Population Pharmacokinetic Model of Amiodarone and N-Desethylamiodarone Focusing on Glucocorticoid and Inflammation

Abstract

Some population pharmacokinetic models for amiodarone (AMD) did not incorporate N-desethylamiodarone (DEA) concentration. Glucocorticoids activate CYP3A4 activity, metabolizing AMD. In contrast, CYP3A4 activity may decrease under inflammation conditions. However, direct evidence for the role of glucocorticoid or inflammation on the pharmacokinetics of AMD and DEA is lacking. The pilot study aimed to address this gap using a population pharmacokinetic analysis of AMD and DEA. A retrospective cohort observational study in adult patients who underwent AMD treatment with trough concentration measurement was conducted at Tokyo Women’s Medical University, Medical Center East from June 2015 to March 2019. Both structural models of AMD and DEA applied 1-compartment models, which included significant covariates using a stepwise forward selection and backward elimination method. The eligible 81 patients (C-reactive protein level: 0.26 [interquartile range; 0.09–1.92] mg/dL) had a total of 408 trough concentrations for both AMD and DEA. The median trough concentrations were 0.49 [0.31–0.81] µg/mL for AMD and 0.43 [0.28–0.71] µg/mL for DEA during a median follow-up period of 446 [147–1059] d. Three patients received low-dose oral glucocorticoid. The final model identified that AMD clearance was 7.9 L/h, and the apparent DEA clearance was 10.3 L/h. Co-administered glucocorticoids lowered apparent DEA clearance by 35%. These results indicate that co-administered glucocorticoids may increase DEA concentrations in patients without severe inflammation.

INTRODUCTION

Amiodarone (AMD) and N-desethylamiodarone (DEA), an active metabolite of AMD, block the potassium channels responsible for cardiac repolarization, thereby suppressing atrial fibrillation and ventricular tachycardia.^1,2) AMD is effective to control arrhythmia (e.g., atrial fibrillation and ventricular/ventricular fibrillation).^3,4) The pharmacokinetics of AMD depends on age and body size.^5–7) A trough concentration of AMD >0.5 µg/mL is required to obtain an antiarrhythmic effect.⁸⁾ Clinicians recently have prescribed low-dose AMD targeting lower concentrations.⁹⁾ AMD frequently causes adverse events in patients with high AMD trough concentrations.¹⁰⁾ For instance, DEA concentration >0.6 µg/mL is a risk factor for interstitial pneumonia.¹¹⁾

Previous population pharmacokinetic analyses focused on covariates for AMD, although these have not examined covariates of DEA that cause lung damage.^5–7,12) Because AMD is metabolized primarily via CYP3A4, glucocorticoids induce CYP3A4 producing DEA from AMD.^13,14) However, drug-drug interaction between AMD (or DEA) and glucocorticoids remains unclear. Furthermore, there is a concern that inflammation inhibits CYP3A4 activity and increases exposure to AMD as with midazolam and perampanel, which positively correlates with C-reactive protein (CRP) levels.^15,16)

Accordingly, this pilot study aimed to conduct a comprehensive assessment of glucocorticoid and inflammation using a population pharmacokinetic analysis of AMD and DEA.

PATIENTS AND METHODS

Study Design

A retrospective cohort observational study was conducted at Tokyo Women’s Medical University, Medical Center East (450 bed, branch hospital of academic medical center) from June 2015 to March 2019. We screened study patients who underwent AMD treatment orally or intravenously with monitored blood concentrations of AMD and DEA. Patients were excluded if they were aged <20 years or had not undergone CRP measurement. The study protocol was approved by the institutional review board of Tokyo Women’s Medical University and followed the Declaration of Helsinki (Approval No. 5151).

Data Collection

We extracted clinical data using electronic medical records from the commencement of AMD treatment to March 2019 or the discontinuation of AMD. All blood samples for trough concentrations were taken before the next administration of AMD. AMD and DEA trough concentrations were quantified by performing LC coupled with ion trap mass spectrometry. The limit of quantification was 20.0 ng/mL for AMD and 50.0 ng/mL for DEA.

