. Increased extracellular fluid and increased total body water result in larger volumes of distribution of most drugs in neonates or pediatric animals, which may decrease plasma drug concentrations and prolong half-life. As the adult ages, fluid content declines, especially in the geriatric cat. The impact of these changes might be greater on water-soluble drugs because the extracellular component represents a greater proportion of total body water in the young animal.

Because the proportion of body fat is smaller in pediatric patients, the distribution of lipid-soluble drugs that accumulate in fat (e.g., organophosphates, chlorinated hydrocarbons, ultrashort thiobarbiturates) may be proportionately decreased. Although drug half-life would decrease, PDCs may become toxic. Many lipid-soluble drugs have a high affinity for and are bound by plasma proteins, thus facilitating their movement through the body. Binding, however, limits their distribution to tissues. Predicting the distribution of highly protein-bound drugs is complicated in the pediatric patient. Serum concentrations of both serum albumin, the protein to which most drugs are bound, and α₁-glycoproteins (to which basic drugs preferentially bind) are decreased in pediatric patients.³⁸ Protein binding of drugs may also be reduced because of differences in albumin structure or because drugs compete with endogenous substrates (e.g., bilirubin) for binding sites.^24,39 As drugs are displaced, the concentration of free, pharmacologically active drugs and the risk of adverse reactions increases. These changes are significant, however, only if the drug is highly (i.e., >80%) protein bound and characterized by a small therapeutic index. Although the concentration of free drug increases, that of total drug in the plasma tends to decrease because unbound drug is free to distribute into tissue.³⁹ Consequently, drug half-life may increase, and longer dosing intervals may be indicated for potentially toxic drugs. Increased clearance of unbound drug may ultimately normalize a half-life that has been lengthened by an increased volume of distribution.

Differences in regional organ blood flow might cause clinically important changes in drug disposition in pediatric animals. Differences in renal blood flow have been documented^40,41 and result in clinically important differences in drug excretion. Blood flow to vessel-rich tissues of the body (i.e., heart and brain) is greater and faster;¹⁹ the pediatric patient is thus more susceptible to drug-induced cardiac and central nervous system (CNS) toxicity. The potential for CNS toxicity is further increased because the blood–brain barrier is poorly developed immediately after birth. Increased permeability protects the neonatal brain from a deficiency of nutritional fuels in stressful states (e.g., hypoglycemia, hypoxia, acidosis) by allowing the movement of oxidizable substrates such as lactate into brain cells.⁴² The status of efflux proteins is not yet established. Drugs normally incapable of reaching the adult brain are, however, also able to reach brain cells, which are very susceptible to their effects, thus increasing the risk of CNS toxicity.^43,44

Metabolism

Drug elimination, including both hepatic metabolism and renal excretion, is limited in neonatal and pediatric patients. Thus many drugs administered to the young animal are characterized by decreased clearance.^20,24 In contrast to human infants, hepatic metabolism of drugs is incompetent in the near-term and neonatal puppy.^45-47 Both phase I (e.g., oxidative) and phase II (e.g., glucuronidation) reactions are reduced. The various pathways of metabolism mature at different rates. Phase I activity may not occur in the neonatal puppy and may not be evident until day 9 if it does. Activity appears to progressively increase after day 25, not reaching adult levels until 135 days postpartum.⁴⁶ Drug metabolism in pediatric patients, however, is quite complex. For example, Ecobichon and coworkers⁴⁸ found that in Beagle puppies, phase I metabolites decreased with age, and for phase II metabolism, glucuronidation was the most predominant, with sulfonation decreasing as puppies got older.

Generally, decreased hepatic drug metabolism is reflected as decreased plasma clearance, increased plasma half-life, and potentially toxic PDCs. Dose reduction, dose prolongation of intervals, or both may be indicated for some drugs. Oral bioavailability of drugs characterized by significant first-pass metabolism in adults (e.g., propranolol) is probably greater in puppies and kittens. Response to prodrugs (e.g., primidone; prednisone; enalapril; and, potentially, methylprednisolone) may be reduced because of decreased formation of active drug products. Pediatric hepatic drug-metabolizing enzymes do appear to be inducible by phenobarbital and other drugs. Nonhepatic drug-metabolizing enzymes also appear to be decreased in pediatric patients. For example, pediatric lower plasma cholinesterase can result in increased sensitivity to organophosphates, succinylcholine, and procaine.

Excretion

Reduced renal excretion, characteristic of the pediatric puppy, results in decreased clearance of renally excreted parent drugs and products of phase II drug metabolism. Although the number of glomeruli remains constant throughout pediatric development, both glomerular filtration and renal tubular function progressively increase.^41,49 Adult values may not be reached until approximately 2.5 months of age. In contrast to glomerular filtration and secretion, renal tubular reabsorption in puppies appears to be similar to that in adults as long as body fluids and electrolytes are maintained.^50,51 The sequelae of developmental changes in pediatric renal function include decreased clearance and prolonged half-life of drugs (primarily water soluble) excreted by the kidneys. Such a pattern has been shown for several drugs. Compared with current recommendations for adults, pediatric patients may require a higher dose (owing to increased volume of distribution) and longer intervals (owing to increased distribution and decreased clearance) for gentamicin administration. More important, modifications should be anticipated in the gentamicin dosing regimen of unhealthy puppies because they are likely to be affected by conditions that increase the potential for gentamicin-induced nephrotoxicity (e.g., dehydration). However, underdeveloped glomeruli may actually protect the pediatric patient from aminoglycoside-induced nephrotoxicity.²¹ Further investigations are needed to establish safe yet effective doses of gentamicin for the neonatal puppy or kitten.

Pregnancy and Lacation

Roles of Species and Breed Differences in Drug Disposition

Differences among species in the kinetics and response to drugs are profound; Riviere⁷² has offered a review in animals. Few studies have compared disposition among species, and when reported, data are generally oriented toward human medicine, with the intent to identify the (laboratory) animal most predictive of human drug disposition. Presumably, species that are physiologically similar tend to have similar drug disposition patterns, and the same dosing regimen can often be used for a particular drug.⁷³ However, pharmacokinetic data (Vd, clearance, and mean residence time) after intravenous administration have been compared among humans, monkeys, dogs, and rats for over 100 molecules (including many drugs).⁷⁴ Correlations between chemical structures of the compounds with pharmacokinetics allowed two categories of compounds to be identified: those characterized by high clearance, for which mathematical models would allow general extrapolation of data from one species to another, and those whose characteristics yielded incorrect extrapolations. Although the authors concluded that such an approach might be useful for extrapolation, more to the point is the reality that drugs may not be extrapolatable even with the most sophisticated of mathematical models. The combination of computation analysis of the molecular characteristics of each compound with in vivo pharmacokinetic data might ultimately be helpful in identifying the most appropriate species for predicting pharmacokinetic behavior of a particular molecule in another species. Thus caution needs to continue when extrapolating information from humans to the dog, and dog to cat. Likewise, pharmacodynamic responses may be profound.

Distribution

Because most determinants of xenobiotic distribution are largely affected by cardiac output, regional blood flow, and xenobiotic chemistry, distribution differences among species might be predictable through allometric scaling.⁸³ However, they must also take into account differences in transport proteins. Differences in drug distribution can result in important differences in drug response. Blood volume of the cat (70 mL/kg) is less than that of the dog (90 mL/kg); PDCs of drugs whose distribution is confined to the plasma compartment may therefore differ between the species. The same amount of drug (on a per-kilogram basis) is diluted less in cats because the plasma volume is smaller. Thus drug concentrations after administration of a mg/kg dose might initially be higher in cats than in dogs. Organs that are well perfused (i.e., heart, brain) may be more susceptible to toxicity. Cats are approximately the same size as the smaller dog breeds. Thus doses determined for medium-size to large-size dogs may not be appropriate for the cat because the smaller animals have a greater body surface area. In larger animals body water makes up a larger proportion of body weight, which tends to dilute the drug. A higher dose may be needed for larger animals. Because the drug half-life may be longer (owing to increased distribution), however, the dosing interval may need to be longer.

Differences in plasma protein-binding characteristics (particularly albumin) may alter the distribution of drugs that are highly protein bound. The degree to which various drugs are protein bound varies dramatically among the species, although the clinical implication is not clear. Although the elimination characteristics of many drugs have been established for cats, few studies have determined the extent of protein binding.

