Impact of Drug Interactions and Adverse Events on Therapeutics
Tep Kang
Andrew M. Peterson
As the quantity and types of pharmacologic agents continue to expand, the likelihood of drug interactions and adverse reactions increases. Currently, more than 8,000 drugs are available to treat various conditions. Each agent is designed to alter the homeostasis of the human body to some degree, and individual responses to these agents can be unpredictable.
In a prospective study, Benard-Laribiere et al. (2014) found that 3.6% (97/2,692) of hospital admissions were due to serious adverse drug reactions (ADRs). Thirty percent of which were preventable and 16.5% of which were potentially preventable. Drug interactions caused 29.9% of ADR-related hospital admissions. According to the Institute for Safe Medicine Practices (ISMP) QuarterWatch (2014), psychiatric adverse drug events, notably suicidal behaviors, represent the major adverse effects reported in children under age 18. In a meta-analysis of observational studies, Martins et al. (2014) reported a 21.3% incidence of adverse drug events among adult inpatients. These data were captured during prospective monitoring whereby the events are detected during the hospital stay and can include interviews of the patient and/or care team and reviews of clinical and laboratory records.
ADRs present an alarming problem that warrants significant attention from health care practitioners. Not only do ADRs affect morbidity and mortality, they also dramatically increase health care costs. In the United States, the impact of ADRs may cost up to $30.1 billion per year. Most of the cost is attributed to increased hospitalization, increased length of stay, and increased cost of performing additional tests (Sultana et al., 2013).
Similarly, drug interactions are potentially preventable ADRs posing a significant problem to the health care community. It has been reported that approximately 10% to 20% of hospital admissions are drug related and about 1% of these are secondary to drug interactions. Others have reported that drug interactions are responsible for up to 3% of hospital admissions (Bjerrum et al., 2008). In addition, the prevalence of a first dispensing of drug-drug interactions in people older than age 70 has been reported to have increased from 10.5% in 1992 to 19.2% in 2005 (Becker et al., 2008). Therefore, a thorough understanding of how drug-drug interactions occur and how they relate to ADRs should help decrease the rate of occurrence and the associated morbidity/mortality. This chapter discusses the mechanisms of drug interactions and their potential consequences. For the purpose of this chapter, these interactions are broken down into four major categories: drug-drug interactions, drug-food interactions, drug-herb interactions, and drug-disease interactions. Each of the interaction categories can affect the drug’s pharmacokinetic or pharmacodynamic profile. The definition, identification, and management of ADRs are discussed at the end of the chapter.
DRUG-DRUG INTERACTIONS
When a person takes two or more medications concomitantly, the potential exists for one or more drugs to change the effect of other drugs. The drug whose effect is altered by another drug is termed the object or target drug. Although minor interactions between drugs probably occur frequently, these interactions may not be significant enough to alter the effect of either drug. However, it is important for the practitioner to understand the mechanisms behind these interactions to predict more accurately when clinically significant (and potentially fatal) drug interactions may occur.
Pharmacokinetic Interactions
Absorption
Because most medications in the ambulatory care setting are administered orally, this route is the focus of discussion. For a drug to exert its effect, it must reach its site of action. Normally, this requires access to the bloodstream. As discussed in Chapter 2, drugs administered orally must be absorbed into the portal vein, through the intestinal wall, to reach the systemic circulation. The oral tablet must dissolve in the gastrointestinal (GI) tract before it can penetrate the intestinal wall.
Acidity (pH)
For some drugs, this process depends on the acidity in the GI tract. Therefore, if a drug that alters the gastric pH is administered concomitantly with a drug that depends on a normal gastric pH for dissolution, the absorption of the object drug will be affected. An example of this type of interaction is the concurrent administration of a histamine-2 (H2) receptor antagonist (e.g., ranitidine) and ketoconazole, an imidazole antifungal agent. Ketoconazole is the object drug that requires an acidic pH for absorption. When ranitidine is administered along with ketoconazole, the increase in gastric pH hinders the dissolution of ketoconazole and, therefore, decreases its absorption. Similarly, this change in pH can increase the absorption of other drugs that require a more alkaline environment for absorption.
