A cross-sectional study on metoprolol concentrations in various biological samples and their inter-correlations

Abstract

Background

Concentrations of metoprolol in exhaled breath condensate (EBC) have not been investigated. Herein, we aim to determine the metoprolol levels in EBC, plasma, and urine samples.

Methods

Biological samples were collected from 39 patients receiving metoprolol. Metoprolol was determined using liquid chromatography mass spectrometery. The obtained metoprolol levels in biological fluids were investigated for possible inter-correlations.

Results

Acceptable linearity was obtained with coefficient of determinations equal to 0.9998, 0.9941, and 0.9963 for EBC, plasma, and urine samples, respectively. The calibration curves were linear in the ranges of 0.6–500, 0.4–500, and 0.7–10,000 µg·L^− 1 regarding EBC, plasma, and urine samples, respectively. The detection and quantification limits were (0.18, 0.12, and 0.21 µg·L^− 1) and (0.60, 0.40, and 0.70 µg·L^− 1) for EBC, plasma, and urine samples, respectively. The relative standard deviations for the intra- and inter-day replications were obtained between 5.2 and 6.1 and 3.3–4.6%, respectively. The obtained mean metoprolol levels in EBC, plasma, and urine samples of 39 patients were 5.35, 70.76, and 1943.1 µg·L^− 1. There were correlations between daily dose and plasma and urinary concentrations of metoprolol in the investigated samples, whereas no significant correlation was observed for daily dose and EBC levels. The correlation among plasma-urine levels was significant, however, the non-significant correlation was obtained between plasma and EBC concentrations.

Conclusion

Metoprolol levels varied widely due to the metabolic pattern of the Azeri population, different dosages received by the patients, formulation effects, age, sex, and interactions with the co-administered drugs. A poor correlation of EBC-plasma concentrations and a significant correlation of plasma-urine concentrations were observed. Further investigations are required to provide the updated services to personalized medicine departments.

Background

Metoprolol, a selective blocker of β₁-adrenergic receptors, is one of the most widely used drugs in clinical practice [1, 2]. The β₁ receptors are mainly found in the heart and affect cardiac function. Therefore, metoprolol is mainly prescribed to manage cardiovascular disorders, including hypertension, heart failure, angina, cardiac arrhythmias, and myocardial infarction [1, 3, 4]. In addition to therapeutic applications, metoprolol is used as a doping agent in sports to enhance the shooting of amateur sportsmen [5].

Considering the clinical importance and high prevalence of cardiovascular disorders and the widespread use and misuse of metoprolol, measuring the therapeutic level of metoprolol is particularly critical. A therapeutic regimen can be easily managed by determining drug levels in biological fluids. Therefore, therapeutic drug monitoring (TDM) of beta-blockers using liquid chromatography-mass spectrometry, which is a relatively expensive process and requires highly skilled personnel, [6] is performed in some hospitals. Metoprolol distributes very rapidly between the blood and various extravascular sites, and only 1 to 2% of the total amount of the drug in the body is localized in the blood at an apparent distribution equilibrium [7]. The blood/plasma concentration of metoprolol is in the range of 5–80 µg·L^− 1 (mean 33 µg·L^− 1) after the 20 mg dose administration, and 14–212 µg·L^− 1 (mean 111 µg·L^− 1) after the 50 mg dose [8].

Drug levels in the blood or blood-derived fluids such as serum and plasma reflect systemic drug exposure, widely accepted in biomedicine, and are currently the most widely used biological samples in clinical analysis. However, these samples have disadvantages such as invasive sampling, the need for a skilled person for sampling, very high matrix effect, and low compliance of the patients [9]. In addition, sample preparation methods are necessary for the analysis of these samples, because the blood concentration range is below the limit of detection (LOD) of direct analysis in most analytical tools. Furthermore, direct analysis is not feasible due to the matrix interferences in these samples [10].

Therefore, alternative biological fluids, including exhaled breath condensate (EBC) and urine may be considered. The EBC sample consists of water vapor present in the breath and very small liquid droplets of the fluid covering the surface of the lung, which are condensed by a cooling collection device [11]. The main advantages of using the EBC samples include non-invasive sampling, a simple matrix (compared to other biological fluids), the possibility of repeating sampling as often as needed, and the feasibility of direct sample injection to the analytical tools [12]. The EBC can be considered as a possible alternative sample for drug concentration monitoring [13,14,15,16], early diagnosis of diseases by checking the level of appropriate biomarkers, and response to drug treatment [12, 17].

