Coping with the issue of drug-drug interactions (DDIs) is one of the most challenging aspects of modern psychopharmacology. Psychiatrists are treating patients with medication regimens of ever-increasing complexity.
Sponsored by CME LLC for 1.5 Category 1 credits.
Original release date 11/06. Approved for CME credit through October 2007.
After reading this article, you will be familiar with:
Who will benefit from reading this article?
Psychiatrists, primary care physicians, neurologists,nurse practitioners, psychiatric nurses,and other mental health care professionals.Continuing medical education credit is availablefor most specialties. To determinewhether this article meets the CE requirementsfor your specialty, please contact yourstate licensing board.
Dr Sandson is clinical assistant professor in the department of psychiatry at the University of Maryland Medical School and director of residency training at Sheppard Pratt Health System in Baltimore. He reports that he has no conflicts of interest concerning the subject matter of this article.
Coping with the issue of drug-drug interactions (DDIs) is one of the most challenging aspects of modern psychopharmacology. Psychiatrists are treating patients with medication regimens of ever-increasing complexity. Grappling with these regimens presents us with the daunting task of anticipating the myriad array of possible DDIs. While exhaustive tracking of all potentially significant DDIs is not currently considered the standard of care, that provides little comfort to the thousands of patients each year who suffer the clinical sequelae.1 Thus, it behooves the concerned clinician to master the essentials of this domain of pharmacotherapy.
Before plunging into this topic, some terms need to be defined. Most drugs are metabolized and inactivated before excretion. Various enzymes catalyze these metabolic processes, most of which occur in the liver. The drugs whose metabolism is catalyzed by enzymes are referred to as substrates of the en zymes. Other drugs may be coadministered with these enzymatic substrates, and these drugs may im pair the ability of enzymes to catalyze the metabolism of their substrates. Such drugs are referred to as inhibitors.
Enzymatic inhibition may occur be cause the inhibitor binds so avidly to the substrate-binding site of the enzyme that it interferes with the ability of the substrate (or cosubstrate) to gain access to the enzyme for metabolism. This form of inhibition is termed competitive inhibition. Conversely, the inhibitor may bind to a nonsubstrate-binding site on an enzyme and impair the efficiency of that enzyme in performing catalytic reactions with its substrates. This form of inhibition is termed non competitive inhibition. On the other hand, there are drugs that induce the liver to increase production of enzymes. The greater quantity and availability of enzymes leads to more efficient metabolism of sub strates of these enzymes. Such drugs are termed inducers.
Aside from their opposite effects on drug metabolism, another important feature that distinguishes enzymatic inhibition from induction is the duration of these 2 processes. Once an inhibitor becomes bioavailable in the liver, inhibition is essentially in stantaneous. However, once an inducer is present, it takes the liver several days to weeks to produce clinically meaningful decreases in associated substrate blood levels.
Phase 1, or oxidative metabolism, is the primary component involved in the metabolism and, usually, inactivation of drugs. The cytochrome P-450 system is the most important contributor to phase 1 metabolism. There are a number of distinct P-450 enzymes, identified by a number-letter-number sequence in a manner analogous to family-genus-species designa tions in the animal and plant kingdoms.
features a P-450 table with common substrates, in hibitors, and inducers for the main P-450 enzymes.
The core principles governing P-450-based DDIs are actually quite straightforward. Coadministration of substrates and inhibitors of the same enzyme(s) leads to increases in substrate levels, while coad min istration of substrates and inducers of the same enzyme(s) leads to decreases in substrate levels. How ever, the devil is in the details. The complex metabolic pathways for the broad array of available drugs and the highly variable abilities of drugs to act as clinically meaningful inhibitors and inducers of the en zymes that catalyze the metabolism of other drugs are the elements that make DDIs easy to understand in theory but difficult to anticipate and prevent in practice.
Much of the work of understanding DDIs in clinical practice deals with the specific drugs in question, how they inhibit and/or induce each other's me tabolism, the potential for toxicity and/or loss of effectiveness, and so on. However, there is another element that is critical, and that is the sequence in which drugs are added or removed from a regimen. This article will address the implications of the various sequential patterns that characterize DDIs.
