Sign up for our e-newsletter

Enduring Material

This CME has expired

Print Friendly 

Antidepressant Drug-Drug Interactions: Clinical Relevance and Risk Management

Charles B. Nemeroff, MD, PhD, Sheldon H. Preskorn, MD, and C. Lindsay DeVane, PharmD


CNS Spectr. 2007;12(5 Suppl 7):1-16

An expert panel review of clinical challenges in psychiatry

This expert roundtable supplement is based on an i3 CME presentation held December 5, 2006, in Hialeah, Florida. 

Both the presentation and this supplement are supported by an educational grant from Bristol-Myers Squibb.

Accreditation Statement/Credit Designation

i3 CME is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians. 

i3 CME designates this educational activity for a maximum of 1 AMA PRA Category 1 Credit.TM Physicians should only claim credit commensurate with the extent of their participation in the activity.   

The University of Florida College of Pharmacy is accredited by the Accreditation Council for Pharmacy Education as a provider of continuing pharmacy education. This activity is approved for 1 contact hour (0.1 CEU) in states that recognize ACPE. 

To receive enduring credit, you must pass the post test with a score of at least 70% and complete the evaluation form. Statements of CE Credit will be issued by the University of Florida College of Pharmacy upon proof of successful completion of the activity. (Enduring 012-999-06-194-L01).

i3 CME is an approved provider of continuing nursing education by the Utah Nurses Association, an accredited approver by the American Nurses Credentialing Center’s Commission on Accreditation. The participant will be awarded 1 contact hour of credit for attendance and completion of supplemental materials.

i3 CME is approved by the American Psychological Association to sponsor continuing education for psychologists. i3 CME maintains responsibility for this program and its content. The participant will be awarded 1 contact hour of credit for completion of the quiz and evaluation form.

This activity has been approved to provide 1 hour of CE for case managers by the Commission for Case Manager Certification (CCMC).

Target Audience

This activity is designed to meet the educational needs of physicians, pharmacists, nurses, psychologists, and case managers.

Learning Objectives

• Identify the prevalence of drug-drug interactions and discuss their clinical relevance in patients with depression.
• Recognize appropriate treatment approaches after weighing the actual benefits of pharmacotherapy, the morbidity and mortality associated with undertreatment or nontreatment, and the risks of unintended drug-drug interactions in patients with depression.
• Indicate appropriate treatment strategies, dosing and titration issues, and the use of appropriate medications to avoid/manage drug-drug interactions in patients with depression.

Faculty Affiliations and Disclosures

Dr. Nemeroff is Reunette W. Harris Professor and Chairman of the Department of Psychiatry and Behavioral Sciences at Emory University School of Medicine in Atlanta, Georgia. Dr. Nemeroff is a currently on the scientific advisory boards of Forest, Janssen/Ortho-McNeil, Quintiles, and NeuroPharmaboost; receives grant/research support from the American Foundation for Suicide Prevention,  the National Alliance for Research on Schizophrenia and Depression, and the National Institute of Mental Health (NIMH); and is a major stockholder or owns equity in Corcept, CeNeRx, Revaax, and NovaDel.

Dr. Preskorn is Professor and Chair of Psychiatry and BehavioralSciences at the University of Kansas School of Medicine, and president and chief executive officer of the Clinical Research Institute in Wichita, Kansas. Dr. Preskorn is a consultant to Bristol-Myers Squibb, Comentis, Cyberonics, Eisai, Eli Lilly, EnVivo, Evotec, Johnson & Johnson, Memory, Otsuka, Pfizer, Predix, Shire, Somerset, and Wyeth; is on the advisory boards of Bristol-Myers Squibb, Eli Lilly, Johnson & Johnson, Pfizer, Shire, Somerset, and Wyeth; is on the speaker’s bureaus of Bristol-Myers Squibb, Cyberonics, Forest, Otsuka, and Pfizer; and receives grants/research support from Bristol-Myers Squibb, Comentis, Cyberonics, Johnson & Johnson, Merck, Memory, the NIMH, Novartis, Organon, Otsuka, Pfizer, Predix, Sepracor, Somerset, and Wyeth.

Dr. DeVane is Professor of Psychiatry and Behavioral Sciences, and  Vice Chair for Research in the Department of Psychiatry at the Medical University of South Carolina in Charleston. Dr. DeVane is a consultant to Janssen, Bristol-Myers Squibb, GlaxoSmithKline, Novartis,  and Theracos; is on the advisory boards of NovaDel and Theracos; and receives grants/research support from the NIMH and the National Institute on Drug Abuse.

Acknowledgment of Commercial Support

This activity has been supported through an educational grant to i3 CME from Bristol-Myers Squibb.

To Receive Credit for this Activity

Read this expert roundtable supplement, reflect on the information presented, and take the CME quiz on page 14. Complete the answer form and evaluation on page 15 and return it to: i3 CME, PO Box 1111, State College, PA 16804.