The interaction between disease and species should be considered when considering species differences in drug distribution. For example, the unhealthy cat does not maintain hydration as well as the dog; fluid imbalances resulting from dehydration or edema alter drug distribution. The obese cat can represent a “sink” for drugs that are lipid soluble, thus lowering PDCs potentially to submaximal levels if the dose is not appropriately increased. Weight loss in a hyperthyroid cat can have the opposite effect.

Sight hounds (e.g., Salukis, Greyhounds; Figure 2-3) offer an example of potential breed differences in drug distribution. Their lean body weight provides little fat tissue for drug distribution. As a result, they are more susceptible to overdosing with drugs that redistribute, such as thiobarbiturates.

Figure 2-3 Differences in breeds may affect body composition (A) as well as other aspects of drug disposition. The sight hounds, represented by the Greyhound (B) is at risk for selected adverse drug events, in part because of a larger lean body mass compared with fat. Drugs that normally accumulate in fat may achieve higher drug concentrations in such breeds.

Distribution is influenced by the multidrug transport protein P-gp. Polymorphism has been demonstrated in the Collie, yielding differences in P-gp content. Two P-gp transporting proteins are encoded by ABCB1 (previously MDR1 or PGY1) and MDR3 (also named MDR2 and PGY3); only the MDR1 gene product is thought to significantly influence drug metabolism. P-gp acts as an efflux pump by translocating drugs from the intracellular to extracellular compartments (Table 2-2) (see Chapter 3). The protein transports a large number of drugs that are chemically divergent; further, these drugs are associated with a specific CYP450 responsible for metabolism of the drugs that have been transported. Polymorphism of the MDR1 gene and P-gp have been reported in humans and are associated with altered drug disposition and thus susceptibility to adverse drug events. Interestingly, polymorphism also has been associated with an increased risk of certain illnesses (e.g., refractory seizures, Parkinson’s disease, inflammatory bowel disease biliary mucoceole in dogs∗). Polymorphism reflecting a mutation deletion of MDR1 that causes nonfunctional P-gp has been documented in Collie and related working-breed dogs. The incidence of the deletion is impressively high: in the U.S., in one study, 35% of Collies were homozygous and another 42% heterozygous for the mutation deletion.⁸⁴ A similarly high incidence was found in dogs in France: 20% of Collies and related breeds were found to be homozygous for the normal allele, 32% heterozygous for the deletion (carrier), and 48% homozygous for the mutant allele (affected dogs).⁸⁵ The impact of the mutation on drug safety in afflicted animals can be profound (see Chapter 3). Substrate specificity for P-pg appears to be similar among species, suggesting that human data can be used to predict which drugs might be more likely to cause adverse effects in these breeds.⁸⁶ However, protein type and amount may vary among species and breed.

Table 2-2 Substrates for P-glycoprotein Transport Pump and Known Inhibitors and Inducers of Drug Transport

Substrates	Inhibitors	Inducers
Antimicrobial Drugs
Erythromycin	Bromocriptine	Clotrimazole
Tetracycline	Carvedilol	Dexamethasone
Itraconazole	Cyclosporine	Morphine
Fluorinated quinolones (selected)	Erythromycin	Rifampin
Anticancer Drugs	Fluoxetine	Phenothiazine
Doxorubicin	Intraconazole	St. John’s Wort
Vinblastine	Ketaconazole
Vincristine	Meperidine
Mitoxantrone	Pentazocine
Anthelmintics	Progesterone
Ivermectin	Quinidine
Cardioactive Drugs	Tacrolimus
Digoxin	Verapamil
Quinidine
Diltiazem
Verapamil
CNS-Active Drugs
Phenothiazines
Amitryptyline	Inhibitor of CYP3a
Morphine	Substrate CYP3a
Endogenous Substrates
Bilirubin
Steroidal Hormones
Cortisol
Aldosterone
Gastrointestinal Drugs
Cimetidine
Loperamide
Ondansetron
Tacrolimus
Immunomodulators
Dexamethasone
Methylprednisolone
Cyclosporine
Tacrolimus
Colchicine

CNS, Central nervous system.

Metabolism

Human polymorphisms in CYP metabolic enzymes have been associated with therapeutic failure resulting from extremely rapid metabolism of a drug and toxic effects caused by decreased metabolism.⁸⁷ Species extrapolation among the CYP450 appears to be predictable in order of most to least: 2E1 (reasonably predictable) greater than 1A1, 1A2, and 4A (cautiously predictable) greater than 2D and 3A greater than 2A, 2B, and 2C (major caution with extrapolation).⁸⁸ Studies attempting to identify similarities between dog and human CYP activity suggest general extrapolation between the species is not prudent.⁸⁷ Although significant differences among species were not detected among rodents, rabbits, and dogs for CYP2E1, 3A2, and 4A1, comparison of animal CYP activity with human found only 2D6 to be most similar to dog.⁸⁹ Other enzymes present in dogs are CYP1A, 2C, and 2D families and subfamilies (Table 2-3).⁹⁰

Table 2-3 Cytochrome P450 Families, Their Contribution to Drug Metabolism in Humans, and Their Orthologs in Other Species

KEY POINT 2-4

The genetic basis for differences in drug transport and metabolism limits extrapolation of dosing regimens among species, and potentially, breeds. However, understanding of these polymorphisms is increasing our ability to predict increased risks associated with drug administration in some breeds.

Species differences have been documented in the handling of racemic isomers of selected drugs. Inversion (from one isomer to the other) patterns differ and are likely to result in different pharmacologic or toxic effects of the drugs. Induction (or presumably inhibition) of drug-metabolizing enzymes also will differentially affect the isomers.⁹¹

As in humans, polymorphism in drug-metabolizing enzymes has been reported in dogs⁹² but is not as well described. Differences in response to anesthesia recognized in sight hounds reflect both differences in drug distribution (to lean versus fat compartments; see Chapter 1) as well as differences in metabolism. Cytochrome-mediated clearance of several anesthetic agents is less in Greyhounds compared with other (nonsight hound) dogs; documented drugs include thiopental, thiamylal, and methohexital. Clearance of propofol by Greyhounds is three times less than that by Beagles. Ketoconazole plasma concentrations were twofold higher than expected in Greyhounds in one study.⁹³ Further, Greyhound disposition of celecoxib, a cyclooxygenase-1 protective NSAID, indicates that breed differences may predispose this breed to adverse drug reactions. Polymorphism also has been described for CYP2C isoenzymes, again in Beagles⁹⁴ and possibly Greyhounds.⁹⁵ Polymorphism in celecoxib metabolism was attributed to CYP2D15, for which three canine variants were found.⁹⁶ In a study of 242 Beagles receiving celecoxib, approximately 50% were considered efficient metabolizers and 50% poor metabolizers, with bioavailability and maximum PDC in the latter group almost twofold higher. The impact of species differences in pharmacokinetic and pharmacodynamic considerations of enantiomers has been described.^91,97 For example, for many NSAIDs the S-isomer has a much greater affinity for cyclooxygenase-2, but the proportion of the S isomer varies among species. Further, species differ in their ability to interconvert S and R isomers.⁹⁷ For example, chiral inversion has been described for ketoprofen in cats.⁹⁸Other polymorphisms that have been described include thiopurine methyltransferase (TPMT), which is one of several enzymes responsible for the metabolism of the active metabolite of the prodrug azathioprine; polymorphisms resulting in deficiencies in humans have been associated with an increase in the toxic bone marrow effects of the drug. Differences in dogs have been demonstrated as well: Kidd⁹⁹ and coworkers have demonstrated that Giant Schnauzers have significantly less and Alaskan Malamutes significantly more TPMT compared with other canine breeds.