SIX PATTERNS OF DDIsPattern 1: inhibitor added to a substrate
As mentioned above, this pattern results in an in crease in substrate blood levels. This increase will oc cur rapidly, within hours to days. If the substrate in question has been titrated to the appropriate dose and blood level before adding the inhibitor and if the substrate has a low therapeutic index, then there is a significant risk of drug toxicity. If one suspects an interaction between the drugs in question, then careful therapeutic drug monitoring (TDM) or lowering of the dose of substrate in anticipation of the DDI may avert drug toxicity. However, if one is un aware of the potential for a DDI between the drugs in question, then toxicity becomes more likely.
Example. Fluoxetine or paroxetine is added to nor triptyline. Nortriptyline is a substrate of P-450 2D6,2 while fluoxetine and paroxetine are both potent in hibitors of this enzyme.3,4 These DDIs have been known to increase nortriptyline levels up to 4-fold.5,6 Thus, if the nortriptyline level had been therapeutic before the addition of fluoxetine or paroxetine, then the DDI arising from this sequence would probably produce clinical tricyclic toxicity.
The rapid onset of side effects and/or frank toxicity make pattern 1 DDIs the easiest to detect, albeit after the fact.
Pattern 2: substrate added to an inhibitor
This combination also results in increases in sub strate blood levels. If a substrate with a low therapeutic index is added to an inhibitor in accordance with pre set dosing guidelines, then clinical toxicity is a significant risk. Awareness of the DDI naturally de creases the likelihood of overaggressive substrate dos ing regardless of whether the DDI follows pattern 1 or pattern 2. However, if substrate doses are titrated gradually, using therapeutic response, emergence of side effects, and/or TDM as guides, then toxicity is much less likely in a pattern 2 scenario, even if the clinician is completely unaware that a DDI is present, although the need for smaller doses of substrate than expected might arouse some suspicion.
Example. Phenytoin is added to fluoxetine or fluvoxamine. Phenytoin is a substrate primarily of P-450 2C9 and 2C19,7 while each of these SSRIs inhibits both of these enzymes.8-10 Despite this DDI leading to higher-than-expected levels of phenytoin at a given dose,11,12 cautious titration of phenytoin in accordance with clinical response and/or TDM is likely to avoid phenytoin toxicity, even if the prescriber is unaware of this DDI. The final phenytoin dose is significantly less than would be needed in the absence of these SSRIs.
Pattern 2 DDIs are easier to miss than pattern 1 DDIs, since they often produce no immediate adverse clinical sequelae. If cautious dose titration in the con text of a pattern 2 DDI, consistent with pru dent clinical practice, is sufficient to avoid most difficulties, then this begs the question of why it is useful to even bother categorizing this DDI sequence. The answer is that an unrecognized pattern 2 DDI creates the potential for a later pattern 5 DDI, reversal of inhibition (see below), which can have severe clinical consequences that are much more difficult to anticipate and prevent.
Pattern 3: inducer added to a substrate
As mentioned above, this pattern results in a decrease in substrate blood levels over the course of several days to 2 to 3 weeks. If the substrate had been titrated to an effective dose and blood level before the addition of an inducer, then this DDI poses a significant risk of producing subtherapeutic substrate blood levels. If one is aware of a DDI, then this outcome can be averted by increasing substrate doses in the days following the addition of the inducer to compensate for this effect. However, if one is un aware of the potential for a DDI between the drugs in question, then loss of effectiveness becomes more likely.
Example. Phenytoin is added to risperidone. Ris peridone is a substrate of both P-450 2D6 and 3A4,13,14 and phenytoin is an inducer of 3A4.15 Thus, the addition of phenytoin will lead to decreases in risperidone blood levels,13 with a significant risk of loss of antipsychotic effectiveness.
Although there is more of a time lag with substrate-inducer DDIs than with substrate-inhibitor DDIs, pattern 3 DDIs are basically straightforward and not too difficult to detect.
