Part II: Cardiac drug and psychotropic drug interactions: significance and recommendations

PART II: Cardiac Drug and Psychotropic DrugInteractions: Significance and Recommendations

James J. Strain, M.D., Gina Caliendo, Pharm. D., R.Ph.,Jeffrey D. Alexis, M.D., R. Sandlin Lowe, III, M.D., Anwarul Karim, M.D.,Mark Loigman, M.D.

James J. Strain, M.D.Division of Behavioral Medicine and

Consultation PsychiatryMount Sinai-NYU Medical Center/Health System1 Gustave L. Levy PlaceNew York, NY 10029

Gina Caliendo, Pharm. D., R.Ph.Department of PharmacyMount Sinai Hospital1 Gustave L. Levy PlaceNew York, NY 10029

Jeffrey D. Alexis, M.D.Cardiovascular InstituteMount Sinai-NYU Medical Center/Health System1 Gustave L. Levy PlaceNew York, NY 10029

R. Sandlin Lowe, III, M.D.Division of Behavioral Medicine and


Anwarul Karim, M.D.Division of Behavioral Medicine and


Mark Loigman, M.D.Division of Behavioral Medicine and


Abstract: Understanding cardiac drug interactions withconcurrent psychotropic prescriptions is essential for thepracticing cardiologist and primary care physician, aswell as for the psychiatrist. There has been an explosiveuse of new drugs in both psychiatry and cardiologywithout widespread knowledge of their potential interac-tions. The increasing tendency toward poly-pharmacy,the use of psychotropic medications by cardiologists andprimary care physicians caring for cardiac patients, andthe growth of the aging population present major chal-lenges for the practitioner. Finally, there is a need to havemodels/paradigms for predicting potential drug interac-tions—e.g., the Cytochrome p450 schema. This paperdescribes a method to identify, understand, and codifythe interactions between psychotropic and cardiac drugs,a systematic approach for updating this key database andspecific cardiac-psychotropic drug interactions. Specifi-cally, this paper 1) details the interactions, 2) addressesthe level of their clinical significance, 3) describes thepotential mechanism(s) of the interactions, and 4) offersrecommendations to the clinician. Since the majority of

the original clinical trials, either for cardiac medicationsor psychotropic drugs, do not include studies comparingthese two drug domains contemporaneously, their inter-actions often become known only with their combineduse in the clinical arena, using the patient as “guineapig,” and through subsequent reporting. © 1999Elsevier Science Inc.

INTRODUCTION

In 1996 we developed a review of drug-psychotropic drug interactions and end organ dys-function and considered the issue of age and phar-macokinetics [1]. It became clear that a morefocused review would be important with the ad-vancement of our knowledge in drug-drug re-actions, for example, in patients with cancer,neurological diseases, diabetes, gastrointestinal dis-orders, stroke, and heart disease [2]. This need tounderstand and communicate psychotropic and

408

medical drug interactions is further underscored bythe aging of the population and the trend towardpolypharmacy when various specialists and sub-specialists manage patients without effective com-munication among the physicians involved in thepatient’s care. The Cardiovascular Institute, the De-partment of Pharmacy, and the Division of Behav-ioral Medicine and Consultation Psychiatry at theMount Sinai-NYU Medical Center/Health Systemhave collaborated in an analysis of psychotropicand cardiac drug interactions. Since Consultation-Liaison (C-L) psychiatrists encounter cardiac pa-tients in cardiac care units, surgical intensive-careunits, medical and surgical inpatient settings, andambulatory clinics, it is mandatory that they under-stand the interaction of psychotropic and cardiacdrugs. Likewise, cardiologists, internists, and pri-mary care physicians frequently prescribe psycho-tropic medications to their heart patients.

In addition, the constant introduction of newpsychotropic and cardiac drugs to the market pre-sents an additional complication in management; inpractice, neither cardiac nor psychotropic medica-tions have been evaluated in patients in whom bothgenres of drugs are administered simultaneously.Drug interactions occur in the “natural experiment”using the patient as guinea pig when two drugs areadministered to the same patient. Such drug-druginteractions occur over time and may not be ob-served or known during initial clinical trials. Theinteraction must be studied in a series of cases, notjust in an isolated case. A given psychotropic drugmay never be given concurrently with a specificcardiac drug and consequently the potential inter-action may never be observed. This need for thenatural experiment and the patient as guinea pig isnot surprising, as it is impossible to test in vivoevery drug against every other drug, food product,and additive that patients ingest. And such testingis not required for Federal Drug Administration(FDA) approval. Because of this “wait and see”phenomenon it is essential that the physician havea methodology to track and monitor the reportingof drug-drug interactions, and to evaluate the po-tential risk for a given patient.

This paper describes those psychotropic and car-diac drug interactions that we have identified. Itdoes not include evaluation of the instruments em-ployed, publications from foreign countries not re-ported in the English-language databases we used,technical reports from pharmaceutical firms, or casereports to drug manufacturers or national informa-tion clearing houses, such as the FDA and Centers for

Disease Control. Furthermore, although we searchedMEDLINE, it does not list every published paper [3].

SEARCH METHODS

A team consisting of a cardiologist (JA),pharmacist-drug information specialist (GC), threepsychiatrists (JJS, RSL, ML) and a research assistant(AK) examined the literature on 65 cardiac drugs vs60 psychotropic drugs (Tables 1 and 2). The drugsselected for review came from two sources: 1) thosemost commonly employed by the CardiovascularInstitute and the Division of Behavioral Medicineand Consultation Psychiatry (Mount Sinai-NYUMedical Center/Health System) and 2) the Red-book list of the 200 most often prescribed drugs inthe United States.

Table 1. Commonly prescribed cardiac drugs

AdenosineAmilorideAmiodaroneAmlodipineAspirinAtenololAtorvastatinAtropineBenazeprilBisoprololBretyliumBumetanideCandesartanCaptoprilCarvedilolCerivastatinCholestyramineClonidineClopidogrelColestipolDigoxinDiltiazenDisopyramideEnalaprilEporostenolFelodipineFlecainideFluvastatinFurosemideGemfibrozilHeparinHydralazineHydrochlorothiazideIbutilide

Isosorbide dinitrateLabetalolLidocaineLisinoprilLosartanLovastatinMetolazoneMetoprololMexiletineMinoxidilNadololNicardipineNifedipineNimodipineNitroglycerinNorepinephrinePhenylephrinePravastatinPrazosinProcainamidePropafenonePropranololQuinidineSimvastatinSotalolSpironolactoneTerazosinTiclopidineTocainideTorsemideTriamtereneVerapamilWarfarin

Cardiac Drug and Psychotropic Drug Interactions

409

The search strategy consisted of searching 1)MEDLINE using the cardiac drug name, the psy-chotropic drug name, and the term “interaction”; 2)Hanston’s Drug Interaction Analysis and ManagementText (quarterly updated version) [4]; 3) Drug Inter-actions Facts [5]; 4) Micromedix Drug-dex [6]; 5) Amer-ican Hospital Formulary Service Drug Information [7];and 8) Facts and Comparisons (quarterly updatedversion) [8]. (Exclusions were case studies (see be-low), animal work, research regarding children, orreports in foreign languages).

Although there are many drug interaction books,journals, software programs, and on-line services,no one source provides complete, exhaustive datafrom all these reference sources, or offers arbitratedconclusions from disparate data reports. Few ofthese sources use a format that offers sound clinicaladvice to the practitioner.

In addition, we attempted to examine the poten-

tial interactions risks when only one of the drugs ina domain/genre had been reported to interact witha common enzyme system. What are the possibleinteractions when two drugs affect the same en-zyme system (or have a similar structure or meta-bolic pathway) and only one shows evidence of anadverse interaction? Should it be assumed that thesecond drug, despite the fact that there had been nospecific report, is also a candidate for producing anadverse response? It can be argued that the similar-ity of action in a particular enzyme system mightindicate hypotheses to what might occur (and oftenhas occurred over sufficient time periods). For ex-ample, nefazadone was a known inhibitor of thecytochrome P3A4 isoenzyme when it first came tomarket. Yet, its use with cisapride was not contra-indicated. Cisapride was a known and establishedcytrochrome P3A4 pathway-utilizing substancethat caused cardiac arrthymias and/or death whengiven with other cytochrome P3A4 inhibitors. Pru-dent practice would have prevented the use of ne-fazadone with cisapride despite the lack of docu-mented interactions. Subsequently, the use ofcisapride was officially contraindicated with ne-fazadone.

Although randomized, double-blind, placebo-controlled clinical trials are regarded as the “goldstandard” for clinical decision making, they arerarely used for studying drug interactions. Norcould studies placing patients at risk in this mannerbe used. Examination of case reports carries thedifficulty of overemphasizing the effect; does notcontrol for patient and sampling bias, confoundingfactors, outcomes, or assessment of adversity; doesnot provide the scope of patients at risk; and thefindings are not generalizeable. Yet some case re-ports may be so compelling, as with thalidomide orphen fen, that they mandate cautious drug utiliza-tion or its discontinuation. In instances such asthese the case report (without control trials) is suf-ficient for clinical decision making to avoid morbid-ity and mortality and were included.

DYNAMICS OF DRUG-DRUGINTERACTION

We developed an algorithm to determine the cir-cumstances in which drugs might interact, andwhen the clinical situation would be protective oraversive, by adding or subtracting Drug A to orfrom Drug B, and vice versa. For example, if DrugA and Drug B were started contemporaneously,

Table 2. Commonly prescribed psychotropicmedications

AmantadineAmitriptylineAmphetamineAlprazolamAmobarbitalBenztropineBuproprionBuspironeButalbitolCarbamazepineChloral HydrateChlordiazepoxideChlorpromazineCitalopramClomipramineClonazepamClozapineDesipramineDextroamphetamineDiazepamDiphenhydramineDivalproex sodiumDonepezilFluoxetineFluphenazineFluvoxamineGabapentinHaloperidolImipramineIsocarboxazidLamotrigineLithium

LorazepamLoxapineMeprobamateMethylphenidateMirtazapineMolindoneNefazadoneNortriptylineOlanzapineOxazepamParoxetinePemolinePerphenazinePhenelzinePhenobarbitalQuetiapineRisperidoneSertindoleSertralineTacrineTemazepamThioridazineThiothixeneTradozoneTranylcypromineTriazolamTrifluoperazineTrihexiphenidylValproic acidVenlafaxineZolpidem

J. J. Strain et al.

410

regular monitoring of the clinical response may beprotective in that the net effect of the interactioncould be adjusted for as the clinician treated thepatient either until improvement or toxicity.