The most significant and best-characterized differences in drug disposition between the dog and cat probably result from differences in drug metabolism. Identification of phase I enzymes and their specific drug substrates is difficult, and few species differences have been described in the cat or dog. However, a recent review of CYP450 activity–based substrate metabolism in cats suggests that cats have very low activity of CYP2C, but activity of CYP2D and CYP3A approximates that of dogs or humans, depending on sex.⁷⁷ Deficiencies in demethylation and hydroxylation have been described in the cat and may be responsible for different patterns of prodrug activation (e.g., primidone; see Chapter 27) or adverse reactions to selected drugs (i.e., chloramphenicol).¹⁰⁰ Deficiencies in phase I demethylation and hydroxylation as well as phase II glucuronidation lead to much slower elimination of phenols and aromatic acids and amines in the cat compared with other species (Figure 2-4).^101,102 The described reaction of cats to diazepam may represent differences in the metabolites produced, as may the susceptibility of the feline liver to metabolite-induced damage.¹⁰³ Polymorphisms in drug-metabolizing enzymes have also been described in cats, which have at least three variants of CYP2E,104, although breeds were not cited.¹⁰⁴

Figure 2-4 Examples of drugs (many containing phenols) whose metabolism in cats is slower owing to deficiencies in drug phase I and phase II drug-metabolizing enzymes.

Deficiencies in phase II metabolism have long been recognized in cats. The deficiency reflects extremely low concentrations of some glucuronyl transferases. Thus many drugs excreted as glucuronide conjugates in other species are characterized by a prolonged clearance rate and half-life in the cat. Toxic levels may accumulate much more quickly in the cat, and exaggerated pharmacologic responses or toxicities occur more easily (Figure 2-4). Dosing regimens must be modified for such drugs by either decreasing the dose (especially for drugs whose dosing interval is shorter than the elimination half-life) or prolonging the dosing interval. The prototypic example is aspirin, whose half-life approximates 36 hours in cats compared with 8 hours in dogs. To prevent toxicity in the cat, aspirin is dosed every 48 to 72 hours, compared with twice daily in dogs.

Not all drugs that are conjugated with glucuronide are predisposed to toxicity in the cat. This is true for several reasons. First, the cat is deficient only in certain families of glucuronyl transferase. Cats can conjugate and excrete endogenous substrates such as bilirubin, thyroxine, and steroid hormones as well as other species. Metabolism of a variety of exogenous drugs, however, particularly phenols and aromatic acids and amines, occurs at a much slower rate in the cat than in other species.^101,102 The degrees of deficiency and potential toxicity depend on the drug substrate. For example, some phenolic compounds are sufficiently conjugated, whereas others are not. Second, glucuronide-conjugated drugs characterized by a wide safety margin are associated with few adverse reactions even if accumulation occurs. Finally, in the absence of glucuronide, drugs may be sufficiently metabolized by an alternative pathway. Some sulfates may be particularly well developed in the cat, and many drugs that are excreted as glucuronide-conjugates by the dog may be excreted as sulfated compounds by the cat. Other sulfate-conjugating systems, however, appear to be easily saturated in the cat. Unfortunately, alternate pathways of drug metabolism may also contribute to the toxicity of some drugs because they may involve phase I enzymes that catalyze the formation of toxic metabolites. Thus drugs shunted to another pathway in the cat may be very toxic to the cat but minimally toxic in other species. Deficiencies in both glucuronide and (potentially) glutathione transferase may also predispose the cat to poor scavenging systems in erythrocytes and hepatocytes, limiting the otherwise protective effects that might be realized by these systems and further contributing to toxicity. Acetaminophen is an excellent example of the potential sequelae of phase II deficiences in the cat (See Chapter 4). Because glucuronide is deficient, excessive acetaminophen is shunted to phase I enzymes, which produce toxic oxygen radicals. More metabolites are produced than can be handled, and the glutathione-scavenging system of feline erythrocytes and hepatocytes is overwhelmed, resulting in life-threatening methemoglobinemia and (potentially) hepatic necrosis. The rationale for cimetidine treatment can be understood in the context of the role of phase I metabolism, as well as that of N-acetylcysteine, a glutathione precursor.

Not all deficiences in metabolism occur in the cat. Although acetylation is not a common route of elimination for xenobiotics, it is an enzyme system whose deficiency in the dog is clinically relevant (e.g., procainamide). For example, the antiarrhythmic procainamide is acetylated in humans to an active metabolite. Procainamide is less potent than its acetylated metabolite and the canine dose for procainamide is considerably higher than that in humans on a mg/kg basis in order to achieve an equivalent pharmacologic response (see Chapter 14, Cardiac). A second example might be sulfonamide elimination. Sulfonamides are detoxified by N-acetylation in humans. In the face of deficient acetylation, shunting of the xenobiotic to an alternative pathway in dogs, with the production of the cytotoxic metabolite hydroxylamine, may be one mechanism of sulfonamide toxicity in dogs, although alternative mechanisms are likely to be responsible.¹⁰⁵

Role of Species Differences in Target Tissues

It is difficult to predict differences in drug reaction that can be ascribed to differences in target tissues because very little is known about cats. Differences in response to selected drugs (e.g., opioids, insulin, chlorinated hydrocarbons) are known to be or are thought to be reflections of differences in tissues; however, often these differences turn out to be pharmacokinetic differences reflecting decreased or increased PDCs of active (including toxic) compounds rather than differences at the receptor level.

Feline erythrocytes (hemoglobin) appear to be more susceptible to oxidation and thus to methemoglobinemia. Drugs reported to cause methemoglobinemia in the cat include urinary antiseptics containing methylene blue¹⁰⁶ or azodyes, acetaminophen^101,107,108 and related compounds, benzocaine,¹⁰⁹ and propylthiouracil.¹¹⁰ Several mechanisms have been postulated to explain the potential increased sensitivity of cats to methemoglobin formation. Lower concentrations or activities of the intracellular repair enzyme methemoglobin reductase have been postulated but not confirmed.^111,112 Faster metabolism of specific drugs to toxic metabolites that overwhelm scavenging systems has already been discussed as a likely cause for some drugs, particularly those whose elimination is shunted to alternate (toxic) pathways (i.e., acetaminophen).¹⁰⁷ Differences in the structure of feline hemoglobin have also been postulated. Feline hemoglobin contains up to 20 sulfhydryl groups compared with a maximum of four in other species. Sulfhydryl groups tend to be reactive and thus are susceptible to interaction with reactive parent drugs or metabolites. Thus more sulfhydryl groups would need to be maintained in a reduced state in cats.¹¹³ Other unique considerations for the cat might include its propensity for drug-induced retinal damage; hepatic lipidosis associated with anorexia; and, potentially, an increased risk of nephrotoxicity.

Canine breed differences in response to drugs may also reflect pharmacodynamic responses. For example, brachycephalic breeds (e.g., Boxers) are more susceptible to cardiac arrhythmias (sinoatrial block) caused by acepromazine.

Miscellaneous

Differences in circadian rhythm (i.e., diurnal versus nocturnal) play a role in some differences between the dog and cat. Aminoglycosides are less likely to cause toxicity if administered during active periods. This also has been established for theophylline, for which clearance occurs more rapidly at night in the dog compared with early morning in cats. Dosing of glucocorticoids at night has been recommended for cats in order to mimic endogenous release patterns. The clinical significance of these differences has not been determined.

Drug Information Sources

Extrapolation of doses of human drugs to dogs and cats and from dogs to cats should be based on a knowledge of the clinical pharmacology of the drug to be administered and on the physiologic differences of the target species. The safer the drug, the safer the extrapolation. Numerous resources are available for human drug information (e.g., Facts and Comparisons, USP Pharmacopeia, the package insert from the product, Physicians’ Desk Reference)¹¹⁴ to determine the safety and the determinants of disposition of a new drug. The veterinary literature and clinicians with expertise in the field, including diplomates of the American College of Veterinary Clinical Pharmacology, are additional sources.

