Introduction
Clozapine is an atypical antipsychotic primarily indicated for the management of treatment resistant schizophrenia and reduction in the risk of recurrent suicidal behaviour in schizophrenia or schizoaffective disorder (Lehman et al. Reference Lehman, Lieberman, Dixon, McGlashan, Miller, Perkins and Kreyenbuhl2010). However, a subset of patients fail to respond or have only a partial response to clozapine, leading clinicians to search for potential augmentation strategies to improve outcomes. Psychotropic medications used in combination with clozapine include antidepressants, antipsychotics and mood stabilisers (Calabrese & Gajwani, Reference Calabrese and Gajwani2000; Fuchs, Reference Fuchs1994; Chan & Sweeting, Reference Chan and Sweeting2007). Non-psychotropic medications are also added for management of associated medical conditions. Smoking and smoking cessation have also been reported to affect clozapine levels (Sandson et al. Reference Sandson, Cozza, Armstrong, Eckermann, Fischer and Phillips2007).
It is proposed that the therapeutic efficacy of clozapine is mediated through antagonism of the dopamine type 2 (D2) and serotonin type 2A (5-HT2A) receptors (Lieberman & Safferman, Reference Lieberman and Safferman1992). In addition, clozapine acts as an antagonist at alpha-adrenergic, histamine H1 and cholinergic which accounts for its side effects (Lieberman & Safferman, Reference Lieberman and Safferman1992). Studies using PET and a variety of ligands show less than 50% dopamine receptor occupancy at normal therapeutic levels of clozapine (Farde et al. Reference Farde, Nordström, Wiesel, Pauli, Halldin and Sedvall1992). Positron emission tomography imaging studies found that clozapine, at doses known to be effective in routine clinical settings, showed a D2 occupancy clearly lower (16%-68%) than that of other atypical antipsychotics (Kapur et al. Reference Kapur, Zipursky and Remington1999). The use of this finding in routine clinical practice is not known. Clozapine is 97% bound to plasma proteins and has a mean half-life of about 12 hours (range 6–33 hours). It is metabolised in the liver by CYP1A2 and CYP3A4 to the relatively inactive compounds norclozapine and clozapine-N-oxide (Bertilsson et al. Reference Bertilsson, Carrillo, Dahl, Llerena, Alm, Bondesson, Lindstrom, Rodriguez de la Rubia, Ramos and Benitez1994; Brøsen, Reference Brøsen1993). There are variety of analytical methods to assay clozapine and its metabolite including high performance liquid chromatography, gas chromatography, gas chromatography/mass spectrometry and radio-immunoassay (Cooper, Reference Cooper1996). Medications interfering with the activity of these cytochrome enzymes results in changes in plasma clozapine levels. The therapeutic drug monitoring (TDM) of clozapine, and of its principal plasma metabolite N-desmethyl clozapine (norclozapine), has been shown to assist management by: identifying or confirming suspected non-adherence; assessing whether clozapine is being used at a therapeutic dose; and minimising the risk of dose-related toxicity (Couchman et al. Reference Couchman, Bowskill, Handley, Patel and Flanagan2013). Plasma clozapine concentrations in the range of 0.35–0.60 mg l−1 are associated with a good antipsychotic response in many adult patients, although the upper limit remains a subject of debate (Hiemke et al. Reference Hiemke, Dragicevic, Gründer, Hätter, Sachse, Vernaleken and Müller2004; Khan & Preskorn Reference Khan and Preskorn2005; Taylor et al. Reference Taylor, Paton and Kapur2009). The threshold for therapeutic response may be lower once a degree of symptom control has been achieved (Yusufi et al. Reference Yusufi, Mukherjee, Flanagan, Paton, Dunn, Page and Barnes2007). Results for both clozapine and norclozapine below the tenth percentile are taken to suggest recent poor adherence, although rapid metabolism (notably in young, male, smokers) cannot always be excluded (Couchman et al. Reference Couchman, Bowskill, Handley, Patel and Flanagan2013). Norclozapine has a longer plasma half-life than clozapine in patients who have been taking the drug chronically, hence plasma norclozapine shows less day-to-day variation than plasma clozapine (Couchman et al. Reference Couchman, Bowskill, Handley, Patel and Flanagan2013). The clozapine:norclozapine ratio has practical significance; for example, a ratio of <0.5 suggests either poor adherence within the last 24 hours or that alterations in dose schedule might be beneficial; and a ratio >3 suggests that either absorption of clozapine from the last dose may not have been completed at the time the sample was obtained, or that clozapine metabolism is saturated either because of the dose prescribed or because of inhibition of clozapine metabolism by a coadministered drug (Yusufi et al. Reference Yusufi, Mukherjee, Flanagan, Paton, Dunn, Page and Barnes2007). Besides medications, other demographic variables can also affect clozapine levels including gender, race, age, smoking behaviour and weight. The gender-associated differences can be a function of hormonal balance, body composition, and/or activity of certain enzymes (Rowland & Tozer, Reference Rowland and Tozer1995). The age-related differences can be a function of drug absorption, distribution, delay in gastric emptying, and/or changes in renal and/or hepatic elimination (Haring et al. Reference Haring, Meise, Humpel, Saria, Fleischhacker and Hinterhuber1989). Changes in body water spaces, muscle mass, organ blood flow, and organ function are related to body weight and can effect volume of distribution and clearance (Rowland & Tozer, Reference Rowland and Tozer1995).