To obtain credit, you should score 70% or better. Termination date: May 31, 2009. The estimated time to complete this activity is 1 hour.


Multiple medication use is a common phenomenon, especially in patients with comorbid conditions and those treated with psychiatric drugs such as antidepressants. Combination treatment may result in potentially harmful drug-drug interactions (DDIs). Results of DDIs range from nuisance side effects to serious adverse consequences. DDIs may also result in improved efficacy. Augmentation strategies, for example, are intentional therapeutic DDIs. Pharmacokinetic DDIs occur when a second drug alters the absorption, distribution, metabolism, or clearance of the first drug. Research has concentrated on the relative effects of antidepressants on cytochrome P450 enzymes and, more recently, on drug transporters as potential mediators of clinically important pharmacokinetic DDIs. The most common, clinically relevant pharmacokinetic DDIs involve alteration in oxidative drug metabolism. Pharmacodynamic DDIs occur when the effects of a second drug quantitatively or qualitatively alters those of the first drug. Pharmacodynamic DDIs are not typically studied in vivo because of the potential for a serious adverse effect. All antidepressants can interact pharmacodynamically with certain other drugs. The risk of harmful DDIs can be reduced by recognizing variables that affect dose-concentration-effect relationships. It is important for physicians to weigh the risks and benefits of potential DDIs against the risks that accompany timid or ineffective disease treatment.


Antidepressant drug-drug interactions (DDIs) can adversely affect patient outcome and thus must be taken into account during treatment planning and management. This review will address the basic principles underlying DDIs, as well as those interactions that are specific to different types of antidepressants including selective serotonin reuptake inhibitors (SSRIs), serotonin norepinephrine reuptake inhibitors (SNRIs), bupropion, mirtazapine, tricyclic antidepressants (TCAs), and monoamine oxidase inhibitors (MAOIs).

Multiple Medication Use: Rationale and Prevalence

Multiple medication use (MMU) is common in clinical practice. In general medicine, the most compelling reason for MMU is treatment of patients who have more than one chronic medical illness. For example, the patient with hypertension, diabetes, and depression will likely be on one or more medication for each of these condition, and these drugs may interact to alter patient outcome. Patients may also have more than one psychiatric syndrome and thus may be treated with more than one psychiatric medication. Development of an acute medical illness such as an upper respiratory tract infection may require additional brief pharmacotherapy. There are also several cyclic illnesses, such as bipolar disorder, and patients may have medications added or modified as they go through the various phases of their illness. Additional reasons why a clinician may use more than one medication include: (A) to treat an adverse effect of a first drug; (B) to augment the desired effect of a first drug in a partial responder; or (C) to accelerate the onset of action of a first drug.1

A 2004 Health and Human Services survey underscored the substantial prevalence of multiple medication use in the United States. Among Americans >18 years of age, 40% were treated with one or more prescription drug, 20% were treated with three or more drugs, and 7% were treated with five or more drugs. These numbers were triple for people >65 years of age, with 20% of elderly Americans receiving five or more prescription drugs (Slide 1).2

As of January 2006 there were 3,500 discrete chemical entities approved as prescription drugs in the Physicians Desk Reference.3 To illustrate the complexity that can potentially occur, consider that 3,500 drugs on the market means that 520 quadrillion different  combinations of up to 5 prescription drugs could be prescribed to a patient. That is more combinations than humans have lived on the earth. That number is, of course, theoretical so a 2005 pharmacoepidemiology study of outpatients being treated within the Veteran Affairs Healthcare System (VA) examined how many patients were in fact treated with unique combinations of medications. For the purposes of this study, a unique regimen was defined as a specific combination of drugs that was being taken by only one person in the surveyed population which numbered 5,000 individuals. The study found that most patients were on unique combinations and that antidepressant treatment and age were risk factors for both a higher incidence of MMU and more complex MMU: 96% of patients >60 years of age on an antidepressant were on a unique regimen versus 75% age-matched patents not on an antidepressant, and 83% of patients <60  years of age who were on an antidepressant were on a unique regimen versus 62% age-matched patients not on an antidepressant (Slide 2).4

These results demonstrated that a substantial percentage of patients in this healthcare system have the potential to experience one or more DDIs and that potential is increased in patients on antidepressants. The study did not explore why patients on antidepressants are at increased likelihood of being treated with multiple medications but the following are possible reasons: Depressed patients are at risk for comorbid medical disorders, including heart disease, stroke, and diabetes, which may account for the increase in polypharmacy among these patients. Depressed patients are high utilizers of healthcare services and thus may see more physicians. Depression frequently presents in primary care with a vague variety of somatic complaints and patients may be treated symptomatically before the diagnosis of depression is made.