Adding Drug A to Drug B might change theconcentration of B, which is already in the patient’ssystem, and initiate an adverse drug response.Three situations may prevail: more of B, less of B, ora chemical interaction with side effects from thesimultaneous presence of the two drugs. Beforeadding A, drug B had been at a stable therapeuticdose. On the other hand, adding B to a steady stateof drug A, which is not affected by Drug B, will notbe aversive, as the clinician may add more or less ofdrug B, depending upon the clinical situation.Thus, the clinician can compensate for the interac-tion by changing the concentration of drug B. Theeffect of A on B is compensated for as its change inpotency is monitored for drug B’s clinical effect ortoxicity. An important strategy is to take into ac-count the clinical situation and correct for the ad-dition of a second drug that may affect its concen-tration, or which renders the patient vulnerable toan aversive interaction. For example,

Affecteddrug

Start A Start B Can be accounted for: adjust B as itis started

Start A On B Must prepare for unexpected changeof B

On A Start B Can be accounted for: adjust B as itis started

Similarly, taking Drug A away from Drug B orDrug B from Drug A has to be considered: remov-ing one drug may raise or lower the concentrationof another, and may alter the clinical effects of theremaining drug and precipitate toxicity. Thus, thepractitioner must carefully note the clinical effectsof the remaining drug during withdrawal of thesecond drug. The following summaries of drug in-teractions use this system to assist the clinicianwhen prescribing or terminating drugs that haveproduced adverse clinical reactions.

FOOD AND DRUG INTERACTIONS

We did not examine the interactions between car-diac or psychotropic drugs and food, food addi-tives, nutritional supplements, or herbs. This issuch a complicated area, often bereft of randomizedcontrolled trials and would require a volume of its

own (much of the data comes from clinical obser-vations and case reports), and in most situations thepharmacokinetic interactions are not known.Grapefruit juice and its effects on the cytochromeP450 enzyme system (which will not be discussed)are an exception.

CYTOCHROME P450 SYSTEM

We did not focus on the cytochrome P450 systemalone for understanding drug-drug interactions be-cause it does not incorporate all possible mecha-nisms of interaction. In order to locate interaction/intervention information with the cytochrome P450system one must know the isoenzymes involvedwith the suspect drug or must review the entiresystem for possibilities. Much of the informationregarding this enzyme system is based on in vitroor animal data. The practicing clinician cannot keepabreast of all drugs, their metabolic pathways, andtheir impact on different isoenzymes. Knowledge ofkey interactions and the impact of key drugs ondifferent isoenzymes are imperative, but one mustknow where to find additional vital information.

MANAGEMENT OF DISCREPANCIES

Our investigational team represented three disci-plines: cardiology, pharmacy, and psychiatry. Thisallowed a consensus for determining the impor-tance of certain interactions, the frequency withwhich drugs are used by diverse specialists, andhow clinically significant the interaction may beconsidered with the need for a specific drug. Couldthe interacting drug be replaced with another agentto achieve the desired clinical effect? The opinionregarding the degree of severity of the reaction andthe potential replacement of one drug with anotherto avoid the reaction, might differ among the car-diologist, the psychiatrist, and the pharmacist. Onespecialist may not understand the entire clinicalsituation when treating a patient’s psychiatric andmedical comorbidities. By having the three disci-plines discussing judgments as to the severity andreplacement possibilities with an alternative drug,there was a more balanced approach to the suggested“advise to the practitioner.”

DRUG-PSYCHOTROPIC DRUGINTERACTION SUMMARIES

Most drug interactions are not contraindicationsbut do require increased monitoring and alteration


411

of doses or administration schedules [5,9,10]. Fewdrugs are contraindicated with end stage organdysfunction, e.g., end stage renal disease or ad-vanced age, but many require dose adjustment andcaution [11,12]. Many available drug interactiondata resources offer ongoing updates with descrip-tions of the interaction, relevance, and references tosalient literature [5,13–15]. (Although many hospi-tals have drug interaction software systems [16–18],Mount Sinai Hospital has not made this section ofinformation available to the hospital-wide system.)Drug interaction software systems may indicate thepotential severity of a response and the need fordocumentation, but not sufficiently consider theimplications for optimal clinical management. Rec-ommendations to “reduce doses” are often nonspe-cific. Finally, the number of interactions can be sooverwhelming in complicated medical situationswhen patients are on multiple medical regimens thatthe house officer may ignore or temporize his/herresponse to this information overload.

At the Mount Sinai Hospital, the pharmacy com-puter system automatically screens for drug inter-actions but does not yet provide adequate, clinicallyrelevant, management information.

Commercial drug interaction databases are oftenupdated only quarterly, and may not include newdrugs or newly recognized interactions. Some sys-tems are unable to screen all forms of a drug or toreport the required additional steps needed to per-form the screening. It has also been shown thatcommercial integration screenings too often do notprovide complete screening for all significant inter-actions. This can provide a practitioner with a falsesense of security.

RATING SYSTEM: SIGNIFICANCELEVELS

Most rating systems indicate 1—major significance,2—moderate significance, and 3—minor signifi-cance. Significant Level 1 interactions require theclinician to document why he/she is prescribingthis combination, and to use both drugs only ifthere is no other way to manage the patient. Drugcombinations producing an interaction with a Sig-nificance Level 1 are combinations that result inserious and potentially life-threatening adverse ef-fects such as arrhythmia. Obviously, if this combi-nation is to be used, the drug(s) in question must beprescribed with an explanation as to the need for itsconcurrent use and cautious monitoring. The opti-

mum choice, if possible, is to use another drug toavoid Significance Level 1 interactions. With Signif-icance Level 2, the potential interaction must bedocumented and the clinical outcome(s) must bemonitored carefully so that unacceptable, perni-cious reactions may be halted as soon as possible.Again, it is often better to use another drug whenpossible. In any event, it is incumbent upon theclinician to provide careful documentation that ad-verse effects were considered when using this drugcombination. It is also mandatory to communicateto the patient’s health care providers the potentialinteractions and adverse outcomes. SignificanceLevel 3 may not preclude the use of a specific drug,but clinical decision making needs to considerwhether the untoward side effects, (e.g., nausea,rash) could be avoided by another drug choice.Again, the potential interaction and its mechanismshould be documented in the patient’s medicalchart and patient’s health care providers need to beinformed.

METHODS TO KEEP CURRENT

In order to keep abreast of the ever-unfolding andexpanding information regarding drug interac-tions, a system for updating medication interac-tions needs to be established by each clinician. Sincethis millennium issue of General Hospital Psychi-atry will be sent to most medical schools in theworld, many of whom will not have access to thejournals in which this information has been presentedby the authors before, it was decided that it needed tobe reiterated in this issue [2].

Drug Information Services

Pharmaceutical. Approved drug information is pri-marily founded on carefully controlled Phase I, II,and III studies [19]. The drug’s package insert datais based on preclinical and clinical studies and de-scribes contraindications, drug interactions, and ad-verse reactions. After a drug is approved, the com-pany continues to collect adverse reaction data andinteraction reports from controlled and uncon-trolled studies, physicians, pharmacists, and pa-tients [19]. The database of individual reports cancause a sampling bias in that only those cases re-ported are documented; other adverse responsesmay not be called in. Questions also arise: Whoreports adverse reactions (the majority remain un-reported)? The nature and reliability of the descrip-tions, and whether a reaction is related to the com-

J. J. Strain et al.

412

bination of drugs in questions or is an associated,but not causally related event all impugn the reli-ability of the information [17,18]. Psychotropic-medical drug interactions require time for reportingand systematic study before their significance canbe adequately assessed. As stated earlier, the hu-man must be a guinea pig; the interaction andexperience must be reported before sufficient con-cern warrants a firm contraindication.

Institutional Systems. Electronic patient recordsystems have become more common in institutionsto forestall inadvertent therapeutic duplication [16].Often these systems do not have a drug-drug com-ponent since great concern remains over the inter-pretation of the information provided, and the pos-sibility that a clinician who did not find thewarning or alert would assume that no potentialproblem exists. Those systems that provide auto-matic drug interaction screenings initially appear tobe excellent for the primary prescriber to ascertainthe presence of drug interactions. However, practi-tioners should be aware of the limitations and dan-gers of these screening tools: some systems screenonly routine medications, leaving interactions withother required medications undetected. Intrave-nous medications may not be included in thescreening. Consequently, the absence of any warn-ing may give the prescriber a false sense of safety.Yet, excessive numbers of cautions may cause theclinician to become insensitive to their significance.If five interactions are detected, the prescriber maynot fully appreciate the relevance of the only im-portant interaction. Some systems permit the settingof a predetermined level of significance for notifica-tion to eliminate incidental interactions with the dan-ger that significant interactions are then missed.

Our use of trained pharmacists to access essentialinteraction data from sources not commonly knownto most clinicians (including C-L psychiatrists)avoids many of the pitfalls described above. Phar-macists’ references are often updated quarterly andprovide a description of drug interactions, mecha-nisms of action and relevance, with a short litera-ture review, and references [15]. Hospital pharma-cies commonly employ computerized patientprofile systems linked to drug interaction screeningsoftware, which automatically screen for drug in-teractions when new drugs are added to a patient’sregimen [15–17,21]. No system screens a medicationregimen for all interactions with complete accuracy.These systems may indicate the potential severityand do not fully consider the implications of proper

management. Such management suggestions as“reduce doses” are often nonspecific [15,22].

This is illustrated by the interaction between flu-conazole and cisapride which is given a significancerating of 3 (intermediate concern) in some systemsbecause published case reports do not exist, eventhough the manufacturer’s package insert contrain-dicates the combination [23]. Cisapride and flucon-azole should not be used in combination because ofthe risk of inhibited cisapride metabolism, in-creased cisapride levels, and resultant cardiac ar-rhythmias.

Drug Interaction Information Sources

1. MEDLINE is a complex searching system thatcan identify myriad articles which must be win-nowed tediously to locate the specifics in ques-tion. (Only 1%–2% of all citations for a specificchemotherapeutic agent address the generaltopic of drug interactions and very few of theseinvolve cardiac drug-psychotropic drug interac-tions [24,25]). The majority of physicians are notsufficiently experienced to conduct an efficientMedline search. Moreover, the Cochrane Collab-oration (Oxford University, Oxford, England)which systematically evaluates random controlclinical trials (RCTs) to identify those that qualifyfor a metanalytic study, reports that about one-third of published papers do not appear in MED-LINE [26]. Therefore, the Cochrane Collabora-tion requires hand searches of the relevantjournals as an essential preliminary step [3,26].(Cochrane evaluations are conducted by expertsin the field using a rigorous protocol to assessthe quality of the science and therefore needaccess to all the data extant.)

