Cytochrome P450‐mediated interaction between perazine and risperidone: implications for antipsychotic polypharmacy

Abstract

Background

Although clinically widespread, scientific evidence for antipsychotic polypharmacy is still limited. Combining different drugs increases the potential for drug–drug interactions, enhancing the risk of adverse drug reactions. We aimed to unravel the potential pharmacokinetic interactions between risperidone (RIS) and perazine.

Methods

Using a therapeutic drug monitoring database containing plasma concentrations of RIS and its active metabolite [9‐hydroxyrisperidone (9‐OH‐RIS)], we considered two groups: a group of patients under antipsychotic monotherapy with RIS (n = 40) and a group of patients that was comedicated with perazine (n = 16). Groups were matched for demographic characteristics and daily dosage of RIS. Plasma concentrations, concentrations corrected for the dose (C/D) of RIS, 9‐OH‐RIS and the active moiety (RIS + 9‐OH‐RIS), as well as the metabolic ratios of concentrations of 9‐OH‐RIS/RIS, were compared using nonparametric tests.

Results

All parameters other than plasma concentrations and the C/D ratio of 9‐OH‐RIS differed between groups. Median values for plasma concentrations of the active moiety and C/D of the active moiety were higher in the perazine group (P < 0.001 and P < 0.001, respectively). Differences were driven by variations in the plasma concentrations and C/D of RIS, which were higher in the perazine group (P < 0.001 and P < 0.001, respectively). Metabolic ratios were lower in the perazine group (P = 0.003).

Discussion

The coadministration of perazine in RIS‐medicated patients leads to significantly higher plasma concentrations and C/D values of RIS and its active moiety, and a lower metabolic ratio, reflecting the cytochrome P450 (CYP) 2D6 phenotype. We suggest that the mechanism underlying the effect of perazine on RIS metabolism is based on an inhibition of CYP2D6 and CYP3A4 activity.

What is Already Known about this Subject

• Pharmacokinetic interactions are of high clinical relevance, even in complex clinical situations.

• Little is known about the pharmacokinetic interactions between risperidone (RIS) and perazine.

• Perazine has not been previously known as a potent inhibitor of distinct cytochrome P450 (CYP) enzymes.

What this Study Adds

• Perazine leads to significantly higher plasma concentrations of the parent compound (RIS) and its active moiety (RIS + 9‐hydroxyrisperidone) during combined treatment.

• Dose‐adjusted plasma concentrations (C/D ratio) of RIS and its active moiety increase significantly under combined treatment.

• The findings underscore evidence for a relevant pharmacokinetic interaction, most likely via CYP 2D6 and/or CYP 3A4, due to CYP inhibitory properties of perazine that clinicians should be aware of.

Tables of Links

TARGETS
G protein‐coupled receptors 2	Enzymes 3
D2 receptor http://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=249	Cytochrome P450
	CYP2 family; CYP2D6 http://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=1479
http://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=2635	CYP1 family; CYP1A2 http://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=1494
	CYP3 family; CYP3A4 http://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=1495

LIGANDS
Risperidone	Dopamine

These Tables list key protein targets and ligands in this article that are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY 1, and are permanently archived in the Concise Guide to PHARMACOLOGY 2015/16 2, 3.

Introduction

Antipsychotic polypharmacy has increased over the years, and current evidence suggests that it may have some clinical benefits, most likely by enabling individualized and tailored antipsychotic treatment regimens for better symptom control 4, 5. Consequently, combined treatment strategies are frequently applied for treatment‐refractory symptoms, even in schizophrenia 6, 7. The efficacy of such treatment strategies is due to pharmacodynamic interactions, including amplified dopaminergic and noradrenergic blockade, although the available data on clinical benefits are still limited 8, 9. Moreover, data regarding the efficacy and safety of combined treatment regimens are occasionally conflicting, and tailored strategies lack evidence from clinical studies 10. Antipsychotic polypharmacy has been linked to an increased prescription of anticholinergic agents, which may be reflected in the enhanced prevalence of adverse reactions 11, and combined treatment strategies may also give rise to pharmacokinetic drug–drug interactions (DDIs) 12, 13, 14. From a clinical perspective, data highlight the unmet need for a better understanding of both pharmacodynamic and pharmacokinetic DDIs. Aside from a scientific challenge, this clinically relevant issue needs to be addressed, especially in the treatment of patient subgroups that are more vulnerable to adverse effects, such as elderly patients 15.

