Simultaneous LC-ESI-MS/MS Quantification of Levosimendan and Its Metabolites for Therapeutic Drug Monitoring of Cardiac Surgery Patients

Levosimendan is used in severe chronic cardiac insufficiency, also within the peri-operative setting. Real-life pharmacokinetic data in surgical patients is lacking, making therapeutic drug monitoring (TDM) of levosimendan, its pharmacologically active metabolite OR-1896, and its intermediate OR-1855 important. A simultaneous highly sensitive quantification of levosimendan and its metabolites in small-volume samples has not yet been described. Here, levosimendan (LLOQ 0.450 nM), OR-1896, and OR-1855 (LLOQ both 1.0 nM) were successfully quantified by LC-ESI-MS/MS after liquid-liquid extraction in 300 µL of blood. A short C8 column under reversed-phase conditions enabled simultaneous and fast quantification of levosimendan in the negative and the metabolites in the positive ionization mode in a single run within 2 min. Interestingly and unexpectedly, constitutional isomers of levosimendan metabolites with identical mass transitions and similar retention times were observed in surgical patients’ samples, which we identified as the metamizole metabolites 4-aminoantipyrine and 4-acetamidoantipyrine. A longer C8 column and a modified mobile phase enabled selective quantification of all analytes in a single run within 7 min. We developed, validated, and applied highly sensitive LC-ESI-MS/MS methods for simultaneous quantification of levosimendan and its metabolites, enabling efficient TDM of cardiac surgery patients even with additional metamizole administration.


Introduction
As a calcium sensitizer, levosimendan (Simdax ® ) has positive inotropic effects and is therefore used for therapy of severe chronic cardiac insufficiency [1]. In total, 4-7% of the administered levosimendan is metabolized by intestinal bacteria to the metabolite OR-1855, which is acetylated by hepatic N-acetyltransferase (NAT2) to the pharmacologically active OR-1896 ( Figure 1) [2]. Whereas levosimendan has a short half-life of one hour, the half-life of the metabolites is 70-80 h. Thus, it is hypothesized that the long-term effects, observed after termination of levosimendan treatment, arise from OR-1855 and OR-1896 [3,4]. Levosimendan is also used in the peri-operative management in cardiac surgery patients, as it is assumed to have a beneficial effect on cardiac function and consequently, postoperative hemodynamics' [5]. Although numerous phase I-III studies exist, characterizing the pharmacodynamic characteristics and the pharmacokinetic profile of levosimendan as such [6][7][8][9], there is still a lack of real-life data regarding actual levosimendan concentrations within this context. Furthermore, levosimendan is often applied in off-label dosages within this particular setting. Moreover, surgery can additionally affect pharmacodynamics [10]. Whereas high levels of levosimendan can lead to an increased incidence of side effects, such as tachycardia, hypotension, or atrial fibrillation [11] and a higher rate as well as longer duration of vasopressor infusion [12], underdosing may fail to induce a therapeutic effect. Thus, therapeutic drug monitoring (TDM) of levosimendan and its metabolites is essential, enabling adaption of pharmacotherapy according to individual pharmacokinetics. For TDM, easy compound extraction and rapid analysis with high specificity and sensitivity is key. One reason for the limited number of pharmacokinetic studies regarding levosimendan and its metabolites in a real-life setting may be the lack of an efficient method for simultaneous quantification of these compounds. Liquid chromatography coupled with electrospray ionization-based tandem mass spectrometry (LC-ESI-MS/MS) is usually the technique of choice for highly sensitive quantification of multiple analytes with a reasonable throughput in the field of TDM [13]. Although a few protocols for quantification of levosimendan and metabolites are described [14][15][16][17], there is unfortunately none enabling quantification of all three analytes in small-volume blood samples by means of LC-ESI-MS/MS in a single chromatographic run. Hence, we aimed to establish a highly sensitive LC-ESI-MS/MS method for quantification of levosimendan and its metabolites in a single run with a short chromatographic cycle time, which is suitable for TDM in patients undergoing cardiac surgery.
Pharmaceutics 2022, 14, x FOR PEER REVIEW 2 of 17 active OR-1896 ( Figure 1) [2]. Whereas levosimendan has a short half-life of one hour, the half-life of the metabolites is 70-80 h. Thus, it is hypothesized that the long-term effects, observed after termination of levosimendan treatment, arise from OR-1855 and OR-1896 [3,4]. Levosimendan is also used in the peri-operative management in cardiac surgery patients, as it is assumed to have a beneficial effect on cardiac function and consequently, postoperative hemodynamics , [5]. Although numerous phase I-III studies exist, characterizing the pharmacodynamic characteristics and the pharmacokinetic profile of levosimendan as such [6][7][8][9], there is still a lack of real-life data regarding actual levosimendan concentrations within this context. Furthermore, levosimendan is often applied in off-label dosages within this particular setting. Moreover, surgery can additionally affect pharmacodynamics [10]. Whereas high levels of levosimendan can lead to an increased incidence of side effects, such as tachycardia, hypotension, or atrial fibrillation [11] and a higher rate as well as longer duration of vasopressor infusion [12], underdosing may fail to induce a therapeutic effect. Thus, therapeutic drug monitoring (TDM) of levosimendan and its metabolites is essential, enabling adaption of pharmacotherapy according to individual pharmacokinetics. For TDM, easy compound extraction and rapid analysis with high specificity and sensitivity is key. One reason for the limited number of pharmacokinetic studies regarding levosimendan and its metabolites in a real-life setting may be the lack of an efficient method for simultaneous quantification of these compounds. Liquid chromatography coupled with electrospray ionization-based tandem mass spectrometry (LC-ESI-MS/MS) is usually the technique of choice for highly sensitive quantification of multiple analytes with a reasonable throughput in the field of TDM [13]. Although a few protocols for quantification of levosimendan and metabolites are described [14][15][16][17], there is unfortunately none enabling quantification of all three analytes in small-volume blood samples by means of LC-ESI-MS/MS in a single chromatographic run. Hence, we aimed to establish a highly sensitive LC-ESI-MS/MS method for quantification of levosimendan and its metabolites in a single run with a short chromatographic cycle time, which is suitable for TDM in patients undergoing cardiac surgery. To reach this goal, we used a triple quadrupole mass spectrometer with fast polarity switching capability and 13 C6-labelled analogues of all three analytes as internal standards to achieve a robust quantification of all compounds while taking the different retention behaviours of levosimendan and its metabolites into account. To reach this goal, we used a triple quadrupole mass spectrometer with fast polarity switching capability and 13 C 6 -labelled analogues of all three analytes as internal standards to achieve a robust quantification of all compounds while taking the different retention behaviours of levosimendan and its metabolites into account.