Several analytical techniques for the determination of metoprolol in pharmaceutical and biological samples have been reported in the literature (for details see supplementary information). These techniques are reliable and accurate methods for therapeutic drug monitoring [10]. Until now, high-performance liquid chromatography (HPLC) [18], gas chromatography [19], mass spectrometer detector [20], as well as capillary electrophoresis [21] have been developed to analyze metoprolol in biological samples. The use of these devices requires the implementation of preparation methods in order to remove impurities and disturbing factors and their pre-concentration is essential. Protein precipitation and dilution of the samples were the most convenient approaches for the analysis of biological samples. These methods do not require complicated steps, compounds, or instruments, and they can be easily used prior to sensitive instruments like liquid chromatography-mass spectrometry (LC-MS) and liquid chromatography-mass spectrometery/mass spectrometry (LC-MS/MS). In addition to separation-based methods, spectroscopic methods were also reported for analysis of metoprolol in EBC, plasma, and urine [6] and in EBC [22].

The major aim of this work was the establishment of a cross-sectional study on metoprolol concentrations in different biological samples, including EBC, plasma, and urine, and presetting their inter-correlations. For this purpose, the extracted samples were analyzed by LC-MS/MS.

Methods and materials

Materials and solutions

The analytical standard of metoprolol was purchased from Daru Pakhsh (Tehran, Iran). The utilized trichloroacetic acid and methanol were bought from Merck (Darmstadt, Germany). The deionized water was prepared by a Milli-Q–Q water system (Millipore, Billerica, MA, USA). A stock solution of metoprolol was prepared in methanol at a concentration of 100 mg·L^–1 and dilution of this solution with deionized water was used for the preparation of working solutions.

Instrumentation

A high-performance liquid chromatography (2695, Waters, Alliance, Milford, MA, USA) coupled to MS/MS (Waters, Quatro micro-mass, triple quadrupole) was applied for the quantification of metoprolol. A Zorbax RR Eclipse C₁₈ column (100 mm × 4.6 mm i.d., and 3.5 μm particle size), adjusted at 30 °C, was utilized for the separation of metoprolol. A mixture of HPLC-grade methanol and formic acid solution (0.1% v/v) at a ratio of 65:35 (v/v) was the mobile phase composition. All experiments were done with this composition without any alteration. Other relevant parameters of the MS/MS were as follows: mobile phase flow rate, 0.6 mL·min^− 1; injection volume, 50 µL; source temperature, 110 °C; desolvation temperature, 350 °C; desolvation gas type, nitrogen (flow rate of 600 L·h^− 1); precursor ion (m/z), 268.1; product ions (m/z), 116.2; cone voltage, 35 V; collision energy, 35 eV; collision gas (flow rate), argon (100 L·h^− 1); and nebulization gas type (flow rate), nitrogen (600 L·h^− 1).

Patients’ samples

In this research, 30 female and 9 male patients were included with mean (± standard deviation) age, weight, and height of 59.08 ± 9.31 years (range 36–76 years), 74.67 ± 14.05 kg (range 50–110 kg), and 157.00 ± 11.13 cm (range 140–189 cm) at admission, respectively. The mean daily metoprolol dose was 82.7 ± 29.9 mg (range 47.5–190 mg). The established method was used for the analysis of metoprolol in plasma, urine, and EBC samples obtained from the patients who consumed the drug. All steps and goals of the project were completely explained to the participating persons and they signed a consent form approved by the Ethics Committee of Tabriz University of Medical Sciences (approval code of IR.TBZMED.VCR.REC.1401.138). Patients EBC, plasma, and urine samples were collected at the same time at different time intervals from admistration of the last dose (varying from 4 to 20 h). The EBC samples were collected after washing the mouth with double distilled water and wearing a nose clip and breathing into the set-up for 10–15 min. Urine samples were collected as spot samples. Analyte-free EBC sample was collected by our laboratory made set-up [23] from healthy volunteers. The analyte-free plasma was kindly provided by the Iranian Blood Transfusion Organization (Tabriz, Iran). Blank urine samples were collected from two volunteers who did not consume the drug.

Sample preparation

The studied biological samples were extracted according to the following steps:

• EBC samples: All of the EBC samples were analyzed directly without further pretreatments.

• Plasma samples: A 0.4 mL of blank plasma spiked with metoprolol at a concentration of 50 ng·mL^− 1 or real plasma sample was mixed with 0.225 mL methanol, and 0.2 mL trichloroacetic acid solution (25% w/v) and the mixture was sonicated for 2 min. After that, it was centrifuged at 13,000 rpm for 10 min and the clear supernatant phase was injected into the determination system.