Note that every drug that has been approved for use in humans has been studied in dogs. The studies generally have focused on safety, however, not efficacy. The information regarding safety (often including pertinent pharmacokinetic data, such as volume of distribution, bioavailability, and drug elimination half-life) may be obtainable through a Freedom of Information Act request, which would be processed by the Food and Drug Administration.¹¹⁵

Extrapolation of dosing regimens should be limited to relatively healthy animals, if possible, to avoid the effects of disease on drug disposition. Likewise, extrapolation to geriatric and pediatric patients is discouraged. Administration by the oral route is generally safer (although gastric irritation may be more likely). Oral administration is less preferred if the drug undergoes first-pass metabolism, however, because this can vary dramatically among animals. A 50% change in first-pass metabolism may double the pharmacologically active dose in a patient. Intravenous administration is not recommended; when it is absolutely necessary, the drug should be administered slowly (over 5 to 10 minutes or more). Drugs with long half-lives (>12 hours) should generally be avoided. If a drug is administered at an interval that is less than the drug half-life, accumulation should be anticipated and accounted for in the dosing extrapolation. Note that maximal adverse effects may not appear until accumulation is complete at steady state. Also, a drug half-life can change (as a result of disease or drug interactions). Thus a drug that initially did not accumulate (and whose dose is based primarily on volume of distribution) may begin to accumulate as disease worsens. Unless the drug can be monitored, a change in drug half-life will be missed. In such instances a dosage reduction is again indicated. On the other hand, as a patient improves, response to therapy may again change disposition, perhaps leading to therapeutic failure. The veterinarian should be prepared to treat adverse effects if they occur. If the drug half-life is long, the time necessary for abatement of the adverse reaction will also be long. In general, extrapolation of lipid-soluble drugs is discouraged because of the risk of too many species differences.

Water-Soluble Drugs

As a general rule, extrapolation of doses for drugs that are water soluble is more appropriate because these drugs are distributed to extracellular fluids (normalizing Vd) (Box 2-2); protein binding is likely to be negligible; and hepatic metabolism is minimized. Drug Vd and renal elimination may be similar among species, and the interval used for such drugs can often be extrapolated among species. The dose administered, however, probably should be reduced to compensate for differences in blood volume among animals. Increased doses are indicated for pediatric patients and for patients with edema; decreased doses are indicated for geriatric and dehydrated patients.

Lipid-Soluble Drugs

Lipid-soluble drugs tend to be distributed to total body water and beyond, leading to a greater risk of differences among species. They are more likely to be highly protein bound, leading to a risk of differences in tissue distribution and in the proportion of pharmacologically active drug. In contrast to water-soluble drugs, lipid-soluble drugs are more likely to require hepatic metabolism (Box 2-3). In general, the clinician should anticipate a longer half-life in cats for drugs that undergo phase I metabolism in other species. Note that species differences in phase I metabolism can be very profound. If acetylation is a major phase II route of elimination, it is likely that the drug may be metabolized faster in cats than in dogs. If phase I metabolism and glucuronidation is the major route of elimination, a longer half-life should be anticipated in cats. Although glucuronidation does not necessarily indicate that elimination of the drug will be slower in cats, until an appropriate study has established the kinetics of the drug in cats, its use is discouraged. An exception might be made if the drug can be monitored or the drug is characterized by a wide therapeutic window. Altered-release preparations are not recommended because rates of absorption among the species can be dramatic. Finally, it might be necessary to refrain from administering preparations containing propylene glycol and other unknown carriers because of adverse reactions in cats.

Renal Disease

Drug toxicity in renal failure may result either from increased sensitivity to the drug owing to uremia-induced alterations in tissue receptors or from decreased or increased PDCs caused by disease-induced changes in pharmacokinetics. Changes induced by renal disease have been best characterized (see Table 2-4).^117-120

Renal blood flow is often profoundly decreased in patients with renal disease. Changes in glomerular filtration and tubular secretion tend to parallel changes in renal blood flow. The effects of changes in renal blood flow (usually decreased) on drug excretion are most profound if renal extraction of the drug is high (e.g., penicillins, sulfates, glucuronide conjugates) but are less significant for drugs that are slowly extracted (e.g., aminoglycosides, diuretics, digoxin).

Glomerular filtration of drugs and other compounds is also adversely affected in renal disease independent of changes in renal blood flow. The determinants of glomerular filtration include protein binding, glomerular integrity, and the number of functional (filtering) nephrons. The molecular size of the drug is also important because drugs with a molecular weight greater than approximately 70,000 usually cannot be filtered. Drugs that are tightly protein bound (such as NSAIDs) are not filtered until they are displaced from the protein. Factors that tend to displace such drugs from protein-binding sites may increase the rate of drug excretion in renal disease and include hypoalbuminemia, competition for protein-binding sites owing to accumulation of uremic toxins, or changes in the conformation and thus binding affinity of the protein (e.g., albumin). Changes in protein binding that have been measured in renal disease include decreased binding of acidic drugs (e.g., furosemide, NSAIDs, selected penicillins and anticonvulsants) and normal or increased binding of basic drugs owing to increased concentrations of inflammatory proteins (e.g., propranolol, diazepam, prazosin). The impact of increased clearance of otherwise protein-bound drug may be offset by the increase in plasma concentration of the free drug displaced from the protein, thus increasing elimination of unbound drug.

Although changes in active tubular reabsorption that may accompany renal disease probably do not profoundly influence the rate of drug excretion, changes in active tubular secretion can be significant. Active tubular secretion occurs in the pars recta (straight segment) of the proximal tubule. A transport system exists for a variety of organic acids (e.g., penicillins, cephalosporins, NSAIDs, sulfonamides, several diuretics) as well as bases (e.g., cimetidine, procainamide, some morphine derivatives). Distal nephron active transport may also be important for some drugs (e.g., digoxin). Excretion of these drugs is most likely to be decreased in the presence of renal disease because of decreased nephron mass, decreased renal blood flow, and decreased tubular function.

In addition to these changes, renal disease can also alter drug disposition because of changes in electrolyte, acid–base, and fluid balance. Changes in electrolytes and acid–base balance may also be important in altering receptor sensitivity to drugs, such as those affecting the cardiovascular system (e.g., hyperkalemia and its effects on responses to digitalis, quinidine, and procainamide).

For drugs whose elimination depends on renal function and whose clearance is known to be decreased in renal disease (Box 2-4), dosing regimens can be appropriately altered to reduce the incidence of adverse reactions.¹¹⁹ Either serum creatinine or creatinine clearance is generally used to estimate glomerular filtration rate; however, because of its ease of measurement, serum creatinine is most commonly used. Decreases in the renal elimination of a renally excreted drug tend to parallel increases in serum creatinine, and dosing regimens can be altered by either lengthening the dosing interval or decreasing the dose by a proportional decrease in creatinine clearance or the increase in serum creatinine (Box 2-5).

The parameter of the dosing regimen that should be altered depends on the drug. Lengthening the interval results in wider swings in PDCs during a single dosing interval and thus may not be desirable for time-dependent antimicrobial therapy but would be acceptable for concentration-dependent drugs or lengthening the interval also should be avoided for anticonvulsant or cardioactive drugs that depend on maintenance of a minimum drug concentration within a specified therapeutic range. Thus decreasing the dose may be more appropriate for these drugs. For drugs with a long elimination half-life or drugs whose effects persist in the absence of detectable drug (e.g., selected antimicrobials [e.g., concentration-dependent], glucocorticoids, nonsteroidal agents), however, it may be more appropriate to prolong the interval. Aminoglycosides previously were dosed at 12-hour intervals. Because their efficacy depends on a high PDC, yet safety is based on allowing PDCs to fall below a recommended trough concentration, prolonging the interval (i.e., from 12 to 18 or 24 hours) was more appropriate than lowering the dose in patients with renal disease sufficient to alter creatinine clearance. However, aminoglycosides are currently administered at 24-hour dosing intervals, precluding the need to prolong intervals in some renal disease patients. For other drugs whose toxicity is related to peak drug concentrations, the dose might be decreased by 50%, or the interval prolonged one half-life. Few clinical studies have addressed the impact of renal disease on the clearance of drugs in dogs or cats. Experimentally induced renal disease sufficient to cause serum creatinine to increase to 145 ± 122 μmole/L in dogs caused a slight decrease in renal clearance but no clinically significant change in elimination half-life of marbofloxacin.¹²¹

Hepatic Disease

The efficiency of hepatic elimination is determined by hepatic clearance and the hepatic extraction ratio of the drug.^122-125 Both, in turn, depend on hepatic blood flow; the extent of drug protein binding; and intrinsic hepatic clearance, which itself consists of hepatic uptake (the rate-limiting step of hepatic clearance), intracellular transport, metabolism, and (if applicable) biliary elimination. Drugs that are eliminated by the liver can be categorized according to their rate of extraction.¹²⁶ Flow-limited drugs (e.g., lidocaine, propranolol, verapamil) are so rapidly extracted by the liver that their rate of elimination depends only on the rate at which it is delivered to the liver (e.g., hepatic blood flow). Such drugs are insensitive to changes in hepatic metabolism but are very sensitive to changes in hepatic blood flow. Capacity-limited drugs (e.g., diazepam, prednisolone, phenylbutazone, phenytoin, theophylline, cimetidine, and antipyrine) are extracted slowly by the liver, and their elimination depends on hepatic uptake and metabolism but is independent of hepatic blood flow. The elimination of such drugs is affected by changes in hepatic metabolism but not by changes in hepatic blood flow. Some drugs are intermediate, being partially dependent on hepatic blood flow and hepatic metabolism.^123,126