Pattern 4: substrate added to an inducer
As with pattern 3, this DDI results in lower blood levels of substrate at given doses. If one is dosing substrate according to preset dosing guidelines, then standard dosing is much less likely to provide a therapeutic response. Awareness of the DDI naturally decreases the likelihood of inadequate sub strate dosing regardless of whether the DDI follows pattern 3 or pattern 4. However, dose titration guid ed by TDM and/or the principle of increasing the dose until a therapeutic response is achieved as long as side effects are not present, is likely to pro duce a positive response at a higher-than-expected dose of the substrate, even if the clinician is unaware of the DDI.
Example. Alprazolam is added to carbamazepine. Alprazolam is a substrate of P-450 3A4,16 while car bamazepine is an inducer of this enzyme.17 Despite this DDI producing lower than expected blood levels of alprazolam at a given dose,18 titration of al prazolam according to clinical response and/or TDM will eventually produce a clinical response, albeit at a higher dose than would have been necessary in the absence of the carbamazepine, even if the prescriber is un aware of the DDI.
It is easier to miss a pattern 4 DDI than a pattern 3 DDI, since the former has less likelihood of ad verse clinical sequelae. As with pattern 2 DDIs, it is reasonable to ask why it is useful to bother categorizing this DDI sequence. The answer is that an unrecognized pattern 4 DDI creates the potential for a later pattern 6 DDI, reversal of induction (see below), which can produce serious sequelae that are especially difficult to anticipate and/or prevent.
Pattern 5: reversal of inhibition
In this scenario, a substrate and an inhibitor have been stably coadministered, producing a therapeutic blood level of the substrate. Then the inhibitor is discontinued, resulting in more efficient metabolism of the substrate and a rapid decrease in sub strate blood levels, possibly to an extent that produces a loss of effectiveness.
Example. Stable coadministration of cimetidine and doxepin is followed by discontinuation of the cimet idine. Doxepin is a tertiary amine tricyclic antidepressant (TCA) that is primarily metabolized via P-450 2D6, 3A4, and 2C19, with 1A2 serving as a secondary enzyme.19 Cimetidine is a potent pan-inhibitor of most P-450 enzymes.20,21 Thus, discontinuation of the cimetidine enables P-450 2D6, 3A4, 2C19, and 1A2 to resume their more efficient baseline levels of catalytic activity. This would lead to a decrease in the doxepin blood level (D. Benedek, personal communication, May 2002), quite possibly below the therapeutic range, with potentially dire clinical consequences.
Pattern 6: reversal of induction
In this scenario, a substrate and an inducer have been stably coadministered, producing a therapeutic blood level of the substrate. Then the inducer is discontinued, resulting in less efficient metabolism of the substrate and an increase in substrate blood levels over the next several days to weeks, possibly to an extent that produces drug toxicity.
Example. An outpatient with schizophrenia has been taking a stable dose of clozapine (producing a therapeutic blood level) while smoking 2 packs of cigarettes each day. This patient is then hospitalized in a "no smoking" unit. Clozapine is primarily metabolized via P-450 1A2, with 2C19 and 3A4 serving as the most significant secondary enzymes.22 Smoking of tobacco products is a potent and ubiquitous inducer of P-450 1A2.23 (It is worth noting that neither nicotine itself nor nonsmoking routes of tobacco use induce 1A2.) Thus, with the dis continuation of smoking, 1A2 gradually returns to its lower baseline levels, resulting in less efficient metabolism of clozapine at the outpatient dose. This leads to a gradual increase in clozapine levels.24 This specific DDI has accounted for many instances of clozapine toxicity, with significant patient morbidity.While smoking cessation is a laudable clinical goal, insufficient attention has been paid to the unintended consequence of this policy. This pattern 6 DDI can lead to adverse events when patients who smoke also take clozapine, olanzapine, warfarin, theophylline, or other P-450 1A2 substrates.24-26
Reversals of Fortune
These case examples are disturbing in terms of the challenges that they pose in recognizing and pre venting DDIs. Reversals of inhibition and induction are especially difficult to detect. Even if one is using a Palm-based DDI software program, unless one compiles a DDI profile generated by a patient's regimen and compares this with the DDI profiles generated by any addition or deletion to the pa tient's regimen, these DDIs (or "un-DDIs") are likely to be missed. The only hope for detecting pattern 5 and 6 DDIs is the emergence of electronic medical records, which can interface with physician order entry systems to alert prescribers when an addition or deletion to a regimen changes a patient's DDI pro file. Even then, there exists an inherent tension around how sensitive and specific to make such programs. If they are too sensitive and not specific enough, then clinicians have to wade through endless "false-positive" alerts and often develop "alert fatigue," in which they become habituated to the alerts and stop paying appropriate attention to them. On the other hand, if programs are too specific and not sensitive enough, then DDIs of im portance to specific patients will often be missed.
