Drug-drug interactions (Proceedings)

Clinically significant drug interactions are rarely reported in veterinary medicine, however the incidence is probably far greater than is reported.  With the introduction of more and more veterinary drugs, as well as the use of more human drugs in animals, the incidence is likely to increase in the next few years.  Additionally, the practice of ‘polypharmacy' or having animals on multiple medications causes a logarithmic increase in the risk of drug-drug interactions.  It is the veterinarian's responsibility to be aware of any potential drug interactions between drugs being used on veterinary patients, and to recognize the potential dangers involved with drug combinations. 

There are three categories of drug interactions that can occur.  Pharmaceutical, pharmacokinetic and pharmacodynamic interactions can all be present. These may result in harmful consequences, additive or synergistic beneficial effects, or inactivation of some drugs, resulting in therapeutic failure. The following presentation will define these categories of interactions, and present specific examples of drug interactions in veterinary medicine. 

Types of Drug Interactions

Pharmaceutical interactions

These interactions commonly occur due to mixing of two drugs with incompatible pH, and usually result in a visible precipitate.  Examples of this include meperidine and thiopental, which have a pH of 3.5 and 10.8, respectively.  In some instances, the interaction occurs not between two drugs, but between the drug and the container, or a drug and its vehicle, particularly with compounded drugs.  Diazepam and the highly lipophilic drug itraconazole have been known to adsorb to plastic or glass containers.  Drugs with strong chelating abilities can be inactivated if they are combined with vehicles containing cations.  Fluoroquinolones can be inactivated in solutions with calcium (such as lactated Ringer's solution) or iron (lixotinic).  However, physical inactivation can also occur for other reasons, and may or may not cause a visible precipitate.  A good example of this is the combination of aminoglycosides and penicillins in vitro.  Although no visible change occurs, the drugs become inactivated, which results in lower concentrations of active drugs, and potential therapeutic failure in the patient.  When these drugs are given in vivo, this interaction does not occur, however, as the drugs are sufficiently diluted in the patient's blood to prevent the interaction.  Pharmaceutical interactions can occur in vivo, however and this fact can be manipulated pharmacologically in the form of an antidote.  Protamine sulfate is an antidote for heparin toxicosis.  It works by combining with heparin in the body to form a stable, inactive salt formulation with heparin. 

Pharmacodynamic interactions

Pharmacodynamic interactions can occur in a variety of different ways.  They can cause synergism between two drugs, resulting in a greater than expected increase in the action of one or both drugs.  This occurs with combinations of β-lactam antibiotics and aminoglycosides, and sulfonamides and dihydrofolate reductase inhibitors.  Additive effects are reported using combinations of barbiturates and benzodiazepines in producing sedation/hypnosis, and combinations of opioids and NSAIDs for analgesia.  Pharmacodynamic effects are the basis of the use of reversal agents.  Good examples of this in veterinary medicine include atipamezole and metdetomidine, and opioids and naloxone. Pharmacodynamic interactions can also lead to increased toxicity, if both drugs adversely affect the same organ system.  For example, co-administration of NSAIDs and steroids increases the risk of gastrointestinal ulceration. Also, concurrent administration of 2 nephrotoxic drugs, such as an NSAID and an aminoglycoside, may increase the chance of nephrotoxicity.

Pharmacokinetic interactions

Pharmacokinetic drug interactions are common in humans and can be a result of changes in drug absorption, distribution, metabolism or excretion. 