Protein binding can affect the elimination of some capacity-limited drugs because only unbound drug can be extracted by the liver. Flow-limited drugs tend to be binding insensitive in that hepatic extraction is so fast that binding to proteins does not alter their rate of elimination. Some capacity-limited drugs are not significantly protein bound and thus are also binding insensitive (e.g., antipyrine). In contrast, some capacity-limited drugs are binding sensitive (e.g., theophylline, phenytoin) because their slow rate of extraction can be increased by decreasing or increasing, respectively, protein binding.¹²⁷

The effects of hepatic disease on drug disposition are very complex, particularly for drugs that are affected by changes in hepatic blood flow, hepatic metabolism, and protein binding (Box 2-6).^123,126 Each of these parameters may be altered in various ways in patients with liver disease. Hepatic blood flow is generally reduced in chronic liver disease because of formation of portosystemic shunts and intrahepatic shunting. Drug delivery bypasses many functional hepatocytes that would otherwise clear the drug. Plasma and tissue drug concentrations are markedly higher when dosing regimens are not appropriately altered. This is particularly important for highly extracted drugs when they are administered orally. The dose of such drugs (e.g., propranolol, verapamil, prazosin, morphine derivatives) in the presence of normal hepatic blood flow is based on decreased bioavailability owing to first-pass extraction: a large percentage of the drug does not reach systemic circulation because it is removed from portal blood by the liver the first time it passes through the liver. Decreased hepatic blood flow and intrahepatic shunting of blood can markedly increase systemic bioavailability of such drugs (Figure 2-5).¹²⁸ Studies in human patients with liver disease suggest that the intrinsic metabolism of highly extracted drugs also is reduced in patients with liver disease. In human patients with liver disease, the dose of many highly extracted drugs is reduced by 50%; such an approach is probably reasonable for the veterinary patient.

Figure 2-5 Portosystemic shunting affects drug disposition at multiple sites. Clearance of flow-limited drugs will be decreased in proportion to the fraction of blood shunted around the liver. Capacity-limited drugs may also be decreased if hepatic mass is sufficiently reduced. Accumulation of fluid in response to sodium retention may be profound, particularly in the abdomen. Albumin may be decreased, increasing the fraction of unbound drug, which may not be cleared as rapidly in the presence of disease.

The effect of liver disease on poorly extracted (capacity-limited) drugs is more difficult to predict, particularly for drugs that are highly protein bound. Generally, hepatic metabolism is reduced in patients with liver disease. Disease is generally quite profound, however, by the time changes in drug disposition become evident. The severity of disease is manifested as a decrease in serum albumin and blood urea nitrogen levels.This has been shown by studies that have measured the clearance of antipyrine and caffeine, which are capacity-limited, binding-sensitive and binding-insensitive drugs, respectively, in dogs with experimentally induced liver disease.^129,130

Because elimination of these drugs depends entirely on hepatic metabolism, their clearance might be used as a hepatic function test. Because their elimination does not necessarily correlate with the elimination of other drugs that also depend on hepatic clearance, however, they are not used to predict changes in dosing regimens for the patient with liver disease. Recommendations regarding dosing regimens for drugs not highly extracted by the liver thus are difficult to make for the patient with liver disease, in part because there is no simple test that will assess or quantitate hepatic function.^125,129-133 In general, clearance of capacity-limited drugs is probably not impaired in patients whose serum albumin and blood urea nitrogen levels are within normal limits. Common sense dictates, however, that discretion be used when administering potentially toxic drugs that depend on hepatic clearance for elimination.

Diseases of the biliary tract alter the disposition of drugs eliminated through the bile. This route of elimination is complex, however, with drugs being eliminated in feces and undergoing enterohepatic circulation. Characterizing these changes is difficult without catheterization of the biliary duct, and recommendations are very difficult to offer. Cholestasis decreases the content or activity of cytochrome P450 drug-metabolizing enzymes and thus can affect the elimination of drugs that are not secreted in bile.^134,135 In general, doses of drugs eliminated principally in the bile should be reduced, particularly if the drug is characterized by a narrow therapeutic window. Examples include selected antimicrobials (doxycycline and clindamycin), digitoxin, and naproxen.

The effects that changes in protein binding may have on hepatic drug clearance contribute to the unpredictable nature of liver disease–induced changes in drug disposition.^127,136,137 Decreased protein binding of drugs that may accompany liver disease (i.e., caused by decreased synthesis of albumin, competition for binding sites by endogenous compounds, or changes in conformation of the binding site) may increase hepatic clearance and thus compensate for reduced hepatic metabolism, although the ability of the diseased liver to compensate is not clear. Increased rather than decreased binding may occur particularly for basic drugs (e.g., lidocaine) as a result of increased production of acute-phase proteins, which may have the effect of decreasing clearance. Changes in protein binding will also affect drug distribution. Protein binding decreases the amount of drug that distributes to peripheral tissues. Drug distribution in the patient with liver disease will be further complicated by the effects of fluid, electrolyte, and acid–base imbalances, which are also likely to occur. For example, an ascitic compartment may lead to overdosing if a patient is not dosed on the basis of lean body weight and the drug does not distribute to the ascitic compartment, which may constitute up to 30% of body weight (Figure 2-6).

Figure 2-6 Disease can affect body composition. The ascitic compartment in an animal with liver or other disease can represent up to 30% or more of body weight. If this compartment is not one to which a drug distributes and the animal is dosed on the basis of total body weight, plasma drug concentrations may be more than 30% higher than expected. If the drug is distributed to the ascitic compartment, prolonged distribution from the compartment back to plasma may result in a longer elimination half-life.

Cardiac Disease

Cardiac disease profoundly affects drug disposition (see Table 2-4). Primary or compensatory disturbances that lead to altered drug disposition include renal sodium and water retention, increased pulmonary and systemic venous pressures, and increased sympathetic nervous system output.⁶ Sodium and water retention can cause profound changes in drug distribution owing to changes in the sizes of body compartments. In addition, increased sympathetic outflow results in redistribution of blood flow such that the heart and brain receive a higher proportion of blood and thus are exposed to more drug. Because other tissues, particularly skeletal muscle, represent a large volume to which drug is normally distributed, reduced distribution of drug to these tissues results in even higher PDCs in blood going to the heart and brain. Thus the heart and brain are more susceptible to toxicity. CNS and cardiac toxicities to lidocaine and cardiac toxicity to digoxin have been described in some patients with cardiac disease; these toxicities have been attributed to blood redistribution, which accompanies cardiac failure.

As cardiac output decreases, reduced blood flow to the kidney and liver will have profound effects on clearance of drugs through either of these organs. Reduced blood flow reflects, in part, redistribution mediated by sympathetic output. As cardiac disease progresses, however, decreases in blood flow to both the liver and kidneys parallel decreases in cardiac output. The effects of decreased hepatic blood flow on drug elimination depend on the drug; as with liver disease, hepatic clearance of flow-limited drugs may be profoundly decreased. In the kidney, sympathetically mediated intrarenal redistribution of blood from cortical to juxtaglomerular tubules increases the likelihood of tubular reabsorption, which prolongs drug half-life.

Tissue hypoxia and decreased delivery of nutrients to the kidney and liver also contribute to decreased clearance by these organs. The metabolic capacity of the liver is reduced; thus clearance of capacity-limited drugs is impaired. Similarly, renal tubular function is impaired.

Drug absorption may be impaired in the patient with cardiac disease. This is particularly true for parenterally administered drugs. The rate of absorption is more likely to be affected than the extent of absorption; hence peak concentrations may be less, although the extent of drug absorption may not be affected. Redistribution of blood away from skeletal muscle and skin decreases the rate of drug absorption after intramuscular and subcutaneous injections. Autonomic disturbances, consisting of increased sympathetic activity and decreased autonomic tone, tissue hypoxia, and mucosal edema may decrease both the rate and the extent of gastrointestinal absorption. Decreased blood flow to the intestinal villus may decrease absorption of drugs that are normally very rapidly absorbed from the gastrointestinal tract. Finally, the effects of cardiac disease on hepatic clearance of flow-limited drugs and thus on systemic bioavailability of orally administered drugs must be considered.

