Adverse Drug Events Defined
An adverse drug event (ADE) refers to any harm caused by a therapeutic or preventative (or diagnostic) intervention. They can be further subdivided to include medication errors and adverse drug reactions (ADR; see last page). Medication errors are ADE that result from a mistake made by the caregiver, including, but not necessarily limited to, administration of or at the wrong drug, dose, interval, or route to the wrong patient.
In contrast to medication errors, an adverse drug reaction (ADR) refers to a noxious and unintended response to a drug or other medication which occurs at a dose given in order to achieve the intended use of the medication. As such, the term ADR implies a reaction that might cause serious harm to the patient and thus reflects a response to the inherent properties of the drug. An ADR may reflect a pharmacodynamic response or a pharmacokinetic effect. An ADR should be contrasted with the term "side effect," which refers to an effect that does not cause harm; indeed, a side effect may not be undesirable, but, in fact, may be desirable or inconsequential to the health of the patient. Like Type A adverse reactions (see below), side effects are generally predictable and dose dependent.
Adverse drug reactions can be further classified as either type A (type I) or type B (type II). Type A ("augmented") or type I adverse events generally result from drug concentrations at the site (generally estimated by plasma drug concentrations [PDC] that either exceed the maximum or drop below the minimum therapeutic range). If the clinician is familiar with the drug and the patient, type A reactions are largely predictable and, as such, avoidable. Generally, type A are manifested as an exaggerated, but normal or expected pharmacologic response to the drug. Like side effect, an ADR can be an exaggerated primary or desired response (e.g., bradycardia in a patient receiving propranolol to slow a sinus tachycardia), but also may reflect an unwanted, secondary response resulting from the drug's pharmacologic effects (e.g., bronchospasms induced by the β-blockade effects of propranolol). Some drugs also cause adverse events unrelated to their pharmacologic response. These reactions usually reflect damage to target cells and are referred to in this chapter as cytotoxic adverse reactions. Cytotoxic adverse reactions are perhaps best exemplified by nephrotoxicity induced by aminoglycosides or hepatic necrosis or methemoglobinemia induced by acetaminophen. Often, it is the metabolite of the drug rather than the drug itself that causes cytotoxicity. In such cases, drugs that induce metabolism, particularly in the liver (e.g., phenobarbital, phenytoin), may increase the risk of toxicity, whereas drugs that decrease metabolism reduce the risk of toxicity (e.g., cimetidine). Cytotoxic drug reactions might be treated with drugs that scavenge radical metabolites (i.e., N-acetylcysteine, a glutathione precursor).
In contrast to type A events, type B ("bizarre") events are not dose or concentration dependent. As a result, these reactions are not predictable and are largely unavoidable. They occur only in a small percentage of the population receiving the drug; in human medicine, they account for about 6 to 12% of all ADR. Generally, their incidence - indeed their existence - often is not documented until the drug is in wide use. In addition, because their cause is not well understood, treatment is generally limited to symptomatic therapy. Examples of type B adverse reactions include drug allergies or "idiosyncrasies." Many of the idiosyncrasies ultimately may be shown to be genetically or otherwise based (i.e., polymorphisms in transport or drug metabolizing proteins [ivermectin toxicity of collie-related breeds]), but the cause has yet to be identified and thus the reaction cannot be predicted. As with type A events, type B events can occur in response to the parent drug or its metabolite. Manifestations of acute type II reactions generally reflect tissues containing mast cells (lungs in cats, gastrointestinal tract and skin in dogs). Treatment focuses on antihistamines, glucocorticoids and, if lungs are involved, bronchodilators.
Organ Predisposition to Toxicity
A variety of factors can influence the likelihood of adverse, and particularly toxic, reactions. The organs most susceptible to type A drug events usually are those subjected to the greatest exposure or concentration of the drug. Thus, the organs with the greatest blood flow and those organs capable of drug concentration, such as the liver and kidney, are the most vulnerable to systemic drugs. Highly metabolically active organs are also more likely to manifest toxic effects for two reasons. First, such organs depend on the presence of energy, and anything that impairs acquisition of energy (including blood flow) can lead to malfunction. Second, if the metabolic activity includes metabolism of compounds, the production of potentially reactive metabolites can increase the likelihood of cytotoxicity if these metabolites interact with cellular structures. The organs most susceptible to damage by type B reactions tend to be the organs that contain tissues that act as haptens for drug-induced allergy (e.g., skin, blood-forming units) or tissues that filter and trap immune complexes (e.g., glomerulus and joints). Organs containing a preponderance of mast cells also are more likely to manifest immune-mediated reactions (i.e., "shock organs; see Allergic Reactions). Not all adverse reactions are clinically relevant. Sometimes the reaction is not detectable unless actively sought. For example, clinical laboratory tests may detect a drug-induced hepatotoxicity (e.g., increased serum alanine transferase activity) that was clinically silent. Many drugs can directly interfere with or indirectly influence clinical laboratory tests, including endocrine function testing. If an adverse drug event is suspected, the importance of reporting it (or suspicion) cannot be overemphasized (see later discussion). Organ predisposition to drug-induced toxicity may show diurnal variations. For example, both aminoglycosides and cisplatin (To 2000) exhibit increased renal toxicity in humans when administered in the evening compared to the morning. For the aminoglycosides, safety in the morning has been attributed, in part, to an increase in glomerular filtration rate which occurs in the morning; an increased sensitivity to interleukin-6-induced inflammation was suggested for cisplatin.
