Impact of amiodarone use on metoprolol concentrations, α‐OH‐metoprolol concentrations, metoprolol dosing and heart rate: A cross‐sectional study

Abstract

Small studies suggest that amiodarone is a weak inhibitor of cytochrome P450 (CYP) 2D6. Inhibition of CYP2D6 leads to increases in concentrations of drugs metabolized by the enzyme, such as metoprolol. Considering that both metoprolol and amiodarone have β‐adrenergic blocking properties and that the modest interaction between the two drugs would result in increased metoprolol concentrations, this could lead to a higher risk of bradycardia and atrioventricular block. The primary objective of this study was to evaluate whether metoprolol plasma concentrations collected at random timepoints from patients enrolled in the Montreal Heart Institute Hospital Cohort could be useful in identifying the modest pharmacokinetic interaction between amiodarone and metoprolol. We performed an analysis of a cross‐sectional study, conducted as part of the Montreal Heart Institute Hospital Cohort. All participants were self‐described “White” adults with metoprolol being a part of their daily pharmacotherapy regimen. Of the 999 patients being treated with metoprolol, 36 were also taking amiodarone. Amiodarone use was associated with higher metoprolol concentrations following adjustment for different covariates (p = .0132). Consistently, the association between amiodarone use and lower heart rate was apparent and significant after adjustment for all covariates under study (p = .0001). Our results highlight that single randomly collected blood samples can be leveraged to detect modest pharmacokinetic interactions.

Abbreviations

AV

Atrioventricular

CYP

Cytochrome P450

DEA

Desethylamiodarone

IM

Intermediate metabolizer

LC–MS/MS

High‐pressure liquid chromatography coupled to electrospray ionization tandem mass spectrometry

LLOQ

Lower limits of quantification

MHI

Montreal Heart Institute

MRM

Multiple reaction monitoring

NM

Normal metabolizer

P‐gp

P‐glycoprotein

PM

Poor metabolizer

UM

Ultrarapid metabolizer

1. INTRODUCTION

Metoprolol is a β1‐adrenoceptor antagonist commonly used to treat cardiovascular conditions such as hypertension and atrial fibrillation and is primarily metabolized by cytochrome P450 (CYP) 2D6. ¹ , ² CYP2D6 is responsible for approximately 80% of the total clearance of metoprolol into α‐hydroxymetoprolol, commonly referred as α‐OH‐metoprolol (10%), O‐desmethylmetoprolol (approximately 65%) and N‐deisopropylmetoprolol (10%). ³ This enzyme presents a highly variable activity profile owing to the genetic polymorphism that influences its expression and function. ⁴ , ⁵ , ⁶ CYP2C9, CYP3A4, and CYP2B6 contribute to the metabolism of metoprolol to a lesser extent. ³

Amiodarone is a potent antiarrhythmic drug used for rhythm and rate control in the treatment of multiple arrhythmias. ⁷ Even though amiodarone is not currently widely used, it remains an antiarrhythmic of choice in certain cases, such as when first‐line agents for rhythm control of atrial fibrillation have failed or in those with heart failure. ⁸ , ⁹ Amiodarone and especially its active metabolite desethylamiodarone (DEA) inhibit the activity of several CYPs, ¹⁰ , ¹¹ including CYP2D6, for which it is a weak inhibitor. Thus, coadministration of amiodarone leads to increases in concentrations of drugs metabolized by CYP2D6, such as metoprolol. ³ , ¹² , ¹³ These increases in metoprolol concentrations, combined with the fact that both amiodarone and metoprolol possess β‐adrenergic blocking properties (i.e., heart rate reduction), and that amiodarone present calcium channel‐blocking properties (negative inotropy), explain the increased risk of bradycardia and atrioventricular (AV) block in patients taking these two agents concomitantly. ¹⁴ , ¹⁵ , ¹⁶

In a recent analysis, we demonstrated that the use of metoprolol plasma concentrations collected at random timepoints as part of a large hospital cohort could be leveraged to identify multiple established determinants of metoprolol concentrations, including metoprolol dosing, CYP2D6 genotype‐inferred phenotype and the use of moderate to strong CYP2D6 inhibitors. ¹⁷ In the current study, we sought to expand on these findings to determine if random plasma metoprolol concentrations in regard to last the dose can be used to identify more modest pharmacokinetic interactions, such as the very well‐known interaction between amiodarone and metoprolol. Moreover, we explored the clinical impact of the interaction exerted by amiodarone on patients' heart rate.