Sample Preparation-Liquid-Liquid Extraction
Human blood samples (serum and plasma) were stored at −80 • C. After thawing under room temperature, 25 µL 25 mM CH 3 COONH 4 (pH 4.75) and 1000 µL MTBE were added to 300 µL sample. The mixture was subsequently vortexed for 30 s, centrifuged (5 min, 1792× g), and finally, 900 µL of the organic phase were collected. Extraction was repeated four times, and the collected organic phase evaporated under nitrogen at room temperature, followed by reconstitution of the obtained residue in 300 µL mobile phase. Sample preparation was performed in plastic tubes (Eppendorf, Hamburg, Germany).

Method Validation
Quantification was validated according to the recommendations of the FDA/CDER [18]. The validation procedures for both protocols were performed three times on three different days. If not specified otherwise, serum was used as matrix.

Matrix Effect
Blank matrix (obtained by LLE) as well as solvent was spiked with levosimendan, OR-1855, OR-1896, 4-AAP, and 4-AAAP in quadruplicates at different concentrations. The matrix effect (ME) was characterized by comparison of peak areas in matrix vs. solvent samples and quantitatively assessed according to literature as corresponding percentages of area ratios [18,19].

Linearity
If not mentioned otherwise, preparation of calibrators was performed as described below. Samples from healthy blood donors were spiked before LLE with analytes and internal standards to yield calibration standards. The peak area ratios of analyte and internal standard (y-axis) were plotted against the concentrations of analyte (x-axis). Calibrators investigated at six concentration levels should be within ±15% of nominal concentrations [18]. For linear regression 1/x weighting was used.

Selectivity
For validation six matrix blanks from different individual sources were prepared and investigated [18].

Carryover
After measuring the highest calibrator concentration, a blank sample was measured. Carryover did not exceed 20% of the LLOQ [18].

Accuracy and Precision
If not stated otherwise, samples from healthy blood donors were spiked before LLE with analytes and internal standards to yield quality controls (QCs) samples. QCs were prepared at four concentrations (LLOQ, Low, Mid, High) in quintuplicates. For within-run and between-run analysis, a deviation up to ±15% of nominal concentrations (except at LLOQ ± 20%) was defined as acceptable. For precision, the within-run and between-run analysis was defined as acceptable within ±15% of CV (except LLOQ ± 20%) [18].