Several recommendations can be made for administering drugs to the patient with cardiac disease (Box 2-7):

Thyroid and Other Diseases

Both hyperthyroidism and hypothyroidism can profoundly affect drug disposition, although the manner is unpredictable.¹³⁸ The effects thus far involve metabolism. In human patients with hyperthyroidism, the activities of some cytochrome P450 enzymes (e.g., hydroxylation) are increased, whereas those of others (e.g., N-demethylation) are decreased. In rats in which hyperthyroid disease had been induced, enzymes that act as cofactors were increased. Thyroidectomized animals have a general decrease in drug metabolism, although the sequelae are not always predictable. The effects of thyroid disease on drug disposition also depend on sex, with male rats having a general decrease in cytochrome P450 enzymes. The clinical sequelae of changes in drug disposition induced by thyroid disease are not well described in scientific studies, although several examples are provided in the human literature. Digoxin doses necessary to induce a clinical response are in general increased for patients with hyperthyroidism, whereas smaller doses than normal are needed for patients with hypothyroidism. Interestingly, propranolol clearance is decreased in cats with hyperthyroidism.¹³⁹

Diseases of other body systems can also dramatically alter drug disposition. For example, gastrointestinal disease (e.g., chronic inflammatory bowel disease) alters absorption of orally administered drugs.¹⁴⁰ Diseases of any system may alter the response of that system to a drug. Nutrition also can alter drug-metabolizing enzymes. The effects of disease, regardless of the system, on drug absorption, distribution, metabolism, and excretion are very complex, often subtle, and very difficult to predict. Finally, as therapy becomes successful and the clinical signs of disease resolve, the sequelae of disease on drug disposition also resolve. Therapy once again may need to be adjusted as the animal responds.

Pharmaceutical Drug Interactions

Pharmaceutical drug interactions occur before the drug is absorbed and may occur before administration (Figure 2-7). Interactions can occur between two drugs or between a drug and a carrier (solvent), a receptacle (including intravenous tubing), or the environment in which it is administered (e.g., gastric environment).¹⁴⁴ In human medicine pharmaceutical interactions most frequently result from the addition of drugs to intravenous fluid preparations. In veterinary medicine they are most likely to occur in the critical care environment, where intravenous administration and multiple drug administration are common, or after inappropriate compounding of drugs. Drug incompatibilities can change the chemical or physical nature of a drug (Figure 2-8). Incompatible reactions can reflect degradation caused by changes in pH, binding by drugs with different charges, or other molecular interactions; changes in temperature; or exposure to ultraviolet radiation.¹⁴⁵ Incompatibilities may occur both with (approved) finished dosing forms or compounded preparations. A number of sources can be used to minimize the risk of drug interactions involving admixtures of drugs.^146-150

Figure 2-7 An example of a drug interaction is exemplified by drug efflux from calcium hydroxyapatite (plaster of Paris) beads. Individually, amikacin or vancomycin is characterized by an initial rapid release, followed by slow release from the beads over several weeks. However, when both drugs are mixed in the beads, efflux of both amikacin and vancomycin is much more rapid, being nearly complete in 1 to several days.

Figure 2-8 Examples of chemical changes in drugs. Transdermal gels are thermoreversible, liquefying at refrigerated temperatures (A) and solidifying at room temperature (B). Discoloration, shown here for an ophthalmic product (C and D) and steroid.

Intravenous Preparations

A number of interactions involve drugs intended for injection (Table 2-6). Drugs that are unstable generally have a short shelf life when in solution. Reconstituted parenteral solutions should always be labeled with the new expiration date and used with strict adherence to the product label instructions after reconstitution. If directed by the label, refrigeration or freezing can prolong the shelf life. It is risky, however, to assume that cold storage will prolong the shelf life of the drug unless efficacy has been documented at the intended conditions. Freezing can increase the degradation (e.g., ampicillin), crystallization (e.g., heparin, dobutamine, furosemide), or precipitation (e.g., insulin) of drugs. Refreezing of a previously frozen and defrosted solution increases the risk of efficacy loss.Often, if specifically queried, the manufacturer of the product can provide specifit information regarding drug stability in conditions beyond those stated on the label. The proper reconstituting fluid should be used to prevent inactivation of drugs. For example, amphotericin B should be diluted only with 5% dextrose because precipitates will otherwise form; whole blood or packed red blood cells should be diluted only with 0.9% saline to prevent damage to infused cells.

Table 2-6 Examples of Drug Interactions in Solution

Changing the pH of a solution by improperly diluting it or mixing it with another drug can be risky. The release of some insulins is pH dependent. Diluting insulin with a solution other than that provided by the manufacturer may change the pH and thus the rate of insulin release. The pH of a solution may be needed to keep the active drug dissolved or stable; changing the pH may result in precipitation or loss of stability. For example, acid-labile drugs (e.g., penicillins) can be destroyed in a low-pH solution. Drugs prepared as an acid salt (e.g., lidocaine hydrogen chloride) or in acidic solutions (i.e., sodium heparin) should not be combined with alkaline solutions (i.e., sodium bicarbonate).

Drugs can bind to and inactivate one another, often because of ionic attractions. Calcium in solutions causes precipitation when combined with solutions containing carbonates (e.g., sodium bicarbonate). Heparin is incompatible with several drugs, such as aminoglycoside and beta-lactam antibiotics. Thus saline rather than heparin should be used whenever possible to maintain patency of catheters through which drugs are administered. When present in sufficiently high concentrations, penicillins inactivate aminoglycosides. In fact, ticarcillin can be used therapeutically to bind gentamicin in cases of overdose. Although plasma concentrations after therapeutic dosing of either drug probably do not achieve concentrations necessary to inactivate aminoglycosides, in the critical care situation, and as a once-daily dosing regimen of aminoglycosides becomes more generally acceptable, the risk of aminoglycoside (or a fluorinated quinolone) antimicrobial inactivation by a penicillin may become greater.

Often, a pharmaceutical drug interaction involving intravenous solutions can be detected by a visual change in the appearance of the drugs. Discoloration, cloudiness, and formation of precipitate generally are indications of an interaction, and with some exceptions use of the drug should be reconsidered. Not all interactions will result in a physical change of the appearance, however. Likewise, the change in physical appearance of a drug combination does not necessarily indicate that the activity of the drug has been changed. For example, diazepam has been mixed with other preanesthetics with no observable change in drug efficacy, despite a cloudy discoloration. Whereas a pink discoloration of dopamine indicates inactivation, discoloration of dobutamine does not preclude efficacy if the drug is used within 24 hours. Slight yellow discoloration of procainamide is acceptable; dark discoloration indicates a loss of efficacy.

Several drugs can bind to receptacles. For example, lipid-soluble drugs (e.g., diazepam) can bind to plastic containers; insulin binds to selected glasses and to many plastics, including polyethylene and polyvinyl; aminoglycosides bind to glass. Binding to catheters and intravenous lines can be minimized by flushing each new system with a sufficient volume of solution (50 mL) before drug administration. Drugs packaged in brown bottles (e.g., diazepam, furosemide) are somewhat protected as such from ultraviolet light, and protection should be continued if they are transferred to another vial.