2. REACTIONS is published biweekly by the Aus-tralian Drug Information System (ADIS) in hardcopy, and quarterly on the CD-ROM version ofMICROMEDEX. REACTIONS screens some 7500health care journals, although it does not rou-tinely review medical malpractice litigation pub-lications. Each article is abstracted by a medicalwriter on staff at ADIS, providing an indepen-dent perspective. The focus of REACTIONS isadverse drug events and drug interactions.

3. AMERICAN HOSPITAL FORMULARY SER-VICE DRUG INFORMATION (AHFS) is an ex-cellent drug information resource on CD-ROMand hard copy. It is published annually by theAmerican Society of Health-System Pharmacistsand updated quarterly. It offers drug interac-


413

tions that include the mechanism, time course,outcome, risk factors (if available), managementand techniques to minimize adverse events. TheCD-ROM version has references at the end ofeach monograph while they are not includedwithin the text of the hard copy; if a reference fora particular statement is required, the user needsto contact the publisher who provides the refer-ence in 24–48 hours.

4. The Physicians Desk Reference (PDR) also avail-able on CD-ROM, can be installed on a portablecomputerized reference tool, and is also avail-able via MICROMEDEX. The primary limitationwith this format, as with the print version is thatit contains only the FDA approved package la-beling. The data is far from complete as it doesnot always offer the clinician practical advicewhen needed. For instance, the pharmacokineticdrug interaction study may be included, butwithout showing the impact on patients, or howto manage the interaction.

5. The MICROMEDEX on line system has sectionsfor different drug information systems [22]. It isthe most frequently employed computerized da-tabase in the inpatient setting with subset data-bases emphasizing special areas.

6. The DRUGDEX system focuses on drugs, andprovides extensive monographs, including druginteractions from references and case reports.DRUGDEX is the primary database for drug in-teraction data. It is organized by drug and con-tains an independent summary of articles in astandardized format. Because it is independentlypublished, it is not limited to FDA agents. There-fore it is an excellent resource for unapprovedagents, unapproved uses, adverse events, anddrug interaction data not described in thepackage insert. DRUGDEX is easier to use thanMEDLINE, as much of the data is already sum-marized. It is well referenced, permitting theuser to locate the original report and review it. Itis not all inclusive and must be read carefully.Errors in coding and data entry occur, but cor-rections appear in the quarterly updates. Thesystem is best used only as an initial guide to becomplemented by additional references.

7. The POISINDEX system concentrates on treatingoverdoses due to common drugs and householditems including plants.

The MICROMEDEX system contains additionaldatabases: reproductive toxicology, patient educa-tion systems, care plan development tools, hazard

communications systems, foreign compendia, phar-macokinetics programs, miscellaneous clinical data,emergency care information, and other indexingand abstracting systems. The availability of itemsdepends on the system each institution uses, itsfinancial limitations, and utilization goals. Each ad-ditional database increases the cost. MICROMEDEXhas three hardware options: CD-ROM, main-frame, and Intranet. Institutions may have variantsof the database in different locations: Whereasthe emergency room could use the POISINDEX,EMERGINDEX, the patient education andDRUGDEX programs, the standard patient careunit would have limited use for the first two, butwould use the patient education, DRUGDEX, andperhaps the reproductive toxicology system.

8. The DRUG-REAX (within the MICROMEDEXsystem) is an interactive drug database that al-lows the practitioner to enter an entire drug reg-imen, to receive a summary of the drug-drug,drug-food, and drug-disease interaction withreferences [22]. Its major shortcoming is thatonly single interactions are presented. For exam-ple, if a patient is on five medications, the effectsof combining all five are not evaluated.

9. MARTINDALES: THE EXTRA PHARMACO-POEIA, is another MICROMEDEX resource avail-able in electronic as well as print formats that offersinformation about medications available in Eu-rope. It is especially helpful for agents new to theUnited States, but which have been in clinical usein Europe for some time. MARTINDALES’ advan-tage is that the reference citation immediately fol-lows the discussion. However, it is not all inclusiveand not every section has a quarterly update.

Monitoring Services and Abstracting Systems

Two drug interaction monograph systems that areavailable, 1) DRUG INTERACTIONS ANALYSISand 2) MANAGEMENT BY APPLIED THERAPEU-TICS AND DRUG INTERACTION FACTS BYFACTS AND COMPARISONS appear in hard copyand are updated quarterly. DRUG INTERACTIONFACTS updated quarterly is on CD-ROM. Interac-tions are coded for significance that takes into ac-count the probability of the interaction occurring,the potential patient outcome, and the quality of thedata documenting the interaction. Each monographbegins with a brief description of the interaction, itsmechanism and severity, the onset time, and anevaluation of the literature. These systems are help-

J. J. Strain et al.

414

ful because they cross-reference agents similar tothose in published reports. Though every drug in aclass may not interact in the same way with others,these services alert practitioners to the possibility ofinteractions occurring with similar agents.

Although they are not computer systems per se,several journals and periodicals publish drug-druginteractions, although they are often derived fromclinical trials and case reports:

Annals of Pharmacotherapy describes drug therapyand therapy outcomes.

Pharmacotherapy published by the American Collegeof Clinical Pharmacy focuses on human thera-peutics.

Clinical Pharmacology and Therapeutics emphasizes onpharmacokinetics and healthy volunteer studies.

Drug Safety, published by the Australian Drug In-formation System, focuses on adverse drugevents, including drug interactions.

Clinical Pharmacokinetics, produced by the Austra-lian Drug Information System, emphasizes phar-macokinetics, including healthy volunteer research.

Drugs, published by the Australian Drug Informa-tion System, describes new drugs, with in-depth,well-referenced reviews.

Current Contents/Adverse Reactions and ExcerptaMedical/Adverse Reactions reviews journals em-phasizing adverse events during drug therapy;offers occasional abstracts.

American Journal of Health-System Pharmacy, CurrentLiterature (a section of the Journal of the AmericanSociety of Health-System Pharmacists) providesbrief abstracts organized by topics. It is not in-dexed, but is an important resource for new cita-tions on drug interactions and adverse eventsand its perusal of drug therapy literature.

Clin-Alert reports from 800–900 journals, includingthose covering medical malpractice litigation. Re-views appear within 1–2 months of an article’spublication; contains a commentary not writtenby the original authors.

Clinical Abstract/Current Therapeutic Findings pub-lishes abstracts, notes, and reports of drug interac-tions, pharmacokinetics, and general therapeutics.

Food and Drug Administration

1. Most commonly, updates are in the form of casereports and letters to the editors, whereas theweekly F-D-and-C Pink Sheets presents regula-tory issues for new drugs, drugs close to ap-proval, and important labeling changes for cur-

rent drugs. The table of contents can be accessedvia the Internet; electronic subscriptions areavailable. The F-D-and-C Pink Sheets reviewsregulatory concerns and discussions; safety is-sues are frequently reviewed in detail and up-coming FDA proceedings are listed. The primarylimitation of this important resource remains itscost and the amount of financial, regulatory, andcorporate information incorporated.

2. The FDA serves drug information centers andany interested health care practitioners whohave e-mail addresses. Thereby drug informa-tion centers have relevant information that theaverage practitioner may not be aware of. Mem-bers receive updated safety information viae-mail as soon as it becomes available. Notifica-tion of Dear Health Care Professional letters,labeling changes, significant recalls, and new ap-provals are also distributed via e-mail. Poten-tially significant adverse reactions and manage-ment guide lines are described. This servicerecently alerted practitioners to changes in thepackage labeling for isotretinoin (accutane) [26]which relate to the occurrence of depression andsuicidal thoughts and attempts. Since the DearHealth Care Profession data is derived from post-marketing adverse-event reports, it does not allowa definitive cause-effect mechanism to be deter-mined. Practitioners can subscribe to the FDA ser-vice by sending an e-mail to [email protected].

USE OF THE INTERNET

The Internet provides ready access to a wealth ofmedical information but unfortunately, is of vari-able quality. Many pharmaceutical companies keepupdated package inserts and allow the submissionof questions to their medical information depart-ments via their websites. The FDA and CDC web-sites allow practitioners to locate safety informationand labeling changes. There are also on-line druginformation centers located on-line, but turn-around time can be long, and there is no real-timediscussion between the center and the inquirer. Ad-ditionally, there is no way to verify the accuracy ofthe information. Many search engines can direct theuser to health care websites, but not all informationon websites is derived from reputable or knowl-edgeable sources. Many free and subscription-required journals and literature monitoring servicesare available on-line. Appropriate use of Internetresources may improve patient care and enhanceprofessional knowledge, but determining the reli-


415

ability of the information acquired remains the cli-nician’s responsibility. World Wide Web sites areCytochrome p-450 www.accp.org; FDA (safety in-formation): www.fda.gov/medwatch; medical jour-nals updates: www.mol.net/medj.htm.

This work was funded by the Green Fund, Inc., New York, New York.

REFERENCES

1. Strain JJ, Caliendo G, Himelein C: Part II: Drug-psychotropic drug interactions and end organdysfunction: Clinical management recommenda-tions, selected bibliography, and updating strategies.Gen Hosp Psychiatry 18:294–375, 1996.

2. Strain JJ, Caliendo G, Himelein C: Using computerdatabases to predict and avoid drug-drug interac-tions in the cancer patient requiring psychotropics.Psychooncology 7:321–332, 1998.

3. Chalmers I, Dickersin K, Chalmers TC: Getting togrips with Archie Cochrane’s agenda. Br Med J 305:786–788, 1992.

4. Hansten PD, Horn JR, Koda-Kimble MA, Young LY(eds): Drug Interactions and Analysis and Manage-ment. Facts and Comparisons. St. Louis, MO, quar-terly Update, 1999.

5. Tatro DS (ed): Drug Interaction Facts. St. Louis: Factsand Comparisons, 1998. Quarterly Update 1999.

6. Micromedex Drugdex, Micromedex, Vol 97, Engle-wood, NJ, 1998.

7. McEvoy GK (ed): American Hospital Formulary Ser-vice Drug Information 1999. American Society ofHealth System Pharmacists. Bethesda, MD.

8. Sewester CS, Olin B, Hebel SK, et al (eds): Drug Factsand Comparisons. St Louis: Facts and Comparisons,1999. Monthly Update March 1999.

9. May JR: Adverse drug reactions and interactions. InDipiro JT, Talbert RL, Hayes PE, et al. (eds),Pharmcotherapy: A Pathophysiologic Approach, 2nd

ed. New York, Elsevier, 1992.10. Slaughter RI, Edwards DJ: Recent advanced: the cy-

tochrome p450 enzymes. Ann Pharacother 29:619–624, 1995.

11. Peck CC, Rooman JH: Analysis of clinical pharmaco-kinetic data for individualizing drug dosage regi-mens. In Evans WE, Schentag JJ, Jusko WJ (eds),Applied Pharmacokinetics: Principles of Therapeutic

Drug Monitoring, 2nd ed. Spokane, Applied Thera-peutics, 1986.