Population Pharmacokinetic Modeling

Pharmacokinetic models for AMD and DEA were constructed using non-linear mixed effect analysis with Phoenix^® NLME Version 8.1 software (Certara USA, Inc., Princeton, NJ, U.S.A.). The Laplace approximation computational algorithm was used. A model including both AMD and DEA was simultaneously used and sequentially fitted to the data, that is, AMD parameters were first estimated, then, fixing those parameters estimated, DEA parameters were estimated. Since only AMD and DEA trough concentrations were available, we could not construct a multi-compartment model and estimate the absorption rate constant (k_a) of AMD. Therefore, a pharmacokinetic model for AMD was constructed as a 1-compartment model with a first-order absorption process. Oral bioavailability (F_oral) was fixed as 54%.¹⁷⁾ As k_a data was not available, we arbitrarily considered k_a as 1.0 h⁻¹. The structural model was constructed considering amiodaron clearance (CL_AMD) and volume distribution (Vd_AMD) as the pharmacokinetic parameters; Vd_AMD was fixed as 14.0 L/kg because we could not obtain peak concentration data.¹⁷⁾ Similarly, the structural model for DEA was constructed as a 1-compartment model with clearance (CL_DEA) and volume distribution (Vd_DEA) in addition to a metabolic fraction of AMD to DEA (f_m). Since we could not obtain detailed DEA pharmacokinetics data, we assumed that Vd_DEA was compatible with Vd_AMD and fixed Vd_DEA as 14.0 L/kg following AMD data. We tested an additive or a proportional error model to describe residual variability, defined as follows:

  

where C_obs and C_pred denote observed and predicted trough concentration (AMD or DEA), and ε denotes measurement error including intraindividual variability, analytical error, and dosing error.

Interindividual variability of AMD and DEA clearance was described using an exponential random-effects model defined as follows:

  

where CL_AMD and CL_DEA denote the parameter for AMD and DEA clearance, tv CL_AMD, and tv CL_DEA are the typical values for CL_AMD and CL_DEA, respectively, and each η represents interindividual variability. We explored covariates for CL_AMD and CL_DEA. The continuous covariate was modelled as follows:

  

where P_tv denotes the typical value of the parameter, θ_tv denotes the parameter for the covariate equal median value, and θ denotes a covariate scale factor for the covariate. In addition, the categorical covariate was modelled as follows:

  

where n is an indicator of the covariate (0 or 1).

The candidate covariates comprised sex, age, body mass index, cardiac comorbidity, CRP, and glucocorticoids. Time-varying CRP values were used in the covariate modeling. The last observation carried forward was used to impute missing values for clinical laboratory data. We checked the presence of multicollinearity of covariates. The stepwise forward selection and backward elimination method determined covariates according to the objective function value (OFV). The significance level for forward selection and backward elimination was set at p < 0.05 (ΔOFV 3.84) and p < 0.01 (ΔOFV 6.63).

Pharmacokinetic Model Validation

Goodness-of-fit plots showed the correlations of observed vs. predicted concentrations, observed vs. individual predicted concentrations, conditional weighted residuals vs. predicted concentrations, and conditional weighted residuals vs. time after dose. We visually compared observed with predicted values based on prediction corrected visual predictive check (pc-VPC) using 1000 hypothetical data sets. Using operating bootstrap methods, we obtained mean estimates with 95% confidence intervals (95% CI) of respective parameters from the generated 1000 resampling datasets. The final model was considered reliable if the parameter estimates were within the 95% CI. The results of bootstrap methods were compared with those obtained from the final model.

Pharmacokinetic Model Application

We performed simulations to confirm an oral dosing schedule for AMD treatment. The initial oral dose was fixed at 200 mg twice daily for a week. Maintenance oral dose was assigned as follows: 1) 50 mg once daily; 2) 50 mg twice daily; 3) 100 mg once daily; 4) 100 mg twice daily. Body weight was fixed as 50, 70, and 90 kg. The results were summarized according to glucocorticoid use. We simulated steady-state trough concentrations at day 150 after AMD treatment.