In this paper we will review the case report literature on clozapine drug interactions and the effect of these interactions on serum level changes in clozapine.
Methods
We reviewed the literature by manual and computerised database search from January 1970 to May 2013 (Medline, Ovid database, PsycARTICLES, PsychINFO and PsychiatryOnline). The following terms were cross-referenced with clozapine in the search: high level, interactions, case report, selective serotonin reuptake inhibitors (SSRI), fluvoxamine, fluoxetine, lamotrigine, valproate, neuroleptic, antipsychotics, antidepressants, proton pump inhibitors (PPIs), antibiotics, infection, smoking and caffeine. The initial process of cross referencing was restricted to antidepressants but cross-referencing of published bibliographies yielded additional reports. Our search was therefore broadened to other classes of psychotropic and non-psychotropic medications, but our selection was not all inclusive. Individual case reports were reviewed for patient's age, sex, clozapine dose, dose of interacting drug, plasma clozapine level pre and post interaction, and the clinical status post interaction. We narrowed our search to case reports since these papers, unlike other studies, reported clozapine levels both pre and post drug-drug interactions. In vitro studies data, post marketing studies or trial data were not included.
Results
A total of 54 articles with a total of 109 individual case reports were found. The results were further divided based on drug interactions with clozapine levels and associated changes in adverse events or psychopathology. We found 27 reports of drug interactions with clozapine levels >1000 ng ml−1 with adverse events (Table 1), 12 reports with clozapine levels <1000 ng ml−1 with adverse events (Table 2), 14 reports with clozapine levels >1000 ng ml−1 with no adverse events (Hiemke et al. Reference Hiemke, Weigman, Härtter, Dahmen, Wetzel and Müller1994; Jerling et al. Reference Jerling, Lindström, Bondesson and Bertilsson1994; Dequardo & Roberts, Reference Dequardo and Roberts1996; DuMortier et al. Reference DuMortier, Lochu, Colen de Melo, Ghribi, Roche-Rabreau, DeGrassat and Desce1996; Pinninti & de Leon, Reference Pinninti and de Leon1997; Carrillo et al. Reference Carrillo, Herraiz, Ramos and Benitez1998; Raaska K 2002; Haack et al. Reference Haack, Bak, Beurskens, Maes, Stolk and Delespaul2003; Sandson et al. Reference Sandson, Cozza, Armstrong, Eckermann, Fischer and Phillips2007; Bugamelli et al. Reference Bugamelli, Madrioli, Kenndler, Bartoletti, Boncompagni and Raggi2007; Sambhi et al. Reference Sambhi, Puri and Jones2007; Stevens et al. Reference Stevens, Freudenreich and Stern2008) and 13 reports with clozapine levels <1000 ng ml−1 with no adverse events (Raitasuo et al. Reference Raitasuo, Lehtovaara and Huttunen1993; Finley & Warner, Reference Finley and Warner1994; Tyson et al. Reference Tyson, Devane and Risch1995; Nebel et al. Reference Nebel, Schneider, Baker and Kroll1999; Conca et al. Reference Conca, Beraus, König and Waschgler2000; Frick et al. Reference Frick, Kopitz and Bergemann2003; Bugamelli et al. Reference Bugamelli, Madrioli, Kenndler, Bartoletti, Boncompagni and Raggi2007; Avari et al. Reference Avari, Mahgoub and Alexopoulos2011). A total of 41 case reports found no association of drug interactions with changes in clozapine levels (Taylor et al. Reference Taylor, Ellison, Ementon, Wickham and Murray1998; Spina et al. Reference Spina, Avenoso, Salemi, Facciolá, Scordo, Ancione and Madia2000; Bergemann et al. Reference Bergemann, Kress, Frick and Kopitz2005; Khan & Preskorn, Reference Khan and Preskorn2005; Abu-Tair et al. Reference Abu-Tair, Kopitz and Bergemann2006). We also reviewed the changes in norclozapine levels with drug interactions and found 16 studies (Table 3). These results were further compared for norclozapine levels pre and post interactions, and clozapine dose to norclozapine level ratios following drug interactions (Table 3). We present these results by drug class.