12. Nies AS: Principles of therapeutics. In Gilman AG,Rall TW, Nies AS, et al. (eds), The PharmacologicBasis of Therapeutics, 8th ed. Elmsford, New York,Pergamon Press, 1990.

13. Collins GE: Searching and organizing the profes-sional literature. In Brown TR, Smith ML (eds),Handbook of Institutional Pharmacy Practice, 2nd ed.Baltimore, Williams & Wilkins, 1986.

14. Jankel CA, Speedie SM: Detecting drug interactions: areview of the literature. DICP. Ann Pharmacother24:282–289, 1990.

15. Poirier TI, Giudici R: Evaluation of drug interactionsmicrocomputer software: an updated comparison.Hosp Pharm 30:888–890, 893–894, 1995.

16. Ryan JL, Rinke R, deLeon R: Selecting a pharmacycomputer system for the future. Pharm Pract ManageQ 15:1–14, 1995.

17. Dasta JF: Enhancing the pharmacists computer inter-face. Ann Pharmacother 26:99–100, 1992.

18. Rosenberg JM, Fuentes RJ, Starr CH, et al: Directoryof pharmacist-operated drug information centers inthe United States. J Pharm Technol 11:1–5, 1995.

19. Colvin CL: Drug information services provided bythe pharmaceutical industry. Am J Hosp Pharmacol47:1989–2001, 1990.

20. Rosenberg JM, Fuentes RJ, Starr CH, et al: Pharma-cist-operated drug information centers in the UnitedStates. J Pharm Technol 11:1–5, 1995.

21. Lee PM: Automation and the future practice ofpharmacy: changing the focus of pharmacy. Pharm-col Pract Manag Q 15:23–25, 1995.

22. Anon. Learning Guide. Englewood, NJ, MicromedexInc. Medical Economics, 1995.

23. MedAson Information Overview. First Data BankInc. San Bruno, CA. 1997.

24. Kivisto KT, Kroemer HK, Eichelbaum M: The role ofhuman cytochrome p450 enzymes in the metabolismof anticancer agents: implications for drug interac-tions. Br J Clin Pharmacol 40:523–530, 1995.

25. Dear Doctor Letter. Important changes to the pre-scribing information for accutane (isotretinoin).Roche Laboratories, Nutley, NJ February 1998.

26. The Cochrane Collaboration. The Cochrane Databaseof Systematic Reviews. BMJ Publishing Group, Lon-don 1999.

APPENDIX

CARDIAC DRUG–PSYCHOTROPICDRUG INTERACTIONSACE Inhibitors (Enalapril)/Tricyclic Antidepres-sants (Clomipramine)

Significance level 2: ModerateAdvice to Practitioner: The addition of clomipramineto long-standing enalapril therapy results in an in-crease of serum clomipramine with signs of toxicitydeveloping (confusion, insomnia, irritability, andmood changes). If a patient is stabilized on thecombination, there is little need to alter therapy.

J. J. Strain et al.

416

However, when one drug is added to the other,serum clomipramine levels should be monitoredand the dosage of the clomipramine adjusted ac-cordingly. Patients should be cautioned about thepotential for toxicity and its signs and symptoms.Action: Monitor clomipramine concentrationswhen using enalapril and clomipramine in com-bination.

ACE Inhibitors/PhenothiazinesSignificance level 2: ModerateAdvice to Practitioner: Concurrent use of an ACEinhibitor and a phenothiazine (particularly chlor-promazine) can result in synergistic additive phar-macologic effect leading to hypotension and/ororthostasis/postural intolerance. If a patient is sta-bilized on the combination, there is little need toalter therapy unless a patient has existing conges-tive heart failure. When one drug, however, isadded to the other, patients should be cautionedabout the potential for orthostasis and/or hypoten-sion and the appropriate monitoring techniquesshould be performed.Action: Monitor for hypotension and orthostasiswhen adding one drug to the other.

Adenosine/CarbamazepineSignificant level 1: MajorAdvice to Practitioner: Adenosine has important in-teractions with a number of drugs but with psycho-tropics they do not have notable interaction exceptfor carbamazepine that has the effect of slowing AVconduction. When these drugs are prescribed con-comitantly, they may act synergistically and lead toa fatal heart block.Action: Action should be judged on a prioritybasis as adenosine is mostly administered in crit-ical situations and its half-life is very short (6minutes). In that case no action need be under-taken. But if administered for a longer time con-comitantly, carbamazepine should be withheldfor 4 days prior to the administration of adeno-sine.

Amiodarone/ClonazepamSignificance level 3 (Case Report): MinorAdvice to Practitioner: Amiodarone is a Class IIIantiarrhythmic which has been shown to inhibit themetabolism of drugs cleared by the oxidative mi-crosomal enzymes. Clonazepam undergoes exten-sive hepatic metabolism, primarily by reductionand acetylation, and is therefore susceptible to al-tered disposition during concomitant administra-tion of agents that inhibit hepatic microsomal en-

zymes. Hypothyroidism, occurring in up to 32% ofpatients treated with amiodarone, can also alterdrug metabolism and sensitize the central nervoussystem, thus increasing the potential for toxicity. Ina published report, a patient taking amiodarone tosuppress sustained ventricular tachycardia devel-oped signs and symptoms of benzodiazepine tox-icity when clonazepam was subsequently initiated.These signs and symptoms cleared upon discontin-uation of clonazepam.Action: The complex pharmacologic profile ofamiodarone and its effects on multiple organ sys-tems necessitates close patient monitoring duringconcurrent administration of medications such asclonazepam which are cleared by oxidative me-tabolism.

Amiodarone/TrazodoneSignificance level 1 (Case Report): MajorAdvice to Practitioner: Trazodone, a second genera-tion antidepressant, has been shown to be equal tostandard tricyclics in clinical efficacy while causingsignificantly fewer cardiac side effects. Althoughclinical trials have demonstrated a minimal arryth-mogenic profile, trazodone has been implicated inthe pathogenesis of ventricular arrythmias. A pub-lished report describes a patient in whom the ad-ministration of trazodone in addition to the previ-ously well-tolerated long-term amiodarone therapywas associated with a marked prolongation of theQT interval and polymorphous ventricular tachy-cardia.Action: Consider an alternative agent. Caution isrecommended in the concurrent use of trazodoneand other medications (i.e., amiodarone) known tocause QT prolongation and polymorphous ven-tricular tachycardia.

Atropine/AmantadineSignificance level 2: ModerateAdvice to Practitioner: Anticholinergic side effectsmay be increased in patients treated with anticho-linergics and amantadine. Although this interactionhas not been reported with the anticholinergic at-ropine, a similar interaction is theoretically possi-ble. The short half-life of atropine, which is usuallyused in emergency situations, may make this inter-action less likely.Action: Atropine may be used in emergency situ-ations in patients treated with amantadine.


417

Atropine/HaloperidolSignificance level 2: ModerateAdvice to Practitioner: Worsening of schizophrenicsymptoms, decreased haloperidol levels, and tar-dive dyskinesia have been reported. Although thisinteraction has not been reported with the anticho-linergic atropine, a similar interaction is theoreti-cally possible. The short half-life of atropine, whichis usually used in emergency situations, may makethis interaction less likely.Action: Atropine may be used in emergency situ-ations in patients treated with haloperidol.

Atropine/PhenothiazinesSignificance level 2: ModerateAdvice to Practitioner: Therapeutic actions of phe-nothiazines may be decreased by anticholinergics.An increase in anticholinergic side effects may benoted. Although this interaction has not been re-ported with the anticholinergic atropine, a similarinteraction is theoretically possible. The short half-life of atropine, which is usually used in emergencysituations, may make this interaction less likely.Action: Monitor for anticholinergic side effects.Adjust dose of phenothiazines as necessary.

Beta Blockers/BarbituratesSignificance level 2: ModerateAdvice to Practitioner: Barbiturates may induce themetabolism of some beta blockers. A reduced ther-apeutic response to beta blockers metabolized bythe liver (propranolol, metoprolol, and alprenolol)is the result of this interaction. Interactions of thistype present problems when the drug that altershepatic metabolism is added after a patient is sta-bilized on the affected drug. A problem may alsoarise if a patient is stabilized on the affected drugwhile receiving an enzyme-altering drug and thatdrug is discontinued. Interactions of this type maytake a few days to become evident and resolvebecause effects of the enzyme-altering agent maytake a few days to become evident or dissipate. If apatient is stabilized on the combination, there islittle need to alter therapy. If a patient is on abarbiturate and a hepatically metabolized betablocker is added, the dose of the beta blocker willbe determined by clinical efficacy at the start oftherapy and over time. If a patient is on a hepati-cally metabolized beta blocker and a barbiturate isadded, serum levels of the beta blocker may de-crease during combination therapy. In this case,blood pressure and heart rate should be monitored.Action: Monitor blood pressure and heart ratewhen a barbiturate is added to or withdrawn from

a hepatically metabolized beta blocker (propran-olol, metoprolol and alprenolol).

Beta Blockers/Calcium Channel BlockersSignificance level 2: ModerateAdvice to Practitioner: Calcium channel blockers areused in psychiatry as augmentation agents, moodstabilizers, and for tremor/movement side effects.Additive negative inotropic effects may be seenwhen this combination is used. However, it is usedfrequently for the treatment of hypertension with-out deleterious effects. If a patient has been stabi-lized on the combination, there is little need to altertherapy. When adding one drug to the other, pa-tient’s blood pressure and heart rate should bemonitored frequently to avoid possible hypotensivestates.Action: Additive negative inotropic effects mayresult when adding one drug to the other. Monitorblood pressure and heart rate frequently whencombining these drugs.

Beta Blockers/PhenothiazinesSignificance level 2: ModerateAdvice to Practitioner: Each of these agents may in-hibit the metabolism of the other. Interactions ofthis type present problems when the drug that al-ters hepatic metabolism is added after a patient isstabilized on the affected drug. A problem may alsoarise if a patient is stabilized on the affected drugwhile receiving an enzyme-altering drug, which isthen discontinued. Interactions of this type maytake a few days to become evident and resolvebecause effects of the enzyme-altering agent maytake a few days to become evident or dissipate. If apatient is stabilized on the combination, there islittle need to alter therapy. Since both of these drugsinhibit the metabolism of the other, enhanced ef-fects of both drugs may be seen when one is addedto the other. Augmented hypotensive effects of eachmay be seen.Action: Monitor clinical efficacy when adding oneto the other. The most significant interactions oc-cur with propranolol in combination with chlor-promazine or thioridazine.