Outcome Analysis

We performed an exploratory analysis examining the association between trough concentration and the typical outcome of overt hypothyroidism associated with AMD using univariate logistic regression analysis. Moreover, the ratio of DEA to AMD concentration and the sum were investigated similarly. We defined hypothyroidism as a thyroid-stimulating hormone level of >10 mU/L.¹⁸⁾

Statistical Analysis

Statistical analysis was conducted using JMP^® pro 14 (SAS Institute Inc., Cary, NC, U.S.A.). Normally distributed continuous data were expressed as the mean ± standard deviation, whereas non-normally distributed data were expressed as median values and interquartile range (IQR). Categorical data were expressed as numbers (%). When we conducted a logistic regression analysis, a two-sided p-value <0.05 was considered statistically significant.

RESULTS

Patient Characteristics

We enrolled 266 patients (>20 years) who underwent AMD therapy and therapeutic drug monitoring at our hospital from June 2015 to March 2019. There were no patients who were younger than 20 years. We excluded 185 patients who had no CRP data. Therefore, we identified 81 patients in this study. Baseline clinical data are summarized in Table 1. The median initial daily dose of AMD [IQR] was 200 [100–400] mg. There were 10 (12%) patients who orally administered AMD every other day. The median CRP level at the baseline was 0.26 [0.09–1.92] mg/dL, which corresponded to the low range. All three patients received low-dose oral prednisolone. The median trough concentrations were 0.49 [0.31–0.81] µg/mL for AMD and 0.43 [0.28–0.71] µg/mL for DEA during a follow-up period of 446 [147–1059] days.

Table 1. Baseline Patient Clinical Data

Variables	Data, N = 81
Demographical data
Male, N (%)	55 (68)
Age, years	64 ± 15
Height, cm	164 ± 11
Body weight, kg	61 ± 15
Body mass index, kg/m2	23 ± 4
Cardiac comorbidity
Congestive heart failure, N (%)	48 (59)
Atrial fibrillation, N (%)	27 (33)
VT or VF, N (%)	39 (48)
Detail of amiodarone exposure
Initial daily dose, mg	200 [100–400]
Oral routea), N (%)	61 (75)
Intravenous route, N (%)	20 (25)
Clinical laboratory data
CRP, mg/dL	0.26 [0.09–1.92]
Concurrent medications
Glucocorticoidb), N (%)	3 (4)

N, number; VT, ventricular tachycardia; VF, ventricular fibrillation; CRP, C-reactive protein. a) 10 patients received amiodarone every other day. b) All patients received oral prednisolone. Normally distributed continuous data are expressed as mean ± standard deviation, whereas non-normally distributed data are expressed as median values and interquartile range (IQR). Categorical data are expressed as numbers and percentages.

Population Pharmacokinetic Modeling

A total of 408 AMD and 408 DEA measurement points were available for the population pharmacokinetic modeling. The median number of measurement points was 6 [2–14] per patient. There was a total of 26 AMD and 26 DEA measurement points in patients who received prednisolone. The final model parameters are summarized in Table 2. The proportional error model was incorporated in the structured model. The stepwise forward selection method determined that co-administered glucocorticoids lowered apparent CL_DEA by 35% (Table 2). We did not conduct the stepwise backward elimination method since the final models included no covariates for AMD and only glucocorticoid for DEA. The overall η-shrinkage was 21.5 and 23.5% for CL_AMD and CL_DEA, respectively. Additionally, the overall ε-shrinkage was 8.7 and 8.9% for AMD concentration and DEA concentration, respectively.