• Urine samples: A 0.4 mL of blank urine spiked with the analyte at a concentration of 50 ng·mL^− 1 or real sample was transferred into a glass test tube and was mixed with 0.425 mL methanol. Then, the mixture was sonicated for 2 min. The obtained mixture was centrifuged for 10 min at 13,000 rpm and the obtained supernatant phase (nearly 650 µL) was used in the determination step.

Results and discussion

Method validation and matrix effect studies

To ensure the method’s validity and accuracy, several figures of merit, including LOD, lower limit of quantification (LLOQ), linear range (LR), coefficient of determination (R²), extraction recovery, accuracy, and repeatability were examined based on the ICH protocol. The LODs (signal to noise (S/N) = 3), LLOQs (S/N = 5), and limit of quantifications (LOQs) (S/N = 10) were (0.18, 0.12, and 0.21 µg·L^− 1), (0.30, 0.20, and 0.35 µg·L^− 1) and (0.60, 0.40, and 0.70 µg·L^− 1) for EBC, plasma and urine samples, respectively (Table 1). Calibration curves of all samples were plotted by performing the method on spiked blank EBC, plasma, and urine samples at various concentrations and plotting the analytes peak area versus concentration. The obtained results showed that the calibration curves were linear in the ranges of 0.6–500, 0.4–500, and 0.7–10,000 µg·L^− 1 regarding EBC, plasma, and urine samples, respectively (Table 2). The relative standard deviations (RSDs), calculated by the peak area of the analyte in replicate analyses on the same day or various days, were obtained between 3.3 and 4.6% and 5.2–6.1, respectively (Table 3). The absolute recoveries of the method were obtained by dividing the amount of extracted analyte by its initial concentration and were 93, 90, and 92 for EBC, plasma, and urine samples, respectively.

In order to study the matrix effect, blank EBC, plasma, and urine samples were obtained from healthy volunteers, and they were spiked with the analytes at two concentrations (5 and 50µg·L^− 1). The mean relative recoveries for these experiments confirmed that the method can determine the analyte accurately and the matrix effect is not significant. The mean relative recovery values were 91–102%, 79–89%, and 86–94%, in EBC, plasma, and urine samples, respectively. To ensure the method’s selectivity, other drugs including bisoprolol, atenolol, daclatasvir, empagliflozin, sofosbuvir, and metformin were added to EBC, plasma, and urine samples and analyzed by the method. The data showed that there were no interfering peaks with the analyte peak. Furthermore, the method stability was checked by analyzing EBC, plasma, and urine samples, containing the analyte at a concentration of 250 µg·L^− 1, kept at room temperature and in the freezer for 24 h. Also, method stability was checked on the samples that were placed under freezing-thawing cycles (n = 3). The RSD values in all cases were ≤ 10% for the analyte compared to fresh samples.

Table 1 LOD, LOQ, and LLOQ values of metoprolol obtained by the developed method

Table 2 Calibration curves parameters

Table 3 Repeatability evaluation

Patient’s sample analysis

The mean EBC, plasma and urine levels of metoprolol were reported respectively as 5.35 µg·L^− 1 (range 1.9–20.9 µg·L^− 1), 70.76 µg·L^− 1 (range 1.1–253.2 µg·L^− 1), 1943.1 µg·L^− 1 (range 3.6–9050.7 µg·L^− 1). The mean plasma concentration of metoprolol lay within the therapeutic range (i.e. 5–80 µg·L^− 1) [8]. All patients received other medications while using metoprolol listed along with other data in Table 4. Typical LC-MS/MS chromatograms of EBC, plasma, and urine samples are shown in Fig. 1. In case of urine samples of all investigated patients, a big peak appeared just after metoprolol peak which apparently belong to the metabolite of the drug, i.e. α-hydroxy metoprolol. Metoprolol concentrations in biological samples of patients are also listed in Table 4. The individual plasma concentrations of 39 patients versus the daily dose of metoprolol are quite varied. As an example, it varied from 1.15 to 253.21 µg·L^− 1 after oral administration of 95 mg of metoprolol per day. Wide variation in metoprolol plasma levels (from < 0.4 to 253.21 µg·L^− 1) is also reported by others [24].