Adsorption
Another mechanism of absorptive drug-drug interactions is adsorption. Adsorption occurs when one agent binds the other to its surface to form a complex. The most common agents associated with this type of interaction are divalent and trivalent cations (Mg2+, Ca2+, Al3+, found in antacids and some vitamin preparations) and anionic-binding resins (colestipol, cholestyramine). This type of interaction occurs when certain medications such as tetracyclines or fluoroquinolones are given with antacids. The metal ions in the antacid chelate (form a complex with) the antibiotic, preventing absorption of both components (ion and antibiotic). Adsorbents can interact with a variety of drugs; therefore, appropriate intervals between doses of the interacting medications are warranted. In general, with agents known to interact in this manner, the object drug should be administered at least 2 hours before or 4 to 6 hours after the interacting agent.
Gastrointestinal Motility and Rate of Absorption
Drugs that affect the motility of the GI tract produce a less common absorption-altering mechanism. These agents tend to affect the rate of absorption and not the amount of drug absorbed. Any agent—for example, metoclopramide—that stimulates peristalsis and increases gastric-emptying time can affect the rate of absorption of other medications. In most cases, an increase in the rate of absorption occurs because the target drug reaches the duodenum faster, allowing absorption to occur sooner. However, in some cases such as with metoclopramide and digoxin, a decrease in digoxin concentrations may occur (American Society of Health-System Pharmacists, 2008).
Conversely, anticholinergic agents and opiates decrease gastric motility, thereby decreasing the rate of absorption of object drugs. Because this type of interaction does not usually affect the amount of drug absorbed, it is usually clinically insignificant.
GI Flora and Absorption
The bacteria present in the GI tract are also responsible for a portion of the metabolism of some agents. An example of this is digoxin; concomitant administration with antibiotics (such as erythromycin or tetracycline) may alter the normal bacterial flora and reduce digoxin metabolism, thereby increasing bioavailability and serum concentrations in some patients (Susla, 2005). To the contrary, GI bacteria produce enzymes that deconjugate inactive unabsorbable ethinyl estradiol metabolites of oral contraceptives that have been excreted into the GI tract via the bile. Deconjugation allows reabsorption of active ethinyl estradiol back into the bloodstream. By disrupting the GI flora, anti-infectives may decrease or eliminate reabsorption of active ethinyl estradiol, thereby decreasing plasma concentrations and the effectiveness of oral contraceptives (Weaver & Glasier, 1999). However, in a case-crossover study of 1,330 failure cases, Toh et al. (2011) did not find an association between concomitant antibiotic use and the risk of breakthrough pregnancy among combined oral contraceptive users.
Table 3.1 summarizes some of the major drug interactions that occur in the absorptive process.
Distribution
After drugs are absorbed into the bloodstream, most of them, to some degree, are bound to plasma protein such as albumin or α1-acid glycoprotein. As described in Chapter 2, only an unbound drug is free to interact with its target receptor site and is therefore active. The percentage of drug that binds to plasma proteins depends on the affinity of that drug for the protein-binding site. If two drugs with high affinity for circulating proteins are administered together, they may compete for a single binding site on the protein. In fact, one drug may displace the other from the binding site with the result being an increase in the unbound (free) fraction of the displaced drug. This increase in free drug may trigger an exaggerated pharmacodynamic response or toxic reaction. However, because the excess unbound drug is now subject to elimination processes, the increases in both free drug fraction and the effects produced are usually transient.
Clinically significant drug displacement interactions normally occur only when drugs are more than 90% protein bound and have a narrow therapeutic index. For example, warfarin is 99% protein bound, and therefore, only 1% of the drug in the bloodstream is free to induce a pharmacodynamic response (inhibition of clotting factors). If a second drug is administered that displaces even 1% of the warfarin bound to albumin, the amount of free warfarin is doubled, to 2% free.
This can result in a significant increase in its pharmacodynamic action, leading to excessive bleeding. Table 3.2 lists examples of several displacement interactions.
This can result in a significant increase in its pharmacodynamic action, leading to excessive bleeding. Table 3.2 lists examples of several displacement interactions.
TABLE 3.1 Drugs Affecting Absorption | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Metabolism
Lipophilicity (fat solubility) enables drug molecules to be absorbed and reach their site of action. However, lipophilic drugs are difficult for the body to excrete. Therefore, they must be transformed by the body to more hydrophilic (water-soluble) molecules. This is accomplished primarily through phase I, or oxidation, reactions. The main sites of metabolism in the body are the liver (hepatocytes) and small intestine (enterocytes). Other tissues, such as the kidneys, lungs, and brain, play a minor role in the metabolism of drug molecules (Michalets, 1998). These sites of metabolism contain enzymes called cytochrome P-450 isoenzymes. This group of isoenzymes has been identified as the major catalyst of phase I metabolic reactions in humans.