Table 2. Population Pharmacokinetic Model, Parameters for Amiodarone and N-Desethylamiodarone, and the Bootstrap Method Results

Parameter	Estimated mean	RSE%	Bootstrap mean	95% CI
AMD
Population value
CLAMD, L/h	7.9	10.9	7.7	7.6 to 7.8
VdAMD, L/kg (fixed)	14.0	NA	14.0	NA
Foral, % (fixed)	54.0	NA	54.0	NA
ka, /h (fixed)	1.0	NA	1.0	NA
Interindividual variability
ωCLAMD, %	43.3	23.3	46.9	46.1 to 47.8
Intraindividual variability
σ, %	31.3	10.8	30.9	30.7 to 31.1
DEA
Population value
CLDEA/fm, L/h	10.3	11.0	10.1	10.0 to 10.2
Glucocorticoids on CLDEA/fm	−0.35	42.5	−0.37	−0.38 to −0.36
VdDEA, L/kg (fixed)	14.0	NA	14.0	NA
Interindividual variability
ωCLDEA, %	39.4	21.1	42.6	41.7 to 43.5
Intraindividual variability
σ, %	23.9	11.8	23.3	23.1 to 23.5

RSE%, relative standard error (%); 95% CI, 95% confidence interval; AMD, amiodarone; CL_AMD, amiodaron clearance; Vd_AMD, amiodarone volume distribution; F_oral, oral bioavailability; k_a, absorption constant; DEA; N-desethylamiodarone, CL_DEA, N-desethylamiodarone clearance; f_m, fraction of metabolism; Vd_DEA, N-desethylamiodarone volume distribution; NA, not applicable. The final model equations were as follows. CL_AMD= 7.9 L/h Vd_{_AMD} = 14.0 × Body weight CL_DEA/f_m = 10.3 × (1−0.35 × [if glucocorticoid]) L/h Vd_DEA = 14.0 × Body weight The difference in objective function value of final model was 5.0. The overall η-shrinkage was 21.5 and 23.5% for CL_AMD and CL_DEA, respectively. The overall ε-shrinkage was 8.7 and 8.9% for AMD concentration and DEA concentration, respectively.

Pharmacokinetic Model Validation

Goodness-of-fit plots are shown in Figs. 1 and 2. Although some tendency was observed, the model has shown some correlation and margin of conditional weighted residual of ±3, indicating that the final model had the acceptable capacity to predict AMD and DEA trough concentrations. The results of pc-VPC also confirmed an acceptable agreement between the observed and predicted values although there were outliers as shown in Fig. 3. Bootstrap methods were 100% successful and appropriately estimated the final model parameters with mean values of the bootstrap parameters closely equivalent to those of the model parameter estimates (Table 2). A summary of three patients who received glucocorticoid has shown in Table 3.

Fig. 1. Goodness-of-Fits Plots for Amiodarone

A: Observed concentrations and population predicted concentrations. Y-axis and X-axis represent observed concentrations and population predicted concentrations, respectively. Black circles represent each point, while the solid line represents the reference line. B: Observed concentrations and individual predicted concentrations. Y-axis and X-axis represent observed concentrations and individual predicted concentrations, respectively. Black circles represent each point, while the solid line represents the reference line. C: Conditional weighted residual and time after the first dose. Y-axis and X-axis represent conditional weighted residual and time after the first dose, respectively. The solid horizontal line represents the reference line, while black circles represent each point. D: Conditional weighted residual and predicted concentrations. Y-axis and X-axis represent conditional weighted residual and predicted concentrations, respectively. Black circles represent each point. CWRES; conditional weighted residual.

Fig. 2. Goodness-of-Fits Plots for N-Desethylamiodarone

A: Observed concentrations and population predicted concentrations. Y-axis and X-axis represent observed concentrations and population predicted concentrations, respectively. Black circles represent each point, while the solid line represents the reference line. B: Observed concentrations and individual predicted concentrations. Y-axis and X-axis represent observed concentrations and individual predicted concentrations, respectively. Black circles represent each point, while the solid line represents the reference line. C: Conditional weighted residual and time after the first dose. Y-axis and X-axis represent conditional weighted residual and time after the first dose, respectively. The solid horizontal line represents the reference line, while black circles represent each point. D: Conditional weighted residual and predicted concentrations. Y-axis and X-axis represent conditional weighted residual and predicted concentrations, respectively. Black circles represent each point. CWRES; conditional weighted residual.

