Basic Principles of Drug Interactions
Michael A. Mancano, PharmD
Dr. Mancano is Associate Professor of Clinical Pharmacy at Temple University School of Pharmacy in Philadelphia,
PA.
The topic of drug interactions is dynamic
and varied in that there are many potential mechanisms by which drugs can interact and new information is always forthcoming in the medical literature. In addition, with the large number of new medications approved by the FDA on a yearly basis, the pharmacist faces a daunting task to remain current with the new information concerning drug interactions.
Currently, the cytochrome P (CYP) 450 enzyme inhibition interactions have attracted much attention because of their prevalence and severity; however, drugs can interact via many mechanisms.
Table 1 contains a list of the many possible mechanisms by which drugs can interact. The list is extensive and represents a challenge for pharmacists in screening and interpreting the myriad drug interaction reports and computer screening program red flags seen in daily practice.
Table 1. Drug Interaction
Mechanisms
While specific knowledge of serious drug interactions is important when screening a patient’s drug profile, it is important for the pharmacist to have a solid grasp on the general mechanisms by which drugs can interact. In this way, pharmacists can anticipate potential drug interactions as new agents are introduced into the market. We can then classify new agents by their potential interaction properties and similarities to agents that have known drug interaction problems.
In a broad sense, drug interactions can be separated into two basic types. The first way drugs can interact is via a pharmacokinetic
mechanism.1 The study of the pharmacokinetics of a drug looks at the absorption, distribution, metabolism, and excretion properties of the drug in question. Drugs can be said to interact by a pharmacokinetic mechanism if the interaction between the two drugs affects the absorption, distribution, metabolism, or excretion of one of the drugs.
Conversely, drugs can interact via a pharmacodynamic mechanism. The detection of pharmacodynamic interactions requires a knowledge of the pharmacology of the drugs involved. For example,
.drugs can interact by producing antagonistic effects on each other. That would be considered a pharmacodynamic interaction. Additionally, drugs can interact by causing synergistic adverse effects or by producing indirect pharmacodynamic effects. Specific examples of these interactions will be included later in the article.
Concerning absorption interactions, drugs can interact by altering the rate or extent of the absorption of another medication. If the rate of absorption of a medication is decreased, typically, the ultimate steady-state concentration of the drug in question is not decreased. However, the slower rate of absorption may be clinically significant if the desired therapeutic effect of the medication is needed rapidly. For example, by slowing the rate of absorption of a medication used as a sleep aid, the therapeutic effect may be delayed and the patient may not experience the desired therapeutic effect as usually encountered.
Second, if two medications interact and the extent of absorption of one of the medications is decreased, the ultimate steady-state concentration of the medication may be decreased. This decrease in steady-state concentration may in fact be clinically significant. For example, if Drug A decreases the extent of absorption of Drug B by 25%, the ultimate steady-state level of Drug B will be decreased. This may lead to decreased therapeutic effects of Drug B and the need for an upward titration of the dose of Drug B.
Absorption interactions typically occur in the small intestine. This makes logical sense because of the fact that the small intestine is the primary site from which orally administered medications are absorbed. The intestine is the primary site of absorption due to its very large absorptive area, increased permeability to drugs, and the large amount of blood flow through the intestinal capillaries, which allows for the absorbed drug to be transported to the systemic
circulation.2,3
Drug binding interactions are a common type of absorption interaction that pharmacists can help to prevent. Drugs that pharmacists should think about as potential interacting drugs are antacids, iron, cholestyramine, and colestipol. These agents possess a large surface area to bind drugs to themselves. This binding prevents the medication in question from passing through the intestine and ultimately reaching the systemic circulation.
An example of this type of interaction occurs when cholestyramine is concomitantly administered with levothyroxine. Cholestyramine can potentially decrease the efficacy of levothyroxine and possibly induce hypothyroidism. The mechanism of this interaction is most likely due to the binding of levothyroxine in the GI tract by cholestyramine. Cholestyramine can bind both thyroxine and triiodothyronine in the intestine.
The usual recommendation that we make to patients to prevent a binding interaction is to separate the drugs by approximately 2 hours. This recommendation is appropriate in most situations, however, it will not minimize the cholestyramine-induced binding of levothyroxine. Due to the enterohepatic recirculation of levothyroxine, cholestyramine has a longer duration of time to bind with levothyroxine. Therefore, the appropriate recommendation is to instruct patients to separate the administration of these medications by approximately 4 hours at a minimum. Some recommendations state that administration should be separated by up to 6 hours.
While I have discussed concomitant cholestyramine as a potential binding problem for levothyroxine therapy, it would be prudent to also be aware of concomitant colestipol administration.