Table 1 Drug interactions with clozapine level >1000 ng ml−1
Clozapine level conversion: 3.06 × ng ml−1 = nmol l−1; 0.003 × ng ml−1 = μmol l−1; NA, not available
Table 2 Drug interactions with clozapine level <1000 ng ml−1
Table 3 N-desmethyl clozapine (norclozapine) level changes with drug interactions
CLZ/NCZ, post-interaction (clozapine dose) : (norclozapine level) ratio
Antidepressants
There are 41 reports on interactions of antidepressants with clozapine (Hiemke et al. Reference Hiemke, Weigman, Härtter, Dahmen, Wetzel and Müller1994; Jerling et al. Reference Jerling, Lindström, Bondesson and Bertilsson1994; Dequardo & Roberts, Reference Dequardo and Roberts1996; DuMortier et al. Reference DuMortier, Lochu, Colen de Melo, Ghribi, Roche-Rabreau, DeGrassat and Desce1996; Koponen et al. Reference Koponen, Leinonen and Lepola1996; Armstrong & Stephans, Reference Armstrong and Stephans1997; Chong et al. Reference Chong, Tan and Lee1997; Pinninti & de Leon, Reference Pinninti and de Leon1997; Kuo et al. Reference Kuo, Lane and Chang1998; Taylor et al. Reference Taylor, Ellison, Ementon, Wickham and Murray1998; Heeringa et al. Reference Heeringa, Beurskens, Shouten and Verduijn1999; Spina et al. Reference Spina, Avenoso, Salemi, Facciolá, Scordo, Ancione and Madia2000; Khan & Preskorn, Reference Khan and Preskorn2005; Sandson et al. Reference Sandson, Cozza, Armstrong, Eckermann, Fischer and Phillips2007). Among selective serotonin reuptake inhibitors (SSRIs), fluvoxamine resulted in marked increased in plasma clozapine level (Hiemke et al. Reference Hiemke, Weigman, Härtter, Dahmen, Wetzel and Müller1994; Jerling et al. Reference Jerling, Lindström, Bondesson and Bertilsson1994; Dequardo & Roberts, Reference Dequardo and Roberts1996; DuMortier et al. Reference DuMortier, Lochu, Colen de Melo, Ghribi, Roche-Rabreau, DeGrassat and Desce1996; Armstrong & Stephans, Reference Armstrong and Stephans1997; Chong et al. Reference Chong, Tan and Lee1997; Kuo et al. Reference Kuo, Lane and Chang1998; Heeringa et al. Reference Heeringa, Beurskens, Shouten and Verduijn1999; Koponen et al. Reference Koponen, Leinonen and Lepola1996). Clozapine is metabolised by cytochrome P450 3A3/4 and 1A2 (Brøsen, Reference Brøsen1993; Bertilsson et al. Reference Bertilsson, Carrillo, Dahl, Llerena, Alm, Bondesson, Lindstrom, Rodriguez de la Rubia, Ramos and Benitez1994), and fluvoxamine inhibits both 3A3/4 and 1A2.
There is one case report for fluoxetine (Sandson et al. Reference Sandson, Cozza, Armstrong, Eckermann, Fischer and Phillips2007), where a fivefold rise in clozapine levels was observed. Ferslew et al. (Reference Ferslew, Hagardorn, Harlan and McCormick1998) reported a case of fatal drug interaction with fluoxetine resulting in death from conpulmonary edema, visceral vascular congestion, paralytic ileus, gastroenteritis and eosinophilia.