Fig. 3. Prediction Corrected Visual Predictive Checks for Amiodarone and N-Desethylamiodarone

A: X-axis and Y-axis represent the dosing period and prediction-corrected observed and predicted values of amiodarone (logarithmic scale). Opened circles represent each point. The solid and dotted lines represent predicted median and 95% confidence interval. The lower limit of predicted concentration is below 0. B: X-axis and Y-axis represent the dosing period and prediction-corrected observed and predicted values of N-desethylamiodarone (logarithmic scale). Opened circles represent each point. The solid and dotted lines represent predicted median and 95% confidence interval. The lower limit of predicted concentration is below 0.

Table 3. Summary of Characteristics in the Three Patients Who Concomitantly Received Glucocorticoid

	Sex	Age year	BW kg	BMI kg/m2	Daily dose mg	Treatment duration day	CRP mg/dL	Mean troughAMD µg/mL	Mean troughDEA µg/mL	Measurement pointsa)
1	M	74	46	19	200	683	0.17	1.08	0.90	34
2	M	88	64	24	200	390	1.43	0.98	1.18	4
3	F	43	45	18	50	1225	0.02	0.16	0.22	14

BW, body weight; BMI, body mass index; CRP, C-reactive protein; trough_AMD, amiodarone trough concentration; trough_DEA, N-desethylamiodarone trough concentration; M, male; F, female. Mean trough_AMD and trough_DEA values were calculated by each value during the study period. a) Measurement points were available for AMD and DEA.

Pharmacokinetic Model Application

The summary of simulations revealed that glucocorticoid co-administration led to approximately 50% higher steady-state DEA trough concentration regardless of body weight and dosing schedule (Table 4).

Table 4. Summary of the Estimated Amiodarone and N-Desethylamiodarone Concentrations According to the Simulation

Daily dose	With glucocorticoid		Without glucocorticoid
	AMD, µg/mL	DEA, µg/mL	AMD, µg/mL	DEA, µg/mL
Body weight, 50 kg
50 mg once	0.08	0.10	0.08	0.07
50 mg twice	0.16	0.20	0.16	0.13
100 mg once	0.15	0.20	0.15	0.13
100 mg twice	0.33	0.41	0.33	0.27
Body weight, 70 kg
50 mg once	0.08	0.10	0.08	0.07
50 mg twice	0.17	0.20	0.17	0.13
100 mg once	0.16	0.20	0.16	0.13
100 mg twice	0.33	0.41	0.33	0.25
Body weight, 90 kg
50 mg once	0.08	0.10	0.08	0.07
50 mg twice	0.17	0.20	0.17	0.13
100 mg once	0.16	0.20	0.16	0.13
100 mg twice	0.34	0.41	0.34	0.27

AMD; amiodarone, DEA; N-desethylamiodarone.

Outcome Analysis

Out of all patients, 13 (16%) experienced overt hypothyroidism associated with AMD. The significant relationship between DEA to AMD ratio and the development of overt hypothyroidism was present in this population (odds ratio = 23.9, 95% CI: 2.19 to 261.2, p < 0.01) (Supplementary Table 1).

DISCUSSION

Although there were only 3 patients concomitantly administered glucocorticoids, we used 26 blood samples of the patients and showed that glucocorticoids lead to the accumulation of DEA concentration in patients without severe CRP elevation.

Several population pharmacokinetic models have identified significant covariates, such as age and body size, in Japanese patients.^5,7) In contrast, no covariates were found in a certain clinical study with a limited sample size.¹⁹⁾ Thus, it is controversial about covariates for AMD. Our final model parameters (e.g., half-lives of about 80 h) were similar to the values reported in a previous study.⁷⁾

Glucocorticoids penetrate cells and bind to nuclear receptors, thereby exerting a 1.5-fold higher CYP3A4 activity than the control.²⁰⁾ Indeed, high dose tacrolimus was required to reach the target trough concentration in patients with low-dose prednisolone.²¹⁾ Thus, we considered that glucocorticoids induce the conversion from AMD to DEA although we could not conclude whether our findings resulted from a decrease in CL_DEA or an increase in f_m. Since Fukuchi et al. suggested that elderly patients had high trough concentrations of AMD⁷⁾ (Table 3), concomitant use of glucocorticoid may have a significant impact on AMD metabolism in elderly patients. Interstitial pneumonia is one of the adverse events associated with AMD.^22,23) The risk of interstitial pneumonia was reported to increase by 1.3 times with the elevation of DEA trough concentration 0.2 µg/mL independent of AMD trough concentration.¹¹⁾ Thus, the effect of glucocorticoids on AMD pharmacokinetics may contribute to the increased risk of interstitial pneumonia. Although we did not collect the data on the onset of interstitial pneumonia, AMD dose adjustment may be an efficient strategy to minimize the risk of interstitial pneumonia in patients who co-administered AMD and glucocorticoids.