Beta Blockers (Lipophilic Hepatically Metabo-lized)/Selective Serotonin Reuptake Inhibitors(Fluoxetine, Sertraline, Fluvoxamine, Paroxetine)Significance level 2: ModerateAdvice to Practitioner: The combination of a selectiveserotonin reuptake inhibitor (SSRI: fluoxetine, ser-traline, fluvoxamine, paroxetine) and a lipophilic,hepatically metabolized beta blocker (propranolol,

J. J. Strain et al.

418

metoprolol, labetalol) may result in an increase inthe beta adrenergic blocking effect and cardiac tox-icity. The mechanism is thought to involve inhibi-tion of oxidative metabolism of the beta blocker bythe SSRI (particularly P450 2D6 by fluoxetine andits metabolite norfluoxetine) and its interactionwith 5-HT receptors in the atrium of the heart.Fluvoxamine and sertraline appear to be less potentP450 2D6 inhibitors. Likewise, the use of a water-soluble, renally cleared beta blocker (sotalol, ateno-lol) is not subject to this interaction. If a patient hasbeen stabilized on the combination, there is littleneed to alter therapy. When adding one drug to theother, patient’s blood pressure and heart rateshould be monitored frequently to avoid possiblehypotensive and bradycardic states and electrocar-diograms should be obtained to monitor for dys-rhythmias.Action: Additive negative inotropic effects mayresult when adding one drug to the other. Monitorblood pressure and heart rate frequently whencombining these drugs. Obtain serial ECGs tomonitor for dysrhythmias and discontinue thecombination if a dysrhythmia develops.

Bretylium/Amphetamines, Methylphenidate, Mono-amine Oxidase InhibitorsSignificance level 1: MajorAdvice to Practitioner: When used in the treatment ofhypertension: the anti-hypertensive effects of bre-tylium are reportedly blocked or reversed by theconcomitant administration of amphetamines,methylphenidate, or monoamine oxidase inhibi-tors. It is not known if the antiarrhythmic effects ofbretylium are affected by concommitant use ofthese drugs.Action: Carefully observe clinical state when add-ing these drugs to bretylium.

Bretylium/Tricyclic AntidepressantsSignificance level 1: MajorAdvice to Practitioner: Concomitant bretylium andtricyclic antidepressant therapy (desipramine, dox-epin, imipramine) has been reported to result ininterference with bretylium’s effects.Action: Tricyclic antidepressant medicationsshould not be used concomitantly with bretyliumif possible.

Calcium Channel Blockers (Dihydropyridinesand Non-Dihydropyridines)/BarbituratesSignificance level 2: ModerateAdvice to Practitioner: Barbiturates may induce themetabolism of calcium channel blockers. A reduced

therapeutic response to calcium channel blockersmetabolized by the liver is the result of this inter-action. Interactions of this type present problemswhen the drug that alters hepatic metabolism isadded after a patient is stabilized on the affecteddrug. A problem may also arise if a patient is sta-bilized on the affected drug while receiving anenzyme-altering drug, and that enzyme-alteringdrug is discontinued. Interactions of this type maytake a few days to become evident and resolvebecause effects of the enzyme-altering agent mayrequire time to become evident or dissipate. If apatient is stabilized on the combination, there islittle need to alter therapy. If a patient is on abarbiturate and a hepatically metabolized calciumchannel blocker is added, the dose of the calciumchannel blocker will be determined by clinical effi-cacy at the start of therapy and over time. If apatient is on a hepatically metabolized calciumchannel blocker and a barbiturate is added, serumlevels of the calcium channel blocker may decreaseduring combination therapy. In this case, bloodpressure and heart rate should be monitored.Action: Monitor blood pressure and heart ratewhen a barbiturate is added to or withdrawn froman hepatically metabolized calcium channelblocker (dihydropyridines and non-dihydro-pyridines).

Calcium Channel Blockers (Non-Dihydropyri-dines)/BenzodiazepinesSignificance level 2: ModerateAdvice to Practitioner: Some non-dihydropyridinecalcium channel blockers (diltiazem and verapamil)may inhibit the hepatic metabolism of some benzo-diazepines (in particular the triazolol benzodiaz-epines, e.g. alprazolam and triazolam) and increasebenzodiazepine concentrations. This mechanismappears to be the result of competition for the cy-tochrome P450 3A4 isozyme. Interactions of thistype present problems when the drug that altershepatic metabolism is added after a patient is sta-bilized on the affected drug. Difficulty may alsoarise if a patient is stabilized on the affected drugwhile receiving an enzyme-altering drug, and thatenzyme-altering drug is discontinued. Interactionsof this type may take a few days to become evidentand resolve because effects this enzyme-alteringagent may require time to become evident or dissi-pate. If a patient is stabilized on the combination,there is little need to alter therapy. If a patient is ondiltiazem or verapamil and alprazolam or triazo-lam is added, the dose of the benzodiazepine will


419

be determined by clinical efficacy, and no addi-tional intervention is required. If a patient is on thebenzodiazepine and diltiazem or verapamil isadded, serum levels of alprazolam or triazolammay be increased, and the patient will experienceincreased and prolonged sedative/hypnotic effects.Action: When combining a non-dihydropyridinecalcium channel blocker (diltiazem or verapamil)with a triazolol benzodiazepine (alprazolam ortriazolam), monitor the patient’s blood pressureand heart rate and watch for increased and pro-longed sedative/hypnotic effects. Because of thecompetitive inhibition in the 3A4 isozyme system,clinical efficacy will occur at lower doses.

Calcium Channel Blockers (Dihydropyridines)/CarbamazepineSignificance level 2: ModerateAdvice to Practitioner: The concurrent use of somedihydropyridine calcium channel blockers (felodip-ine and nimodipine) with carbamazepine results indecreased serum levels of the dihydropyridineagents. Carbamazepine is a potent inducer of thehepatic mixed function oxidase system, and sincefelodipine and nimodipine are extensively metabo-lized by this enzyme system, there is an increase inthe first pass metabolism of these agents resultingin decreased bioavailability and efficacy. Interac-tions of this type present problems when the drugthat alters hepatic metabolism is added after a pa-tient is stabilized on the affected drug. A problemmay also arise if a patient is stabilized on the af-fected drug while receiving an enzyme-alteringdrug, and that enzyme-altering drug is discontin-ued. Interactions of this type may take a few days tobecome evident and resolve because effects of theenzyme-altering agent may require time to becomeevident or dissipate. If a patient is stabilized on thecombination, there is little need to alter therapy. Ifa patient is on carbamazepine and felodipine ornimodipine is added, the dose of the calcium chan-nel blocker will be determined by clinical efficacy atthe start of therapy and over time. If a patient is onfelodipine or nimodipine and carbamazepine isadded, clinical efficacy of the calcium channelblocker should be monitored.Action: Monitor clinical efficacy of the dihydro-pyridine calcium channel blocker (felodipine ornimodipine) with concurrent carbamazepinepharmacotherapy.

Calcium Channel Blockers (Non-Dihydropyri-dines)/CarbamazepineSignificance level 2: Moderate

Advice to Practitioner: Some non-dihydropyridinecalcium channel blockers (diltiazem and verapamil)may inhibit the hepatic metabolism of carbamaz-epine and increase carbamazepine concentrations.Interactions of this type present problems when thedrug that alters hepatic metabolism is added after apatient is stabilized on the affected drug. A problemmay also arise if a patient is stabilized on the af-fected drug while receiving an enzyme-alteringdrug, which is discontinued. Interactions of thistype may require time to become evident and re-solve because effects of the enzyme-altering agentmay take a few days to become evident or dissipate.If a patient is stabilized on the combination, there islittle need to alter therapy. If a patient is on dilti-azem or verapamil and carbamazepine is added,the dose of carbamazepine will be determined byserum concentrations and clinical efficacy, and noadditional intervention is required. If a patient is oncarbamazepine, and diltiazem or verapamil isadded, serum levels of carbamazepine may be in-creased. Serum carbamazepine levels and clinicalefficacy should be monitored.Action: Monitor serum carbamazepine levelswhen a non-dihydropyridine calcium channelblocker (diltiazem or verapamil) is added.

Calcium Channel Blockers (Non-Dihydropyri-dines)/LithiumSignificance level 2: ModerateAdvice to Practitioner: Although this combination isused frequently with no untoward effects, there arereports where some non-dihydropyridine calciumchannel blockers (diltiazem and verapamil) whenadministered concurrently with lithium have re-sulted in both a reduction in lithium levels andlithium toxicity, as well as cardiac toxicity. Associ-ated signs and symptoms range from worsening ofmania to psychosis and neurotoxicity. The mecha-nisms are unknown. If a patient is stabilized on thecombination, there is little need to alter therapy. Ifa patient is on diltiazem or verapamil and lithium isadded, the dose of the lithium will be determinedby serum concentration and clinical efficacy, and noadditional intervention is required. If a patient is onlithium, and diltiazem or verapamil is added, theconcentration of lithium may be either increased ordecreased, and the patient may experience neuro-toxicity, psychosis, lithium toxicity and/or worsen-ing of mania. Lithium serum levels may not bepredictive of the toxicity.Action: When using a non-dihydropyridine cal-cium channel blocker (diltiazem or verapamil)

J. J. Strain et al.

420

concurrently with lithium, monitor closely the pa-tient’s mental state and the serum lithium level.Lithium serum levels may not be predictive oftoxicity. Watch for signs of neurotoxicity.

Central Antiadrenergic Agents (Clonidine)/BetaBlockersSignificance level 1: MajorAdvice to Practitioner: Concurrent use of clonidineand beta blockers can result in synergistic additivepharmacologic effect leading to severe hypoten-sion. Extreme caution should be used when admin-istering these drugs in combination or adding oneto the other. Monitor closely for hypotension andorthostasis.Action: Monitor for hypotension and orthostasiswhen using in combination or adding one drug tothe other.

Central Antiadrenergic Agents (Clonidine)/Cal-cium Channel Blockers (Verapamil)Significance level 1: MajorAdvice to Practitioner: Concurrent use of clonidineand verapamil can result in synergistic additivepharmacologic and toxic effects leading to severehypotension and atrioventricular (AV) block. Ex-treme caution should be used when administeringthese drugs in combination or adding one to theother. Monitor closely for hypotension and performserial ECGs to watch for the developement of anyaberrant rhythms. If a dysrhythmia develops, dis-continue both agents.Action: Monitor for hypotension and orthostasiswhen using in combination or adding one drug tothe other. Perform serial ECGs to monitor for de-velopment of dysrhythmias (AV block) and dis-continue combination therapy if dysrhythmias oc-cur.