Extraction Recovery
Matrix samples spiked with levosimendan, OR-1855, OR-1896, 4-AAP, and 4-AAAP before LLE were compared with those spiked after LLE. Extraction recovery was determined as the area ratio of matrix samples spiked before LLE vs. after LLE [18].

Bench Top Stability
Different concentrations of levosimendan and its metabolites were tested at room temperature and 4 • C. Measurements were performed either directly after sample preparation or 6 h later.

Human Samples
Experiments with serum and plasma were carried out following the rules of the Declaration of Helsinki of 1975, revised in 2013 [20] and was approved by the clinical ethics committee at the medical faculty, LMU (identification code 20-665, approved 11 February 2021; identification code 20-1089, approved 31 August 2020). Human serum and plasma from healthy voluntary blood donors were prepared by centrifugation for 10 min at 1000 g. For plasma samples, the tripotassium salt of ethylenediamine tetraacetic acid ("K3EDTA") was used as an anticoagulant. Anonymized serum samples from patients receiving levosimendan within the peri-operative management at the Department of Cardiac Surgery of the LMU Klinikum were measured retrospectively without preparing replicates due to limited sample volumes. Patients were receiving either 12.5 mg levosimendan or different off-label dosages between 1.25 and 5 mg.

Data Analysis
Log D and pK a values were calculated with MarvinSketch 18.1, ChemAxon (Basel, Switzerland). Data are always shown as means ±SD. Data analysis was performed using Excel 2019, Microsoft (Redmond, WA, USA), and Prism 5.0, Graphpad (San Diego, WA, USA).

LC-MS
Levosimendan, OR-1855, and OR-1896 were recorded at the mass transitions (m/z) 279→227 (negative mode), 246→204 (positive mode), and 204→120 (positive mode), respectively, based on previously reported methods [9,16]. For the 13 C 6 labelled internal standards, the mass transitions corresponding to the six mass units' heavier ions (parent as well as fragment ions) in comparison to the non-labelled compounds were found to be the most intensive ones and therefore used for detection. With respect to chromatography, we tried to find isocratic conditions leading to retention factors of at least one for all analytes and sufficient intensities under ESI detection at a short run time. Starting with chromatographic conditions close to those described in previous studies, we could not achieve these aims. Therefore, we investigated various stationary (RP and polar) and mobile phases (varying pH, additives, kind, and portion of organic solvents) and found a short (30 mm) Luna C8 column with 76.6% 5 mM NH 4 HCO 3 (pH 8.5) and 23.3% CH 3 CN (v/v), as mobile phase, at a flow rate of 350 µL/min suitable to reach our goal (Figure 2A).

Liquid-Liquid Extraction
Before starting efforts to improve LLE protocols described in literature [2,16], we calculated pK a and log D values for levosimendan and its metabolites. Taking this information into account, we hypothesized that a pH of 4.75 in the aqueous phase should be a good compromise for the compounds' different dissociation behaviours. We obtained, employing MTBE as organic solvent, extraction efficiencies of 39.6% (±8.2%, n = 23), 67.9% (±12.3%, n = 23) and 84.0% (±8.6%, n = 23) for levosimendan, OR-1855, and OR-1896, respectively ( Figure 2B).

Matrix Effect
The comparison of spiked serum and solvent samples showed a mean peak area ratio of 114.2% for levosimendan, i.e., an enhancement under matrix conditions. For OR-1855 and OR-1896, peak mean area ratios were 65.7% and 79.1%, respectively, indicating suppression under matrix conditions for both metabolites (Table S2).

Linearity
Blank matrix samples spiked after LLE with analytes and internal standards were used. Linearity was examined in a range from 150 pM to 500 nM for levosimendan and from 500 pM to 1000 nM for the two metabolites, respectively. All calibration functions were characterized by r 2 values ≥ 0.9995 (Supplementary Materials).

Selectivity
No interfering signals were detectable in matrix blanks ( Figure 2C).

Liquid-Liquid Extraction
Before starting efforts to improve LLE protocols described in literature [2,16], we calculated pKa and log D values for levosimendan and its metabolites. Taking this information into account, we hypothesized that a pH of 4.75 in the aqueous phase should be a good compromise for the compounds' different dissociation behaviours. We obtained, employing MTBE as organic solvent, extraction efficiencies of 39.6% (±8.2%, n = 23), 67.9%

Accuracy and Precision
Quality controls (QC) were prepared from blank matrix spiked after LLE with analytes and internal standards. Representative MRM-chromatograms of the analytes at the LLOQ (levosimendan: 150 pM, OR-1855 and OR-1896: 500 pM) as well as for 50 nM are shown in Figure 2D,E. The criteria for accuracy and precision within and between runs were met in all validation series (Table 1).