Changes in drug absorption can occur following pharmaceutical interaction in the stomach or small intestine.  The classic example of this is tetracycline chelation in the stomach by calcium containing solutions, such as a milk diet in newborns.  Cation containing drugs or solutions, such as antacids and sucralfate, can also bind tetracyclines and fluoroquinolones.  Drug interactions can also occur with drugs that alter the pH of the stomach, when they are co-administered with drugs that have a pH dependent solubility.  For example, proton pump inhibitors have been shown to reduce the oral absorption of the azole antifungal drugs itraconazole and ketoconazole, by decreasing their solubility.  Some drugs also alter gastric emptying and intestinal motility, which may affect drug absorption by delaying delivery of the drug to the site of absorption, which is typically the proximal small intestine.  Opioid drugs and the anticholinergic drugs, such as atropine or butylscopolamine, alter gastric motility to the extent that orally administered drugs may exhibit delayed absorption.  Alteration of absorption may also occur following intramuscular or subcutaneous routes.  Epinephrine added to local anesthetics delays drug absorption from the injection site, resulting in prolonged effects.  Two inhalant gases administered together can alter the rate of uptake at the alveolar level.   

Changes in drug absorption can also be brought about by changes in drug uptake or metabolism in the intestine.  This is due to alterations in P-glycoprotein (P-gp) and cytochrome P450 (CYP450) enzymes in the small intestinal mucosa.  P-glycoprotein is a drug efflux protein found enterocytes in the small intestinal mucosa.  It acts as a pump, pumping drugs out of the cell and back into the intestinal lumen.  This typically results in a decrease in drug absorption. Drugs that inhibit P-gp may result in an increased absorption of other drugs.  Cyclosporine is a P-gp inhibitor that has been used in clinical situations combined with the anticancer drug docetaxel, to increase the absorption of this drug, making it possible to treat animals orally, rather than using the injectible formulation, which has been known to cause severe reactions due to the vehicle used.  CYP450 enzymes are present in the intestinal mucosa in most species.  These enzymes can contribute significantly to a first-pass effect, metabolizing drugs prior to absorption and significantly decreasing bioavailability.   CYP450 enzymes are also present in many other tissues of the body, and will be discussed in greater detail in the section on metabolism (see below). 

Alteration in drug distribution by drug-drug interactions can also occur.  The classic example of altered drug distribution between 2 drugs has always been protein binding interactions.  The most cited one being the interaction between the 2 highly protein bound drugs phenylbutazone and warfarin.  Phenylbutazone can displace warfarin from its binding site on albumin, resulting in an increase in free (active) warfarin concentrations, and bleeding abnormalities in the patient.  However, it is now widely assumed that protein binding interactions are not as common as they were previously thought to be.  This is due to several reasons, the first one being that there are several different binding sites on albumin, so that the drugs must share a common binding site and be very specific for that receptor in order for displacement to happen.  Another reason is that free (unbound) drug is available for excretion, therefore clearance of the drug may increase, so that the free fraction of the drug changes, but the free drug concentration is minimally affected.  In order for a clinically significant drug interaction to occur due to protein binding, the 2 drugs must be highly protein bound, exhibit a high clearance, and have a low therapeutic index, therefore protein binding interactions are very rare in veterinary medicine.  Co-administration of thiamylal followed by phenylbutazone in anesthetized ponies did show significant difference in the protein binding percentages of phenylbutazone and thiamylal at select timepoints, however there was no significant difference in the duration of anesthesia.  Dexamethasone caused a decrease in alpha-1 acid glycoprotein in dogs, which did affect the distribution of quinidine and caused a 2-fold increase in quinidine volume of distribution.

Another way drug-drug interactions can affect drug distribution involves those drugs that alter the blood flow to tissues.  Injectible and inhalant anesthetics can decrease tissue perfusion, resulting in a slower absorption of drugs injected intramuscularly or subcutaneously.  P-gp interactions can also affect tissue distribution, as P-gp efflux pumps are found in tissues such as the CNS and the eye.  The importance of these pumps in drug distribution is best exemplified in animals that are P-gp deficient.  These animals (collies, English Sheepdogs, Australian Shepherds, etc) have an increased risk of CNS toxicity due to drugs such as the avermectins, ivermectin and moxidectin due to an increased concentration of the drug in the CNS.  Presumably, if a P-gp inhibitor drug is co-administered with a drug that is also a P-gp substrate, a similar increase in drug concentrations in the CNS could also be expected to occur. 