Central Antiadrenergic Agents (Clonidine)/Phe-nothiazinesSignificance level 2: ModerateAdvice to Practitioner: Concurrent use of a centralacting antiadrenergic agent and a phenothiazinewith its alpha-adrenergic blocking activity (partic-ularly fluphenazine) can result in a synergistic ad-ditive pharmacologic effect leading to hypotensionand in the case of fluphenazine, the development ofan acute organic brain syndrome may appear. If apatient is stabilized on the combination, there islittle need to alter therapy unless an acute organicbrain syndrome develops whereby the clonidineshould be discontinued. When one drug, however,is added to the other, patients should be monitored

for orthostasis and/or hypotension and followedclosely in case an acute organic brain syndromedevelops.Action: Monitor for hypotension and orthostasiswhen adding one drug to the other. Observe pa-tient for development of an acute organic brainsyndrome and discontinue clonidine if deliriumoccurs.

Central Antiadrenergic Agents (Clonidine)/Tricy-clic AntidepressantsSignificance level 1: MajorAdvice to Practitioner: Tricyclic antidepressants mayinhibit central alpha-2 adrenergic receptors, result-ing in a loss of blood pressure control in patientsreceiving clonidine. This may result in severe hy-pertension, including hypertensive crises. Thiscombination should be avoided whenever possible.An alternative antihypertensive or a nontricyclicantidepressant should be used. (This interactiondoes not occur with the peripheral acting agentssuch as prazosin).Action: Avoid this combination whenever possi-ble.

Cholestyramine/Lorazepam and NeomycinSignificance level 3: MinorAdvice to Practitioner: Cholestyramine alone has noimpact on the pharmacokinetics or pharmacody-namics of lorazepam. However, when cholestyra-mine and oral neomycin are administered withlorazepam, reductions in lorazepam serum levelshave been reported. The exact mechanism of thisinteraction is not known, but may be due to bindingin the GI tract and altered GI flora resulting inchanges in enterohepatic recirculation. This is ararely seen drug interaction with unknown clinicalsignificance.Action: In general, no action is needed. If encoun-tered, the dose of lorazepam can be adjustedbased on the patient’s clinical response.

Cholestyramine/PhenobarbitalSignificance level 2: ModerateAdvice to Practitioner: Cholestyramine is a bile-binding resin that can decrease absorption of cer-tain drugs. Administration with phenobarbital hasbeen associated with decreased extent or rate orphenobarbital absorption. Separating the doses ofphenobarbital and cholestyramine may decreasethe extent of the interaction.Action: Administer phenobarbital 2 hours beforeor 6 hours after cholestyramine.


421

Cholestyramine/Valproic AcidSignificance level 2: ModerateAdvice to Practitioner: Cholestyramine may reducethe absorption of valproic acid due to a direct bind-ing interaction. Although not reported with colesti-pol, a similar interaction may occur since bothdrugs share a common mechanism of action. Spac-ing the doses of valproic acid and cholestyraminemay decrease the magnitude of the interaction.Action: Administer valproic acid 2 hours before or6 hours after cholestyramine.

Colestipol/CarbamazepineSignificance level 2: ModerateAdvice to Practitioner: Colestipol can decrease theabsorption of carbamazepine. The magnitude ofthis interaction is small and an interaction has notbeen reported for cholestyramine.Action: Use cholestyramine if possible. If colesti-pol is necessary, administer carbamazepine 2hours before or 6 hours after colestipol.

Digoxin/Anticholinergic Agents, Agents with An-ticholinergic Side Effects (Tricyclic Antidepres-sants)Significance level 3: MinorAdvice to Practitioner: Anticholinergic agents (e.g.,benztropine) may decrease the dissolution ofdigoxin tablets and decrease gastrointestinal motil-ity resulting in increased absorption. It is unlikelyto be clinically significant, however, an occasionalpatient may require dose adjustments after addi-tion of one of these agents to a previously stabledigoxin regimen. In particular, if a patient is man-ifesting significant anticholinergic effects or is re-ceiving multiple agents with anticholinergic prop-erties, this drug may present a problem. If a patientis already receiving these agents and digoxin isadded, no intervention would be required, as thedose of digoxin will be determined by serum levelsand the patient’s clinical response.Action: Adjust dose and monitor serum levels asdescribed above.

Digoxin/BenzodiazepinesSignificance level 2: ModerateAdvice to Practitioner: Some benzodiazepines havebeen reported to increase the serum digoxin con-centration. This drug interaction has been reportedwith diazepam and alprazolam. With alprazolam, adose response has been noted, with doses less than0.5 mg per day reported to have no impact onserum levels. The mechanism of this interactionmay be increased protein binding and decreased

tissue binding. Interactions of this type are unlikelyto be a problem if the digoxin is added to thebenzodiazepine. If the benzodiazepine is added tothe digoxin, increases in digoxin levels may be seen.Since benzodiazepines have differing pharmacoki-netic properties, a class interaction cannot be pro-vided.Action: Monitor digoxin concentrations whenadding diazepam or alprazolam (over 0.5 mgdaily) in a patient previously stabilized ondigoxin.

Digoxin/Non-dihydropyridine Calcium ChannelBlockers (Verapamil, Diltiazem)Significance level 2: ModerateAdvice to Practitioner: Addition of diltiazem or ve-rapamil to a previously stable digoxin regimen hasresulted in increases of up to 50%–75% in serumdigoxin concentrations. When adding verapamil ordiltiazem to a digoxin regimen, digoxin serum con-centrations should be monitored. When addingdigoxin to a non-dihydropyridine calcium channelblocker, the digoxin dose will be determined byserum levels and clinical response. There is conflict-ing data regarding this interaction with nifedipine.Some studies have suggested that digoxin serumconcentrations will be increased by 15%–45%,while others suggest no change in pharmacokinet-ics. All calcium channel blockers can decrease AVnodal conduction and lead to AV block withdigoxin. This is unlikely to be a significant problemwhen the combination is used for treatment of atrialarrhythmias but may be a problem when a calciumchannel blocker is added to digoxin for a psychiat-ric diagnosis.Action: Monitor digoxin concentrations whenadding a non-dihydropyridine calcium channelblocker to digoxin. Monitor AV nodal conductionand heart rate when adding a non-dihydro-pyridine calcium channel blocker to digoxin.

Digoxin/Fluvoxamine, SertralineSignificance level: Not SpecifiedAdvice to Practitioner: Studies have suggested thatfluvoxamine and sertraline do not alter serumdigoxin concentrations. However, anecdotal re-ports suggest that increases in digoxin concentra-tions may occur and may be more likely with con-current warfarin therapy. (See Digoxin/Nefazodonefor additional information).Action: No action needed.

J. J. Strain et al.

422

Digoxin/Nefazodone, ParoxetineSignificance level 2: ModerateAdvice to Practitioner: When nefazodone or parox-etine was added to digoxin therapy, increases in thearea under the concentration vs time curve, mini-mum concentration, and maximum concentrationof digoxin have been reported. The exact mecha-nism of this interaction is unknown, however, in-creases of 15%–30% have been reported. When ne-fazodone or paroxetine is added to digoxin therapy,digoxin levels should be closely monitored. Ifdigoxin is added to nefazodone or paroxetine ther-apy, no intervention would be required as the doseof digoxin will be determined by serum levels andclinical response.Action: Monitor digoxin levels when adding ne-fazodone or paroxetine to digoxin or stopping ne-fazodone or paroxetine in a patient stable ondigoxin.

Digoxin/PhenobarbitalSignificance level 2: ModerateAdvice to Practitioner: Phenobarbital may decreasedigoxin levels by inducing the cytochrome p4503A4 isoenzyme, which is responsible for somedigoxin metabolism. Interactions of this typepresent problems when the enzyme-altering drug isadded to a previously stable patient receiving thetarget drug. Interactions due to this mechanismmay take a few days to weeks to become evident ordissipate as it may take some time before theenzyme-alteration occurs or resolves. When pheno-barbital is added to a previously stable regimen ofdigoxin, the serum levels of digoxin may be de-creased. When digoxin is added to a regimen con-taining phenobarbital, no intervention would berequired, as the dose of digoxin would be deter-mined by serum concentrations and clinical response.Action: Monitor serum digoxin levels as indicatedabove.

Disopyramide/Anticonvulsant (phenobarb, car-bamazepine)Significance level 2: ModerateAdvice to Practitioner: Anticonvulsants such as phe-nobarbital and carbamazepine induce hepatic en-zymes which enhance the metabolism of disopyr-amide during concurrent administration. In suchsettings, disopyramide’s dose needed to be in-creased in order to achieve the effective pharmaco-logic action. But one potential complication associ-ated with this maneuver is the exaggeratedanticholinergic effect of disopyramide. Disopyr-amide is converted to mono-N-dealkylated metab-

olite that has 20 times stronger anticholinergic ef-fects and only one-fourth of the anti-arrhythmiceffect than that of the parent drug. Thus, increasingthe disopyramide dose to attain an adequate car-diac effect may increase anti-cholinergic side ef-fects.Action: Monitor therapeutic efficacy of disopyr-amide with combined therapy. Consider obtain-ing disopyramide level within first 3 weeks whenphenobarbital is added or discontinued from theregimen.

Diuretics/LithiumSignificance level 1: MajorAdvice to Practitioner: In general, both thiazide andloop diuretics can decrease the renal excretion oflithium and result in lithium toxicity. This interac-tion is well established with thiazide diuretics al-though it is also possible with loop diuretics.Similar interactions with triamterene and spirono-lactone have been hypothesized. Serum lithiumlevels should be monitored during concomitanttherapy. If a patient is stabilized on the combina-tion, there is little need to alter therapy. If a patientis receiving a diuretic and lithium is added, thelithium dose will be determined by serum levelsand clinical response, with little additional inter-vention required. If a diuretic is added to lithiumtherapy, serum lithium levels may increase andtoxicity may result.Action: Monitor lithium serum levels when a di-uretic is added to lithium therapy.

Enalapril/LithiumSignificance level 2: ModerateAdvice to Practitioner: Lithium toxicity may result inconcomitant administration with enalapril. Moder-ate renal insufficiency or acute renal failure hasoccurred with this combination. The mechanism isnot certain, but it is hypothesized that enalaprilincreases sodium excretion secondary to decreasedaldosterone secretion or by altering renal functionsecondary to ACE inhibition.Action: Monitor lithium concentrations and de-crease lithium dosage if necessary.

Flecainide/SSRIsSignificance level 2: ModerateAdvice to Practitioner: Flecainide is metabolized bycytochrome P450 2D6 enzyme which is found to beinhibited by fluoxetine, paroxetine, sertraline and ifthey are administered concomitantly, flecainidelevel may be raised higher than the expected level(but there is no clinical data available to support this).


423

Action: When co-administered, doses may need tobe reduced. Monitoring heart rate and serial ECGcan be advised as a precaution.