Perazine is a first‐generation antipsychotic agent in the phenothiazine class, commonly prescribed over the last decades as an adjuvant to an ongoing antipsychotic treatment, especially in elderly patients. However, data regarding its efficacy and tolerability still remain limited 13 perhaps because its use is limited to only a few countries, such as Germany, Poland, the Netherlands and some others 16. Its main effects include a moderate dopamine D₂ blockade and weak blockade of serotonin (5‐HT_2A) receptors, and it also has antiadrenergic (α₁) and antimuscarinic (M₁) properties 17. Its major metabolic pathway comprises a cytochrome P450 (CYP) 1A2‐ and, to a lesser extent, a CYP3A4‐mediated 5‐sulphoxidation 17. Alternative metabolic pathways may include a CYP2C19‐catalysed N‐demethylation. An increasing amount of preclinical and clinical data demonstrates perazine to have a considerable potential for DDIs in combination with various agents, including antidepressants and other antipsychotic drugs 18, 19, 20. Proposed mechanisms mediating such DDIs include inhibition of CYP1A2, CYP3A4 and CYP2C6 activity 21.

Risperidone (RIS), a benzisoxazole derivative, is a second‐generation antipsychotic agent with antagonistic properties at serotonin 5‐HT₂ and dopamine D₂ receptors 22. RIS has been established over recent years in the treatment of schizophrenia and a broad spectrum of other psychiatric diseases 23, 24, 25. Its primary metabolic pathway comprises a CYP2D6‐catalyzed 9‐hydroxylation, leading to a main active metabolite, 9‐hydroxyrisperidone (9‐OH‐RIS). In vitro data suggest that CYP3A4 and CYP3A5 might be also involved 26, 27, 28. As 9‐OH‐RIS is pharmacologically active and has a far longer half‐life than RIS, clinicians consider the combined concentration of RIS and 9‐OH‐RIS (active moiety) as the most relevant measure of pharmacological activity. According to the AGNP (Arbeitsgemeinschaft für Neuropsychopharmakologie und Pharmakopsychiatrie) consensus guidelines, a so‐called therapeutic reference range is suggested as 20–60 ng ml⁻¹ for the active moiety 29.

Therapeutic drug monitoring is a specific method in clinical pharmacology that can be used successfully to guide psychopharmacological treatment in clinical routine by determining the plasma concentrations of applied drugs. Therapeutic drug monitoring databases are a valuable source for promoting the safety and tolerability of pharmacotherapy in a clinical setting. Furthermore, the evaluation of databases of drug concentrations allows many interesting pharmacokinetic and/or clinical issues to be addressed, such as DDIs 12, 13, the effect of smoking on drug concentrations 30, adverse drug reactions 11, 31 and clinical topics such as treatment responses in correlation with pharmacokinetic patterns of drug concentrations in patients' blood 32, 33, 34. As data on the pharmacokinetics of perazine (e.g. inhibitory effects on distinct isoenzymes of CYP P450) are mainly deriving from preclinical studies and case reports, we leveraged a large therapeutic drug monitoring database of patients whose antipsychotic treatment with RIS was individually optimized using therapeutic drug monitoring to investigate further the potential pharmacokinetic effects of perazine on RIS metabolism.

Materials and methods

A large database as part of KONBEST, a web‐based laboratory information management system for therapeutic drug monitoring laboratories 35, containing the plasma concentrations of RIS and 9‐OH‐RIS of 2293 patients, was analysed. Data collection took place between 2006 and 2015 as part of the clinical routine in different institutions of the AGATE (Arbeitsgemeinschaft Arzneimitteltherapie bei psychischen Erkrankungen). AGATE is a cooperative for drug safety and efficacy in the treatment of psychiatric diseases (http://www.amuep‐agate.de) 36. Retrospective analysis of clinical data for this study was in accordance with the local regulatory authority.

In this naturalistic database, patients were under medication with RIS for different reasons, with only patients with organic mental disorders being excluded. Also excluded from the analysis were patients receiving depot formulations and those receiving concomitant medication with possible CYP2D6 inhibitory or CYP3A4 inhibitory or inducing properties, according to the suggestions by the US Food and Drug Administration 29, 37. We considered two groups: a group of patients receiving concomitant medication with perazine (group R_PER) and a group receiving RIS monotherapy matched for age, gender and daily dosage of RIS with the R_PER group (control group, R₀). Additional matching processes (i.e. for diagnoses, severity of illness, or length or time of onset and the length or duration of illness) were not undertaken.