Kahn & Preskorn (2005) reported that paroxetine use was associated with marked increase in clozapine levels. Similarly Pinninti & de Leon (Reference Pinninti and de Leon1997) found high doses of sertraline (300 mg daily) was associated with marked rise in clozapine levels. Other reports on sertraline (Chong et al. Reference Chong, Tan and Lee1997; Spina et al. Reference Spina, Avenoso, Salemi, Facciolá, Scordo, Ancione and Madia2000), paroxetine (Chong et al. Reference Chong, Tan and Lee1997) and citalopram (Taylor et al. Reference Taylor, Ellison, Ementon, Wickham and Murray1998) found no effect on clozapine levels.
No case reports were found that specifically addressed the interaction of tricyclic antidepressants with changes in plasma clozapine levels, but it has been suggested that tricyclic antidepressants with clozapine could exacerbate the adverse effects related to increased anticholinergic effect (Wetzel et al. Reference Wetzel, Anghelescu, Szegedi, Wiesner, Weigmann, Hartter and Hiemke1998).
Anticonvulsants
There are nine reports on potential interactions of anticonvulsants with subsequent changes in clozapine levels (Raitasuo et al. Reference Raitasuo, Lehtovaara and Huttunen1993; Finley & Warner, Reference Finley and Warner1994; Conca et al. Reference Conca, Beraus, König and Waschgler2000; Kossen et al. Reference Kossen, Selten and Kahn2001; Egger et al. Reference Egger, Muehlbacher, Grohmann and Stuppaeck2010). Among anticonvulsants, lamotrigine (Kossen et al. Reference Kossen, Selten and Kahn2001; Egger et al. Reference Egger, Muehlbacher, Grohmann and Stuppaeck2010) resulted in a four- to fivefold increase in plasma clozapine levels. Clozapine and lamotrigine do not share a common cytochrome P450 pathway. The active compounds clozapine and norclozapine are further glucuronidated by uridine 5′-diphosphate glucuronosyltransferases (UGT) 1A4 (and 1A3). Lamotrigine was reported to be metabolised by UGT1A4. As both clozapine and lamotrigine use UGT1A4, the rise in clozapine and norclozapine may be due to a competitive inhibition of the conjugate (Wynn et al. Reference Wynn, Oesterheld, Cozza and Armstrong2009, pp. 423–460).
Discontinuation of carbamazepine resulted in two fold rise in clozapine levels (Raitasuo et al. Reference Raitasuo, Lehtovaara and Huttunen1993). A single case of neuroleptic malignant syndrome was reported by Müller et al. (Reference Müller, Becker and Fritze1998), who concluded that combining clozapine and carbamazepine is not considered safe for risk of neutropenia and decreased seizure threshold.
Doubling of clozapine serum levels after the discontinuation of valproic acid was reported by Conca et al. (Reference Conca, Beraus, König and Waschgler2000). Centorrino et al. (Reference Centorrino, Baldessarini, Kando, Frankenburg, Volpicelli, Puopolo and Flood1994) reported a weak increase of clozapine serum levels (about 6%) with valproic acid in 11 patients. In contrast Longo & Salzman (Reference Longo and Salzman1995) reported a 15% decrease in clozapine blood levels in seven patients after addition of valproic acid. The coexistence of two mechanisms of interaction (CYP 1A2 enzyme inhibition and protein binding displacement) leading to opposite changes in total clozapine concentration may explain the opposite findings (Finlay & Warner, 1994). This supports the clinical importance of therapeutic monitoring of serum clozapine.
Two cases reports indicated a worsening of psychosis after phenytoin was added for a decrease in plasma clozapine concentrations (Miller, Reference Miller1981).
Antipsychotics
Since a second antipsychotic is often added as an augmentation strategy, it is important to consider the safety of these combinations. Clozapine interactions with antipsychotics have only been reported with aripiprazole and risperidone (Tyson et al. Reference Tyson, Devane and Risch1995; Abu-Tair et al. Reference Abu-Tair, Kopitz and Bergemann2006; Avari et al. Reference Avari, Mahgoub and Alexopoulos2011). There were six reports of aripiprazole and one report of risperidone interaction with clozapine. The addition of aripiprazole resulted in clozapine level reduction (Avari et al. Reference Avari, Mahgoub and Alexopoulos2011) and was associated with severe exacerbation of delusions and hallucinations in one case report (Avari et al. Reference Avari, Mahgoub and Alexopoulos2011). The other five cases (Abu-Tair et al. Reference Abu-Tair, Kopitz and Bergemann2006) reported improvement in negative symptoms after combination with clozapine and no changes in clozapine levels were noted. The mechanism for this interaction is not well understood.