Basic Principles of Drug-Drug Interactions

When a drug is ingested, it produces a measurable plasma drug concentration, which in turn, results in occupation of a percentage of the available transporter receptor sites in the body depending on the nature of the drug (Slide 3). The receptor sites for the therapeutic effects of psychiatric drugs are located in the central nervous system (CNS). The pharmacologic effects can be direct and immediate or indirect and delayed, but eventually the effect translates into behavioral, cognitive, or emotional changes in the patient. The vast majority of psychiatric drugs are metabolized in the liver and eventually excreted as metabolites in the urine. This latter process is called clearance.

When a patient is prescribed a second drug, there is the potential for a DDI to occur. A pharmacokinetic DDI occurs when the second drug alters the absorption, distribution, metabolism, or clearance of the first drug. The most common, clinically relevant pharmacokinetic DDIs involve alteration in oxidative drug metabolism. Alternately, pharmacodynamic DDIs can occur when the effect(s) of a second drug alters the effect(s) of a first drug quantitatively (magnitude or duration of the effect) or qualitiatvely (nature of the effect).5 A change in the nature of the effect can be sudden and catastrophic. Examples include hypertensive crisis, serotonin syndrome, seizures, arrhythmias, and even sudden death. While most physicians think of DDIs exclusively in terms of these types of outcomes, they are relatively rare. Instead, most DDIs—particularly pharmacokinetic DDIs—present as a change in the magnitude or duration of the effect of a given dose of a given drug. Such interactions can present in a myriad of ways including the patient being “sensitive” or “resistant” to the effect of a drug and thus are easily missed even though they are clinicially meaningful.

Pharmacokinetic Mechanisms

Possible pharmacokinetic mechanisms include effects on protein binding, changes in the activity of Phase I or Phase II enzymes, ABC drug  transporters, and nuclear receptors.6 It is well known that most drugs do not freely circulate in the blood; rather, drugs are avidly bound  to plasma proteins, principally albumin and lipoproteins such as a1 acid-glycoprotein. Displacement  of  protein binding of one drug by another is a mechanism commonly thought of as a meaningful cause of clinically relevant DDIs, but this not correct. Theoretically, it is true that a second drug can displace a first drug from a plasma binding site, resulting in increased free drug concentration that could lead to increased pharmacologic effects. However, that increased free drug concentration is also subject to elimination, so the effect is often short-lived and rarely, if ever, of clinical significance.6 In addition, drugs may be highly protein bound but that does not mean that they have saturated their binding sites and therefore may not be easily subject to displacement by another drug.

A much more significant mechanism for clinically relevant pharmacokinetic DDIs are those mediated by changes in the activity of drug-metabolizing hepatic enzymes. These enzymes can be divided into Phase I and Phase II types. Phase I enzymes can be further divided into cytochrome P450 (CYP) enzymes and non CYP enzymes. The former are responsible for the bulk of oxidative drug metabolism that breaks drugs into component parts which are more water soluble and susceptible to phase II enzymes which add moieties to them such as glucuronides or sulfates,  producing a compound which is readily excreted in urine.6

There are several other potential mechanisms for DDIs, such as drug transporters. Many drugs permeate cellular membranes by simple passive diffusion. However, there are ~40 different genes and their associated proteins that have demonstrated importance in drug disposition, including transporters that actively carry drug molecules across cell membranes. This opens up new opportunities to discover mechanisms of pharmacokinetic drug interactions that were unknown 5 or 6 years ago.6

Understanding Pharmacokinetic Interactions

Despite its complexity, the best way to understand pharmacokinetic interactions is to simplify. For example, Slide 4 shows that Drug A has an effect on a single CYP enzyme (enzyme X), either increasing or decreasing its activity.7 That enzyme in turn is responsible for the metabolism of several different drugs: B, C, D, E, and F.  Therefore, the addition of drug A to a patient on drug  B or one of the others metabolized by enzyme X will alter the rate of the metabolism of drug B and hence its accumulation in the body and hence the magnitude or duration of its effects on the patient.7

DDIs can be particularly problematic when the patient has several prescribing healthcare practitioners who are not aware of what the other practitioners are prescribing. The consequences can range from clinically trivial to devastating and catastrophic. The aforementioned pharmacoepidemiology study in the VA found a significant relationship between the number of prescribers a patient was seeing and their likelihood of being on eight or more drugs.4 With each additional prescriber (up to four) the patient was seeing, the patient’s likelihood of being on eight or more drugs doubled. With 3,500 medications in the pharmacopoeia it is difficult for a busy practitioner to keep track of potential DDIs, and communication between prescribers is critical.

The example of drug A influencing enzyme X is probably most important for those drugs that are predominantly eliminated by one metabolic pathway. Examples include the antidepressant desipramine, or the antihypertensive metoprolol, which have often been used as model substrates in pharmacokinetic DDI studies exploring the potential effects of drugs on CYP2D6 because these two drugs are primarily dependent on the CYP2D6 enzyme  for their biotransformation. Although some drugs are similar to desipramine and metoprolol in terms of being principally dependent on a single CYP enzyme, others are metabolized by multiple CYP enzymes. For example, there are four different pathways involved in the clearance of mirtazapine and each pathway contributes approximately equally to its clearance.  For this reason, a drug like mirtazapine is not as prone to a clinically meaningful CYP-enzyme–mediated DDI as are desipramine and metoprolol.