2. METHODS

2.1. Study design

We performed an analysis of a cross‐sectional study, conducted as part of the Montreal Heart Institute (MHI) Hospital Cohort. ¹⁷ The MHI Hospital Cohort collects information on the medical, genealogical, psychological, biological, pharmacological, and genetic profile of participants. The methods of the MHI Hospital Cohort and of the cross‐sectional study have been previously described. ¹⁸ , ¹⁹

2.2. Study population

The selection of individuals has been previously reported. ¹⁷ Briefly, this study included self‐reported “White” males and females aged 18 years or older with metoprolol as part of their daily pharmacotherapy regimen at the time of their baseline visit and blood sampling.

2.3. Study endpoints

The primary objective of this study was to assess the association between the use of amiodarone and concentrations of metoprolol. We also investigated the effects of amiodarone on concentrations of α‐OH‐metoprolol, on metoprolol daily dose and on patients' baseline heart rate as secondary endpoints.

2.4. Measurement of metoprolol concentrations

Plasma samples were collected at random timepoints with regard to drug intake. Concentrations of metoprolol and α‐OH‐metoprolol were quantified at the Université de Montréal's bioanalytical laboratory of the Platform of Biopharmacy as previously reported. ¹⁷ Quantification of metoprolol concentrations and its metabolite was completed using high‐pressure liquid chromatography coupled to electrospray ionization tandem mass spectrometry (LC–MS/MS) based on selective multiple reaction monitoring (MRM), with a validated quantification range of 1 to 1000 ng/mL for both metabolites.

2.5. Genotyping

The genotyping of all participants was performed using a combination of Agena's iPLEX ADME PGx Pro Panel 1.0 or MHI ADME Panel V3.0 (Agena Bioscience) and the MassARRAY® Analyzer Compact 384 system as previously described. ¹⁷ All manipulations were completed at the Université de Montréal Beaulieu‐Saucier Pharmacogenomics Centre. CYP2D6 genotype‐inferred phenotypes were constructed according to the current classification guidelines and participants were categorized into poor metabolizer (PM), intermediate metabolizer (IM), normal metabolizer (NM) and ultrarapid metabolizer (UM) phenotypes. ²⁰

2.6. Statistical analyses

Descriptive statistics of the clinical and demographic characteristics were obtained, and differences between amiodarone users and non‐users were tested using Kruskal‐Wallis for continuous variables and Chi‐Square test or Fisher exact test for categorical variables. Categorical variables were expressed as counts and percentages while continuous variables were presented using means and standard deviations. A series of linear regression models were fitted to investigate the associations between amiodarone use and the study endpoints. For the analyses of concentrations and metoprolol dosing, the model adjustment variables were CYP2D6 genotype‐inferred phenotypes, age, sex, metoprolol dose, weight and use of CYP2D6 inhibitors. For the endpoint of heart rate, history of atrial fibrillation/flutter and the use of heart rate lowering drugs other than amiodarone were included as additional adjustment variables. All endpoints were log‐transformed in order to satisfy the normality assumption. Concentration values below the lower limits of quantification (LLOQ) were set to 0. All statistical comparisons were two‐tailed and a P‐value lower than 0.05 was considered statistically significant. Statistical analyses were performed using SAS version 9.4 (SAS Institute, North Carolina, USA).

2.7. Ethics statement

This study was conducted in accordance with the MHI Hospital Cohort protocol conforming to the Management Policy of the MHI Hospital Cohort. This study was approved by the institution's Scientific and Ethics Committees. All included participants had previously signed an informed consent form prior to their enrollment in the MHI Hospital Cohort biobank.

2.8. Nomenclature of targets and ligands

Key protein targets and ligands in this article are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY, ²¹ and are permanently archived in the Concise Guide to PHARMACOLOGY 2021/22. ²² , ²³

3. RESULTS

3.1. Study population

A total of 999 patients were included in this study. Of these, 36 participants reported being treated with oral amiodarone at the time of plasma sampling. In general, characteristics of amiodarone users were similar to those of non‐users. Descriptive results are detailed in Table 1.

TABLE 1.

Baseline characteristics of users and non‐users of amiodarone.