The addition of risperidone resulted in mild increase in plasma clozapine with transient lightheadedness (Tyson et al. Reference Tyson, Devane and Risch1995). This may represent a pharmacokinetic interaction of competitive cytochrome P450 2D6 enzyme metabolism.
Levomepromazine discontinuation resulted in four- and tenfold rises in clozapine levels (Bugamelli et al. Reference Bugamelli, Madrioli, Kenndler, Bartoletti, Boncompagni and Raggi2007). The most likely explanation is that levomepromazine can be an inducer of CYP2D6 isoenzymes (Hals & Dahl, Reference Hals and Dahl1994), which are in turn involved in the biotransformation of clozapine.
The atypical antipsychotic drug amisulpride is not known to have any drug interactions with clozapine (Bergemann et al. Reference Bergemann, Kress, Frick and Kopitz2005).
Antibiotics, infections and anti-tuberculosis medications
An association of antibiotics and/or infection with the changes in plasma clozapine levels has also been reported. Both ciprofloxacin (Sandson et al. Reference Sandson, Cozza, Armstrong, Eckermann, Fischer and Phillips2007; Sambhi et al. Reference Sambhi, Puri and Jones2007; Brownlowe & Sola, Reference Brownlowe and Sola2008; Brouwers et al. Reference Brouwers, Söhne, Kuipers, van Gorp, Schellens, Koks, Beijnen and Huitema2009) and erythromycin (Funderburg et al. Reference Funderburg, Vertrees, True and Miller1994; Cohen et al. Reference Cohen, Chesley, Eugenio, Flood, Fisch and Goff1996) resulted in clozapine levels >1000 ng ml−1, but clozapine levels prior to introduction of antibiotics were not available.
Ciprofloxacin is a potent inhibitor of CYP450 1A2 and 3A4 (Batty et al. Reference Batty, Davis, Ilett, Dusci and Langton1995), which impairs clozapine metabolism resulting in an increase in clozapine level (Sambhi et al. Reference Sambhi, Puri and Jones2007; Brouwers et al. Reference Brouwers, Söhne, Kuipers, van Gorp, Schellens, Koks, Beijnen and Huitema2009).
Erythromycin is a selective inhibitor of CYP 3A enzyme (Murray, Reference Murray1992), resulting in decreased metabolism of clozapine with increased serum concentration, increased somnolence, difficulty in coordination and tonic-clonic seizure (Funderburg et al. Reference Funderburg, Vertrees, True and Miller1994; Cohen et al. Reference Cohen, Chesley, Eugenio, Flood, Fisch and Goff1996).
Infection (Uges et al. Reference Uges, Boom, Wentjex and Versteege2000; Raaska et al. Reference Raaska, Raitasuo, Arstila and Neuvonen2002; de Leon & Diaz, Reference de Leon and Diaz2003; Haack et al. Reference Haack, Bak, Beurskens, Maes, Stolk and Delespaul2003; Jecel et al. Reference Jecel, Michel, Gutknecht, Schmidt, Pfuhlmann and Jabs2005; Darling & Huthwaite, Reference Darling and Huthwaite2011) was associated with an up to tenfold rise (up to 14505 nmol−1) in plasma clozapine level (Darling & Huthwaite, Reference Darling and Huthwaite2011). An increase in clozapine and norclozapine levels in the early course of acute infection could be due to different infection-related, not fully understood, alterations of specific CYP 450 enzymes (Jecel et al. Reference Jecel, Michel, Gutknecht, Schmidt, Pfuhlmann and Jabs2005). During the acute phase of inflammation, the cytochrome P450 enzymes, including CYP1A2, are down-regulated by up to 90% with the cytokine interleukin 6 (Siewert et al. Reference Siewert, Bort, Kluge, Heinrich, Castell and Jover2000).
Isoniazid (Van Strater & Bogers, Reference Van Strater and Bogers2012) and rifampin (Nebel et al. Reference Nebel, Schneider, Baker and Kroll1999) are still the first line for treatment of tuberculosis. Rifampin is a known inducer of CYP 1A2 and 3A resulting in decreased clozapine concentrations (Nebel et al. Reference Nebel, Schneider, Baker and Kroll1999), and isoniazid inhibits CYP 1A2, responsible for a rise in clozapine levels (Van Strater & Bogers Reference Van Strater and Bogers2012). When considering the effects of antibiotics on clozapine levels, duration of treatment should also be borne in mind. For instance, ciprofloxacin and erythromycin are prescribed for 5–7 days whereas rifampin is prescribed for longer, making it necessary to monitor clozapine serum concentrations with rifampicin.