Quantification of RIS and 9‐OH‐RIS

Clinicians were asked to draw blood just before drug administration (trough concentration) at steady state (>5 elimination half‐lives under the same drug dose). RIS and 9‐OH‐RIS concentrations were determined by high‐performance liquid chromatography with ultraviolet detection (HPLC/UV) 38. The limit of detection (LOD), defined as a signal‐to‐noise ratio of 3:1, was 5 ng ml⁻¹ for both RIS and 9‐OH‐RIS. The interday precision, determined as duplicates on three different days at 5 ng ml⁻¹, was 5.56% of the mean and 5.21% of the mean, respectively. For further details on the analytical method, see Paulzen et al. 34.

Statistical analysis

The analysis included the comparison of two study groups, as defined above. We compared the distributions of the plasma concentration of the parent compound (RIS), the active metabolite (9‐OH‐RIS) and the active moiety (RIS + 9‐OH‐RIS) between the groups. Additional comparisons included the dose‐corrected plasma concentrations (C/D) of RIS, 9‐OH‐RIS and active moiety as well as the metabolite‐to‐parent drug ratios of 9‐OH‐RIS/RIS for identification of the CYP2D6 metabolizer status phenotype. Histograms yielded evidence of non‐normal distributions, so a nonparametric Mann–Whitney U test (MWU) with a significance level of 0.05 was conducted. Statistical analysis was carried out using IBM SPSS version 18.0 (SPSS Inc., Chicago, IL, USA).

Results

After exclusions for confounding comedications, we detected 16 RIS‐medicated patients who were comedicated with perazine (R_PER group). Out of the 852 patients receiving RIS monotherapy, we considered a subgroup matched with the R_PER group for age (P = 0.364), gender (P = 0.733) and daily dosage of RIS (P = 0.97) (n = 40, R₀ group). The demographic data are summarized in Table 1.

Table 1.

Patients' demographic characteristics

Group	Number	age (±SD)	Gender		DD RIS (mg day–1)
			% Females	% Males	Median (range)
R0	40	42.0 ± 8.2	57.5	42.5	6.0 (4.00–10.0)
RPER	16	41.6 ± 15.4	62.5	37.5	5.5 (2.00–10.0)

DD, daily dose; R₀, group receiving RIS monotherapy (control group) matched for age, gender and daily dosage of RIS with the R_PER group; R_PER, group of patients receiving concomitant medication with perazine; RIS, risperidone; SD, standard deviation

The median plasma concentrations (ng ml⁻¹) of RIS, 9‐OH‐RIS, RIS + 9‐OH‐RIS, as well as the 9‐OH‐RIS/RIS ratios, are displayed in Table 2.

Table 2.

Median plasma concentrations (range) of risperidone (RIS), 9‐hydroxyrisperidone (9‐OH‐RIS) and active moiety (RIS + 9‐OH‐RIS), and metabolic ratios in the study groups

Group	RIS [ng ml−1]	9‐OH‐RIS [ng ml−1]	RIS + 9‐OH‐RIS [ng ml−1]	9‐OH‐RIS/RIS
R0	5.25 (0.3–71.0)	23.0 (1.0–61.0)	30.0 (7.1–104.0)	3.22 (0.05–110.0)
RPER	25.4a (5.0–157.0)	30.0 (7.0–61.0)	64.6a , b (34.0–185.0)	1.15a (0.17–8.26)

R₀, group (control group) receiving RIS monotherapy matched for age, gender and daily dosage of RIS with the R_PER group; R_PER, group of patients receiving concomitant medication with perazine

^aPlasma concentrations of RIS and active moiety were significantly higher, while metabolic ratios were lower in the R_PER group than in the control group (P < 0.001, P < 0.001 and P = 0.003, respectively, in the Mann–Whitney U test)

^bAs the therapeutic reference range is suggested as 20–60 ng ml⁻¹ for the active moiety (RIS + 9‐OH‐RIS), patients receiving combined treatment are more likely to have supratherapeutic plasma concentrations

Table 3 shows the C/D [(ng ml⁻¹)/(mg day⁻¹)], for RIS, 9‐OH‐RIS and RIS + 9‐OH‐RIS for both groups.

Table 3.