Variable	Non‐users of amiodarone n = 963 (96.4%)	Users of amiodarone n = 36 (3.6%)	All n = 999 (100.0%)	p‐value a
Socio‐demographic variables
Age, n (%)	66.47 ± 8.79	68.06 ± 7.15	66.52 ± 8.74	.4552
Female, n (%)	264 (27.41%)	5 (13.89%)	269 (26.93%)	.0725
Self‐reported “White”, n (%)	963 (100.0%)	36 (100.0%)	999 (100.0%)	1.0000
Lifestyle and physical measures
Smoking status, n (%)
Never‐smoker	268 (27.83%)	10 (27.78%)	278 (27.83%)	.9291
Past‐smoker	614 (63.76%)	24 (66.67%)	638 (63.86%)
Current‐smoker	81 (8.41%)	2 (5.56%)	83 (8.31%)
Weight (kg), mean ± SD	84.49 ± 17.11	81.28 ± 14.61	84.37 ± 17.03	.3553
BMI, mean ± SD	30.11 ± 5.40	28.09 ± 4.62	30.03 ± 5.39	.0660
Heart rate, mean ± SD	64.03 ± 10.62	57.58 ± 13.37	63.80 ± 10.79	3.73 × 10 −5
Cardiovascular chronic conditions
Hypertension, n (%)	755 (78.40%)	29 (80.56%)	784 (78.48%)	.7574
Diabetes, n (%)
No	669 (69.47%)	27 (75.00%)	696 (69.67%)	.1644
Type 1	6 (0.62%)	1 (2.78%)	7 (0.70%)
Type 2	288 (29.91%)	8 (22.22%)	296 (29.63%)
Dyslipidemia, n (%)	817 (84.84%)	30 (83.33%)	847 (84.78%)	.8049
Myocardial infarction, n (%)	411 (42.68%)	17 (47.22%)	428 (42.84%)	.5886
Atrial fibrillation or flutter, n (%)	326 (33.99%)	32 (88.89%)	358 (35.98%)	2.96 × 10 −11
Medications, n (%)
Aspirin	768 (79.75%)	20 (55.56%)	788 (78.88%)	.0005
Other antiplatelet agents	148 (15.40%)	1 (2.78%)	149 (14.94%)	.0318
ACE inhibitors	324 (33.68%)	18 (50.00%)	342 (34.27%)	.0428
Angiotensin II receptor blockers	273 (28.35%)	8 (22.22%)	281 (28.13%)	.4221
Calcium channel blockers	253 (26.27%)	8 (22.22%)	261 (26.13%)	.5871
Warfarin	175 (18.17%)	24 (66.67%)	199 (19.92%)	8.51 × 10 −13
Novel oral anticoagulants	31 (3.22%)	5 (13.89%)	36 (3.60%)	.0007
Digoxin	40 (4.15%)	3 (8.33%)	43 (4.30%)	.1987
Other antiarrhythmic agents	14 (1.45%)	0 (0.00%)	14 (1.40%)	1.0000
Diuretics	298 (30.94%)	23 (63.89%)	321 (32.13%)	3.24 × 10 −5
Statins	792 (82.24%)	26 (72.22%)	818 (81.88%)	.1254
Oral hypoglycemic agents	248 (25.75%)	7 (19.44%)	255 (25.53%)	.3940
Interactions
Moderate to strong CYP2D6 inhibitors, n (%)
Moderate	7 (0.73%)	0 (0.00%)	7 (0.70%)	1.0000
Strong	28 (2.91%)	1 (2.78%)	29 (2.90%)	1.0000
Daily amiodarone dose
Mean daily dose (mg), mean ± SD	NA	214.71 ± 50.04	NA	NA
Daily dose (mg), by categories, n (%)
200	NA	31 (91.18%)	NA	NA
300	NA	1 (2.94%)	NA
400	NA	2 (5.88%)	NA
Daily metoprolol dose
Mean daily dose (mg), mean ± SD	84.61 ± 57.53	77.08 ± 42.31	84.33 ± 57.05	.7590
Daily dose (mg), by categories, n (%)
≤12.5	16 (1.66%)	0 (0.00%)	16 (1.60%)	.7303
>12.5–25	139 (14.46%)	5 (13.89%)	144 (14.44%)
>25–50	338 (35.17%)	13 (36.11%)	351 (35.21%)
>50–100	304 (31.63%)	14 (38.89%)	318 (31.90%)
>100–150	60 (6.24%)	3 (8.33%)	63 (6.32%)
>150–200	90 (9.37%)	1 (2.78%)	91 (9.13%)
>200	14 (1.46%)	0 (0.00%)	14 (1.40%)
Plasma concentrations
Plasma metoprolol (ng/ml), mean ± SD	111.39 ± 143.26	101.56 ± 76.54	111.04 ± 141.40	.1313
Plasma α‐OH‐metoprolol (ng/ml), mean ± SD	48.62 ± 53.00	50.61 ± 38.45	48.69 ± 52.53	.3133
CYP2D6 genotype‐inferred phenotypes (n = 996), n (%)
Poor metabolizer (PM)	44 (4.58%)	0 (0.00%)	44 (4.42%)	.4847
Intermediate metabolizer (IM)	326 (33.96%)	16 (44.44%)	342 (34.34%)
Normal metabolizer (NM)	524 (54.58%)	18 (50.00%)	542 (54.42%)
Ultrarapid metabolizer (UM)	66 (6.88%)	2 (5.56%)	68 (6.83%)