Proton pump inhibitors, antacids and histamine receptor antagonists
Omeprazole (Frick et al. Reference Frick, Kopitz and Bergemann2003) resulted in a slight decrease in plasma clozapine levels, whereas the addition of esomeprazole (Wagner et al. Reference Wagner, Varet-Legros, Fabre, Montastruc and Bagheri2011) resulted in a fourfold rise in plasma clozapine levels. Omeprazole is a mixed inducer of CYP450 1A2 and 3A4, resulting in decreased concentration of clozapine, whereas drug interaction studies with esomeprazole found no potential interaction with drugs metabolised by CYP1A2, 2A6, 2C9, 2D6, or 2E1 isoenzymes (Andersson et al. Reference Andersson, Hassan-Aline, Hasselgren and Röhss2001). The rise of clozapine plasma levels after the substitution of omeprazole by esomeprazole may be the result of the removal of the induction of clozapine metabolism caused by omeprazole.
The addition (Szymanski et al. Reference Szymanski, Lieberman, Picou, Masiar and Cooper1991) and/or increase in cimetidine dose (Sandson et al. Reference Sandson, Cozza, Armstrong, Eckermann, Fischer and Phillips2007) resulted in a rise in both clozapine and norclozapine levels, associated with a worsening of adverse side effects including increasing sialorrhea. Cimetidine is considered a ‘pan- inhibitor’ (Martínez et al. Reference Martínez, Albet, Agúndez, Herrero, Carrillo, Márquez, Benítez and Ortiz1999) of the entire range of P450 enzymes (1A2, 2C9/19, 2D6, and 3A4), resulting in a rise of clozapine and norclozapine levels despite the constant dose of clozapine. These changes were not observed with ranitidine, which could be a better alternative (Szymanski et al. Reference Szymanski, Lieberman, Picou, Masiar and Cooper1991).
Only one case report (Allard et al. Reference Allard, Légaré and Millaud2008) was found suggesting the possible association of aluminum hydroxide discontinuation with a doubling of clozapine concentrations, with resultant excessive sleepiness and drooling. The mechanism of action is not known.
Others: ethinylestradiol, amiodarone, modafinil and lisinopril
The addition of oral contraceptive pills containing ethinylestradiol to clozapine resulted in a threefold increase in plasma clozapine levels with adverse events including marked drowsiness, anergy and dizziness (Gabbay et al. Reference Gabbay, O'Dowd, Mamamtavrishvili and Asnis2002; Sandson et al. Reference Sandson, Cozza, Armstrong, Eckermann, Fischer and Phillips2007). This is a greatly under recognised drug interaction (Sandson et al. Reference Sandson, Cozza, Armstrong, Eckermann, Fischer and Phillips2007). Ethinylestradiol is an inhibitor of CYP450 1A2 and 2C19 (Granfors et al. Reference Granfors, Backman, Laitila and Neuvonen2005), which contribute significantly to clozapine metabolism.
Amiodarone was associated with six–sevenfold rise in clozapine serum levels (Stevens et al. Reference Stevens, Freudenreich and Stern2008). Although amiodarone primarily inhibits CYP 3A4, its major active metabolite desethylamiodarone, is a potent inhibitor of 1A1/2, 2B6, and 2D6 (Ohyama et al. Reference Ohyama, Nakajima, Suzuki, Shimada, Yamazaki and Yokoi2000).
Modafinil resulted in a doubling of clozapine levels and the onset of dizziness and unsteady gait (Dequardo, Reference Dequardo2002). Modafinil metabolism involves P450 1A2, 2B6, 3A4/5, 2C9, and 2C19 isoenzymes and inhibits P450 2C19 activity (Huang et al. Reference Huang, Lasseter, Janssens, Verhaeghe, Lau and Zhao2002). CYP450 2C19 inhibition by modafinil interfered with clozapine clearance, elevating serum clozapine levels and thereby producing signs of toxicity.
Lisinopril does not influence the cytochrome systems but instead causes reversible renal impairment. A case of high clozapine and norclozpaine levels were reported with lisinopril, with resultant disorganised behaviour and angry outbursts (Abraham et al. Reference Abraham, Grunberg and Gratz2001).