Factors Increasing the Risk of ADE: Pharmacogenetic Diversity
Increasing genetic diversity is being identified as a cause for individual variation in response to toxins and drugs, leading to the fields of toxicogenetics and pharmacogenetics, respectively. Two areas are described: transport proteins and drug-metabolizing enzymes.
Transport proteins occur in many tissues where they are responsible for moving drugs in or, more commonly, out of cells. An important transport system is the P-glycoprotein (P-gp) system, a MDR-1 gene product best known for imparting multidrug resistance to cancer cells (and to microbes). However, this transport system occurs in several tissues in the body, including portals of drug entry (GI tract, skin, lungs) and sanctuaries (brain, prostate, eye). P glycoprotein transports a large number of drugs that are chemically divergent. 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. Polymorphism reflecting a mutation deletion of MDR1 which causes non-functional P-gp has been documented in collie and related working breed dogs. The incidence of the deletion is impressively high: in the US, in one study, 35% of collies were homozygous and another 42% heterozygous for the mutation deletion. A similarly high incidence was found in dogs in France: 20% 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). The impact of the mutation on drug safety in afflicted animals can be profound. Substrate specificity for P-gp 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. Like CYP, P-gp is subject to induction or inhibition by different drugs, many of which are substrates for the transport protein. Retinal degeneration in cats caused by fluoroquinolones is another example of toxicity resulting from differences in efflux proteins, and specific species differences in retinal proteins. Cats lack the protein responsible for efflux of fluoroquinolones. Accumulated drugs become phototoxic upon exposure to ultraviolet radiation, causing retinal damage. Of the drugs approved for use in cats in the USA and Europe, enrofloxacin is the most toxic with marbofloxacin/pradofloxacin followed by orbifloxacin least. Renal disease may worsen the risk.
Human polymorphisms in CYP metabolic enzymes have been associated with therapeutic failure resulting from extremely rapid metabolism of a drug and toxic effects due to decreased metabolism. As in humans, polymorphism in drug-metabolizing enzymes has been reported in dogs, but not as well described. Differences in response to anesthesia recognized in sighthounds reflects both differences in drug distribution (to lean versus fat compartments), as well as differences in metabolism. Cytochrome-mediated clearance of several anesthetic agents is less in Greyhounds compared to other (non-sighthound) dogs; documented drugs include thiopental, thiamylal, and methohexital. Clearance of propofol by Greyhound is three times less than that of beagles. Ketoconazole plasma concentrations were two-fold higher than expected in Greyhounds in one study. Further, Greyhound disposition of celecoxib, a cyclooxygenase-1 protective nonsteroidal antiinflammatory, indicates breed differences may predispose this breed to adverse drug reactions. Polymorphism also has been described for CYP2C isoenzymes, again in Beagle dogs and possibly Greyhounds. Polymorphism in celecoxib metabolism was attributed to CYP2D15, for which three canine variants were found (Paulson et al. 1999). In a study of 242 Beagles receiving celecoxib, approximately 50% were considered efficient metabolizers and 50% poor metabolizers, with bioavailability and maximum plasma drug concentration in the latter group almost two-fold higher.
Example of Adverse Drug Events
Acetaminophen
Acetaminophen offers an example of how drug metabolism increases the risk of toxicity. It is more toxic in cats compared to dogs because cats are not able to glucuronidate the drug. The drug thus undergoes phase I metabolism, which produces toxic metabolites (oxidizing). Feline hemoglobin has more sulfate groups which are oxidized, causing methemoglobinemia. Treatment focuses on scavenging the oxygen radicals (N-acetylcysteine) and preventing metabolism and thus formation of toxic metabolites (cimetidine).
Sulfonamides
Type I: Sulfonamides act as a substrate for and thus competitively inhibit thyroid peroxidase which interferes with hormone iodination. Thyroid hormone synthesis can be suppressed at high doses (48 to 60 mg/kg/day for 6 weeks). Antimicrobials are not the only sulfonamides to suppress the thyroid gland; zonisamide at 12 mg/kg/day caused suppression in dogs.