Unfortunately, the wide degree of intraspecies variation in terms of pharmacokinetic efficiency and pharmacodynamic hardiness defies the ability to set firm sensitivity-specificity parameters that will ad dress patients at both ends of the "hardiness" spectrum.
Progress is being made on sophisticated pro grams that will provide graded feedback on the likelihood and severity of DDIs, and some of these programs can tailor these outputs based on patient-specific parameters, such as P-450 enzyme polymorphisms revealed by earlier genotyping. How ever, until we have systems and programs that enable clinicians to process DDI information in an optimally accurate and efficient manner, there is no substitute for the application of our attention and intelligence to the problem of DDIs.
Until that day arrives, here are some DDI "survival tips" that will help clinicians minimize the potential impact of DDIs on their patients:
These prudent measures will go a long way toward preventing DDIs, which is usually easier than detecting and correcting them.
If you are a first-time user, you must update your profile.
If you are not on the Lifelong Learning site, click here.
You must keep your own records of this activity. Copy this information and include it in your continuing education file for reporting purposes.
CME LLC is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians. CME LLC designates this educational activity for a maximum of 1.5 category 1 credits toward the AMA Physician's Recognition Award. Each physician should claim only those credits actually spent on the educational activity.
CME LLC is an approved provider of continuing medical education for physicians, nurses and physician assistants in the State of Florida and is registered with CE Broker.
CME LLC is approved by the California Board of Registered Nursing, Provider No. CEP12748, and designates this educational activity for 1.5 contact hours for nurses. The American Nurses Credentialing Center (ANCC) accepts AMA category 1 credit toward recertification requirements.
References1. Grymonpre RE, Mitenko PA, Sitar DS, et al. Drug-associated hospital ad missions in older medical patients. J Am Geriatr Soc. 1988;36:1092-1098.
2. Venkatakrishnan K, von Moltke LL, Greenblatt DJ. Nortriptyline E-10-hydroxylation in vitro is mediated by human CYP2D6 (high affinity) and CYP3A4 (low affinity): implications for interactions with enzyme-inducing drugs. J Clin Pharmacol. 1999;39:567-577.
3. Ozdemir V, Naranjo CA, Herrmann N, et al. Paroxetine potentiates the central nervous system side effects of perphenazine: contribution of cyto chrome P4502D6 inhibition in vivo. Clin Pharmacol Ther. 1997;62:334-347.
4. Stevens JC, Wrighton SA. Interaction of the enantiomers of fluoxetine and norfluoxetine with human liver cytochromes P450. J Pharmacol Exp Ther. 1993;266:964-971.
5. Leucht S, Hackl HJ, Steimer W, et al. Effect of adjunctive paroxetine on serum levels and side-effects of tricyclic antidepressants in depressive inpatients. Psychopharmacology (Berl). 2000;147:378-383.
6. Preskorn SH, Alderman J, Chung M, et al. Pharmacokinetics of desip ramine coadministered with sertraline or fluoxetine. J Clin Psycho phar macol. 1994;14:90-98.
7. Mamiya K, Ieiri I, Shimamoto J, et al. The effects of genetic polymorphisms of CYP2C9 and CYP2C19 on phenytoin metabolism in Japanese adult patients with epilepsy: studies in stereoselective hydroxylation and population pharmacokinetics. Epilepsia. 1998;39:1317-1323.