Smoking and smoking cessation
There are several reports in which patients with a nicotine based smoking history were admitted to smoke-free hospitals, with a resultant rise in plasma clozapine levels (McCarthy, Reference McCarthy1994; Zullino et al. Reference Zullino, Delessert, Eap, Preisiq and Baumann2002; Derenne & Baldessarini, Reference Derenne and Baldessarini2005; Sandson et al. Reference Sandson, Cozza, Armstrong, Eckermann, Fischer and Phillips2007; Jain et al. Reference Jain, Chawla, Theethira, Strassnig and Chengappa2008). Meyer (Reference Meyer2001) found that after a no-smoking policy was implemented at a state psychiatric hospital, serum concentrations of clozapine increased by an average of 71% in 11 patients, one of whom developed aspiration pneumonia at a serum concentration of over 3 mg ml−1. Smoking decreases the clozapine level by induction of CYP450 1A2, with a resultant increase in clozapine level on smoking cessation (McCarthy, Reference McCarthy1994). Many components found in tobacco smoke belong to the polycyclic aromatic hydrocarbons, which are the classical inducers of the pathway involving the aryl hydrocarbon (Ah) receptor. The binding of inducers to the intracellular Ah receptors, together with another protein, the Ah receptor nuclear translocator, increases enzyme expression by binding to an enhancer/promoter region (Zullino et al. Reference Zullino, Delessert, Eap, Preisiq and Baumann2002). N-demethylation by CYP1A2 is the main metabolic pathway for clozapine metabolism. The literature search did not reveal reports of clozapine interactions with nicotine replacement therapies.
Caffeine
Two case reports (Vainer & Chouinard, Reference Vainer and Chouinard1994; Al Hadithy et al. Reference Al Hadithy, Leeffers and Bruggeman2012) indicated that caffeine may significantly inhibit clozapine metabolism, by competition for cytochrome P450 1A2. Carrillo et al. (Reference Carrillo, Herraiz, Ramos and Benitez1998) observed a 50% reduction of clozapine concentrations after the removal of caffeine. This signifies the importance of monitoring consumption of coffee, cola, or other caffeinated beverages in patients treated with clozapine (Al Hadithy et al. Reference Al Hadithy, Leeffers and Bruggeman2012).
St John's Wort (Hypericum perforatum)
St John's wort (Hypericum perforatum) is a herbal remedy, with a favourable effect on depression (Van Strater & Bogers, Reference Van Strater and Bogers2012). Van Strater & Bogers (Reference Van Strater and Bogers2012) reported a patient with schizophrenia whose psychiatric condition deteriorated after self-medication with St John's Wort began. While St John's Wort is known to induce CYP3A4, it also induces P-glycoprotein (Nebel et al. Reference Nebel, Schneider, Baker and Kroll1999) resulting in a decrease in clozapine serum concentration.
Discussion
We found that psychotropic medications likely to increase clozapine levels included fluvoxamine, lamotrigine, carbamazepine, and aripiprazole. Non-psychotropic medications associated with clozapine level changes included erythromycin, ciprofloxacin, omeprazole, cimetidine, oral contraceptive pills containing ethinylestradiol, and amiodarone. Smoking cessation also increased clozapine levels.
The current literature doesn't suggest at what point in time these changes in plasma clozapine levels occur (Préterre, Reference Préterre1995; Vailleau et al. Reference Vailleau, Jeanny, Chomard and Vincent1996). Only two articles addressed the association of specific clozapine levels with clinical response and adverse side effects (Perry et al. Reference Perry, Bever, Arndt and Combs1998; Mitchell, Reference Mitchell2000). VanderZwaag et al. (Reference VanderZwaag, McGee, McEvoy, Freudenreich, Wilson and Cooper1996) reported the superior efficacy of the 200–300 ng ml−1 and 350–450 ng ml−1 serum clozapine level range over the 50–150 ng ml−1 range, with no advantage for 350–450 ng ml−1 over 200–300 ng ml−1. According to VanderZwaag et al. (Reference VanderZwaag, McGee, McEvoy, Freudenreich, Wilson and Cooper1996) giving a single dose in the evening will cause a higher elevation in plasma clozapine levels in blood drawn the next day as compared with divided dosing schedules. In contrast, Liu et al. (Reference Liu, Chang, Wei, Lin and Jann1996) reported an upper therapeutic limit, with clozapine plasma levels greater than 700 ng ml−1 producing diminished response rates. Freeman & Oyewumi (Reference Freeman and Oyewumi1997) reported that serum clozapine level above 1000 mg ml−1 is linked to higher rates of adverse events.