The nomenclature of the cytochrome P-450 system classifies the isoenzymes (designated CYP) according to family (greater than 36% homology in amino acid sequence), subfamily (77% homology), and individual gene (Brosen, 1990; Guengerich, 1994; Nebert et al., 1987). For example, the isoenzyme CYP3A4 belongs to family 3, subfamily A, and gene 4. As one moves down the classification system from family to gene, the structures of the isoenzymes become more similar.
TABLE 3.2 Distribution Drug Interactions | |||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| |||||||||||||||||||||||||||||
This enzyme system has evolved to form new isoenzymes that metabolize foreign substrates (i.e., drugs) that are presented to the body. These enzymes are structured to recognize and bind to molecular entities on substrates. Many different substrates may have molecular structures that differ only slightly; therefore, an isoenzyme can bind to any one of these substrates. Although several different substrates may compete for the same enzyme receptor, the substrate with the highest affinity binds most often. The converse of this is also true. Two isoenzymes can bind to the same substrate (Figure 3.1), but the substrate binds more often to the isoenzyme to which it has the most affinity. However, not every drug molecule (“substrate”) can be metabolized by every enzyme with which it binds; therefore, it is not a true substrate. These concepts form the backbone for the drug interactions that are expanded on later.
Six isoenzymes have been determined to be responsible for most metabolism-related drug interactions. They are the isoforms CYP1A2, CYP2C9, CYP2C19, CYP2D6, CYP2E1, and CYP3A4. The CYP3A4 isoform is responsible for 40% to 45% of drug metabolism, the CYP2D6 for the next 20% to 30%, CYP2C9 about 10%, and CYP2E1 and CYP1A2 each responsible for about 5% (Ingelman-Sundberg, 2004). The remaining 5% to 20% is accounted for by several lesser important isoforms. Because there are so few enzymes that transform a multitude of substrates, it is easy to see how there would be a great potential for interactions.
 FIGURE 3.1 Substrate binding. A. Different substrates. Although Enzyme X (Ex) can bind to both Substrate 1 (S1) and Substrate 2 (S2), S2 has greater affinity for Ex than S1. Therefore Ex will bind to S2 most often. B. Different enzymes. Although Substrate X (Sx) can bind to both Enzyme 1 (E1) and Enzyme 2 (E2), E1 has a greater affinity for Sx than E2. Therefore, Sx will bind to E1 most often. |
There are some genetic variations with respect to the distribution of the enzymes. For example, about 10% of Europeans lack the CYP2D6 enzyme and are therefore considered poor metabolizers of drugs using this pathway for biotransformation. These individuals are at greater risk for ADRs related to drugs metabolized by CYP2D6. In addition, prodrugs requiring this enzyme for activation (e.g., codeine, tamoxifen) may be less effective or have no effect. In contrast, about 5% of this population are considered ultrametabolizers, have too rapid metabolism, and may show little to no response related to drugs metabolized by the CYP2D6 pathway (Ingelman-Sundberg, 2004). Similarly, there is variability within the CYP2C19 isoform, with about 14% of Chinese, 2% of Whites, and 4% of Blacks being poor metabolizers (Scott et al., 2011). The effectiveness of certain prodrugs (e.g., clopidogrel) that require metabolic activation by this enzyme system may be reduced (Holmes et al., 2010).
There has been increasing interest in genetic testing to identify strategies to reduce the risk of ADRs and to optimize therapy for individuals. Pharmacogenomic information has been incorporated into about 10% of labels for drugs approved by the U.S. Food and Drug Administration (FDA) in an effort to identify responders and nonresponders, avoid toxicity, and adjust doses of medications to optimize efficacy and ensure safety (FDA Table of Pharmacogenomic Biomarkers in Drug Labeling accessed July 16, 2015). In addition, regulatory authorities have recently recommended genetic testing to aid the clinician in determining if an agent is safe and effective in certain individuals (e.g., abacavir) (Highlights of Prescribing Information: Ziagen (abacavir sulfate) Tablets and Oral Solution accessed July 16, 2015). While commercial assays are available for genetic testing, turnaround time for the results varies, testing can be expensive, data on the validation of techniques used and reliability and reproducibility are limited, and, to date, there is limited evidence-based data on which to develop specific recommendations on the role of genetic testing in routine care (Holmes et al., 2010).
There are two types of metabolic drug interactions: drugs that inhibit the action of an enzyme and those that induce the activity of the enzyme.