Furosemide/FluoxetineSignificance level 1: MajorAdvice to Practitioner: Two patients receiving furo-semide and fluoxetine died unexpectedly. A causalrelationship was not established. The patients weresick, elderly women who developed significant hy-ponatremia. Fluoxetine has been associated withhyponatremia. Furosemide and ACE inhibitorsmay have also contributed to the hyponatremia inthese patients.Action: Monitor for hyponatremia in patients re-ceiving both furosemide and fluoxetine. If hypo-natremia develops, furosemide or fluoxetine doseshould be discontinued or dose decreased.

Hydralazine/MAOIsSignificance level 2: ModerateAdvice to Practitioner: Use of hydralazine and anMAOI may result in significant decrease in bloodpressure. These agents act synergistically to in-crease the magnitude of blood pressure reduction.Action: Monitor blood pressure and adjust dosesas necessary.

Hydrochlorothiazide/CarbamazepineSignificance level 2: ModerateAdvice to Practitioner: Hyponatremia may developin patients treated with hydrochlorothiazide andcarbamazepine.Action: Monitor for hyponatremia in patients re-ceiving both hydrochlorothiazide and carbamaz-epine. If hyponatremia develops, hydrochlorothi-azide or carbamazepine should be discontinuedor dose decreased.

Ibutilide/Phenothiazines/HaloperidolSignificance level 1: MajorAdvice to Practitioner: Ibutilide increases the QT in-terval, and use with other agents that increase theQT interval is not recommended. Newer antipsy-chotic agents (respiridine, quetiapind) can increasethe QT interval and should be used cautiously if atall with ibutilide. (See Ibutilide/Tricyclic/Antidepres-sants for additional information.)Action: Avoid the combination if possible.

Ibutilide/Tricyclic antidepressantsSignificance level 1: MajorAdvice to Practitioner: Ibutilide increases the QT in-terval, and use with other agents that increase theQT interval is not recommended. Use of alternative

antiarrhythmic agents should be considered when apatient presents with a history of therapeutic orpossible excessive use of a tricyclic antidepressantor phenothiazine.Action: Avoid the combination if possible.

Indomethacin/LithiumSignificance level 2: ModerateAdvice to Practitioner: Simultaneous administrationof indomethacin or any other nonsteroidal(NSAIDs) and lithium may result in an increase insteady state plasma level up to 40%, and a decreasein renal lithium elimination by 23%. Salicylates didnot have this effect when in combination with lith-ium.Action: Use aspirin rather than indomethacin orother NSAIDs when the patient is concurrently onlithium.

Lidocaine/PhenobarbitalSignificance level 3: MinorAdvice to Practitioner: Phenobarbital may decreaseserum levels and clinical efficacy of lidocaine. Li-docaine is hepatically metabolized by enzymes thatmay be induced by phenobarbital. Interactions ofthis type present problems when the enzyme-altering drug is added to a previously stable patientreceiving the target drug. Interactions due to thismechanism may take a few days to weeks to be-come evident or dissipate as it may require timebefore the enzyme alterations occur or resolve. Thisdrug interaction is unlikely to present a clinicalchallenge as lidocaine will be initiated and admin-istered for a short time and dosage will be based onclinical response and serum levels. In general, lido-caine would be expected to be added to phenobar-bital or phenobarbital started during a course oflidocaine, with the period of concomitant therapybeing short lived and unlikely to result in clinicallysignificant changes in lidocaine concentration.Action: No action is necessary, as lidocaine dosagewill be based on concentrations and clinical re-sponse. Increases in the loading dose of lidocainewould not be necessary, however, infusion ratesmay need to be higher than expected. Infusionrates of lidocaine should be based on serum con-centrations and clinical response.

Lisinopril/LithiumSignificance level 2: ModerateAdvice to Practitioner: Lisinopril may result in morereabsorption of lithium from the renal tubule, andtherefore increases serum lithium levels. The mech-anism may emanate from the ACE inhibitor-

J. J. Strain et al.

424

induced sodium depletion. Angiotensin-convertingenzyme results in a loss of sodium and fluid vol-ume. Given the proposed mechanism of these in-teraction mechanisms it is likely that all ACE inhib-itors may result in lithium toxicity. One patient onthis combination developed right bundle branchblock (lithium level 3.8 mmol/L).Action: If possible, avoid the simultaneous ad-ministration of ACE inhibitors and lithium. If thetwo drugs are used concurrently, then cautiousmonitoring of lithium levels must be undertaken,with a decrease in the lithium dose as necessary.Use an agent that does not interfere with lithiummetabolism.

Loop Diuretics/BarbituratesSignificance level 3: MinorAdvice to Practitioner: In one study, a reduced di-uretic effect was seen in patients receiving furo-semide, phenytoin, and phenobarbital. It was notclear if this was due to furosemide, phenobarbital,phenytoin, or the combination.Action: Adjust diuretic dose as necessary.

Loop Diuretics/Chloral hydrateSignificance level 3: MinorAdvice to Practitioner: Transient diaphoresis, hotflashes, hypertension, and weakness have been re-ported in patients treated with furosemide whoreceived chloral hydrate within the previous 24hours.Action: Monitor for adverse effects after treatmentwith furosemide following chloral hydrate.

Losartan/LithiumSignificance level 2: ModerateAdvice to Practitioner: Losartan may increase lithiumlevels by decreasing lithium’s renal excretion andits enhanced absorption. Losartan may reduce therenal elimination of lithium. Although this action isnot established with other angiotensin-2 receptorantagonists, theoretically they may be expected tohave a similar reaction.Action: Monitor lithium levels, monitor clinicalstatus, and adjust dosage downward as needed.

Losartan/PhenobarbitalSignificance level 2: ModerateAdvice to Practitioner: Phenobarbital decreases thearea under the concentration time curve of losartanabout 20%. Therefore, decreased losartan levelsmay result and onset may be rapid. However, theclinical significance of this action is not of seriousconsequences. The mechanism is purported to bethe induction of hepatic microsomal enzymes.

Action: Monitor clinical efficacy of losartan levelsand dosage.

Nimodipine/CarbamazepineSignificance level 2: ModerateAdvice to Practitioner: Plasma concentration of nimo-dipine may decrease in patients treated with nimo-dipine and carbamazepine.Action: Closely monitor nimodipine levels in pa-tients treated with nimodipine and carbamaz-epine.

Nimodipine/PhenobarbitalSignificance level 2: ModerateAdvice to Practitioner: Plasma concentration of nimo-dipine may decrease in patients treated with nimo-dipine and phenobarbital.Action: Closely monitor nimodipine levels in pa-tients treated with nimodipine and phenobarbital.

Nimodipine/Valproic AcidSignificance level 2: ModerateAdvice to Practitioner: Concurrent use of valproicacid and nimodipine can result in an increase inserum nimodipine concentrations leading to nimo-dipine toxicity (dizziness, flushing, headache, pe-ripheral edema). The mechanism of interaction maybe the result of valproic acid inhibition of the firstpass metabolism of nimodipine with the resultingincrease in nimodipine oral bioavailability.Action: With concurrent use, patients should bemonitored for altered nimodipine effects as val-proic acid is initiated, discontinued, or changed indosage.

Nitroglycerin/Calcium Channel BlockersSignificance level 3: MinorAdvice to Practitioner: Use of nitroglycerin and cal-cium channel blockers may result in additive bloodpressure reduction. This drug interaction is un-likely to be clinically significant as both agents areused together frequently.Action: Monitor blood pressure when initiatingeither nitroglycerin or calcium channel blockertherapy to a patient who is currently receiving theother agent.

Nitroglycerin/PhenothiazineSignificance level 2: ModerateAdvice to Practitioner: Use of nitroglycerin and phe-nothiazine may result in additive blood pressurereductions. This drug interaction may result inproblems when a phenothiazine is added to nitro-glycerin without anticipating blood pressure reduc-tions.


425

Action: Monitor blood pressure when initiatingthe combination.

Peripheral Antiadrenergic Agents (Prazosin)/Cal-cium Channel Blockers (Verapamil)Significance level 2: ModerateAdvice to Practitioner: Concurrent use of prazosinand verapamil can result in an increase in serumprazosin concentrations with this synergistic addi-tive pharmacologic effect leading to hypotensionand orthostasis. When administering these drugs incombination or adding one to the other, monitor forhypotension and/or orthostasis and advise the pa-tient appropriately regarding precautions for pos-tural hypotension.Action: Monitor for hypotension and orthostasiswhen using in combination or adding one drug tothe other. Advise patients to take precautions re-garding postural hypotension.

Propafenone/Tricyclic AntidepressantsSignificance level 2: ModerateAdvice to Practitioner: Propafenone inhibits cyto-chrome P450-2D6 which results in decreased tricy-clic antidepressant metabolism and increased tricy-clic blood levels (toxicity). In one report, a patientdeveloped toxic levels of desipramine at doses pre-viously well tolerated when the antiarrhythmicpropafenone was added to his medication regimen.Although toxic blood levels of other tricyclic anti-depressants have not been documented, similartoxicity may result with these agents by the mech-anism described above.Action: Clinicians should use lower doses ofdrugs that are metabolized through the 2D6 isoen-zyme when propafenone is given concurrently.Antidepressant levels in plasma should be mea-sured routinely and are especially indicated whenunexpected side effects or other findings suggesta possible interaction. Dosages should be adustedaccordingly.

Quinidine/BarbiturateSignificance level 2: ModerateAdvice to Practitioner: Phenobarbital is a well-knownmicrosomal enzyme inducer which increases themetabolism of several drugs. When administeredwith quinidine it causes a 50% reduction of its(quinidine) half-life. This response usually takes1–3 weeks to occur and if the drugs are used to-gether for this long, the quinidine dose should beadjusted accordingly.

Action: Monitor quinidine level and adjust thedose during combination therapy if given for ex-tended periods.

Quinidine/SSRISignificance level 1: MajorAdvice to Practitioner: Management of the interac-tion between quinidine and SSRI should be individ-ualized because different SSRIs have a specific en-zyme affinity for each CYP enzyme. For example,fluoxetine and paroxetine have greater inhibitoryeffect on CYP2D6 whereas fluvoxamine exerts astronger inhibitory effect on CYP1A2 and CYP3A4.Overall enzyme studies reflect that fluvoxamine,fluoxetine, and sertraline inhibit the enzymeCYP3A4 and may decrease the metabolism of quin-idine which then can result in quinidine toxicity(arrhythmia). On the other hand, quinidine inhibitsthe metabolism of fluvoxamine, paroxetine, andvenlafaxine by inhibiting the enzyme CYP2D6 andthereby can increase their side effects. Even though,theoretically, it may be assumed that significantinteraction of quinidine with different SSRIs mayoccur, documentation of the clinical impact is yet tobe proven by control studies. No case reports areavailable at present.Action: Appropriate caution should be under-taken when such a combination is used, e.g., mon-itoring the patient for quinidine and SSRI toxicityand follow higher blood levels.