Remington et al. (Reference Remington, Agid, Foussias, Ferguson, McDonald and Powell2013) discussed very limited evidence for an upper threshold related to clinical response, a ‘ceiling effect’, in the range 600–838 ng ml−1. The upper threshold for serum clozapine levels that will consistently result in serious side effects remains unclear. Remington et al. (Reference Remington, Agid, Foussias, Ferguson, McDonald and Powell2013) noted that clozapine doses greater than 500–600 mg per day are associated with a higher risk of seizures, however there is no well-defined plasma threshold for seizures. Remington et al. (Reference Remington, Agid, Foussias, Ferguson, McDonald and Powell2013) commented that a safety-related threshold cannot be established based on increased risk of other potentially serious side effects such as cardiac and gastrointestinal side effects.
In a recent paper focusing on medical reasons that lead to termination of clozapine Nielsen et al. (Reference Nielsen, Correll, Manu and Kane2013) concluded that many potentially serious side effects (such as agranulocytosis and myocarditis) are idiosyncratic with no dose dependency, while other potentially serious side effects, such as tachycardia and seizures, appear to be related to dose. Nielsen et al. (Reference Nielsen, Correll, Manu and Kane2013) made the general statement that several pharmacokinetic interactions may change plasma levels and contribute to toxic clozapine levels but they did not elaborate on this comment.
The major metabolites of clozapine (N-desmethyl clozapine (norclozapine) and clozapine N-oxide) are the result of N-oxidation of clozapine mainly catalyzed by CYP3A4 (Eiermann et al. Reference Eiermann, Engel, Johansson, Zanger and Bertilsson1997) and demethylation by CYP3A4 and CYP1A2 (Eiermann et al. Reference Eiermann, Engel, Johansson, Zanger and Bertilsson1997) respectively. In our review, drug interactions resulted in wide variability in the increased plasma norclozapine levels from 101 to 1559 ng ml−1. Most of the case reports that we reviewed with a ratio of clozapine dose to norclozapine level (CLZ/NCZ) < 2.7, were associated with either worsening side effects or discontinuation of medication (Table 3). The practical use of the clozapine:norclozapine ratio was reported by Couchman et al. (Reference Couchman, Bowskill, Handley, Patel and Flanagan2013). A ratio of ≤ 0.5 suggests either poor adherence within the last 24 hours, or that alterations in dose schedule might be beneficial. A ratio of ≥ 3 suggests that either absorption of clozapine from the last dose may not have been completed at the time the sample was obtained, or that clozapine metabolism is saturated either because of the dose prescribed or because of inhibition of clozapine metabolism by co-administered medications.
From our review of the literature, it appears that the preponderance of serious side effects occurs at clozapine serum levels above 1000 ng ml−1 (Tables 1–3). While there are reported side effects at serum levels less than 1000 ng ml−1, generally the adverse effects are not as serious as those reported above this threshold, with the exception of one case of neuroleptic malignant syndrome. While certainly not definitive, these reports suggest that a clozapine serum level above 1000 ng ml−1 necessitates very close monitoring of the clinical status of the patient, and that the clinician may be well advised to take steps to lower the serum clozapine level.
Conclusions
We suggest that clozapine levels should be closely monitored in the context of potential interactions of medications with clozapine. In the absence of guidelines, it may be prudent to obtain clozapine levels at the first sign of adverse events or worsening of clinical status. Clozapine dose should probably be lowered, with plasma clozapine levels above 1000 ng ml−1 or CLZ/NCZ > 2.7. Due to the difficulty in interpreting the clinical relevance of clozapine levels and absence of recommendations for routine clozapine level monitoring (Eiermann et al. Reference Eiermann, Engel, Johansson, Zanger and Bertilsson1997), clinical judgment should always be used in addition to diagnostic testing. Despite the role of checking medications and smoking status, consideration should be given to age (faster metabolism in younger patients), sex (slower metabolism in females), body weight (higher doses in heavier people) and co-morbid medical conditions as contributory factors. Future research should also focus on the role of both norclozapine and clozapine-N-oxide levels in therapeutic drug monitoring, and the relevance of specific cytochrome P450 enzymes monitoring in drug-drug interactions.
Acknowledgements
We would like to thank Dr Faisil Sethi for his contribution in editing the manuscript and reviewing the tables.
Open access