Antidepressants: Effect on the CYP450 Enzymes

In the early 1990s, the relative effects of antidepressants on CYP enzymes focused on CYP2D6. That research then expanded to include their relative effects on the other major drug metabolizing CYP enzymes including:  1A2, 2C9/10, 2C19, and 3A4. More recent research has started to explore the importance of drug transporters as potential mediators of clinically important pharmacokinetic DDIs.

A list of antidepressants, along with other drugs, that have been documented to substantially inhibit CYP enzymes is provided in Slide 5.  The following enzymes are substantially inhibited by following drugs at their usual therapeutic doses: CYP2D6 by bupropion, fluoxetine, paroxetine, and terbinafine; CYP3A3/4 by clarithromycin, erythromycin, fluconazole, fluvoxamine, indinavir, itraconazole, ketoconazole, nelfinavir, nefazodone, and ritonavir; CYP1A2 by fluvoxamine and omeprazole; and CYP2C by fluoxetine, fluvoxamine, and omeprazole.8

Contrary to popular belief, bupropion does inhibit CYP2D6. At 300 mg/day it produces almost the same magnitude of effect as paroxetine or fluoxetine, which is significant because bupropion is the antidepressant most often combined with an SSRI. Combining bupropion with paroxetine or fluoxetine would likely completely saturate 2D6 and result in complete loss of functional activity of this enzyme.8 As seen in this table, some antidepressants at their usual therapeutic doses can substantially inhibit more than one enzyme which increases the potential for these drugs to cause CYP enzyme mediated DDIs.

Not all antidepressants affect CYP enzymes. Antidepressants that do not substantially inhibit any CYP enzyme at their usual therapeutic doses include the SSRIs citalopram, escitalopram, and sertraline; the SNRIs duloxetine and venlafaxine; mirtazapine; TCAs; and the selegiline transdermal patch (Slide 6). Even so, most of these drugs produce minimal to mild inhibition of this enzyme and the magnitude of such inhibition is dependent on the concentration and hence the dose of the drug.  For example, sertraline has been the best studied in this regard and produces a 20%, 30%, and 64% increases in the area under the curve (AUC) of a CYP2D6 dependent substrate at daily doses of  50, 100, and 150 mg, respectively. The same dose-dependent increase in CYP2D6 would be seen with higher doses of citalopram, duloxetine, and escitalopram. To put these numbers in perspective, fluoxetine and paroxetine at 20 mg/day produce approximately a 500% increase in the AUC of CYP2D6 dependent drugs.  Formal studies demonstrated that coadministration of transdermal selegiline did not alter the clearance of alprazolam, ibuprofen, olanzapine, risperidone, or warfarin.8

Certain antidepressants, such as all of the TCAs, are potential victims of CYP-mediated DDIs which are particularly important because of their narrow therapeutic index.  Bupropion is another antidepressant that has a narrow therapeutic index but to date is not known to be the victim of any specific pharmacokinetic DDIs. Other antidepressants, such as MAOIs, mirtazapine, SNRIs, and SSRIs are generally not victims of pharmacokinetic DDIs.8

In a study of transdermal selegiline, coadministration of the substantial CYP3A4/5 inhibitor ketoconazole did not alter selegiline levels in patients receiving transdermal selegiline 6 mg/24 hours.9 Carbamazepine, a CYP3A4 inducer, at a dose of 400 mg/day increased the plasma concentrations of selegiline two-fold.9 That finding is not consistent with CYP3A4 induction so another, as yet unknown, mechanism likely mediates this DDI. 

The most significant DDIs with transdermal selegiline, as with the classic irreversible, non-selective  MAOIs, are pharmacodynamically rather than pharmacokinetically mediated.  These involved the coadministration of transdermal selegiline with other drugs capable of increasing either the functional release or activity of norepinephrine and/or serotonin in the brain and carries the risk of causing either a hypertensive crisis or the serotonin syndrome, respectively.

Variables that Determine Drug Response

There are three main variables that determine drug response, as illustrated in Slide 7. First, the drug has to work on a site of action that is physiologically capable of producing the effect. Second, the concentration of the drug achieved at the site of action must be sufficient to produce its pharmacodynamic effect. Clinical trials that lead to drug registration are in essence attempts to determine the usual dose that will achieve a concentration that engages the proposed therapeutic mechanism of action to achieve the desired response in the average patient.  The third variable involves the biology of the patient, which poses the greatest challenge for the practitioner because patients differ based on their genetics, age, disease, and internal environment (Slide 7). The internal environment changes as a function of what the patient consumes, including the drugs he/she takes, and can result in a DDI.7 A DDI occurs when the presence of a drug alters the biology of the patient and thus alters their response to a co-prescribed medication(s).