Changes in drug metabolism are probably the most common pharmacokinetic drug interaction reported.  This is due to the fact that many drugs can either induce or inhibit the CYP450 enzymes in the liver.  Enzyme induction can only occur following repeated administration of the drug, therefore it is actually a less common cause of drug interactions than enzyme inhibition.  Drugs that induce CYP450 enzymes include barbiturates, such as phenobarbital, primidone, and rifampin.  CYP450 inhibitors include fluoroquinolones, macrolides, azole antifungals (ketoconazole>itraconazole>fluconazole), calcium channel blockers (diltiazem, verapamil), cimetidine, omeprazole, and propofol.

Several clinically relevant drug-drug interactions have been attributed to alterations in CYP450 metabolism in veterinary medicine. 

• The fluoroquinolone antimicrobials marbofloxacin and enrofloxacin have been shown to decrease the clearance of theophylline resulting in an increase in plasma theophylline concentrations. This was thought to be due to inhibition of CYP 1A2, based on an in vitro study that demonstrated an inhibitory affect of multiple fluoroquinolones on the hepatic CYP1A2 activiteis in hepatic microsomes from beagle dogs.  However, the concentrations used in that study were > 100x the concentrations found following in vivo administration. A more recent study indicated that enrofloxacin, ofloxacin, orbifloxacin and ciprofloxacin did not have any effect on midazolam metabolism in liver microsomes in vitro.  Additionally, neither enrofloxacin nor ofloxacin had any effect on the in vivo pharmacokinetics of quinidine in dogs.  Enrofloxacin does not alter the pharmacokinetics of digoxin in dogs, either, suggesting that the interaction between enrofloxacin and theophylline may be due to a mechanism other than CYP450 inhibition. An interaction between enrofloxacin and the NSAID flunixin in dogs has been shown, resulting in a prolonged half-life of both drugs.  Flunixin is not considered a substrate for CYP450 enzymes, which may suggest an alternate cause of interaction between enrofloxacin and other veterinary medications.

• Chloramphenicol is a CYP450 inhibitor that causes prolonged elimination and prolonged effects of barbiturates and propofol.  Topical administration of chloramphenicol may cause this as well, as the drug is rapidly absorbed from the eye following ocular administration of the ointment.  This has been manipulated clinically in dogs that show a decreased anticonvulsant response to phenobarbital.  Chloramphenicol has also been associated weth prolonged sedation (up to 5 hours) in a young horse being treated with therapeutic doses for 5 days.  Severe gastrointestinal distention also occurred, requiring percutaneous trocharization.  The effects of the xylazine were reversible using yohimbine.

• Ketoconazole increases the whole blood concentrations of cyclosporine and can be used to decrease the dose of cyclosporine by up to 38%.  This results in a substantial cost reduction for the owner.  Ketoconazole has also been shown to decrease the clearance of nifedipine.  Midazolam elimination half-life is prolonged following ketoconazole administration.  Quinidine pharmacokinetics are also affected by co-administration of ketoconazole, resulting in a prolonged elimination half-life and an increased maximum oral concentration.

• Cimetidine has been shown to decrease the clearance of verapamil as well as decrease the clearance and prolong the half-life of theophylline. 

• Phenobarbital decreases propranolol bioavailability and half-life. It has also been shown to decrease the half-life of digoxin acutely, but prolonged co-administration has no effect.

• Rifampin decreases the bioavailability of clarithromycin by up to 90% after coadministration for 11 days to healthy foals.

Interestingly, many drugs that are substrates for CYP450 enzymes are also substrates for P-gp.  This suggests that multiple interactions can occur with these drugs.  For instance, both cyclosporine and ketoconazole are substrates/inhibitors of P-gp, and the extreme changes in kinetics associated with co-administration of these compounds may be related to a dual inhibition. 