Inhibition
Inhibition of drug metabolism occurs through competitive and noncompetitive inhibition. When two drugs, administered concurrently, are metabolized by the same isoenzyme, they are defined as competitive inhibitors of each other. In essence, they compete for the same binding site on an enzyme to be metabolized.
Noncompetitive inhibition also occurs when both drugs compete for the same binding site, but one drug is metabolized by that isoenzyme and the other drug is not. The best known example of a noncompetitive inhibitor is quinidine. Quinidine
is metabolized by the CYP3A4 isoenzyme but can also bind to the CYP2D6 enzyme. Therefore, although quinidine does not compete for metabolism by the CYP2D6 isoenzyme, it does compete for the CYP2D6 isoenzyme binding site.
is metabolized by the CYP3A4 isoenzyme but can also bind to the CYP2D6 enzyme. Therefore, although quinidine does not compete for metabolism by the CYP2D6 isoenzyme, it does compete for the CYP2D6 isoenzyme binding site.
In both competitive and noncompetitive inhibition, the drug with the greatest affinity for the isoenzyme receptor is usually the inhibiting drug because it binds in the receptor site, preventing the other drug from being bound and metabolized (Figure 3.2). The significance of the drug interaction depends on several characteristics of the inhibiting drug.
Affinity The first characteristic, affinity, has already been mentioned. Many drugs may inhibit the same isoenzyme but not to the same extent. The greater the affinity of an inhibiting drug for an enzyme, the more it blocks binding of other drug molecules.
Half-Life Along with affinity, the half-life (t½) of the inhibiting drug determines the duration of the interaction. The longer the half-life of the inhibiting drug, the longer the drug interaction lasts. For example, after a regimen of ketoconazole (t½ = 8 hours) is discontinued, its ability to inhibit the CYP3A4 enzyme lasts until it is eliminated, in three to five half-lives or approximately 1 day. However, the inhibiting effect of amiodarone, with a t½ of approximately 53 days, lasts for weeks to months after its discontinuation.
Concentration The third major factor contributing to a drug’s ability to inhibit hepatic enzymes is the concentration of the inhibiting drug. A threshold concentration must be reached or exceeded to inhibit an enzyme. This is similar to the threshold concentration discussed in Chapter 2 regarding minimally effective concentrations and therapeutic responses. This minimally effective threshold concentration, or concentration-dependent inhibition, is exhibited by a variety of drugs. The dose yielding this concentration-dependent inhibition varies based on volume of distribution, drug and receptor affinity, and characteristics of the individual patient. An example of a dose- or concentration-dependent inhibitor is cimetidine. In most patients, a dose of 400 mg/d results in only weak enzyme inhibition. However, at higher doses, it interacts significantly with both the CYP2D6 and CYP1A2 isoenzymes (Shinn, 1992).
 FIGURE 3.2 Inhibition. A. Competitive inhibition. Drug X (DX) and Drug Y (DY) are both metabolized by Enzyme 2 (E2). B. Noncompetitive inhibition. Although DX and DY compete for the binding site on E2, only DY is metabolized by E2. Therefore, DX noncompetitively inhibits DY. |
Some enzyme inhibitors may affect one enzyme at a smaller concentration and more than one isoenzyme at higher concentrations. These enzyme inhibitors demonstrate that some isoenzymes have differing thresholds. For example, fluconazole at a dose of 200 mg/d significantly inhibits only the 2C9 isoenzyme, but as the dose increases above 400 mg/d, it also inhibits the 3A4 isoenzyme (Hansten & Horn, 2015; Kivisto et al., 1994).
Toxic Potential Another consideration with regard to inhibition interactions is the toxic potential of the object drug. For example, nonsedating antihistamines (i.e., terfenadine and astemizole) are metabolized by the CYP3A4 isoenzyme to a nontoxic metabolite. The parent compound of both drugs is cardiotoxic. If a potent CYP3A4 inhibitor (e.g., erythromycin) is administered concurrently with these agents, the parent compound accumulates in the body and causes a potentially fatal arrhythmia, such as torsades de pointes (Mathews et al., 1991; Woosley, 1996). Because of their serious toxic potential, both terfenadine and astemizole have been removed from the market.