Drug concentration is the dosing rate divided by the clearance and hence drug concentration can be altered either by changing the dose of the drug administered, or by changing the clearance. The clinical implication of this fact can be illustrated with risperidone as follows: Slide 8 illustrates the distribution curve of D2 dopamine receptor occupancy in the striatum of 100 patients treated with 4 mg of risperidone.10 As can readily be seen, there is variability among these patients because they achieve different concentrations of risperidone in the plasma and in the brain despite the fact that they are all taking the same dose. That is a result of the differences in their ability to clear risperidone. The individuals in the extreme right portion of the curve are most likely individuals who have a genetic deficiency or a concomitant drug-induced deficiency in CYP2D6 activity and hence achieve higher than usual concentrations of risperidone for the dose given. Recall that ~7% of Caucasians are genetically deficient in CYP2D6 and that a number of widely used antidepressants such as bupropion, fluoxetine, and paroxetine at their usual therapeutic doses can produce phenocopies of CYP2D6 deficiency. 

Based on positron emission tomography (PET) scans, D2 receptor occupancy exceeding 80% is associated with an increased risk of extrapyramidal symptoms (EPS), whereas 50% to 70% is needed for antipsychotic efficacy. Thus, a pharmacokinetic DDI that increases the accumulation of risperidone will increase the functional blockade of D2 receptors in the striatum and can result in EPS. That may be simply interpreted as the patient being sensitive to D2 receptor blockade without understanding that increased sensitivity is due to a pharmacokinetic DDI. 

The risk of CYP-mediated antidepressant drug interactions can be minimized by taking the following steps: First, the physician initiates the potential “victim” drug at a low starting dose and increases the dose gradually after assessing the patient’s response (essentially a bioassay approach to dose titration). Second, the prescriber may keep the dose of the inhibitor low which will minimize the risk because the magnitude of enzyme inhibition is a function of the potency of the inhibitor times its concentration. Third, the prescriber may prefer to use drugs which have multiple pathways, which all contribute approximately equally to the elimination of the potential victim drug.11   

Pharmacodynamic Drug-Drug Interactions

A hypothetical scheme that is useful for reviewing some of the mechanisms of pharmacodynamic interactions is illustrated in Slide 9.12 Drug A affects three different receptors. For the sake of this discussion, these receptors will be termed B, C, and D. Receptor B is responsible for therapeutic effects, receptor C is responsible for minor adverse events, and receptor D is responsible for much more medically serious adverse events. At the lowest concentrations, the effects of drug A will be restricted to its most potent site of action, receptor B. In the case of drug A, that will result in optimum efficacy with minimal risk of nuisance adverse effects and no meanginful risk of serious adverse effects.  However, if the dose is increased, then the drug may engage receptor B and potentially even C. To the degree that the drug binds to these receptors, the patient is at risk for experiencing nuisance and then more serious adverse effects.12

Although formal in vivo human studies can be conducted to test for pharmacokinetic DDIs, that is not as easily done for many pharmacodynamic DDIs, particularly when the potential consequence of the suspected DDI may be serious (eg, hypertensive crisis or serotonin syndrome). Pharmacokinetic interactions are easier to study because the endpoint is a measurable change in the concentration of the drug rather than a potentially serious adverse event. For these reasons, guidance on pharmacodynamic DDIs are frequently based on extrapolation from mechanism of action, and/or isolated/unreplicated case reports. As a result, such guidance may be overly restrictive.1

All antidepressants can interact with certain other drugs pharmacodynamically. Pharmacodynamic interactions with SSRIs are principally limited to other drugs that affect serotonergic neurotransmission. Pharmacodynamic interactions with SNRIs (duloxetine, venlafaxine) are principally limited to other drugs that affect serotonergic and noradrenergic neurotransmission. Tertiary amine TCAs and mirtazapine have the most varied pharmacodynamic interactions because of their action at multiple receptors and transporters. MAOIs can interact with other drugs that affect noradrenergic, dopaminergic, and serotonergic neurotransmission.

Based on pharmacovigilance studies, venlafaxine apparently has pharmacodynamic effects indicating that it has higher overdose death liability than the SSRIs. In this way, venlafaxine, in addition to its shared neurochemical actions as a reuptake inhibitor of serotonin and norepinephrine, is TCA-like. This is a pharmacodynamic effect that took a long time to become evident, just as with bupropion, which was in use for a long time before its 2D6 inhibition was discovered. 