Another example of a clinically useful drug-drug interaction involves the inhibition of imipenem metabolism by the drug cilastatin.  Cilastatin inhibits the renal tubular dipeptidase enzyme responsible for metabolism of imipenem to its nephrotoxic metabolite.  This helps in 2 ways.  It decreases the nephrotoxicity of imipenem and it incr eases the urinary concentrations of the parent compound, increasing the efficacy of the drug in urinary tract infections.

Behavior modifying drugs are being used more and more often in veterinary medicine.  The FDA approved product selegiline is a monoamine oxidase (MAO) inhibitor that inhibitis metabolic degradation of catecholamines, dopamine, and, at high levels, serotonin.  Inhibition of MAO has been linked with serious, sometimes fatal drug interactions in humans, although there are no reports of this in veterinary medicine at this time.  MAO inhibitors co-administered with other drugs that increase circulating serotnonin levels (ie SSRIs, opioids, tricyclic antidepressants, and other MAO inhibitors), can lead to serotonin syndrome, a potentially fatal disease characterized by autonomic dysfunction, neuromuscular hyperactivity, altered mental status and hyperthermia.  MAO inhibitors co-administered with other sympathomimietic agents may also lead to a severe hypertensive reaction.  Therefore, these drugs should not be administered within two weeks of starting therapy with phenylephrine, phenylpropanolamine, ephedrine, and pseudoephedrine.

The tricyclic antidepressants, such as clomipramine, are CYP450 substrates, and co-administration with CYP450 inhibitors has resulted in an up to 4-fold increase in plasma concentrations of the tricyclic antidepressant.  Since these drugs have a low therapeutic index and can affect a wide variety of central nervous system receptors, they should not be combined with CYP450 inhibitors.  These drugs may also can an increased response to epinephrine and norepinephrine, as well as pancuronium and ketamine.  Cardiac arrhythmias can result from co-administration of these drugs with the tricyclic antidepressants.

Drug Interactions and Nutraceuticals

Alternative and herbal medicines are being used more and more commonly in our veterinary patients.  Many of these drugs are over-the-counter medications and as such, can be given by the animal's owner without the knowledge of the attending veterinarian.  These medications have not been evaluated by the FDA for safety or efficacy.  Because of this, they have also not been evaluated for their effects when administered concurrently with other drugs.  Although there are very few reports of drug-nutraceutical interactions in veterinary medicine, we can estimate the risk based on reports from human medicine.

• Vitamin E at very high doses (400IU/day) can have an effect on coagulation, possibly through antagonism of Vitamin K.  When co-administered with warfarin at these high doses, it caused a severe coagulopathy in dogs.

• Mineral supplements containing calcium magnesium and other multivalent cations can decrease the absorption of drugs such as the fluoroquinolones, penicillamine and tetracycline.  Calcium supplementation decreases the effect of thyroxine. 

• Chloride interferes with the excretion of bromide, and may lead to decreased control of seizures.

• Potassium supplementation given concurrently with spironolactone or angiotensin converting enzyme inhibitors may result in hyperkalemia.

• St. John's wort is a MAO inhibitor, and therefore should be considered dangerous when administered with any of the drugs discussed above. St. John's wort also induces the expression of CYP450 and P-gp in humans, which can lead to a decrease in plasma concentrations of drugs which are substrates. 

• Silymarin has been shown to inhibit CYP450 enzymes in vitro, and may cause disturbance in drug metabolism.

• Gingko biloba induces some human CYP450 enzymes, while it inhibits others.  It may also inhibit P-gp in the intestine.  It also affects platelet function in people, and may increase the risk of bleeding with other drugs that affect platelets, such as non-specific cyclooxygenase inhibitors.

At this point in time, much more research is necessary to determine the safety of these supplements in veterinary medicine, and to determine their effects on other drugs.