Type II: Drug allergies are among the adverse effects associated with sulfonamides (Trepanier JVPT 2004). Onset of hypersensitivity to sulfonamides in dogs ranges from 5 to 36 (mean 12) days. Although it is the most common antimicrobial associated with hypersensitivity, the incidence is nonetheless low. All commercially available sulfonamides, including generic preparations, have been associated with hypersensitivity. The "potentiator" may also be responsible for some reactions: trimethoprim has been associated with skin eruptions or hepatopathy. Hypersensitivity causes reactions in many body tissues in any species (see below), although arthropathy appears to occur more commonly in large-breed dogs (with Dobermans overrepresented). Keratoconjunctivitis sicca (KCS) is a more common side effect of potentiated sulfonamides in dogs but may reflect direct cytotoxicity rather than an allergic reaction. The incidence is more common, occurring in up to 15% of dogs, but time of onset may be months to years after therapy is initiated. Deficiencies in N-acetylation, as occurs in canine drug metabolism, may generally predispose dogs to sulfonamide toxicity. Whether this is due to accumulation of the parent compound or production of a potentially toxic alternative metabolite remains to be revealed. Oxidation of sulfonamides in dogs yields hydroxylamine, a cytotoxic metabolite which also might be the cause of toxicity. The potential role of arylamine as a cause of sulfonamide toxicity is supported by the lack of toxicity by other sulfonamide drugs used in dogs (e.g., deracoxib, furosemide, zonisamide, and acetazolamide) which, lacking a primary arylamine, are not converted to hydroxylamine. The mechanism of hypersensitivity may reflect haptenization of the drug (or its metabolite) and a subsequent T-cell response, although other mechanisms (e.g., humoral response or cytotoxicity) may be responsible (Trepanier 2004).
NSAIDs
NSAIDs offer an example of how package inserts can be useful for predicting ADE.
Type I: Three sources of data are available for consideration regarding the safety of drugs, using NSAIDs as an example. Preclinical studies may involve in vitro studies which may or may not be relevant to the in vitro situation (e.g., Cox1:Cox2 ratios). Preapproval data (toxicity studies [usually limited to 6–12 dogs] and field trials [generally ranging from 100–1000 patients]) generated to support safety during the approval process are available on package inserts. Comparison of these data might offer some discrimination among the products regarding relative safety, although the data reflect overdosing (a situation that is not commonly clinically relevant). The data (Table below) suggest that tepoxalin is the safest of the NSAID (based on GI effects), followed by carprofen, meloxicam, firocoxib, deracoxib and finally etodolac. Of the field trial data, carprofen (n = 900) and deracoxib (n = 194) were the only NSAIDs for which vomiting was not greater than placebo. Limited post-market clinical trials are available; many are sponsored by drug companies and must be considered potentially biased. One independent study (Reimer and coworkers J Vet Intern Med. 1999;13:472–477) found carprofen and etodolac (recommended dose for 28 days) were not, but buffered aspirin (15.6 mg/kg) was (within 5 days) associated with ulceration in the GI tract of dogs. Among the most important sources of data is post-market surveillance which is provided by you, the practitioner (please report adverse drug events which can be reviewed on the FDA website [www.fda.gov/cvm search adverse event reporting]) (VIN editor: as of 10/17/15 link is redirected to www.fda.gov/AnimalVeterinary/default.htm). As an example of the importance of reporting and using these drugs wisely, Novartis reported (J Am Vet Med Assoc. 2005;224:1112) that of 29 dogs that developed perforated ulcers while receiving deracoxib, 90% received either an overdose or another NSAID or glucocorticoid within the last 24 h of presentation.
Type II? Adverse event data reveal that all NSAIDs will cause liver disease in dogs (potentially increased risk of renal disease in cats, particularly with meloxicam?). Because older animals may have less protective ability against NSAIDs, hepatoprotectants such as SAMe, N-acetylcysteine (especially for acute hepatopathy), or milk thistle should be considered. The FDA has asked for Client Information tear sheets to accompany package inserts. Clinicians would be prudent to make sure all clients receive this information and are well counseled regarding the use of NSAIDs in their pet.
Aminoglycosides
Aminoglycosides offer an example of how dosing regimens can impact toxicity. These drugs are toxic because the renal tubular cell actively takes the drug up, concentrates it in lysozymes, which explode and release their destructive enzymes. Toxicity is related to how low drug concentrations become because the kidneys must have time to excrete accumulated drug. Because damage will continue after BUN and creatinine increase, toxicity is best prevented: once daily rather than multiple (every 8 to 12 h) therapy, hydration with sodium-containing fluids, treating dogs in the morning, avoiding nephroactive/toxic drugs, and using N-acetylcysteine for acute toxicity.