Bench Top Stability
Extensive studies covering frozen storage, multiple freeze/thaw cycles, autosampler stability, and batch storage before injection regarding levosimendan, OR-1855, and OR-1896 have already been described by Zhang et al. [16], showing that all three analytes are stable under the investigated conditions. Therefore, we additionally only tested the bench top stability at room temperature and 4 • C. The storage for six hours at room temperature or 4 • C did not show any difference to freshly prepared samples (Tables S3-S5).

LLE and LC-ESI-MS/MS with Serum and Plasma Samples
After LC-MS-method validation, we also examined the complete procedure including sample preparation by adding the analytes and their internal standards to 300 µL serum before LLE. LLOQ for levosimendan and the metabolites was 0.45 nM and 1.0 nM, respectively. The calibrators were between 89.1-104.2%, 91.8-107.4%, and 93.2-114.8% of nominal concentrations for levosimendan, OR-1855, and OR-1896, respectively. As plasma samples are also used in daily clinical practice, sample preparation and LC-MS quantification were additionally investigated for plasma samples with the same concentrations for calibrators and quality controls, as described for serum. The results obtained were also consistent with the FDA requirements (Table 1).

Quantification of Levosimendan and Its Metabolites in Patient Serum
To test if this method was suitable for the detection of levosimendan, OR-1855, and OR-1896 in a clinical sample, serum from a cardiac patient receiving levosimendan, taken before surgery, was first studied. As seen in Figure 3A, all three analytes were detectable at their respective m/z traces 24 h after levosimendan infusion, whereby the concentration of levosimendan was found to be below the limit defined as LLOQ, i.e., <0.126 ng/mL (450 pM LLOQ). For OR-1855 and for OR-1896, we determined 0.224 ng/mL (1.10 nM) and 2.45 ng/mL (10.0 nM), respectively. To evaluate the suitability of our method for TDM in patients receiving several medications during surgery, serum samples taken directly after surgery from cardiac patients receiving levosimendan were investigated (n = 45; data not shown). The chromatograms in the positive mode, however, showed distinctly distorted peak shapes at the mass transitions (m/z) 246→204 and 204→120 and slightly shifted retention times. As an example, a chromatogram from one of these samples is shown in Figure 3B, wherein a levosimendan concentration of 0.390 ng/mL (1.39 nM) could be determined. This phenomenon possibly suggests interferences due to an insufficient mass spectrometric selectivity for the levosimendan metabolites. A database search revealed two metamizole metabolites (4-aminoantipyrine, 4-AAP and 4-acetamidoantipyrine, 4-AAAP), which are constitutional isomers of OR-1855 and OR-1896 ( Figure 1) [21]. Investigation of 4-AAP and 4-AAAP with our LC-MS settings showed that these compounds are actually detected at the mass transitions (m/z) 246→204 and 204→120, used to record OR-1855 and OR-1896, at retention times of 0.51 and 0.79 min, respectively ( Figure 3C). Indeed, metamizole is routinely used in cardiac surgery patients for pain therapy in our hospital. Investigation of additional mass transitions based on other fragmentations of OR-1855 (m/z 204→159 and 204→92) and OR-1896 (m/z 246→111, 246→120, 246→159, and 246→77) showed distinctly lower intensities (in comparison to m/z 246→204 and 204→120) in all cases and revealed only a notable selectivity at m/z 246→120 for OR-1896 vs. 4-AAAP. Due to the high concentrations of 4-AAAP present in some patient samples, however, an even 50 times higher response for OR-1896 vs. 4-AAAP did not guarantee a sufficiently selective quantification of OR-1896. Therefore, a chromatographic separation of the constitutional isomers was required.

Liquid-Liquid Extraction for Levosimendan Samples Containing Metamizole Metabolites
Using the established extraction protocol described above, mean extraction efficiencies of 63.9% for 4-AAP and 10.8% for 4-AAAP, respectively, were determined (Table S6).