Monoamine Oxidase Drug-Drug Interactions

As already mentioned above, MAOIs can interact pharmacodynamically with drugs that affect the norepinephrine, dopamine, and/or serotonin systems. MAOIs were one of the first classes of antidepressants, discovered over 50 years ago. They act by preventing the breakdown of serotonin, norepinephrine, and dopamine. A seminal paper published in the Archives of General Psychiatry revealed a 35% increase in the monoamine oxidase activity in the brains of depressed patients using a novel imaging method developed at the University of Toronto.

Slide 10 is a schematic of a monoamine neuron, including a nerve terminal, which could either contain norepinephrine, serotonin, or dopamine, and monoamine oxidase is shown in yellow.13 There is substantial evidence to suggest that depression is associated with a relative deficiency of all three of these neurotransmitters. However, it is difficult to determine in a given patient whether the problem is due to a deficiency in the functional activity of  norepinephrine, dopamine, or serotonin. Research into safe and effective ways to increase the availability of all three of these monoamines has been ongoing for many years.

There are two major forms of monoamine oxidase in humans: monoamine oxidase A (MAO-A), found primarily in the gut, and monoamine oxidase B (MAO-B), found both in the periphery and in the central nervous system. Both subtypes have the ability to metabolize tyramine and dopamine. MAO-A also has a high affinity for norepinephrine and serotonin, whereas MAO-B has a high affinity for phenylethanolamine and benzylamine (Slide 11).14

The older MAOIs, like phenelzine and tranylcypromine, were effective antidepressants but could cause serious adverse effects when patients treated with them consumed food rich in tyramine (ie, a drug-food interaction). There were also concerns about multiple daily dosing, which had an adverse effect on adherence. Those issues led to the development of the selegiline transdermal system, which provides an example of the difference that route of administration can make in both pharmacokinetics and pharmacodynamics. If selegiline is taken orally it undergoes an extensive first-pass effect in which the drug is biotransformed in the liver into many metabolites. The transdermal system results in a slower release of selegiline, higher and longer-lived plasma concentrations of the parent drug and lower accumulation of metabolites, as well as no significant initial exposure to the gastrointestinal tract, thereby preserving functional activity of MAO-A there. That is important because MAO-A in the gut is an enzymatic barrier preventing tyramine in food from entering the systemic circulation. Nevertheless, transdermal delivery of  selegiline allows the drug to reach the brain in sufficiently high enough concentration to inhibit both B and A in the brain and thus produce an antidepressant effect. Transdermal delivery of selegiline thus provides a margin of safety in terms of avoiding an interaction with dietary tyramine that has not existed before.14  That is the reason why the Food and Drug Administration approved the use of this product at the 6 mg/day dose without any dietary restriction.

Although the 6 mg/day dose of transdermal selegiline avoids the dietary problems associated with oral irreversible MAOIs, it still is susceptible to the DDIs associated with the oral MAOIs because it still reaches the brain where it inhibits both MAO-A and MAO-B. That effect leads to an increase in synaptic concentrations of norepinephrine, dopamine, and serotonin. When the increased availability of any one of these three monamines occurs in conjunction with other antidepressant mechanisms (eg, reuptake blockade), the effects of these neurotransmitters on their target cells can be significantly amplified. For this reason, transdermal selegiline like the traditional oral, irreversible MAOIs, can potentially have additive or even further synergistic effects when co-prescribed with other drugs that promote monoaminergic  transmission in the brain. For example, MAOIs enhance the effects of L-DOPA. Transdermal selegiline, like oral MAOIs, can intensify the action of CNS active drugs such as general anesthetics, sedatives, alcohol, antihistamines, analgesics, and anticholinergics.13,15

Those interactions can result in two different potentially life-threatening effects, the serotonin syndrome and a hypertensive crisis (Slide 12).13,15 Different symptoms can occur when there is too much serotonin released in the brain stem including both psychiatric and neurological effects and potentially fatal cardiovascular effects, as well as gastrointestinal effects, and others, as shown in Slide 13.16


When transdermal selegiline is taken in combination with another drug that potentiates serotonergic neurotransmission (eg, SSRIs, SNRIs, St. John’s Wort, tryptophan, or 5-hydroxytryptophan), there is the potential for a serotonin syndrome. Other drugs that have been known to have effects on the serotonin system include buspirone, mirtazapine, dextromethorphan, propoxyphene, meperidine, tramadol, and methadone (Slide 14).9

Although there is no need for a tyramine-restricted diet with the lowest dose of the transdermal selegiline patch, there is still the potential for a hypertensive crisis when other norepinephrine active drugs are taken together with transdermal delivered selegiline. Norepinephrine reuptake inhibitors that can theoretically produce a hypertensive crisis when combined with a MAOI include the SNRIs; TCAs; reboxetine; atomoxetine; a number of cold, diet, and local anesthetic preparations containing sympathomimic amines such as epinephrine, phenylephrine, and phenylpropanolamine; and pseudoephedrine, amphetamines, methylphenidate, cocaine, and related psychostimulants.9,17,18


The frequency and complexity of MMU is enormous and sets the stage for potentially harmful DDIs. MMU is more frequent and complex in patients treated with psychiatric medications such as antidepressants, which increases the risk of patients experiencing potentially harmful DDIs. DDIs can be either pharmacodynamically and/or pharmacokinetically mediated. The risk of harmful DDIs can be reduced by thoughtful clinical practice and recognition of variables affecting dose-concentration-effect relationships. Some risks for DDIs are recognized, such as the number of medications taken, extremes of age, and any disease of organs that could impair drug clearance.