Median dose‐corrected plasma concentrations (C/D) of risperidone (RIS), 9‐hydroxyrisperidone (9‐OH‐RIS) and active moiety (RIS + 9‐OH‐RIS) in the different groups

Group	C/D RIS [(ng ml−1)/(mg/day)]	C/D 9‐OH‐RIS [(ng ml−1)/(mg/day)]	C/D RIS + 9‐OH‐RIS [(ng ml−1)/(mg day−1)]
R0	1.13 (0.05–11.83)	4.25 (0.17–13.75)	5.45 (0.89–17.33)
RPER	4.65a (1.06–21.6)	5.21 (1.75–15.0)	10.21a (7.66–27.6)a

R₀, group receiving RIS monotherapy (control group) matched for age, gender and daily dosage of RIS with the R_PER group; R_PER, group of patients receiving concomitant medication with perazine

^aPlasma concentrations corrected by dose values (C/D) for RIS and RIS + 9‐OH‐RIS in the R_PER group were significantly higher than in the control group (P < 0.001 and P < 0.001, respectively, in the Mann–Whitney U test)

The comparison of the plasma concentrations of RIS, 9‐OH‐RIS and RIS + 9‐OH‐RIS between the groups yielded significant differences in all cases except the plasma concentration of the active metabolite (P = 0.182). Plasma concentrations of active moiety and RIS were both higher in the R_PER than R₀ group (P < 0.001 for both). Likewise, differences reached statistical significance for the C/D of active moiety and RIS (P < 0.001 for both), but not for 9‐0H‐RIS (P = 0.11) (see Figure 1). In both cases, patients comedicated with perazine demonstrated higher values. Furthermore, the metabolic ratio of 9‐OH‐RIS to RIS was significantly lower in the R_PER than the R₀ group (P = 0.003) (see Figure 2).

Figure 1.

Median dose‐corrected plasma concentrations (C/D) of risperidone (RIS) (A) and its active moiety (RIS + 9‐hydroxyrisperidone) (B) in [(ng ml⁻¹)/mg day⁻¹)]CI, confidence interval

Figure 2.

The metabolic ratio (9‐OH‐RIS/RIS) in the perazine group was significantly lower than in the control group. CI, confidence interval

Discussion

Therapeutic drug monitoring helps to control for interindividual variabilities in plasma RIS and 9‐OH‐RIS concentrations, even in conditions without potentially confounding comedications. Differences may occur for various reasons, such as genetic polymorphisms in drug metabolizing enzymes, differences in drug transporters or may be a due to environmental factors 39, 40, 41.

The addition of a first‐generation antipsychotic drug to an ongoing antipsychotic treatment is a widespread psychopharmacological strategy, offering various therapeutic benefits 42. However, this can be at the expense of increasing the risk of adverse effects due to DDIs. Consequently, the challenge of unravelling the pharmacodynamic and pharmacokinetic background of such treatment is an absolute necessity. Knowledge about the pharmacodynamics and pharmacokinetics of simultaneously prescribed drugs is mandatory to minimize the risk of adverse events commonly reported by patients under such therapeutic regimens 43. However, this need remains unmet for various widely prescribed psychotropic (e.g. perazine) and nonpsychotropic drugs 12.

Comparisons of plasma concentrations of RIS, 9‐OH‐RIS and active moiety were conducted between a group receiving combination therapy with RIS plus perazine and a group receiving RIS monotherapy; both groups were matched for demographics as well as for daily dosage. Differences in pharmacokinetic patterns would therefore not have resulted from differences in the applied dosage. Our major observation was that the addition of perazine obviously altered the metabolism of RIS significantly, showing differences in the concentrations of the parent compound, the active moiety, and for the C/D and the 9‐OH‐RIS/RIS ratios. Patients receiving combined treatment had a fivefold higher plasma concentration of RIS and a twofold higher concentration of active moiety under concomitant medication with perazine compared with patients under monotherapy. Moreover, 9‐OH‐RIS/RIS ratios were significantly lower in comedicated patients. These findings suggest a strong and previously undescribed inhibiting effect of perazine on the CYP‐mediated metabolism of RIS. As yet, there are no human data on the potential pharmacokinetic interactions between perazine and RIS, but animal data suggest that phenothiazines inhibit the metabolism of RIS owing to a blockade of the CYP2D subfamily in rats 44. Furthermore, preclinical data have shown perazine to have an inhibiting effect on CYP2D6 substrates such as fluoxetine and imipramine 18. Nonetheless, pharmacokinetic interactions between perazine and antipsychotic agents in humans are mostly described in case reports demonstrating enhanced plasma clozapine concentrations, most likely via a blockade of CYP1A2‐mediated metabolism, when perazine is added to an ongoing treatment with clozapine 19, 20, 29. However, the metabolism of clozapine is also mediated by CYP3A4, and to a lesser extent by CYP2D6, so blockade of CYP3A4 and CYP2D6 may also be involved. The suggested mechanism ties in well with other in vitro and preclinical data implying an inhibiting effect of perazine on CYP3A4 45, 46. Both animal data and clinical observations corroborate the current findings. We therefore suggest that the mechanism underlying the inhibition by perazine of RIS metabolism may be the result of a blockade of CYP2D6, and possibly CYP3A4, activity.