Oral Preparations

Oral preparations may involve drug–diet or drug–drug interactions (see Box 2-7 and Tables 2-5 and 2-6). Many drugs bind luminal contents (drug–diet interactions; Table 2-7), and oral absorption is impaired.¹⁵¹ Food can alter splanchnic blood flow, gastric motility (and thus mixing and drug dissolution as well as gastric emptying), and gastric secretions. Changes in gastric secretions can alter gastric pH, which can change the percentage of ionized and thus diffusible drug. The net effect of food on drug absorption depends on the pK_a of the drug, whether the drug is labile to the effects of pH and enzymes, and the site of absorption of the drug (i.e., stomach versus intestine).¹⁵¹ The sequelae may affect either the rate or extent of drug absorption. The effect of food on the rate of drug absorption is most important for drugs with a narrow therapeutic window and for drugs with a steep dose–response curve, for which a small change in PDC can cause profound differences in response to the drug. Predicting drug–diet interactions on the basis of chemical structure should be done cautiously. Whereas food will impair the oral absorption of most tetracyclines, doxycycline is minimally affected. Food reduces the absorption of ampicillin but not amoxicillin;¹⁵² food prolongs the time to peak plasma concentrations of cefadroxil (and increases time above the minimum inhibitory concentration) but has no impact on the absorption of cephalexin.¹⁵³ Oral absorption of selected drugs is enhanced rather than decreased in the presence of food (see Table 2-2). Interactions between drugs and foods were recently well reviewed using a body-systems approach.¹⁵⁴

Table 2-7 Examples of Sites of Drug–Drug Interactions

NSAIDs, Nonsteroidal antiinflammatory drugs.

Drug–drug interactions (see Table 2-7) in the lumen include, but are not limited to, changes in the diffusibility, dissolution rate, and particle size of orally administered drugs. Drug–drug interactions in the gastrointestinal tract can inactivate or prevent absorption of drugs. Sucralfate, cimetidine, aluminum hydroxide, and attapulgite (previously Kaopectate®) are examples of drugs that bind to and prevent the absorption of many drugs. Other drugs alter the rate of absorption by altering gastric motility (see Table 2-4; see also the discussion of pharmacokinetic interactions later in this chapter). To minimize these effects, none of these drugs should be given simultaneously with another orally administered drug.

Topical Preparations

Pharmaceutical interactions in topical preparations may occur between drugs or between drugs and the vehicle in which they are carried.^155,156 Both the rate and extent of drug absorption can be affected adversely. For example, macromolecular additives may bind chemically with the active drug. Methyl, ethyl, hydroxyethyl, and carboxymethyl cellulose frequently form complexes with drugs that can lead to drug precipitation. Vehicles are often selected because of their effect on drug absorption. For example, retardant vehicles (e.g., polyethylene glycol 300) interact with the drug, decreasing its absorption, whereas dimethyl sulfoxide (DMSO) is well recognized for its ability to enhance absorption of many topically applied drugs. The pluronic-lecithin vehicle of PLO gels (PLO) designed for transdermal drug delivery is thermoreversible, meaning they are semisolid at room temperature but become liquefied when refrigerated (see Figure 2-8).

The concentration of drug available for skin penetration depends on its dissolution in the vehicle. Drugs must generally dissolve in an aqueous layer of fluids that collects under the ointment base before percutaneous absorption can occur. Few drugs (and particularly water-soluble drugs) are sufficiently soluble to be dissolved in petrolatum bases; most drugs mixed in such bases are present as particles. Because few particles are located at the vehicle–skin interface, and dissolution is very slow for the particles, drugs in such preparations are likely to be ineffective. As such, lipid-soluble drugs tend to be better absorbed, particularly when prepared in lipid vehicles. Because only dissolved drug can move, the solubility of a drug in the vehicle is an important determinant of drug movement into the skin. However, a drug that has too great an affinity for the vehicle also may not be well absorbed simply because it will not leave the vehicle.¹⁵⁶

Compounded Preparations

Many U.S. pharmacies now cater specifically to veterinarians and thus are prepared to address special problems in drug therapy through compounding. As important as compounding can be to the safe and effective administration of drugs to animals, compounded products present more risks of therapeutic failure than encountered with approved drugs.^157,158 Among the risks are the potential failure of the drug at the pharmaceutical phase because of interactions among the active or inactive ingredients. These risks are addressed in more depth in Chapter 4.

Pharmacokinetic Drug Interactions

Drug interactions that occur inside the body may lead to life-threatening adversities because of either therapeutic failure or increased side effects. Pharmacokinetic interactions occur when one drug alters the disposition of another drug.¹⁵⁹ Each stage of disposition of a drug—absorption, distribution, metabolism, or elimination—can be altered by another drug. The majority of drug interactions leading to serious adverse events in humans reflect pharmacokinetic interactions, involving either transport proteins or drug-metabolizing enzymes.¹⁴³ Identifying the role of drug interactions in causing adverse effects is difficult; large numbers of patients generally must be studied to identify even common interactions. Predictions of drug–diet or drug–drug interactions might be facilitated by evaluation of chemistry, but caution is necessary.¹⁵⁴ Models that predict drug interactions are complicated by complex interactions (e.g., species, disease, and multiple drug interactions).¹⁶⁰ Identification of drug interactions involving dietary supplements is handicapped by the lack of a mandated surveillance mechanism; information in human medicine generally is obtained through reviews, case reports, or case series or based on open, uncontrolled studies.¹⁶¹ Evidence-based information in veterinary medicine is even further limited. Several human-medicine reviews have described interactions between drugs and botanicals or herbs¹⁶¹ and foods.¹⁵⁴ Mechanisms are often difficult to identify, in part because of broad substrate specificity. Some examples of drug–drug or drug–supplement interactions follow.

Metabolism

Pharmacokinetic drug interactions frequently alter the metabolism of a concurrently administered drug (see Table 2-4).^166-170 When administering a drug metabolized by the liver, it is wise to anticipate a drug interaction if a second drug also metabolized by the liver is added to therapy. Most of the interactions result from modulation of hepatic (phase I) drug-metabolizing enzymes (see Table 2-8).

Induction of drug-metabolizing enzymes should be considered as a protective mechanism that facilitates excretion of potentially toxic compounds. However, induction is a double-edged sword: although the elimination of a potentially toxic drug increases, so does potential formation of toxic or carcinogenic metabolites. Most CYP enzymes are inducible, with response to inducers varying within and among species, age, and gender. Human CYP known to be influenced by inducers include CYP 1A1/2, 2A6, 2C9, 2C19, 2E1, and 3A4 (see Table 2-8). Inducers generally act as substrates for the induced enzymes, and induction generally is dose dependent.¹⁷¹ Inducers often induce more than one CYP and may significantly increase the activity of an enzyme that otherwise (constitutively) is either absent or present only in very low concentrations. Induction is accompanied by an increase in the transcription and thus intracellular concentration of the induced enzyme. Maximal induction of transcription generally requires 10 to 12 hours of exposure to a drug. Although transcription may return to baseline 18 to 24 hours (depending on dose) after the inducer is discontinued, the impact of the inducer may persist. The duration of the effect varies with the rate of enzyme degradation, which normally results in half-lives that range from 8 to 30 hours. Although not common, induction also may reflect a decrease in the rate of CYP degradation—that is, a prolongation of enzyme half-life.¹⁷²

Barbiturates are recognized inducers of CYP; indeed, the observation that “tolerance” to hypnotics developed in dogs chronically exposed to barbiturates led to the recognition of the phenomenon of induction.¹⁷² Phenobarbital is one of the most potent microsomal enzyme inducers known and can enhance the hepatotoxicity of other hepatotoxic drugs. Likewise, it increases the formation of and response to prodrugs and decreases the effects of itself and other drugs metabolized by the liver as clearance of these drugs is increased.^73,173,141 The CYP2B9 family, responsible for metabolism of a large number of drugs in rodents, is induced by phenobarbital. Therapeutic doses of phenobarbital have been associated with induction of CYPIA activity, as well as alpha-glycoproteins in dogs.¹⁷⁴ The clinical impact of this effect is demonstrated in epileptic dogs treated with phenobarbital: initial concentrations may decrease within several months of therapy despite no dose change. Interestingly, oral pentobarbital also increases the amount of CYP2C present in the intestinal tract of dogs when administered at doses consistent with that consumed by dogs fed commercial dog foods prepared from animals euthanized by pentobarbital (10 to 60 μg/gm). The impact on the metabolism of selected subtrates varies and is greatest for higher doses, but changes are also present at low doses, suggesting a potential clinical impact on therapeutic drugs.¹⁷¹