Clinical significance of DDIs can range from nuisance side effects to serious adverse effects, as well as potentially improved tolerability and/or efficacy. The latter is the case because all augmentation strategies are, in essence, intentional therapeutic DDIs. There are absolute and relative contraindications when using drugs in combinations and these are based on a knowledge of their pharmacodynamics and pharmacokinetics. It is therefore important for prescribers to know and weigh the risk and benefit of potential DDIs and balance those risks against the risk inherent in not effectively treating what can be very lethal diseases such as untreated or ineffectively treated major depression. To put this matter in perspective, there were 28,500 suicides in the United States last year, largely due to untreated or refractory depression.19

DDIs can be divided into pharmacodynamic and pharmacokinetic interactions. Only a few antidepressants have a substantial risk of causing pharmacokinetic DDIs via CYP enzyme inhibition. For this reason, the prescriber should consider which drugs within a functional class, such as SSRIs, should be used as first-line agents. To specifically avoid CYP enzyme mediated DDIs, for example, citalopram, escitalopram, and sertraline have the lowest potential for causing substantial inhibition of any CYP enzyme whereas fluoxetine, fluvoxamine, and paroxetine all inhibit one or more CYP enzymes at usual therapeutic doses.


1. Preskorn S, Lacey R. Polypharmacy: When is it rational? J Pract Psychiatr Behav Health. 1995;1:92-98.
2. HHS Report. Center for Disease Control. United States, 2004.
3. Preskorn SH. Drug approvals and drug withdrawals over the last 60 years. J Psychiatr Pract. 2002;8(1):41-50.
4. Silkey B, Preskorn SH, Golbeck A, et al. Complexity of medication use in the Veterans Affairs Healthcare System: Part II. Antidepressant use among younger and older outpatients. J Psychiatr Pract. 2005;11(1):16-26.
5. Preskorn SH. Drugs are an acquired source of biological variance among  patients. J Psychiatr Pract. 2006;12(6):391-396.
6. Levy RH, Thummel KE, Trager WF, Hansten P, Eichelbaum M. In: Metabolic Drug Interactions. Philadelphia, Lippincott, Williams & Wilkins; 2000.
7. Preskorn SH. Clinical Pharmacology of Selective Serotonin Reuptake Inhibitors. Caddo, Okla: Professional Communications, Inc; 1996.
8. Preskorn SH, Flockhart DA. 2006 guide to psychiatric drug interactions. Primary Psychiatry. 2006;13(4):35-64.
9. Somerset Pharmaceuticals. Selegiline transdermal system (EMSAM). Available at: Accessed August 16, 2006.
10. Frankie WG, Gil R, Hackett E, et al. Occupancy of dopamine D2 receptors by the atypical antipsychotic drugs risperidone and olanzapine: theoretical implications. Psychopharmacol Berl. 2004;175:473-480.
11. DeVane CL. Antidepressant-drug interactions are potentially but rarely clinically significant. Neuropsychopharmacology. 2006;31(8):1594-1604.
12. Preskorn SH. Classification of neuropsychiatric medications by principal mechanism of action: a meaningful way to anticipate pharmacodynamically mediated drug interactions. J Psychiatr Pract. 2003;9(5):376-384.
13. Hardman JG, Limbird LE, Gilman AG. Goodman and Gilman’s The Pharmacological Basis of Therapeutics. 10th ed. New York, NY: McGraw Hill; 2001.
14. Holschneider DP, Shih JC. In: Neuropsychopharmacology: The Fourth Generation of Progress. Available at: Accessed August 16, 2006.
15. Holschneider DP, Shih JC. Monamine oxidase: basic and clinical perspectives. In: Neuropsychopharmacology: The Fourth Generation of Progress. Chapter 46. Available at: Accessed September 1, 2006.
16. Sternbach H. The serotonin syndrome. Am J Psychiatry. 1991;148(6):705-713.
17. Strolin Benedetti M, Dostert P. Monoamine oxidase: from physiology and pathology to the design and clinical application of reversible inhibitors. Adv Drug Res. 1992;23:65-125.
18. Hoffman BB, Lefkowitz RJ. Catecholamines and sympathomimetics. In: Gilman AG, Rall TW, Nies AS, eds. The Pharmacological Basis of Therapeutics. 8th ed. New York, NY: Pergamon Press; 1990:187.
19. Preskorn S, Werder S. Detrimental antidepressant drug-drug interactions: are they clinically relevant? Neuropsychopharmacology. 2006;31(8):1605-1612.