8. Kobayashi K, Yamamoto T, Chiba K, et al. The effects of selective serotonin reuptake inhibitors and their metabolites on S-mephenytoin 4’-hydrox ylase activity in human liver microsomes. Br J Clin Pharmacol. 1995;40:481-485.
9. Olesen OV, Linnet K. Fluvoxamine-clozapine drug interaction: inhibition in vitro of five cytochrome P450 isoforms involved in clozapine metabolism. J Clin Psychopharmacol. 2000;20:35-42.
10. Sayal KS, Duncan-McConnell DA, McConnell HW, Taylor DM. Psychotropic interactions with warfarin. Acta Psychiatr Scand. 2000;102:250-255.
11. Mamiya K, Kojima K, Yukawa E, et al. Phenytoin intoxication induced by fluvoxamine. Ther Drug Monit. 2001;23:75-77.
12. Nelson MH, Birnbaum AK, Remmel RP. Inhibition of phenytoin hydroxylation in human liver microsomes by several selective serotonin re-uptake inhibitors. Epilepsy Res. 2001;44:71-82.
13. Bork JA, Rogers T, Wedlund PJ, de Leon J. A pilot study on risperidone metabolism: the role of cytochromes P450 2D6 and 3A. J Clin Psychiatry. 1999;60:469-476.
14. Fang J, Bourin M, Baker GB. Metabolism of risperidone to 9-hydroxy risperidone by human cytochromes P450 2D6 and 3A4. Naunyn Schmiedebergs Arch Pharmacol. 1999;359:147-151.
15. Raucy JL. Regulation of CYP3A4 expression in human hepatocytes by pharmaceuticals and natural products. Drug Metab Dispos. 2003;31:533-539.
16. Yasui N, Otani K, Kaneko S, et al. A kinetic and dynamic study of oral alprazolam with and without erythromycin in humans: in vivo evidence for the involvement of CYP3A4 in alprazolam metabolism. Clin Pharmacol Ther. 1996;59:514-519.
17. Spina E, Pisani F, Perucca E. Clinically significant pharmacokinetic drug interactions with carbamazepine: an update. Clin Pharmacokinet. 1996;31:198-214.
18. Arana GW, Epstein S, Molloy M, Greenblatt DJ. Carbamazepine-induced reduction of plasma alprazolam concentrations: a clinical case report.J Clin Psychiatry. 1988;49:448-449.
19. Kirchheiner J, Meineke I, Muller G, et al. Contributions of CYP2D6, CYP2C9 and CYP2C19 to the biotransformation of E- and Z-doxepin in healthy volunteers. Pharmacogenetics. 2002;12:571-580.
20. Martinez C, Albet C, Agundez JA, et al. Comparative in vitro and in vivo inhibition of cytochrome P450 CYP1A2, CYP2D6, and CYP3A by H2-receptor antagonists. Clin Pharmacol Ther. 1999;65:369-376.
21. Pies R. Cytochromes and beyond: drug interactions in psychiatry. Psychiatric Times. May 2002:48-51.
22. Olesen OV, Linnet K. Contributions of five human cytochrome P450 isoforms to the N-demethylation of clozapine in vitro at low and high concentrations. J Clin Pharmacol. 2001;41:823-832.
23. Zevin S, Benowitz NL. Drug interactions with tobacco smoking: an update. Clin Pharmacokinet. 1999;36:425-438.
24. Zullino DF, Delessert D, Eap CB, et al. Tobacco and cannabis smoking cessation can lead to intoxication with clozapine or olanzapine. Int Clin Psychopharmacol. 2002;17:141-143.
25. Faber MS, Fuhr U. Time response of cytochrome P450 1A2 activity on cessation of heavy smoking. Clin Pharmacol Ther. 2004;76:178-184.
26. Miller LG. Recent developments in the study of the effects of cigarette smoking on clinical pharmacokinetics and clinical pharmacodynamics. Clin Pharmacokinet. 1989;17:90-108.