The previous report has shown that the inhibition constant of AMD for CYP3A4 is 271.6 µM and DEA for CYP3A4 is 12.1 µM.²⁴⁾ The present study confirmed that the median trough concentrations were 0.49 [0.31–0.81] µg/mL (0.76 µM) for AMD and 0.43 [0.28–0.71] µg/mL (0.66 µM) for N-DEA, respectively. Therefore, AMD and DEA could not exceed the inhibition constant for CYP3A4. Accordingly, we consider that the effect of AMD and DEA on glucocorticoids might be limited.

CRP had little or no effect on AMD pharmacokinetics because no patients had elevated CRP of more than 10.0 mg/dL in our study. A significant correlation between CRP level and medications (e.g., voriconazole and tramadol) has been reported in a wide range of CRP levels.^25–27) Notably, AMD and DEA have long half-lives with high interindividual variability in long-term treatment, which supports our findings (half-lives: about 80 h).²⁸⁾ Conversely, the half-life of CRP was relatively short (about 6 to 12 h).²⁹⁾ Therefore, the transient and mild inflammatory reaction has limited effects on AMD treatment.

Preliminarily, our finding indicated a significant relationship between DEA to AMD ratio and the onset of overt hypothyroidism, which has been a controversial issue.^30–32) The fundamental experimental data suggest that DEA plays an essential role in exerting cytotoxicity.³³⁾ On the contrary, AMD tends to accumulate in thyroid tissue.³⁴⁾ These controversial data should be resolved by additional research.

This study has several limitations. First, this study was a single-center retrospective analysis, thus, unknown confounding factors were inevitable. Second, the sample size was small; therefore, we could not assess the dose–response of glucocorticoids. Additionally, no patients developed hyperthyroidism (thyroid-stimulating hormone <0.35 mU/L).¹⁸⁾ Third, we could not investigate the role of di-N-desethylamiodarone, a metabolite of DEA, exerts hepatotoxicity.³⁵⁾ Finally, inflammation status was only evaluated according to the CRP levels because there was no data on the other inflammatory marker (e.g., interleukin 6).

CONCLUSION

This pilot study suggested that glucocorticoid increases DEA exposure in the absence of severe inflammation. Our population pharmacokinetic model might represent a step toward adequate dose adjustment of AMD in clinical practice. Further study is required to examine the effect of glucocorticoid on AMD pharmacokinetics.

Conflict of Interest

The authors declare no conflict of interest.

Supplementary Materials

This article contains supplementary materials.