Fig. 1

Typical LC-MS/MS analysis of metoprolol in (A) EBC, (B) plasma, and (C) urine after injection of 50 µL of the samples obtained from patients by the presented method

Table 4 Details of patients under metoprolol maintenance treatment and metoprolol concentrations in biological samples

To explain the wide variations of metoprolol levels in the investigated samples, possible affecting factors were studied in more details. In men, absorption of metoprolol is rapid and complete, however, 40–50% of the orally administered metoprolol reaches the systemic blood circulation [25]. There are two types of formulations of metoprolol; the immediate-release type, metoprolol tartrate, and the extended-release type, metoprolol succinate [1]. The bioavailability of oral administration of metoprolol is about 50% for the tartrate form and 40% for the succinate form. The drug’s bioavailability increases with dose (but not in a linear fashion). Moreover, eating food increases the absorption of quick-release tablets of metoprolol with tartrate salt while metoprolol with succinate salt should be consumed without food considerations [26]. The administration of metoprolol succinate extended-release tablet is marked by lower peaks and prolonged time to reach the peak concentration [4]. Therefore, the variation between administrated dose and plasma concentrations of metoprolol may be linked to the two different salts of metoprolol (tablets with 50 mg refer to tartrate salt, while 47.5 and 95 mg are succinate salt in Table 4). In addition, some patients may not have followed the correct method of taking the drug with or without food. There is also wide variations from the last dose administration of metoprolol to the sampling time as has been reported in Table 4. Due to the short half-life of metoprolol (3–4 h) [4, 27], this time difference could be another source of variations in plasma levels and consequently in EBC and urine levels.

The reported volume of distribution of metoprolol is 4.2 L·kg^− 1. It is not highly bound to plasma proteins; only about 11% of the administered dose is bound, mainly to albumin [4, 27]. Metoprolol demonstrates the stereoselective metabolism that is reliant on oxidation phenotype and patients could be classified as poor or extensive metabolizers [28]. Patients’ metabolizing behavior and gender could alter the plasma concentrations too [4]. In our study, 57% of the patients showed a plasma concentration of < 60 µg·L^− 1 which is in agreement with the pharmacogenetic findings from the Azeri population in which 63% of the cases were extensive/ultra-rapid metabolizers [29].

The elimination of metoprolol is mainly renal as metabolites that appear to have no beta-blocking functions, and less than 5% of an oral dose is recovered unmodified in the urine. Urinary concentration of metoprolol in our investigated patients widely varied. Spot sampling in our study could be considered as a source of variation in urinary levels of metoprolol, since the volume of urine depends on water intake of the patients, and the more urine volume is the less urinary concentration of metoprolol. Furthermore, the higher urinary concentrations of metoprolol tartrate compared to plasma concentrations, over a 24-hour duration, which is reported in Table 4 were confirmed in a previous report [30]. The cumulative urinary excretion of total metabolites after oral administration of 50 mg metoprolol tartrate for 48 to 72 h varied between 29 and 89% in patients with renal failure while it was reported to be about 95% in healthy volunteers [31]. We have calculated the urinary concentration of metoprolol in a range of 0.0007–10.14% of the administered dose. These results are consistent with our reported data which indicated the mean excretion of 3.0% for metoprolol for patients with 36–76 years old (Table 4).

Some drugs may cause interactions and lead to fluctuations in the concentrations of metoprolol, so it is necessary to consider co-administrated medications. Patients reported their other medications which are listed in Table 4 and their interactions with metoprolol are listed in Table S1. It should be noted that since patients reported this list, some drugs may not have been declared. As represented in Table S1, sertraline, citalopram, escitalopram, and fluoxetine could surge the plasma concentration of metoprolol while spironolactone declines it.

Results for correlations

The scattered graph was obtained for metoprolol concentrations in EBC against its daily dose as shown in Fig. 2. A very weak/no association was calculated for EBC concentrations of metoprolol and the daily dose (the correlation coefficient (R)=-0.03, N = 23, p = 0.904) and its weight-adjusted daily dose $\left( {\frac{{daily\;dose\;(mg)}}{{weight\;(kg)}}} \right)$ (R=-0.05, N = 23, p = 0.815). Some probable reasons could be considered for these wider variations in EBC data; including the differences in the humidity of the EBC sample collection area, variations in the distribution of metoprolol into lung lining fluid, variations in the content of exhaled water vapor, smoking habits by some patients, and the individual variations which also apply to plasma and urine samples. In addition, other factors could play a role in the rate and extent to which metoprolol is excreted in the urine and EBC, including the pH of the blood, urine, and EBC, and the solubility of metoprolol in these matrices. While plasma pH stays consistently at 7.4, urine and/or EBC pH can fluctuate. The variation in metoprolol concentrations in EBC may partly be attributable to the fact that EBC is highly diluted with water vapor condensed in the collection setup. Despite the fact that the exact degree of EBC dilution is still unknown, the attenuation of analytes in EBC was estimated to reach up to 12,000 [32].