The dearth of clinical or observational studies in the literature hampers the further clarification of the precise mechanism underlying this finding and/or its clinical significance. Yet, our findings indicate that the addition of perazine to an ongoing treatment with RIS leads to considerable changes in the pharmacokinetics of RIS, leading to increased plasma concentrations of RIS and the active moiety. Given the clinical significance of the active moiety of RIS 29, this notion may be of crucial importance when combining different antipsychotic agents. Clinicians may consider antipsychotic agents other than perazine when the therapeutic regimen includes RIS, in order to minimize any pharmacokinetic interactions, presumably mediated via the CYP pathway. If the use of perazine cannot be avoided for clinical reasons, a combination of RIS and perazine should be handled with caution and therapeutic drug monitoring can be employed successfully to guide psychopharmacological strategies.

Limitations

As we retrospectively studied a large population of naturalistic nature, patient information could be considered less reliable than in prospective studies. The application of an experimental design including matching procedures introduces an additional shortcoming as these designs often suffer from reduced statistical power. Nevertheless, matching does not necessarily alter the findings 47. Many clinical parameters, including the onset and duration of illness, clinical rating scales and knowledge about adverse effects, comorbidities, the duration of prior RIS exposure, renal function parameters and smoking habits, was not available, so further analyses of confounding effects could not be undertaken. Moreover, the lack of clinical data considerably hampers the interpretation of this evidence. Furthermore, we could not control for potentially large individual variations in sampling time (although clinicians were asked to draw blood at trough level times) as a result of the clinical setting, which may have partially accounted for the pronounced interindividual variation in plasma concentrations and metabolic ratios. A large interindividual variability in RIS and 9‐OH‐RIS concentrations has been already reported in the literature 40. In the case of multiple plasma concentration determinations, we minimized patient bias by including only one analysis per patient (the most recent one). In order to eliminate the effect of confounding pharmacokinetic factors on plasma concentration, we excluded patients receiving concomitant potent modulators of CYP activity from the analysis. As perazine is only available in some countries (Germany, Poland, the Netherlands and some others), the utility of our results is limited to clinicians in these countries.

Competing Interests

Ekkehard Haen has received speaker's or consultancy fees from the following pharmaceutical companies: Servier, Novartis and Janssen‐Cilag. He reports no conflict of interest with this publication. Christoph Hiemke has received speaker's or consultancy fees from the following pharmaceutical companies: Janssen‐Cilag and Servier. He reports no conflict of interest with this publication. Gerhard Gründer has served as a consultant for Boehringer Ingelheim, Cheplapharm, Eli Lilly, Lundbeck, Ono Pharmaceuticals, Roche, Servier and Takeda. He has served on the speakers' bureau of Eli Lilly, Gedeon Richter, Janssen Cilag, Lundbeck, Roche, Servier and Trommsdorf. He has received grant support from Boehringer Ingelheim and Roche. He reports no conflict of interest with this publication. Georgios Schoretsanitis received a grant from the bequest ‘in memory of Maria Zaoussi’, State Scholarships Foundation, Greece, for clinical research in psychiatry for the academic year 2015–2016. All other authors also declare no conflicts of interest.

The authors wish to express their gratitude to the many people who contributed with excellent professional technical and pharmacological competence to building up the KONBEST database (ranked among the professional groups in historical order). A. Köstlbacher created the KONBEST software in his PhD thesis, based on an idea by E. Haen, C. Greiner and D. Melchner. A Köstlbacher and his colleague A. Haas continuously maintain the KONBEST software and its data‐mining platform (Haas & Köstlbacher GbR, Regensburg, Germany). The authors also thank Mrs Paraskevi Giannakaki for her expertise in statistics. Thanks too to the laboratory technicians who performed the quantitative analysis: D. Melchner, T. Jahner, S. Beck, A. Dörfelt, U. Holzinger, and F. Pfaff‐Haimerl. The research study did not receive funds or support from any source.

Contributors

M.P., G.S., G.G., C.H., E.H. and B.S. participated in research design; M.P., G.S. and S.E.L. performed data analysis; M.P., G.G., C.H., E.H., B.S., S.E.L. and G.S. wrote or contributed to the writing of the manuscript.

Paulzen, M. , Haen, E. , Hiemke, C. , Stegmann, B. , Lammertz, S. E. , Gründer, G. , and Schoretsanitis, G. (2017) Cytochrome P450‐mediated interaction between perazine and risperidone: implications for antipsychotic polypharmacy. Br J Clin Pharmacol, 83: 1668–1675. doi: 10.1111/bcp.13255.