The response to a medication can be altered because of a second medication changing the GI motility of the
patient.4 As mentioned earlier, the small intestine is the primary site of absorption for orally administered medications. If the GI transit time is increased or decreased, the time available for absorption can be limited or maximized.
Medications such as anticholinergic agents and opiates can slow GI transit time. This slowing of transit time can allow for a second medication to be absorbed to a greater extent. This increased extent of absorption can lead to a clinically significant increase in pharmacologic effect.
Certain drugs can be expected to decrease GI transit time
(ie, metoclopramide, erythromycin, and laxatives). This decrease in transit time may carry a second medication through the GI tract at a faster rate than is ordinarily encountered, and it may not allow a second medication to be absorbed to its fullest extent. This decrease in the extent of absorption may lead to a decrease in the second drug’s pharmacologic effect.
This type of interaction usually cannot be avoided by separating the
medications by several hours. Typically, this type of interaction can be resolved by titrating the second medication to its appropriate therapeutic effect.
Many patients are currently receiving therapy with medications that alter the GI pH, therefore, it is prudent to be aware of the potential type of interaction this pH alteration can cause. The use of histamine-2 blockers, proton pump inhibitors, and antacids is common among today’s patients.
To understand the mechanism of the interactions we must understand that many of the currently available medications are weak acids or weak bases. The nonionized form of these medications can cross intestinal membranes and reach the systemic circulation. However, specifically speaking, drugs that are classified as weak acids are absorbed to a greater extent in an acid medium. Conversely, drugs that are classified as weak bases are better absorbed in a basic medium.
An example of a medication that can exhibit a GI pH alteration interaction is
ketoconazole.5 Ketoconazole is a weak acid and therefore requires an acidic environment to be absorbed. If a patient were to initiate therapy with a medication that increases the pH of the stomach
(eg, ranitidine), the amount of ketoconazole absorbed will be decreased. This decrease in absorption could lead to a treatment failure associated with ketoconazole therapy.
Over the years, a number of case reports have appeared in the medical literature reporting contraceptive failure in patients receiving concurrent
ampicillin6 or tetracycline7 with oral contraceptives. In addition, a variety of antibiotics have been reported to cause oral contraceptive failure. The suspected mechanism of interaction is that the antibiotic may suppress the intestinal flora that provide hydrolytic enzymes essential for enterohepatic recirculation of necessary contraceptive hormone conjugates.
Many women may receive an antibiotic and an oral contraceptive product concomitantly and not experience this interaction; however it is not possible to predict which patients will be affected. In light of the unpredictability of this interaction, it is prudent to counsel your female patients receiving oral contraceptives to use supplementary contraception while receiving antibiotic therapy.
In light of the unpredictability of this type of interaction, the recommendation to adjust the dosing times of these medications would not be expected to prevent this interaction. The effect on GI flora occurs gradually and would be expected to correct itself gradually with the regeneration of intestinal flora.
A number of different types of metabolism are known to occur in the wall of the intestine. A partial list of the types of metabolism includes glucuronidation, sulfation, P-450 oxidation, and monoamine
oxidation.1
The simplest type of example to discuss concerning this type of interaction is the ingestion of a meal high in tyramine by a patient receiving a monoamine oxidase inhibitor. In this situation, the monoamine oxidation that is to occur in the wall of the intestine will be inhibited by the monoamine oxidase inhibitor. Therefore, the protective properties of monoamine oxidase in the intestine will be inhibited and the tyramine will reach the systemic circulation, potentially leading to a hypertensive response.
Protein binding is a dynamic process in which there is an equilibrium between bound and unbound drug. The binding of drugs to proteins can be affected by the addition of other medications, specific diseases, and also the accumulation of endogenous substances. Drugs bound to serum proteins are physically inactive, with only the unbound drug being available to reach its site of action. The bound portion of the drug can be thought of as a reservoir to be released when needed. The primary serum protein related to drug binding is albumin. However, other major serum proteins are alpha-1 acid glycoprotein, lipoprotein, and
transcortin.8
The classic drug interaction that has been used to describe this type of interaction is a patient stabilized on warfarin who then initiates aspirin
therapy.9 While warfarin is highly protein-bound, initially the mechanism of this interaction was thought to be primarily due to the displacement of warfarin from albumin. Although aspirin increases the risk of bleeding in patients receiving warfarin, this primarily occurs by inhibition of platelet function by aspirin and possibly by producing gastric
erosion.10 The ability of the highly protein-bound aspirin to displace warfarin from plasma protein binding sites may play a minor role in the interaction mechanism.