Fig. 2

Metoprolol concentrations in EBC versus administered daily dose

Better correlations were observed between total plasma concentrations with daily dose (R = 0.579, p < 0.0005), and the weight-adjusted daily dose (R = 0.600, p < 0.0005) of metoprolol in the studied patients. A scatterplot of daily dose versus plasma concentrations of metoprolol is shown in Fig. 3. The computed R-value (in this work) for previously reported data for the steady-state metoprolol plasma concentration and the adjusted daily dose was (R = 0.445, N = 16, p = 0.084) [33].

Fig. 3

Metoprolol concentrations in plasma versus administered daily dose

Figure 4 illustrates the urinary concentration of metoprolol against daily dose. The R for metoprolol concentrations in the urine samples with daily dose was 0.325 (p = 0.07), while for weight-adjusted dose, it was increased to 0.365 (p = 0.04).

Fig. 4

Metoprolol concentrations in urine versus administered daily dose

The present study was the first report to simultaneously determination of metoprolol concentrations in EBC, plasma, and urine samples of the same patients. However, some studies have been conducted on other drugs. For instance, Heiderich et al. concluded a significant correlation between propofol concentrations in EBC and calculated plasma concentrations using the Kataria model and the Narcotrend index in pediatric patients during induction and maintenance anesthesia [34]. In addition, other studies found significant correlations between EBC and plasma propofol concentrations in adult patients [35,36,37]. Hüppe et al. concluded that the EBC concentration of propofol in single-lung ventilation does not match the expected value [38]. Arvidsson et al. studied the EBC and plasma concentrations of methylphenidate and concluded that the correlation between plasma concentrations and EBC for different enantiomers of the drug varied in different patients [39]. Brinkman et al. concluded a significant correlation between volatile organic compounds in EBC and urinary levels of salbutamol and oral corticosteroids [40]. In another study, Khoubnasabjafari et al. concluded that there was no significant correlation between the EBC, plasma, and urine concentrations of methadone [10].

Calissendorff reported metoprolol plasma concentration of 15 patients in the range of 23 to 340 (mean of 172.73) nmol·L^− 1 or 6.15 to 90.90 (mean of 46.18) µg·L^− 1 with the corresponding concentrations of aqueous humour with the minimum, maximum and mean values of 48 nmol·L^− 1 (12.83 µg·L^− 1), 570 nmol·L^− 1 (152.40 µg·L^− 1) and 201.53 nmol·L^− 1 (53.88 µg·L^− 1), respectively [41]. The Pearson correlation coefficient of plasma and aqueous humour levels was calculated (in this work) as R = 0.254 (N = 15, p = 0.362) which was not a significant correlation. In the following work published by the same group, slightly different results were reported. Metoprolol plasma concentrations of 28 patients were in the range of 0.80 to 275.38 (mean of 52.57) µg·L^− 1 with the corresponding concentrations of aqueous humour with the minimum, maximum, and mean values of 3.74, 144.38, and 32.72 µg·L^− 1, respectively [42]. The Pearson correlation coefficient of plasma and aqueous humour levels was calculated (in this work) as R = 0.695 (N = 28, p < 0.0005) which was a significant correlation.

EBC as an emerging biological sample has attracted more attentions in recent years, and the analytical methods have been reported for quantification of very low concentration of the analytes in EBC due to the dilution with water condensed during low temperature of the EBC trap system. As evidenced from the literature [43,44,45], it could contain almost all of analytes exist in lung lining fluid [11] and so far EBC levels of some drugs have been reported (9–10, 12, 13, 14, 15, 16, 17, 45).

Conclusion

Metoprolol was determined in EBC, plasma, and urine samples by high-performance liquid chromatography-tandem mass spectrometry. Using this method as a high-degree-of-confidence approach could lead to reliable, and reproducible results as confirmed in this report. Based on the results, LR and LOD had similar ranges for measured biological samples.

Metoprolol levels in the investigated biological samples were widely varied from patient to patient in which the observed variations were confirmed by the literature. Pharmacogenetic findings of a metabolizing pattern of the Azeri population, different dosages received by the patients, formulation effects, age, sex, and interactions with the co-administered drugs could be considered as the sources of such variations.

This is the first study reporting metoprolol levels in EBC, plasma, and urine samples of the patients. Despite of poor correlation of EBC-plasma concentrations, due to the advantages of EBC samples, further investigations using more patients’ samples are recommended, especially for chiral separation of metoprolol. A significant correlation was observed for plasma-urine concentrations.

Therapeutic monitoring of metoprolol is required in clinical practice for better management of the patients. In this work, we used LC-MS/MS analysis for metoprolol determination in biological samples which is an expensive method requiring highly skilled personnel. The development of a low-cost, easy to use, and portable analytical tool is highly in demand to provide updated services to personalized medicine departments.