Question-and-Answer Forum

Q: With 3,500 medications available in the United States pharmacopoeia leading to the potential for 520 quadrillion combinations of up to 5 drugs, how can the busy practitioner keep track of the many drug-drug interactions (DDIs) that may have clinical relevance?

Dr. Preskorn: It is helpful to think in terms of the pharmacodynamics and the pharmacokinetics of the drugs and group drugs by those functional classes. It tends to be easier to remember things when they are grouped together rather than trying to remember each isolated interaction.

Also, a number of software packages exist. However, there are a number of limitations with most of these programs which limits how user-friendly they are. Most of these programs function as an alert rather than an information system. They warn that there may be a problem but often do not provide useful information about what the prescriber can do to minimize the risk.  They also tend to be binary, in that they deal with one drug’s effect on another drug and yet the patient may be on multiple medications which can interact in complex ways. Some of these software packages only deal with pharmacodynamic or pharmacokinetic mediated DDIs.

Finally, there is fairly limited knowledge about DDIs in the real world of clinical practice and clinically relevant DDIs are often discovered months to years after the drugs have been marketed. In fact, many of the drugs removed from the market over the last decade were removed because of DDIs discovered after they were marketed.

Dr. DeVane: It is almost impossible for any practitioner to be so conscientious that they can memorize all these facts about drugs and inhibitors. It is better to rely on basic principles of understanding.

Q: If you were to recommend a software package for a practitioner to purchase, which one would it be?

Dr. Preskorn: My preference is Drug Facts & Comparisons.

Dr. DeVane: I would agree. Absolutely. 

Q: How effective is the selegiline transdermal system in treating depression at the 6 mg low dose?

Dr. Nemeroff: The selegiline transdermal system is available in 6 mg, 9 mg, and 12 mg. Most of my patients are prescribed the 6 mg dose, and the vast majority have stayed on that dose, whereas only ~25% of the patients have gone on to one of the higher doses. I tend to treat relatively refractory patients, so I probably have had an inordinate number of patients who have required 9 mg. I have had hardly any patients who required the 12 mg dose.

Dr. Preskorn: Of the pivotal trials that led to the approval of selegiline transdermal patch, one was a positive placebo-controlled, fixed-dose study that only used the 6 mg/day dose, another was an ascending dose study starting all patients on 6 mg/day and allowing titration up to 12 mg/day, and finally the relapse-prevention study which also only used the 6 mg/day patch. Based on these studies, 6 mg/day is clearly an effective dose for many patients. Parenthetically, the package label warns against exceeding 6 mg without using the diet.

Dr. DeVane: Oral selegiline given in a high enough dose could be an effective antidepressant, but the problem is that the dose would have to be so high that it poses a safety danger. The transdermal system avoids the first pass effect, so at the lower dose, 6 mg/day, you get much higher concentrations of selegiline in the blood, and therefore in the brain, than you do if you gave 10 mg of selegiline orally. It is a pharmacokinetic planned consequence of the transdermal delivery system.

Dr. Nemeroff: Another point of interest is that in the pivotal studies for approval of the selegiline transdermal system there was a remarkably low level of sexual dysfunction. It is a conundrum because the selegiline transdermal system increases the availability of serotonin and of course the selective serotonin reuptake inhibitors (SSRIs) have relatively high rates of sexual dysfunction. The answer may well lie in its effect on dopamine. Increasing dopamine neurotransmission likely alleviates sexual dysfunction. I have been very heartened in treating patients who have responded to selegiline, who have either not responded to SSRIs or have been unwilling to continue taking SSRIs because of the associated sexual dysfunction.

Q: Is bupropion a cytochrome P450 (CYP) 2D6 inhibitor?

Dr. Preskorn: Yes. At 300 mg/day bupropion will cause a similar degree of inhibition of CYP2D6 as fluoxetine or paroxetine, and hence a similar degree of increase in concentration of co-prescribed CYP2D6 substrates. That information is in the package insert which gives the results of a study in which participants were treated with bupropion 300 mg/day and then a challenge dose of desipramine as a CYP2D6 probe. The blood levels of desipramine went up five-fold comparable to what is seen when desipramine is co-prescribed with fluoxetine or paroxetine 20 mg/day.

Q: Can the transdermal patch be used as a first-line treatment for depression?

Dr. Nemeroff: I have met with dozens of patients since the introduction of this form of selegiline. With all of my patients, I outline the advantages and disadvantages of each of the available antidepressants, including my belief that this particular agent and monoamine oxidase inhibitors in general are helpful in patients with the symptom complex of atypical depression: hypersomnia, severe anxiety, fatigue, and reverse diurnal mood variation.  I have had more and more patients say to me, “I really think that I would like to try the selegiline transdermal system. I do not like taking pills.”