Note: Significant p‐values (<.05) are highlighted in bold.

Abbreviations: ACE, angiotensin‐converting enzyme; BMI, body mass index; LLOQ, lower limit of quantification; NA, Not applicable; SD, standard deviation.

^a p‐value for continuous variables: Kruskal‐Wallis test, categorical variables: Chi‐Square test or Fisher exact test.

3.2. Amiodarone and plasma concentrations of metoprolol

The unadjusted results showed no significant difference in metoprolol plasma concentrations between amiodarone users and non‐users (Table 1; Figure S1A), but metoprolol concentrations:dose ratio was significantly higher in amiodarone‐treated participants than in non‐users of amiodarone (p < .04, Figure S1B). The association between metoprolol concentrations and amiodarone use was apparent in all models of adjusted analyses but reached significance level when adjusting for metoprolol dose (p = .0095, Model 4 in Table 2). This association remained significant after adjustment for all other potential covariates, including the use of moderate to strong CYP2D6 inhibitors (all p < .02, Table 2).

TABLE 2.

Association between metoprolol plasma concentrations and use of amiodarone.

Effect	Model 1		Model 2		Model 3		Model 4		Model 5		Model 6
	Estimate (SE)	p‐value	Estimate (SE)	p‐value	Estimate (SE)	p‐value	Estimate (SE)	p‐value	Estimate (SE)	p‐value	Estimate (SE)	p‐value
Use of amiodarone	0.366 (0.233)	.1155	0.353 (0.225)	.1176	0.382 (0.222)	.0860	0.476 (0.183)	.0095	0.446 (0.182)	.0146	0.449 (0.181)	.0132
CYP2D6 genotype‐inferred phenotype	–	–	−0.505 (0.062)	1.4 × 10 −15	−0.511 (0.061)	2.9 × 10 −16	−0.631 (0.051)	5.0 × 10 −33	−0.634 (0.050)	1.3 × 10 −33	−0.630 (0.050)	1.1 × 10 −33
Age	–	–	–	–	0.016 (0.005)	.0006	0.015 (0.004)	.0001	0.012 (0.004)	.0030	0.012 (0.004)	.0029
Female sex	–	–	–	–	0.408 (0.094)	1.5 × 10 −5	0.404 (0.077)	2.0 × 10 −7	0.311 (0.081)	.0001	0.299 (0.080)	.0002
Metoprolol daily dose	–	–	–	–	–	–	0.013 (0.001)	1.4 × 10 −85	0.013 (0.001)	3.1 × 10 −88	0.013 (0.001)	1.3 × 10 −89
Weight	–	–	–	–	–	–	–	–	−0.008 (0.002)	.0003	−0.008 (0.002)	.0003
Coadministration of moderate to strong CYP2D6 inhibitors	–	–	–	–	–	–	–	–	–	–	0.386 (0.098)	.0001

Note: Model 1: crude model; model 2: model adjusted for CYP2D6 inferred phenotype; model 3: model adjusted for CYP2D6 inferred phenotype, age and sex; model 4: model adjusted for CYP2D6 inferred phenotype, age, sex and metoprolol dose; model 5: model adjusted for CYP2D6 inferred phenotype, age, sex, metoprolol dose and weight; model 6: model adjusted for CYP2D6 inferred phenotype, age, sex, metoprolol dose, weight, and CYP2D6 inhibitors. Intercepts for model 1: 3.980, model 2: 4.807, model 3: 3.616, model 4: 2.800, model 5: 3.680 and model 6: 3.643.

