A review on liquid chromatographic methods for the bioanalysis of atorvastatin

The unsatisfied clinical need has encouraged the development and validation of bioanalytical procedures for the quantification of drugs in biological samples because the monitoring of drug concentrations helps in personalizing the patient’s pharmacotherapy, assessing the adherence to therapy, and is also extensively useful for pharmacokinetics and drug-drug interactions studies. The present review aimed to provide insightful information about the various liquid chromatographic methods developed till 2019 for the analysis and quantification of atorvastatin, its metabolites, and co-administered drugs in the various biological matrices like the serum, plasma, and urine with special emphasis on sample preparation techniques applied before chromatographic analysis along with different chromatographic conditions and their validation data. A total of 88 published papers that have used liquid chromatography techniques to quantify atorvastatin in biological fluids are included in the study. Out of the total reported liquid chromatographic methods, 34% used UV spectrophotometer as a detector, and 55% used MS/MS as a detector. Whereas 38% of them used protein precipitation procedure, 33% applied liquid-liquid extraction approach, and 12% employed solid-phase extraction technique for sample preparation. In the last decade, numerous bioanalytical procedures have been developed for the quantification of atorvastatin in different biological samples using liquid chromatographic techniques. Moreover, advancement in technology developed several new and advanced sample preparation approaches like dispersive liquid-liquid extraction, microextraction by packed sorbent, which have high recovery rates than conventional procedures. Thus, the summarized review may be consulted as an informative tool to support the optimization of new bioanalytical methods for the quantification of atorvastatin.


Background
Statins (HMG CoA reductase inhibitors) are considered as the drug of choice for the treatment of hypercholesterolemia because of their extremely beneficial effect in reducing total cholesterol and low-density lipoprotein cholesterol levels in the human body [1]. Statins competitively inhibit the HMG-CoA reductase which further depletes the intracellular supply of cholesterol [2]. Atorvastatin (AT) (Fig. 1) is one of the oldest and major drug under the class of statins, which is chemically (3R,5R)-7-2-(4-Fluorophenyl)-3-phenyl-4-(phenyl arbamoyl)-5-propan-2-ylpyrrol-1-yl-3,5-dihydroxyheptanoic acid [3] and shows the highest low-density lipoprotein-cholesterol lowering efficacy in human at a maximum daily dose of 80 mg. With a molecular formula of C 33 H 15 FN 2 O 5 and a molecular weight of 558.064 g mol −1 , AT exists in two different forms, one is open-structured hydroxyl acid while another one is ring-structural lactone which degrades in both high alkaline and low acidic conditions [4]. AT has a significantly longer plasma half-life of 18-24 h compared to other statins [5]. AT is administered in its active acid form which afterward metabolized in its two active hydroxyl metabolites, i.e., ortho-hydroxy atorvastatin (o-OH AT) and para-hydroxy atorvastatin (p-OH AT) and three inactive lactone metabolites [6]. Runny nose, sneezing, and coughing are the most common side effects of AT whereas muscle problems, liver problems, loss of appetite, and upper stomach pain are certain serious side effects of AT [7]. AT is also clinically used to reduce the risk of stroke [8]; myocardial infarction [9], which further reduces the risk for angina [10]; and revascularization procedures [11].
Although plasma drug monitoring is not crucial for the AT, an increasing number of patients taking AT have drawn the attention of healthcare professionals for mandatory monitoring of plasma concentration in various conditions. Moreover, monitoring the plasma concentrations of AT helps personalize the patient's pharmacotherapy and to assess the adherence to therapy. Since AT is also co-administered with many other drugs as polypharmacy therapy, the bioanalytical assays are also extensively used for its pharmacokinetics and drug-drug interactions' studies. Thus, the unsatisfied clinical need has encouraged the development and validation of several liquid chromatography methods as bioanalytical procedures for the determination of AT in body fluids with high accuracy, precision, and better reproducibility. The present review encapsulates all reported liquid chromatography-based bioanalytical procedures available in the literature that are used for the quantification of AT alone, with its metabolites and with co-administered drugs in different biological samples. Also, a succinct tabular outline of the bioanalytical procedures has been included in the review which consists of chromatographic conditions, sample preparation strategies, validation data, etc.