While small doses of aspirin can interact with warfarin, larger doses are likely to have an intrinsic hypoprothrombinemic effect, thus enhancing the anticoagulant effect of warfarin. The risk of clinically important bleeding is increased when high doses of aspirin are used in combination with high-intensity warfarin therapy (ie, international normalized ratio [INR] target of 3.0 to 4.5).
A key point to remember when considering protein-binding displacement drug interactions is that these interactions tend to self-correct with time. This occurs due to the fact that, although the unbound portion of the drug exerts an increased pharmacologic effect, the unbound portion of the drug is also more readily eliminated because it is now unbound. Typically, these types of interactions manifest quickly, and if they have not occurred by the second week of concomitant therapy, an adverse effect will probably not occur.
Most drug renal excretion interactions primarily deal with the inhibition of the clearance of one medication by another. The inhibition of clearance leads to an accumulation of one or both of these medications and potentially adverse effects.
The kidney has three primary methods of eliminating drugs and their metabolites from the body: glomerular filtration, active tubular secretion, and passive tubular
reabsorption.11 Glomerular filtration is primarily affected by changes in filtration pressure. Filtration pressure can be affected by changes in systemic blood pressure or glomerular hydrostatic pressure. The simplest example of this type of interaction is the administration of a medication that causes nephrotoxicity in a patient (ie, an aminoglycoside). If the patient is also receiving digoxin, which is excreted by glomerular filtration, then the nephrotoxicity caused by the aminoglycoside will cause an accumulation of digoxin. Digoxin toxicity can potentially occur if the nephrotoxicity is not detected and appropriate dosage adjustments are not initiated.
Many medications that undergo glomerular filtration may also undergo active tubular secretion. Active tubular secretion is a process in which carrier proteins actively transport drugs through the basolateral and brush border membranes of the kidney. Anytime an active transport mechanism is employed, there is a possibility of saturation of this process. A saturation of this process by two drugs competing for the same carrier proteins can lead to accumulation of one or both of these medications. An example of this type of interaction is the interaction between probenecid and
penicillin.12 Probenecid blocks the excretion of penicillin, thereby increasing penicillin’s duration of effect in the body. We exploit this interaction therapeutically to achieve increased therapeutic effects in the treatment of certain types of infections.
The process of passive tubular reabsorption primarily focuses on the concentration of drug, lipid solubility, and pH of the urine. Non-ionized drugs in the tubular fluid are preferentially reabsorbed over ionized drugs. Simply, in an acidic urine, weakly acidic drugs tend to be reabsorbed while basic drugs are excreted in the urine. In an alkaline urine, basic drugs will be reabsorbed and acidic drugs will be excreted in the urine.
Although interactions can occur via interference with passive tubular reabsorption, this is not a common mechanism of interaction. However, the alkalinization or acidification of the urine is utilized therapeutically in various toxicologic emergencies to enhance the elimination of ingested toxins or drugs.
As discussed earlier, drugs may be excreted into the GI tract and undergo enterohepatic reabsorption. An example of this type of interaction, as cited earlier, is the interaction between levothyroxine and cholestyramine due to enterohepatic recirculation of levothyroxine.
It is essential for pharmacists to understand the issues surrounding cytochrome CYP450 interactions. Because of the large volume of information published concerning this issue, it is easy to be overwhelmed by the data.
CYP450 enzymes are a group of heme-containing enzymes located on the membrane of the smooth endoplasmic reticulum of hepatocytes in the liver and in high concentrations on enterocytes of the small intestine. CYP450 enzymes are also located in small amounts in other tissues, specifically the kidney, lungs, and brain. The CYP450 name is derived from analysis of these enzymes on a spectrophotometer. The spectral absorbancy peak at or near 450 nm when carbon monoxide binds to the isoenzyme in its reduced
state.13
The CYP450 system includes more than 30
isoenzymes. The isoenzymes are grouped into families and subfamilies according to their amino acid sequence. The current method of naming CYP450 enzymes involves the designation of the specific enzyme with the prefix “CYP,” followed by an Arabic number indicating the P-450 family, followed by a capital letter denoting the subfamily, and then an Arabic numeral representing the individual enzyme. For example, the CYP3A4 isoenzyme by its designation indicates that it is a member of family “3,” subfamily “A,” and it is the fourth enzyme in subfamily “A.” It is generally agreed that members of a CYP450 family have
greater than or equal to 40% amino acid sequence homology. Likewise, cytochrome enzymes within the same subfamily have >55% amino acid sequence
homology.14
CYP450 enzymes are important in the metabolism of endogenous substances, such as steroids, hormones, prostaglandins, lipids, and fatty acids. They are also important in the detoxification of exogenous compounds, such as drugs, especially after oral
ingestion.15 Basically, CYP450 enzymes are important in the oxidation, reduction, and hydrolysis reactions that make a drug more water-soluble and more readily excreted in the urine or bile. The majority of CYP450 enzymes involved in drug metabolism appear to belong to three distinct families, CYP1, CYP2, and CYP3.