Significant p‐values (<.05) are highlighted in bold.

3.3. Amiodarone and plasma concentrations of α‐OH‐metoprolol

The unadjusted results showed no difference between amiodarone users and non‐users in α‐OH‐metoprolol concentrations, the main metabolite of metoprolol, and in α‐OH‐metoprolol concentration:dose ratio (Figure S1C,D). In adjusted analyses, the association between concentrations of α‐OH‐metoprolol and use of amiodarone reached significance level in all models that adjusted for metoprolol daily dose (all p < .009, Table S1).

3.4. Amiodarone and daily metoprolol dosing

No significant association was observed between the use of amiodarone and metoprolol daily dose (Table S2).

3.5. Amiodarone and heart rate

The association between amiodarone and heart rate was significant in all models, independently of CYP2D6 genotype‐inferred phenotype, age, female sex, metoprolol daily dose, weight, history of atrial flutter or atrial fibrillation, use of non‐amiodarone heart rate lowering drugs and use of moderate to strong CYP2D6 inhibitors (all p < .0003, Table 3 ).

TABLE 3.

Association between heart rate and use of amiodarone.

Effect	Model 1		Model 2		Model 3		Model 4		Model 5		Model 6		Model 7		Model 8
	Estimate (SE)	p‐value	Estimate (SE)	p‐value	Estimate (SE)	p‐value	Estimate (SE)	p‐value	Estimate (SE)	p‐value	Estimate (SE)	p‐value	Estimate (SE)	p‐value	Estimate (SE)	p‐value
Use of amiodarone	−0.114 (0.028)	.0001	−0.113 (0.028)	.0001	−0.107 (0.028)	.0001	−0.107 (0.028)	.0001	−0.104 (0.028)	.0002	−0.112 (0.028)	.0001	−0.111 (0.028)	.0001	−0.111 (0.028)	.0001
CYP2D6 genotype‐inferred phenotype	–	–	0.029 (0.008)	.0002	0.028 (0.008)	.0003	0.028 (0.008)	.0003	0.028 (0.008)	.0003	0.028 (0.008)	.0003	0.028 (0.008)	.0004	0.028 (0.008)	.0003
Age	–	–	–	–	−0.002 (0.001)	.0013	−0.002 (0.001)	.0012	−0.002 (0.001)	.0081	−0.002 (0.001)	.0049	−0.002 (0.001)	.0047	−0.002 (0.001)	.0047
Female sex	–	–	–	–	0.023 (0.012)	.0457	0.024 (0.012)	.0436	0.033 (0.012)	.0080	0.031 (0.012)	.0108	0.031 (0.012)	.0121	0.031 (0.012)	.0133
Metoprolol daily dose	–	–	–	–	–	–	−4.7 × 10−5 (9.1 × 10−5)	.6058	−7.5 × 10−5 (9.2 × 10−5)	.4153	−6.7 × 10−5 (9.2 × 10−5)	.4689	−7.1 × 10−5 (9.3 × 10−5)	.4486	−6.9 × 10−5 (9.3 × 10−5)	.4572
Weight	–	–	–	–	–	–	–	–	7.7 × 10−4 (3.3 × 10−4)	.0199	7.3 × 10−4 (3.3 × 10−4)	.0261	7.3 × 10−4 (3.3 × 10−4)	.0260	7.4 × 10−4 (3.3 × 10−4)	.0255
History of atrial flutter or atrial fibrillation	–	–	–	–	–	–	–	–	–	–	0.015 (0.011)	.1760	0.014 (0.011)	.2161	0.014 (0.011)	.2088
Coadministration of non‐amiodarone heart rate lowering drugs	–	–	–	–	–	–	–	–	–	–	–	–	0.007 (0.021)	.7413	0.007 (0.021)	.7532
Coadministration of moderate to strong CYP2D6 inhibitors	–	–	–	–	–	–	–	–	–	–	–	–	–	–	0.013 (0.015)	.3819