Drug interactions that reflect inhibition of drug-metabolizing enzymes may be at greater risk in causing serious adverse events. A number of inhibitors of CYP enzymes have been identified. As with inducers, the extent of inhibition is dose dependent,^172,175 with inhibitors often acting as substrates at the inhibited enzyme, and either characterized by broad enzyme interaction, being inhibitory for a number of different CYPs, or acting selectively for a single enzyme (see Table 2-8).¹⁷² Currently, three types of drug-metabolizing enzyme inhibitors have been described: reversible, quasi-reversible, and irreversible, with reversible inhibitors being the most commonly involved in drug–drug interactions. Reversible inhibition is transient, resolving when therapy with the inhibitor is discontinued. Like induction, reversible inhibition appears to be dose dependent.¹⁷⁵ Reversible inhibition most commonly reflects interactions that are competitive or noncompetitive. Competitive inhibition is exemplified by a drug that blocks access to the catalytic site of the enzyme of another structurally similar drug. The drug may or may not be a substrate for the enzyme. Noncompetitive inhibition occurs when substrate binding occurs at a different site but nonetheless changes the catalytic activity of the enzyme such that it is inactivated.¹⁷²

Generally, clearance of a concurrently administered drug metabolized by the liver is prolonged in the presence of inhibitors, increasing the potential for toxicity or for an exaggerated pharmacologic response. Additionally, prodrugs (e.g., enalapril, primidone) are less likely to be activated. Chloramphenicol, cimetidine, and imidazole antifungal drugs are examples of potent microsomal enzyme inhibitors.^73,173,141 Co-administration with potentially toxic drugs that are also metabolized by the liver should be done cautiously. Fluorinated quinolones such as enrofloxacin and marbofloxacin can increase theophylline plasma concentrations to toxic levels, presumably because of impaired hepatic clearance of theophylline.¹⁷⁶ Ketoconazole inhibits cytochrome P450 enzymes, including CYP3A4 in the dog.¹⁷⁷ However, it does not necessarily clinically affect all drugs metabolized by the liver. For example, morphine clearance was not changed in Greyhounds treated with an average of 12.7 mg/kg ketaconazole per day.⁹³ Although area under the curve was greater in the presence of morphine, the change was neither statistically nor clinically significant. This may reflect, however, the flow-limited nature of morphine clearance or, potentially, breed differences in drug metabolism. Other imidazoles (fluconazole and, to a lesser degree, itraconazole) also have broad enzyme inhibitory activity (see chapter 9).

Drug-induced inhibition of drug metabolism can be used for therapeutic benefit. Both cyclosporine and ketaconazole are substrates and inhibitors of both P-gp and CYP3A. As such, the combined use of the drugs may result in marked prolongation of the half-life of either drug. Ketoconazole has been used to increase blood drug concentrations of cyclosporine (owing to both inhibition of drug-metabolizing enzymes and competition with P-gp) in dogs.¹⁷⁸ A twofold increase in expected PDCs may occur, as indicated by a 50% dose reduction in Beagles¹⁷⁹ or a twofold increase in C_max in Greyhounds.⁹³ Indeed, the impact may be greater in some dogs: our laboratory has documented an elimination half-life of over 150 hours in dogs simultaneously receiving ketoconazole, yielding concentrations that exceed 4500 ng/mL. In cats, cyclosporine (4 mg/kg orally) concentrations can be expected to increase approximately twofold when administered with ketoconazole (10 mg/kg).¹⁸⁰ The macrolides also have broad inhibitory effects. The inhibitory effect of erythromycin may reflect inhibition of CYP3A macrolides, which appear to inhibit cyclosporine clearance. Again, in our laboratory azithromycin co-administration was associated with marked prolongation of cyclosporine half-life in a cat, leading to cyclosporine concentrations that exceeded 4500 ng/mL (therapeutic range 800-1400 ng/mL). The effect of cimetidine, another drug-metabolizing inhibitor with broad substrate specificity, is variable, ranging from no significant effect in Beagles (n=10; 15 mg/kg cimetidine every 8 hours for 8 days and 5 mg/kg once daily for cyclosporine) to 50% decrease in clearance in humans.¹⁸¹ Cimetidine-induced enzyme inhibition, however, has been used to prevent metabolism of acetaminophen in humans and cats into potentially lethal toxic metabolites.^182,183 Cilastatin inhibits renal tubular drug metabolism of imipenem; the net effect may prolong the half-life of imipenem, but hepatotoxicity or renal toxicity resulting from metabolites might also be reduced (see Chapter 4). Nutrition, sex, age, and other factors can influence the way drug-metabolizing enzymes respond to drugs. Alcohol and 4-methylpyrazole competitively inhibit alcohol dehydrogenase, the drug-metabolizing enzyme that converts ethylene glycol to its lethal metabolite.

Although less common, drug clearance may also be affected by drugs that change hepatic blood flow, solely due to substrate delivery, not a limitation on metabolic rate. This interaction is significant, however, only for drugs that are characterized by extensive and rapid hepatic clearance (e.g., propanol, lidocaine) and probably is not clinically relevant.

Among the most commonly identified drug–diet (nutrient or supplement) interactions that alter drug pharmacokinetics are those reflecting drug metabolism, particularly CYP3A4.¹⁵⁴ Several dietary supplements are known to interact with drugs.¹⁶¹ These include, but are not limited to, St. John’s wort (induction of CYP3A4, particularly intestinal), echinacea (induction and inhibition of intestinal CYP), ginkgo biloba (induction of CYP219A), and grapefruit (inhibition of CYP3A4 and inhibition of P-gp).

Pharmacodynamic Drug Interactions

Pharmacodynamic drug interactions occur when one drug alters the chemical or physiologic response to another drug (see Table 2-6). Pharmacodynamic interactions may increase response to a drug in an additive or synergistic fashion at the same receptor (e.g., the permissive effect of glucocorticoids on alpha-adrenergic receptors; phenobarbital and clorazepate at gamma-aminobutyric receptors), at an intracellular site (e.g., epinephrine and theophylline in bronchial smooth muscle), or at different sites but with the same physiologic reaction (e.g., hypokalemia induced by cardiac glycosides and diuretics; many interactions of antimicrobials). Pharmacodynamic interactions may also decrease the response of some drugs owing to competitive antagonism at the same receptor site (e.g., atropine and anticholinesterases or atropine and metoclopramide) or antagonistic responses mediated at distant but physiologically related sites. Antagonistic pharmacodynamic interactions have been used therapeutically: oxymorphone or other mu agonistic effects are reversed with naloxone, xylazine and other chemical sedatives are reversed with tolazoline or yohimbine, and medatomidine is reversed with antipamazole.

The most familiar pharmacodynamic interactions are probably those that act in an additive or synergistic manner to augment response to a drug. Augmentation can occur through different mechanisms of action (i.e., controlling vomiting by combining a drug active at the chemoreceptor triggering zone with a drug that acts peripherally, controlling seizures by combining phenobarbital with bromide, controlling tachycardia by combining diltiazem with digoxin or atenolol). Less commonly, augmentation may occur through similar actions at a receptor site; more often, drugs will compete with one another at the same receptor, thus resulting in antagonism. Unfortunately, often forgotten is the fact that augmentation of the desired pharmacologic response may be accompanied by augmentation of an undesirable adverse drug event (ADE). For example, drugs that impair renal prostaglandin synthesis (e.g., NSAIDs, angiotensin-converting enzyme inhibitors, aminoglycosides) should be used in combination cautiously because their combination increases the risk of renal failure. Likewise, ulcerogenic drugs (NSAIDs, glucocorticoids) enhance the risk of gastrointestinal ulceration when used in combination.

Pharmacodynamic interactions may decrease the response to a drug because of competitive antagonism at the same receptor site. Most commonly, these actions are desirable and are frequently the target of combined drug therapy: atropine to treat organophosphate toxicity, reversal agents for opioids and anesthetic agents. Antagonistic pharmacodynamic responses can also occur through different receptor sites or different mechanisms. The combination of a bacteriostatic antimicrobial (one that slows the growth of an organism) with a bactericidal antimicrobial (one whose efficacy depends on rapid growth) might be considered as an example. Also, the prokinetic effects of cisapride and metoclopramide are prevented by anticholinergics; calcium-containing solutions should not be combined with blood or blood components because the loss of anticoagulant effects increases the risk of microthrombi formation in the transfused blood.

Hu¹⁶¹ has reviewed pharmacodynamic interactions between drugs and herbs or botanicals. These can be clinically significant, as is exemplified for selected organosulfur components of garlic, which act as anticoagulants. Increased bleeding times have been documented in patients taking garlic supplements and subsequently treated with warfarin compared with increases with warfarin alone.