Knowledge of the substrates, inhibitors, and inducers of CYP450 enzymes is essential and can assist in the prediction and prevention of clinically significant drug interactions. A substrate is defined as a substance upon which an enzyme acts. Many drugs use CYP450 enzymes as substrates for their metabolism. Drugs can inhibit or induce metabolism of other drugs, and they can also compete for metabolism for the available CYP450
isoenzymes. A list of inducers, substrates, and inhibitors of the major CYP450 isoenzymes is included in
Table 2.
Table 2. CYP450 Inducers, Substrates, and Inhibitors
To further complicate the issue, some drugs may be substrates for more than one CYP450
isoenzyme. When one enzyme system is inhibited or induced by an interacting drug, a clinically significant interaction may or may not occur. Inhibition or genetic absence of one isoenzyme can lead to compensation through the secondary isoenzymes’ metabolic pathway. Therefore, metabolism can be preserved, and a clinically significant interaction may or may not occur. Additionally, a drug may inhibit or induce the activity of a specific isoenzyme although it is not a substrate of that specific
isoenzyme.
Enzyme induction can be considered an adaptive process due to the body’s perception that there is an overabundance of lipid-soluble compounds that need to be eliminated from the body. Enzyme induction interactions occur when an enzyme-inducing drug is administered and it gradually begins to stimulate the synthesis of additional CYP450
isoenzymes. This gradual increase in isoenzymes leads to an increase in metabolism of substrate drugs primarily metabolized by the induced
isoenzyme.
Enzyme induction can be detected within the first 2 days of drug administration; however, it generally takes a week or more before maximal enzyme induction occurs. The onset of induction depends on the half-life of the inducing agent, with longer half-life inducers having a slower onset of
induction.16 An example of this is
phenobarbital, which may take up to 1 month to reach maximal induction. However, rifampin tends to have a more rapid onset of enzyme induction with maximal induction occurring after about 6 to 10 days. As a guide,
Table 3 contains a list of commonly encountered enzyme-inducing drugs.
Table 3. Commonly
Encountered Enzyme Inducing Drugs
Amiodarone is an example of a medication that inhibits the clearance of
S-warfarin and
R-warfarin.20 Warfarin’s stereoisomers differ in that S-warfarin is approximately five times more potent than its stereoisomer
R-warfarin. S-warfarin is primarily metabolized via CYP2C9, while R-warfarin is partially metabolized by CYP1A2 and CYP3A4. Amiodarone primarily reduces the metabolism of warfarin via inhibition of CYP2C9, but it also has inhibitory effects on CYP3A4. Because of this dual inhibition of warfarin’s
stereoisomers, amiodarone can enhance the anticoagulant effect of warfarin by 50% to 100%, while reducing warfarin’s clearance by 35% to 65%. This interaction may be concentration-dependent.
Following the initiation of amiodarone therapy to a patient receiving warfarin therapy, the increased anticoagulant effect of warfarin can begin in approximately 1 week. This interaction stabilizes approximately 1 month after initiation of concomitant therapy. However, because of amiodarone’s long half-life, this interaction may persist for several months after amiodarone is discontinued.
In a patient who is to receive amiodarone therapy and who has been receiving
warfarin, a pharmacist can make a significant intervention to minimize the patient’s risk of bleeding complications. Typically, a 30% to 50% reduction in the warfarin dose is required when amiodarone is initiated. We can also recommend monitoring of the patient’s INR closely for the first 2 to 4 weeks of amiodarone therapy, and the dose of warfarin should be adjusted accordingly.
If amiodarone is to be discontinued in a patient who was stabilized on the combination of warfarin and
amiodarone, additional monitoring is also indicated in this situation. The onset and offset of this interaction is delayed in some patients due to amiodarone’s long half-life. Therefore, in this situation, close monitoring should continue for several months following the discontinuation of amiodarone with appropriate adjustments made to the patient’s warfarin dose.
Antagonistic drug interactions can occur when two drugs that have antagonistic pharmacologic properties are given concomitantly. Pharmacists can avoid these interactions by understanding the pharmacologic properties of the medications in question. A simple example of an antagonistic interaction is the administration of a nonspecific beta blocker and a systemic
beta2-agonist.21 The antagonistic pharmacologic properties of these two agents are evident and could lead to a decreased therapeutic effect of one or both of these agents.