REFERENCES

• 1) Vassallo P, Trohman RG. Prescribing amiodarone: an evidence-based review of clinical indications. JAMA, 298, 1312–1322 (2007).
• 2) Nattel S, Davies M, Quantz M. The antiarrhythmic efficacy of amiodarone and desethylamiodarone, alone and in combination, in dogs with acute myocardial infarction. Circulation, 77, 200–208 (1988).
• 3) Singh BN, Singh SN, Reda DJ, Tang XC, Lopez B, Harris CL, Fletcher RD, Sharma SC, Atwood JE, Jacobson AK, Lewis HD Jr, Raisch DW, Ezekowitz MD. Amiodarone versus sotalol for atrial fibrillation. N. Engl. J. Med., 352, 1861–1872 (2005).
• 4) Connolly SJ, Dorian P, Roberts RS, Gent M, Bailin S, Fain ES, Thorpe K, Champagne J, Talajic M, Coutu B, Gronefeld GC, Hohnloser SH. Comparison of beta-blockers, amiodarone plus beta-blockers, or sotalol for prevention of shocks from implantable cardioverter defibrillators: the OPTIC study: a randomized trial. JAMA, 295, 165–171 (2006).
• 5) Araki R, Yukawa E, Nakashima MN, Fukuchi H, Sasaki H, Yano K, Nakashima M. Population pharmacokinetic investigation for optimization of amiodarone therapy in Japanese patients. Ther. Drug Monit., 33, 750–756 (2011).
• 6) Vadiei K, Troy S, Korth-Bradley J, Chiang ST, Zimmerman JJ. Population pharmacokinetics of intravenous amiodarone and comparison with two-stage pharmacokinetic analysis. J. Clin. Pharmacol., 37, 610–617 (1997).
• 7) Fukuchi H, Nakashima M, Araki R, Komiya N, Hayano M, Yano K, Sasaki H, Yukawa E. Effect of obesity on serum amiodarone concentration in Japanese patients: population pharmacokinetic investigation by multiple trough screen analysis. J. Clin. Pharm. Ther., 34, 329–336 (2009).
• 8) Campbell TJ, Williams KM. Therapeutic drug monitoring: antiarrhythmic drugs. Br. J. Clin. Pharmacol., 52 (Suppl. 1), 21S–34S (2001).
• 9) Aonuma K, Shiga T, Atarashi H, et al. Guidelines for therapeutic drug monitoring of cardiovascular drugs clinical use of blood drug concentration monitoring (JCS 2015)—Digest Version. Circ. J., 81, 581–612 (2017).
• 10) Haffajee CI, Love JC, Canada AT, Lesko LJ, Asdourian G, Alpert JS. Clinical pharmacokinetics and efficacy of amiodarone for refractory tachyarrhythmias. Circulation, 67, 1347–1355 (1983).
• 11) Yamada Y, Shiga T, Matsuda N, Hagiwara N, Kasanuki H. Incidence and predictors of pulmonary toxicity in Japanese patients receiving low-dose amiodarone. Circ. J., 71, 1610–1616 (2007).
• 12) Martin WJ 2nd, Rosenow EC 3rd. Amiodarone pulmonary toxicity. Recognition and pathogenesis (Part 2). Chest, 93, 1242–1248 (1988).
• 13) Matsunaga T, Maruyama M, Matsubara T, Nagata K, Yamazoe Y, Ohmori S. Mechanisms of CYP3A induction by glucocorticoids in human fetal liver cells. Drug Metab. Pharmacokinet., 27, 653–657 (2012).
• 14) Barnes PJ. Corticosteroid effects on cell signalling. Eur. Respir. J., 27, 413–426 (2006).
• 15) Vet NJ, Brussee JM, de Hoog M, Mooij MG, Verlaat CW, Jerchel IS, van Schaik RH, Koch BC, Tibboel D, Knibbe CA, de Wildt SN. Inflammation and organ failure severely affect midazolam clearance in critically ill children. Am. J. Respir. Crit. Care Med., 194, 58–66 (2016).
• 16) Yamamoto Y, Takahashi Y, Horino A, Usui N, Nishida T, Imai K, Kagawa Y, Inoue Y. Influence of inflammation on the pharmacokinetics of perampanel. Ther. Drug Monit., 40, 725–729 (2018).
• 17) Riva E, Gerna M, Latini R, Giani P, Volpi A, Maggioni A. Pharmacokinetics of amiodarone in man. J. Cardiovasc. Pharmacol., 4, 264–269 (1982).
• 18) Batcher EL, Tang XC, Singh BN, Singh SN, Reda DJ, Hershman JM. Thyroid function abnormalities during amiodarone therapy for persistent atrial fibrillation. Am. J. Med., 120, 880–885 (2007).