Quinidine/TCASignificance level 1: MajorAdvice to Practitioner: Co-administration of thesetwo drugs should be done with appropriate precau-tion, otherwise it may lead to serious clinical con-sequences. One of the alarming outcomes of thiscombination is the similar cardiotoxic effect of thesetwo drugs. Both quinidine and TCA cause prolon-gation of Q-T interval, so concomitant use of thesedrugs could result in a fatal arrhythmia. Quinidineincreases the blood level of TCA by inhibiting theenzyme cytochrome P450 2D which is responsiblefor the metabolism of TCAs. Imipramine clearancecan be reduced by 35% and desipramine by 85% ifit is administered concurrently with quinidine. Itusually requires about 2 weeks to demonstrate clin-ically evident TCA side effects, but in the case ofnortriptyline it may occur after a single dose of 50mg of quinidine.Action: Combination of these drugs should beavoided unless there is no alternative. If used incombination, cardiac effects should be monitoredby serial ECG (approximately every month for 3

J. J. Strain et al.

426

months or so) and TCA side effects (dry mouth,sedation, urinary retention) should also be moni-tored. Dose of the drugs should be titrated accord-ingly.

Salicyclates/FluoxetineSignificance level 3: MinorAdvice to Practitioner: Aspirin may result in the re-appearance of hives initially caused by fluoxetine.The mechanism for this allergic reaction is un-known and can be moderately severe.Action: Be aware of potential reactions.

Salicyclates/Valproic AcidSignificance level 2: ModerateAdvice to Practitioner: Salicylates may displace val-proic acid from protein binding sites. This mayresult in increased free valproic acid concentrationwithout any change in the total concentration. Asecond mechanism may be that salicylates may al-ter the metabolic pathways of valproic acid. Be-cause salicylate is highly protein bound, it can bedisplaced or displace other drugs from bindingsites. The acetylation of albumin by aspirin couldalter protein binding of other drugs and this mustbe a caution in its use.Action: Monitoring of valproic acid, total and freeconcentrations, should be conducted if salicylatesare prescribed contemporaneously. Changes inliver enzymes and evidence of valproic acid tox-icity should also be monitored.

Simvastatin, (Lovastatin)/NefazodoneSignificance level 2: ModerateAdvice to Practitioner: The combination of simvasta-tin (or lovastatin) and nefazodone has resulted inthe development of myopathy and rhabdomyoliy-sis. Simvastatin (and lovastatin) are metabolized bythe cytochrome p450 3A4 isoenzyme, which is in-hibited by nefazodone. Interactions of this typepresent problems when the enzyme-altering drug isadded to a previously stable patient. Interactionsdue to this mechanism may take a few days toweeks to become evident or dissipate as it mayrequire time before the enzyme alteration occurs orresolves. Use of an HMG-CoA inhibitor that is me-tabolized by the p450 3A4 isoenzyme, (simvastatinor lovastatin) with nefazodone may result in in-creased serum levels of the HMG CoA inhibitor andlead to toxicity. Agents such as pravastatin, fluva-statin, and atorvastatin may be less likely to causecomplications, as they are not widely metabolizedby the p450 3A4 isoenzyme. Although this interac-tion is not well reported with lovastatin, it would be

expected given the mechanism of the interactionand the metabolic pathway of lovastatin. Dose re-ductions of lovastatin or simvastatin may decreasethe risk of the interaction.Action: Use an alternative HMG CoA inhibitorwhenever possible.

Sympathomimetics/MAO InhibitorsSignificance level 1: MajorAdvice to Practitioner: The combination of MAO in-hibitors and sympathomimetics is contraindicated.Such sympathomimetics include dobutamine, do-pamine, amphetamines, ephedrine, and phenyleph-rine. Severe headaches, hypertensive crises, nausea,vomiting, and cardiac arrhythmias have been re-ported in patients taking MAO inhibitors and sym-pathomimetics.Action: In general, one should avoid concomitantuse of MAO inhibitors and sympathomimetics.However, for patients taking an MAO inhibitor,in emergent situations when sympathomimeticsare required, they may be used, and hypertensivecrisis should be anticipated and treated appropri-ately.

Sympathomimetics/Tricyclic AntidepressantsSignificance level 1: MajorAdvice to Practitioner: Tricyclic antidepressants canpotentiate the pressor effects of direct acting sym-pathomimetics such as dobutamine, norepineph-rine, epinephrine, and phenylephrine and may de-crease the pressor response to indirect-actingsympathomimetics such as dopamine. When theseagents are administered concomitantly, close mon-itoring of blood pressure and heart rhythm is re-quired. However, patients receiving these sympa-thomimetics normally receive close monitoring andcareful dosage titration based on response as part oftheir treatment, and additional monitoring will notbe needed. For a patient located outside of an ICUsetting, additional monitoring should be employed.Action: Anticipate enhanced pressor effects withthis combination.

Thiazides/AnticholinergicsSignificance level 3: MinorAdvice to Practitioner: Anticholinergics may delay GImotility thus producing a delayed peak effect ofthiazide diuretics. Thiazide diuretic action may bestimulated.Action: No specific change required. Monitor di-uretic pattern.


427

Thiazides or Triamterene/AmantadineSignificance level 2: ModerateAdvice to Practitioner: Risk of developing adverseeffects from amantadine may be increased. In onepatient receiving amantadine and triamterene/hy-drochlorothiazide, adverse effects from amantadinewere noted.Action: If adverse effects from amantadine arenoted, diuretic or amantadine dose should be de-creased or discontinued.

Ticlopidine/CarbamazepineSignificance level 2: ModerateAdvice to Practitioner: Neurologic symptoms havebeen seen in association with increased carbamaz-epine levels because of possible interaction withticlopidine. The mechanism may be active circulat-ing metabolites that complicate the relationship ofdose and clinical response.Action: Monitor clinical state and carbamazepinelevels to insure that carbamazepine is not higherthan acceptable levels.

Warfarin/CarbamazepineSignificance level 2: ModerateAdvice to Practitioner: Carbamazepine may increasethe metabolism of warfarin, resulting in reducedwarfarin activity. (See Warfarin/Phenobarbital for ad-ditional information.)Action: Monitor prothrombin times (PT) or inter-national normalized ratios (INR) as indicated.

Warfarin/Chloral hydrateSignificance level 2: ModerateAdvice to Practitioner: Chloral hydrate may increasethe PT when used with warfarin. The mechanism ofthis interaction is unknown. When chloral hydrateis added to a previously stable regimen of warfarin,warfarin activity may be decreased. When warfarinis added to a regimen containing chloral hydrate,no intervention would be required, as the dose ofwarfarin would be determined by prothrombintimes or international normalized ratios.Action: Monitor PT when adding chloral hydrateas recommended.

Warfarin/ChlordiazepoxideSignificance level 2: ModerateAdvice to Practitioner: Chlordiazepoxide has beenreported to decrease prothrombin times when usedwith warfarin. (See Warfarin/Chloropromazine for ad-ditional information.)Action: Monitor PT when adding chlordiazep-oxide as recommended above.

Warfarin/ChlorpromazineSignificance level 2: ModerateAdvice to Practitioner: Chlorpromazine has been re-ported to decrease prothrombin times when usedwith warfarin. The mechanism of this interaction isunknown. When chlorpromazine is added to a pre-viously stable regimen of warfarin, warfarin activ-ity may be decreased. When warfarin is added to aregimen containing chlorpromazine, no interven-tion would be required, as the dose of warfarinwould be determined by prothrombin times or in-ternational normalized ratios.Action: Monitor PT when adding chlorpromazine,as recommended above.

Warfarin/Fluvoxamine, Fluoxetine, SertralineSignificance level 2: ModerateAdvice to Practitioner: These SSRIs may decrease themetabolism of warfarin via the cyp 450 2C9 isoen-zyme, resulting in increased warfarin activity. (SeeWarfarin/Phenobarbital for additional information.)Action: Monitor PT or INR as indicated.

Warfarin/ParoxetineSignificance level 2: ModerateAdvice to Practitioner: Use of paroxetine with war-farin has been associated with increased bleeding,without alterations in the prothrombin times. Whenparoxetine is added to a previously stable regimenof warfarin, warfarin activity may be decreased.When warfarin is added to a regimen containingparoxetine, no intervention would be required, asthe dose of warfarin would be determined by pro-thrombin times or international normalized ratios.Action: Use cautiously and inform the patient tobe aware of increased bruisability. Consider analternative SSRI that is more easily managed.

Warfarin/PhenobarbitalSignificance level 2: ModerateAdvice to Practitioner: Phenobarbital may increasethe metabolism of warfarin resulting in reducedwarfarin activity. Interactions of this type presentproblems when the enzyme-altering drug is addedto a previously stable patient. Interactions due tothis mechanism may take a few days to weeks tobecome evident or dissipate, as it may take sometime before the enzyme alteration occurs or re-solves. When phenobarbital is added to a previ-ously stable regimen of warfarin, warfarin activitymay be decreased. When warfarin is added to aregimen containing phenobarbital, no interventionwould be required, as the dose of warfarin would

J. J. Strain et al.

428

be determined by prothrombin times or interna-tional normalized ratios.Action: Monitor PT or INR as indicated above.

Warfarin/TrazodoneSignificance level 2: ModerateAdvice to Practitioner: Trazodone can decrease thePT and PTT when used with warfarin. The mecha-nism of this interaction is unknown and it is notwidely reported.Action: Monitor PT when adding trazodone towarfarin or stopping trazodone while receivingwarfarin.

Warfarin/Tricyclic AntidepressantsSignificance level 2: ModerateAdvice to Practitioner: Use of some tricyclic antide-pressants has been associated with increases in PTand increased bleeding. The mechanism of this in-teraction is unclear: reduced absorption and de-creased metabolism have been suggested. When aTCA is added to a previously stable regimen ofwarfarin, warfarin activity may be decreased.When warfarin is added to a regimen containing a

TCA, no intervention would be required, as thedose of warfarin would be determined by pro-thrombin times or international normalized ratios.Action: Monitor PT when adding a tricyclic anti-depressant as recommended above.

Warfarin/Other agentsNo Interaction: Since warfarin interacted with somany CNS medications it is important to note thatstudies to date have suggested that warfarin doesnot interact with the agents listed below.

AlprazolamDiazepamMeprobamateNefazodoneOlanzapine

(Since the literature regarding drug interactions isconstantly changing, additional data may eventually be-come available suggesting that interactions with warfa-rin do exist.)


429

Documents

Part II: Cardiac drug and psychotropic drug interactions: significance and recommendations