Note: Model 1: crude model; model 2: model adjusted for CYP2D6 inferred phenotype; model 3: model adjusted for CYP2D6 inferred phenotype, age and sex; model 4: model adjusted for CYP2D6 inferred phenotype, age, sex and metoprolol dose; model 5: model adjusted for CYP2D6 inferred phenotype, age, sex, metoprolol dose and weight; model 6: model adjusted for CYP2D6 inferred phenotype, age, sex, metoprolol dose, weight, and history of atrial flutter or atrial fibrillation; model 7: model adjusted for CYP2D6 inferred phenotype, age, sex, metoprolol dose, weight, history of atrial flutter or atrial fibrillation and non‐amiodarone heart rate lowering drugs; model 8: model adjusted for CYP2D6 inferred phenotype, age, sex, metoprolol dose, weight, history of atrial flutter or atrial fibrillation, non‐amiodarone heart rate lowering drugs and moderate to strong CYP2D6 inhibitors. Intercepts for model 1: 4.145, model 2: 4.098, model 3: 4.220, model 4: 4.225, model 5: 4.138, model 6: 4.144, model 7: 4.145 and model 8: 4.144.

Significant p‐values (<.05) are highlighted in bold.

4. DISCUSSION

The aim of this study was to expand on our previous studies and investigate whether biobanks can be leveraged to identify modest pharmacokinetic interactions of drugs, such as the one between metoprolol and amiodarone. Our results are consistent with previous pharmacokinetic studies which showed that metoprolol's plasma concentrations are increased in patients treated with amiodarone. ²⁴ , ²⁵ Consistent with previous investigations, we found lower heart rates in patients taking both medications concurrently, but the scope of this observation is limited by the cross‐sectional nature of our study.

This current investigation expands on results from prior investigations in which we found that randomly collected blood samples in observational cohorts can be leveraged to identify important determinants of drug pharmacokinetics, such as weight, age, renal function, genotype‐inferred phenotypes, and the use of moderate to strong inhibitors of the agent of interest. ¹⁷ Our results further support that this approach can be used to detect the effect of weak inhibitors. Of note, the daily dosage of the substrate drug of interest is a key covariate to include as part of such investigations. Indeed, associations between metoprolol/α‐OH‐metoprolol concentrations and use of amiodarone were revealed when controlling for metoprolol daily dose.

We selected amiodarone as part of this investigation because both the pharmacokinetics and pharmacodynamics of this interaction with metoprolol are well established. ¹⁵ , ¹⁶ Our future investigations will focus on whether these findings apply to other minor drug interactions and, more importantly, whether the cumulative impact of these demographic, clinical, genetic, and pharmacological factors can be leveraged to further improve drug and dosage selection in a clinical setting.

One unexpected observation is that the use of amiodarone was associated with higher concentrations of α‐OH‐metoprolol. The use of moderate to strong CYP2D6 inhibitors has been previously shown to be a predictor of decreased α‐OH‐metoprolol concentrations, ¹⁷ as we observed in the current study. The increased α‐OH‐metoprolol concentrations in users of amiodarone point toward a modulation of the pharmacokinetics of this metabolite, perhaps unrelated to the inhibition of the CYP2D6 by amiodarone.

Amiodarone has been reported to be involved in a significant number of drug interactions. ²⁶ It exerts a major inhibition of CYP2C8 and CYP3A4 and a more modest inhibition of CYP1A2, CYP2C19, and CYP2D6. ¹⁰ , ¹¹ It has been speculated that the active metabolite DEA of amiodarone is responsible for most of the inhibiting potency of the antiarrhythmic drug. ¹⁰ In addition, amiodarone may interact with other drugs via inhibition of the P‐glycoprotein (P‐gp) membrane transporter system (ABCB1). ²⁷

Deciphering the mechanisms underpinning the association between the use of amiodarone and higher α‐OH‐metoprolol concentrations in humans could prove to be difficult for several reasons. First, several CYPs are contributing to the metabolism of metoprolol into α‐OH‐metoprolol and other metoprolol metabolites such as O‐demethylmetoprolol and N‐deisopropylmetoprolol. ³ Of these CYPs, both CYP3A4 and CYP2D6 are inhibited by amiodarone. ¹⁰ , ¹¹ Considering the multiple pathways involved in metoprolol metabolism, it should be noted that preferential inhibition of the production of metabolites other than α‐OH‐metoprolol could explain the increase in α‐OH‐metoprolol concentrations in amiodarone users. Second, the lack of data regarding the distribution, metabolism, and elimination of α‐OH‐metoprolol metabolite makes it difficult to target a pathway that may be involved in this interaction. A hypothesis has been proposed for which renal function may be one of the routes of elimination of the α‐OH‐metoprolol, ²⁸ yet conflicting data exist and further investigations are needed. Third, both amiodarone and its active metabolite DEA appear to have inhibitory properties, ¹⁰ which further complicates the evaluation of the modest interaction with metoprolol. Fourth, the elimination half‐life of amiodarone varies between 40 and 50 days. ²⁹ , ³⁰ , ³¹ The half‐life of its active metabolite DEA is thought to be even longer. ³¹ Thus, extrapolating results of short‐term pharmacokinetic studies should be done with great caution.