• 19) Pollak PT, Bouillon T, Shafer SL. Population pharmacokinetics of long-term oral amiodarone therapy. Clin. Pharmacol. Ther., 67, 642–652 (2000).
• 20) Matoulková P, Pávek P, Malý J, Vlček J. Cytochrome P450 enzyme regulation by glucocorticoids and consequences in terms of drug interaction. Expert Opin. Drug Metab. Toxicol., 10, 425–435 (2014).
• 21) Anglicheau D, Flamant M, Schlageter MH, Martinez F, Cassinat B, Beaune P, Legendre C, Thervet E. Pharmacokinetic interaction between corticosteroids and tacrolimus after renal transplantation. Nephrol. Dial. Transplant., 18, 2409–2414 (2003).
• 22) Papiris SA, Triantafillidou C, Kolilekas L, Markoulaki D, Manali ED. Amiodarone: review of pulmonary effects and toxicity. Drug Saf., 33, 539–558 (2010).
• 23) Pollak PT. Clinical organ toxicity of antiarrhythmic compounds: ocular and pulmonary manifestations. Am. J. Cardiol., 84 (9A), 37R–45R (1999).
• 24) Ohyama K, Nakajima M, Suzuki M, Shimada N, Yamazaki H, Yokoi T. Inhibitory effects of amiodarone and its N-deethylated metabolite on human cytochrome P450 activities: prediction of in vivo drug interactions. Br. J. Clin. Pharmacol., 49, 244–253 (2000).
• 25) Vreugdenhil B, van der Velden W, Feuth T, Kox M, Pickkers P, van de Veerdonk FL, Blijlevens NMA, Brüggemann RJM. Moderate correlation between systemic IL-6 responses and CRP with trough concentrations of voriconazole. Br. J. Clin. Pharmacol., 84, 1980–1988 (2018).
• 26) Tanaka H, Naito T, Sato H, Hiraide T, Yamada Y, Kawakami J. Impact of CYP genotype and inflammatory markers on the plasma concentrations of tramadol and its demethylated metabolites and drug tolerability in cancer patients. Eur. J. Clin. Pharmacol., 74, 1461–1469 (2018).
• 27) Theuretzbacher U, Ihle F, Derendorf H. Pharmacokinetic/pharmacodynamic profile of voriconazole. Clin. Pharmacokinet., 45, 649–663 (2006).
• 28) Latini R, Tognoni G, Kates RE. Clinical pharmacokinetics of amiodarone. Clin. Pharmacokinet., 9, 136–156 (1984).
• 29) Stanke-Labesque F, Gautier-Veyret E, Chhun S, Guilhaumou R. Inflammation is a major regulator of drug metabolizing enzymes and transporters: Consequences for the personalization of drug treatment. Pharmacol. Ther., 215, 107627 (2020).
• 30) Takeuchi D, Honda K, Shinohara T, Inai K, Toyohara K, Nakanishi T. Incidence, clinical course, and risk factors of amiodarone-induced thyroid dysfunction in Japanese adults with congenital heart disease. Circ. J., 79, 1828–1834 (2015).
• 31) Yamato M, Wada K, Fujimoto M, Hosomi K, Hayashi T, Oita A, Takada M. Association between N-desethylamiodarone/amiodarone ratio and amiodarone-induced thyroid dysfunction. Eur. J. Clin. Pharmacol., 73, 289–296 (2017).
• 32) Lafuente-Lafuente C, Alvarez JC, Leenhardt A, Mouly S, Extramiana F, Caulin C, Funck-Brentano C, Bergmann JF. Amiodarone concentrations in plasma and fat tissue during chronic treatment and related toxicity. Br. J. Clin. Pharmacol., 67, 511–519 (2009).
• 33) Beddows SA, Page SR, Taylor AH, McNerney R, Whitley GS, Johnstone AP, Nussey SS. Cytotoxic effects of amiodarone and desethylamiodarone on human thyrocytes. Biochem. Pharmacol., 38, 4397–4403 (1989).
• 34) Holt DW, Tucker GT, Jackson PR, Storey GC. Amiodarone pharmacokinetics. Am. Heart J., 106, 840–847 (1983).
• 35) Zahno A, Brecht K, Morand R, Maseneni S, Török M, Lindinger PW, Krähenbühl S. The role of CYP3A4 in amiodarone-associated toxicity on HepG2 cells. Biochem. Pharmacol., 81, 432–441 (2011).