Efficacy An additional consideration related to inhibition interactions is the effectiveness of the object drug. This is particularly important for prodrugs that require cytochrome P-450 metabolism to the active metabolite in order for the drug to be effective. An example of this is clopidogrel, an antiplatelet agent, which requires CYP2C19 enzymes to be metabolized to the active form. When administered with omeprazole, a CYP2C19 inhibitor, a reduction in plasma concentrations of the active metabolite of clopidogrel as well as reduced platelet function occur (clopidogrel [Plavix] prescribing information). In addition, tamoxifen, an agent used for breast cancer, is converted to its active metabolite by CYP2D6. Women may have a higher risk of breast cancer recurrence if they take tamoxifen in combination with CYP2D6 inhibitors such as the serotonin reuptake inhibitors paroxetine, fluoxetine, or sertraline (The Medical Letter, 2009).
Not all inhibition reactions result in harmful effects, however. Some interactions may be inconsequential or even beneficial. For example, ketoconazole (a potent inhibitor of the CYP3A4 isoenzyme) can be given with cyclosporine. The consequent interaction enables practitioners to give less cyclosporine to achieve the same immunosuppressive response (Hansten & Horn, 2015).
The cytochrome P-450 system is complex, but an understanding of the basic concepts of inhibitory interactions leads to the ability to anticipate which agents are likely to interact. The affinity, half-life, and drug concentration determine the potency of the inhibiting drug. A potent enzyme inhibitor inhibits most drugs metabolized by that enzyme. A clinically significant drug interaction also depends on the toxic potential of the drug being inhibited. Table 3.3 lists several enzyme inhibitors.
TABLE 3.3 Drugs Affecting Metabolism through Cytochrome P-450 Isoenzymes | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Induction
Drug-drug interactions can also result from the action of one drug (inducer) stimulating the metabolism of an object drug (substrate). This enhanced metabolism is thought to be produced by an increase in hepatic blood flow or an increase in the formation of hepatic enzymes. This process, known as enzyme induction, increases the amount of enzymes available to metabolize drug molecules, thereby decreasing the concentration and pharmacodynamic effect of the object drug.
Some of the more common enzyme inducers are rifampin, phenobarbital, phenytoin, and carbamazepine. Enzyme induction, like enzyme inhibition, is substrate dependent. Therefore, any drug that is a potent inducer of a cytochrome P-450 system increases the metabolism of most drugs metabolized by that enzyme. Also, in a manner similar to that of enzyme inhibitors, inducers may affect more than one cytochrome P-450 isoform; for example, phenobarbital is a potent inducer of the CYP3A4, CYP1A2, and CYP2C isoforms.
Some enzyme inducers, such as carbamazepine, also increase their own metabolism. Over time, carbamazepine stimulates its own metabolism, thereby decreasing its half-life and frequently resulting in an increased dose requirement to maintain the same therapeutic drug level. This process is termed autoinduction.
The onset and duration/cessation of enzyme induction depend on both the half-life of the inducer and the half-life of the isoenzyme that is being stimulated. For example, rifampin (t½ = 3 to 4 hours) results in enzyme induction within 24 hours, whereas the enzyme induction capacity of phenobarbital (t½ = 53 to 140 hours) is not evident for approximately 7 days. The level of induction remains constant while the drugs are being administered. However, on discontinuation of the respective inducers, the inducing action of rifampin ends more rapidly because of its shorter half-life. This occurs because rifampin is removed from the body at a faster rate than phenobarbital and therefore is not available to inhibit hepatic enzymes for as long.
The initiation and duration of enzyme induction also depend on the half-life of the induced isoenzyme. It takes anywhere from 1 to 6 days for a cytochrome P-450 enzyme to be degraded or produced. Therefore, even if a drug achieves a high enough concentration to produce induction of liver enzymes, the increase in metabolism of an object drug may not be evident until more liver enzymes have formed. The effect of rifampin on warfarin metabolism is a good example of this. Although induction begins within 24 hours of rifampin administration, its effect on warfarin metabolism is not evident for approximately 4 days. On discontinuation of rifampin, the remaining drug is metabolized to negligible levels before the effect on warfarin metabolism dissipates. This occurs because the half-life of the liver enzymes is greater than the half-life of rifampin, and therefore, the enzymes remain to metabolize warfarin after rifampin is eliminated from the body.
These concepts are important to remember when monitoring laboratory values that demonstrate the effectiveness of the object medication. For example, the international normalized ratio (INR), which is a surrogate marker of warfarin levels, fluctuates significantly within a couple of days of the initiation or discontinuation of rifampin. However, no change in the INR is evident for approximately 1 week after administration of phenobarbital. This is also true when measuring levels of certain antibiotics and other agents that may be affected by inducers. Table 3.3 lists several enzyme inducers and the drugs they affect.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