Matrix Effect for 4-AAP and 4-AAAP
The comparison of spiked serum and solvent samples showed mean peak area ratios of 85.0% for 4-AAP and 97.4% for 4-AAAP, indicating a slight suppression under matrix conditions. For levosimendan, OR-1855, and OR-1896, similar matrix effects as found for the first chromatography method were detected (Tables S7 and S8).

Linearity
Linearity was examined for levosimendan from 450 pM to 500 nM, for OR-1855 as well as OR-1896 from 1.5 nM to 500 nM, for 4-AAP from 50 nM to 5000 nM, and for 4-AAAP from 1.35 nM to 5000 nM. Calculated calibration functions were always characterized by r 2 values ≥ 0.9980 (Supplementary Materials).

Selectivity
No interfering signals were detectable in matrix blanks ( Figure 4B).

Liquid-Liquid Extraction for Levosimendan Samples Containing Metamizole Metabolit
Using the established extraction protocol described above, mean extraction eff cies of 63.9% for 4-AAP and 10.8% for 4-AAAP, respectively, were determined (Tab

Matrix Effect for 4-AAP and 4-AAAP
The comparison of spiked serum and solvent samples showed mean peak area of 85.0% for 4-AAP and 97.4% for 4-AAAP, indicating a slight suppression under m conditions. For levosimendan, OR-1855, and OR-1896, similar matrix effects as foun the first chromatography method were detected (Tables S7 and S8).

Linearity
Linearity was examined for levosimendan from 450 pM to 500 nM, for OR-18

Quantification of Levosimendan, OR-1855, OR-1896, 4-AAP, and 4-AAAP in Patient Serum
After validation of the second protocol, its suitability for TDM was confirmed studying samples from surgical patients receiving simultaneous levosimendan and metamizole therapy. As exemplified in Figure 4D for the same sample already shown in Figure 3B under the conditions of the first LC-MS method, the analytes were now clearly separated at their corresponding mass transitions. For the representative sample shown in Figure 4D, serum concentration of levosimendan directly after surgery was 0.40 ng/mL (1.41 nM), for OR-1896 it was 0.88 ng/mL (3.59 nM), and for 4-AAAP it was 1.67 ng/mL (6.81 nM). The concentrations of OR-1855 and 4-AAP were below the LLOQs, i.e., <0.308 ng/mL (1.5 nM) and <10.2 ng/mL (50.0 nM).