An interaction that can be classified as causing synergistic adverse effects can occur when two medications with similar adverse-effect profiles are administered. An example of this type of interaction would be the administration of a benzodiazepine with concomitant ethanol
ingestion.22 The combination of these central nervous system depressant agents will lead to excess sedation. This interaction is dangerous and could lead to the death of the patient.
Indirect pharmacodynamic interactions occur when the pharmacologic effect of one drug affects another drug’s therapeutic effect. An example of this type of interaction would be the coadministration of a diuretic and
digoxin. Although this combination is commonly used in many patients, we must to monitor for diuretic-induced hypokalemia and
hypomagnesemia. Hypokalemia and hypomagnesemia may place a patient receiving digoxin at risk of
dysrhythmias.
The elderly are at increased risk for drug interactions due to expected age-related changes in renal and hepatic function. Interactions that perhaps would not be clinically significant in a younger patient may manifest with serious consequences in an elderly patient with decreased renal function. Decreased renal function may cause the accumulation of a drug or a drug metabolite and perhaps shorten the length of time to a drug interaction or prolong the adverse effect.
Patients of any age who see multiple prescribers may be at increased risk for experiencing drug interactions. Oftentimes, patients may not give a complete drug history to a physician who may prescribe an interacting drug or even duplicate therapy. In addition, patients who utilize various pharmacies may also be at increased risk because of possibly incomplete drug profile information at numerous locations. Since the pharmacist may be the last line of defense for patients such as these, it is imperative for pharmacists to take a complete medication history from their patients. Pharmacists should ask patients about the medications they are currently using, including seasonal medications, over-the-counter medications, and herbal medications, as well as drug allergies and past medications they have not
tolerated.
Pharmacists should be vigilant for patients with specific disease states who may experience drug interactions. Some disease states require treatments that may necessitate the use of drugs prone to interaction. High-risk disease states include: HIV/AIDS and the opportunistic infections associated with this disease; connective tissue disorders; GI disorders (ie, peptic ulcer disease and GI motility disorders); cardiovascular disorders (ie, hyperlipidemia); disorders requiring anticoagulation in the short and long term; chronic respiratory disorders (ie, asthma and chronic obstructive pulmonary disease); and seizure disorders.
In addition, other variables may affect drug interactions in patients, such as their specific rate of metabolism, which can be related to pharmacogenetic differences in patients or patient race. The patient’s diet may impact drug interactions
(ie, a patient taking vitamin K and receiving warfarin, or a patient drinking grapefruit juice with their medication). Since grapefruit juice can inhibit the intestinal CYP3A4
isoenzyme, medications such as felodipine and nifedipine, which require this isoenzyme for metabolism, may have their levels increased.
The reasons are numerous. Initially, a drug may be underdosed or initially administered in a lower than usual dose and slowly titrated to its full therapeutic effect. Perhaps that initial lower dosage did not allow the interaction to manifest quickly or perhaps allowed the precipitant and the interacting drug to equilibrate during the dosage titration. Additionally, a patient may only be on an agent for a short period of time, such as with antibiotics. However, if the interaction with the antibiotic is P-450 enzyme inhibition, the interaction may manifest within 24 hours of adding the offending agent to the patient’s drug regimen. Finally, the manifestation of the interaction may take a long time to reach full effect. For example, enzyme induction may take up to 1 month in certain scenarios to fully manifest.
Awareness of drug interactions is a key issue. Pharmacists must stay current with the ever-expanding information generated on this topic. Pharmacists should have a mental list of what I like to refer to as “Red Flag Drugs”
(Table 5) that they should look closely at when they receive an order to initiate or discontinue these agents for one of their patients. Pharmacists must anticipate potential interactions and may have to recommend the withdrawal or dosage adjustment of the precipitant drug. In addition, it may be necessary for the pharmacist to recommend appropriate alternative therapy or increased clinical monitoring with regard to the ordering of appropriate clinical lab tests or drug levels, if suitable. The pharmacist should be ready to make these types of recommendations to a physician who orders an interacting drug.
Detection and interpretation of interaction information is essential for pharmacists. Pharmacists can utilize the drug-interaction checking programs that most dispensing systems are equipped with; however, not all programs supply the pharmacist with additional information on how the reaction manifests, how to monitor therapy, or appropriate alternate therapy recommendations. For the pharmacist who wants to have additional
information at their fingertips, I recommend that they subscribe to one of the drug interaction texts currently available. These references are updated regularly and contain a wealth of information about the prevention and management of many drug interactions.