A strength of our approach consists of the representative “real‐life” setting of our population that comprises polymedicated patients presenting multiple morbidities. As observed in our study and in previous reports, this representation is of interest since patients on amiodarone are usually taking a variety of medications to treat a multitude of diseases. ³² , ³³ Another strength of our study is that patients were taking amiodarone on a regular basis which is a substantial advantage considering that steady‐state plasma concentrations of oral amiodarone and DEA take months to reach. ³⁴ To our knowledge, no other study has investigated the pharmacokinetic of the interaction between amiodarone and metoprolol with a cohort as large as ours.

The cross‐sectional nature of our study limits the interpretation of some of our results, such as changes of medications over time or the evaluation of the pharmacodynamics (i.e., heart rate) of the interaction between amiodarone and metoprolol. Nevertheless, we were able to observe consistent associations between the use of amiodarone and metoprolol concentrations, independently of covariables. Another limitation is that we assumed that patients had adequate adherence to amiodarone and metoprolol. The blood sampling taken at random times in regard to the last dosing may also have limited our approach. However, previous research has demonstrated that this constraint does not preclude the observation of consistent association. ¹⁷ , ³⁵

In conclusion, our results suggest that randomly collected blood samples as part of large biobanks could be leveraged to detect the effect of weak CYP inhibitors on concentrations of selected drugs.

AUTHOR CONTRIBUTIONS

Wrote Manuscript: S.R., M.‐O.P., E.O., and S.dD. Designed Research: S.R., M.‐O.P., E.O., M.M., G.L., M.J., I.M., J.‐C.T., M.‐P.D., and S.dD. Performed Research: S.R., M.‐O.P., E.O., M.M., G.L., M.J., M.‐J.G., I.M., D.B., J.‐C.T., M.‐P.D., and S.dD. Analyzed Data: S.R., M.‐O.P., E.O., I.M., M.‐P.D., and S.dD. Contributed New Reagents/Analytical Tools: G.L., M.J., D.B. and M.‐P.D.

CONFLICT OF INTEREST STATEMENT

S. de Denus reports grants outside the submitted work from AstraZeneca, and RMS/Dalcor. M.‐P. Dubé has minor equity interest in Dalcor Pharmaceuticals. M.‐P. Dubé has a patent Methods for Treating or Preventing Cardiovascular Disorders and Lowering Risk of Cardiovascular Events issued to Dalcor Pharmaceuticals, no royalties received; a patent Genetic Markers for Predicting Responsiveness to Therapy with HDL‐Raising or HDL Mimicking Agent issued to Dalcor Pharmaceuticals, no royalties received; and a patent Methods for using low dose colchicine after myocardial infarction, assigned to the Montreal Heart Institute. J.‐C. Tardif has received research grants from Amarin, AstraZeneca, Ceapro, DalCor Pharmaceuticals, Esperion, Ionis, Novartis, Pfizer, RegenXBio, and Sanofi; honoraria from AstraZeneca, DalCor Pharmaceuticals, HLS Pharmaceuticals, Pendopharm, and Pfizer; and minor equity interest in DalCor Pharmaceuticals. All other authors declared no competing interests for this work.

Supporting information

Figure S1.

Table S1.

Table S2. A

Robert S, Pilon M‐O, Oussaïd E, et al. Impact of amiodarone use on metoprolol concentrations, α‐OH‐metoprolol concentrations, metoprolol dosing and heart rate: A cross‐sectional study. Pharmacol Res Perspect. 2023;11:e01137. doi: 10.1002/prp2.1137

DATA AVAILABILITY STATEMENT

Data available on request due to privacy/ethical restrictions: The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.