Discussion
TDM of levosimendan is of particular importance in patients with conditions or treatments that can change its bioavailability, such as intensive care or surgical patients, where real-life data regarding its pharmacodynamics and pharmacokinetics is missing at large. In addition, due to frequent off-label dosing in these patients, TDM is an essential tool, which could indeed improve individual adjustments of therapy. To predict the duration of the effect of an applied levosimendan dose, it is necessary to record levels of levosimendan and, additionally, of its pharmacologically active metabolite OR-1896, which is assumed to be responsible for the long-term effect [9]. As generation of OR-1896 is dependent on acetylation of the inactive metabolite OR-1855 and its conversion rate depends on different alleles of NAT2, generating slow or rapid acetylators, respectively [8], measurements of OR-1855 have to be considered relevant as well. Here, we established, validated, and applied two protocols for TDM of levosimendan and its metabolites OR-1855 and OR-1896 in blood samples based on highly sensitive, simultaneous, and fast quantification by LC-ESI-MS/MS in a single chromatographic run employing 13 C 6 labelled analogues of all three analytes as internal standards, with the first designed for surgical patients without co-administration of metamizole and the second one for those receiving metamizole. As LC-MS/MS is the favored detection technique for sensitive quantification of small molecule analytes in biological matrices, several groups have tried to make use of it for quantification of levosimendan and its metabolites [2,9,16]. Zhang et al. [16] and Puttonen et al. [9] already reached markedly low LLOQs for levosimendan (0.7 nM), OR-1855 (2.5 nM), and OR-1896 (2.0 nM), but both protocols suffer from some disadvantages impeding implementation for TDM. The LC-ESI-MS/MS approach of Zhang et al. [16] is highly sophisticated and requires two separate chromatographic runs for data acquisition in the negative (levosimendan) and in the positive mode (metabolites). Puttonen et al. published a method based on APCI, able to record all three analytes in a single run, which however consumes high volumes (5-10 mL) of blood samples [9]. As LC-MS detection of levosimendan by APCI was recently found to be less sensitive as compared to ESI by Vlase et al. [17], and furthermore the type of APCI interface employed by Puttonen et al. [9] is no longer available, we focused on an LC-ESI-MS/MS method enabling rapid and highly sensitive quantification of levosimendan and its metabolites in a single chromatographic run, benefiting from a triple quadrupole mass spectrometer allowing fast polarity switching. Realizing this idea, we were able to keep the chromatographic cycle time within 2.0 min and reached even lower LLOQs of 0.45 nM for levosimendan and 1.0 nM for both metabolites in plasma and serum as described so far. Furthermore, we achieved higher extraction efficiencies with LLE compared to the so far most efficient procedure described by Zhang et al. [16] which is in the case of levosimendan with about 40% still only moderate. It has to be considered, however, that this is due to the approach of a single sample preparation step for levosimendan and its metabolites possessing distinctly different acid-base-behavior (levosimendan acidic, metabolites basic) and, therefore, requiring a compromise regarding the extraction conditions. As the intensities observed for equimolar concentrations of our three analytes in LC-ESI-MS chromatograms were sufficiently high for levosimendan but by far lower for the metabolites, a moderate extraction efficiency was more tolerable for levosimendan as compared to its metabolites. Finally, and most importantly in this context, it has to be mentioned that this surely comparably low extraction efficiency does not impair reliable quantification of levosimendan, as low or (even worse) fluctuating extraction efficiencies are in any case fully compensated by the use of 13 C-labeled levosimendan as internal standard.
Application of this validated method to a sample taken from a cardiac patient receiving levosimendan preoperatively revealed that the established protocol is well-suited for TDM in this field and thus represents an improved tool to monitor levosimendan application. Although the primary intention of the present work was to establish a LC-ESI-MS/MS quantification method for the selected analytes in human serum samples, we additionally used our protocol for human plasma to explore the possible extension of this method to plasma samples. However, a further validation is still needed, before clinical use. At the next step, applying the method on samples collected after cardiac surgery, unexpected interferences at the traces recorded in the positive mode for OR-1855 and OR-1896 were observed. These interferences were shown to originate from two metamizole metabolites, 4-AAP and 4-AAAP, possessing the same molecular formulas as OR-1855 and OR-1896 (i.e., constitutional isomers) as well as fragmentation patterns, leading to product ions with the same m/z ratios as the ones to be recorded for detection of the levosimendan metabolites in the MRM mode. Surely, interferences of isobaric compounds can never be excluded in a polypharmaceutical context of surgical patients [22]. However, the situation observed here, where metabolites with identical molecular formulas are generated from different drugs, which in turn fragment-despite distinctly different chemical structures (Figure 1)-to product ions with the same m/z ratios as the metabolites of the investigated drug is rather uncommon. Nevertheless, a "mass-selective" quantification of the levosimendan metabolites in the presence of the metamizole metabolites based on different mass transitions was impossible. Therefore, we started a second attempt based on chromatographic separation of OR-1855 from 4-AAP and OR-1896 from 4-AAAP, respectively. After validating this method, we applied it to samples of cardiac surgery patients, and were now able to quantify levosimendan, OR-1855, OR-1896, 4-AAP, and 4-AAAP reliably in a single LC-MS run still in an acceptable chromatographic cycle time of 7.0 min. However, as the measurements were performed with anonymized patient rest samples, there was unfortunately not enough serum to prepare sample replicates, especially since the same samples were used for both protocols. The so far available studies measuring serum or plasma concentrations of levosimendan and its metabolites in cardiac patients do not report an interference with metamizole metabolites or any other applied drugs [23,24]. It can be assumed that this is due to the fact that these studies did not examine surgical patients and that they were performed in countries where metamizole is not approved [25]. Considering the administration of metamizole (dipyrone) as an antipyretic and nonopioid analgesic drug frequently approved for severe pain upon surgery [26] in 19 European as well as in several Latin American and Middle East countries clearly demonstrates the urgent need for a method enabling TDM for levosimendan including surgical patients and those receiving metamizole. Finally, the TDM protocols established here may be used for clinicians and researchers to obtain data on the therapeutic serum or plasma levels of levosimendan and its metabolites, investigating the effect of different doses in children, patients with myocardial dysfunction undergoing cardiac surgery, sepsis, and weaning failure. This in turn may contribute to the development of treatment guidelines and dosing recommendations for the use in clinical practice. Moreover, immediate post-surgery TDM of especially the metabolites OR-1896 and OR-1855, exhibiting long half-lives, in combination with pharmacokinetic modelling may represent a tool to predict therapeutic effects and enable individualized dose adjustments to ensure an optimal therapeutic outcome in patients with reduced ejection fraction undergoing major surgeries.