Main text
Indeed, the best approach for the accurate analysis of drug concentration in biological samples is the use of high-resolution liquid chromatographic techniques such as HPLC and UPLC coupled with UV spectrophotometer or MS/MS detectors. Liquid chromatography techniques are not only considered as an important tool to support the therapeutic drug monitoring of drugs in the biological fluids but also during quality control of pharmaceutical formulations as well as during nonclinical and clinical drug development. From 1998 to 2019, different liquid chromatographic procedures to analyze AT in biological fluids were developed, validated, and applied in pharmacokinetic studies and therapeutic drug monitoring. Out of the total reported liquid chromatographic methods, 34% used UV spectrophotometer as a detector, and 55% used MS/MS as a detector.

Analysis using HPLC
HPLC is the preferable method for the analysis of drugs and currently the widest accepted method for separation technique. Table 1   enlist the various HPLC methods coupled with UV, DAD, and PDA detectors used to detect AT and co-administered drugs in biological fluids. Usually, the reverse phase HPLC method     has been used to quantify AT using C18 analytical columns because of the presence of pyrrole and phenyl as aromatic functional groups that make analysis suitable by the reverse phase method [52]. However, some published work has used the C8 column as the stationary phase. The mobile phase that has been used in the quantification of AT in biological matrices, primarily comprised of acetonitrile (ACN), methanol (MeOH), water, and different buffer solutions that have either isocratic or gradient elution with a flow rate of 1 mL min −1 . Retention of analyte in the analytical column hinge upon the pH of elution, so in reported bioanalytical assays, pH of the mobile phase has been maintained between 2 and 4 to make better retention of analyte in the analytical column and enhanced the resolution of the peak because these pH values are lesser than pKa value of AT and AT remain in unionized form for a longer time and interact more with the stationary phase [52]. The retention time (R t ) of AT reported in Table 1 variegates from 1.7 to 19.8 min, but in most of the developed HPLC methods, R t has ranged 2 to 5 min. The wavelength of UV detector used to detect AT has ranged from 210 to 270 nm, but commonly, 247 nm has been used as detection wavelength. Conversely, Tekkeli et al. quantify AT in human plasma using fluorimetric detection at excitation and an emission wavelength of 282 and 400 nm, respectively [37]. For the accuracy in quantification, diclofenac [20,26,31,34] and ibuprofen [29,39] have been often employed as internal standards in various developed HPLC methods to determine AT. Considering the validation of developed methods, all the developed methods have been validated as per ICH Q2 (R1) guidelines for system suitability, sensitivity, selectivity, linearity, limit of detection (LOD), limit of quantification (LOQ), and inter-day and intra-day accuracy and precision. LOD of the AT has been determined based using signal intensity to baseline noise ratio (S/N) of 3:1 by comparing test results from samples of a known concentration of analyte with the blank sample while LOQ determination has been based on signal intensity ten times more than baseline noise, i.e., S/N = 10. Least LOQ and LOD value of 0.2033 ng mL −1 and 0.0608 ng mL −1 respectively were observed in the method proposed by Dastkhhon et al. [43]. whereas the method developed by Gholami and Ahmadi was highly sensitive with LOQ value of 0.38 ng mL −1 and LOD value of 0.10 ng mL −1 with linearity in concentration calibration range of 1 to 1000 ng mL −1 and %RSD less than 4.2% [49].

Analysis using LC-MS/MS and UPLC-MS/MS
The liquid chromatography-based bioanalytical procedures using MS/MS detectors are listed in Table 2  . On the other hand, few studies used UPLC MS/MS method for the estimation of AT in the biological fluids and are also indexed in Table 2 [55,56,65,71,74,81,87,91,94,95,100]. Similar to HPLV-UV methods, C18 analytical columns have been extensively used in LC-MS/MS assay as a stationary phase because their long alkyl chain provides more efficient separation of nonpolar compounds in MS/ MS detection [101], whereas a column packed with bridge ethyl hybrid (BEH) has been used in UPLC method as a stationary phase [71,81,87,95] because of their high mechanical resistance [102]. Although MeOH, ACN, and water are used as mobile phases for quantification in LC-MS/MS, certain additives like formic acid (FA), ammonium acetate (AA), and acetic acid (AcA) have been added in the mobile phase to enhance ionization which further increases the sensitivity of the method. Isocratic elution has been predominantly preferred over gradient elution in the analysis of AT with an elution rate of 0.2-0.5 mL min −1 . R t has found to be quite small, ranged between 0.66 and 6.3 min.
AT and its metabolites can be detected in both positive and negative ionization modes of MS, yet most of the published work has performed LC-MS/MS analyses of AT in positive ionization mode using electrospray ionization (ESI), but few reported studies used negative ions [54,56,58,67,80,94] of AT for the analysis even though, polarity switching within run is required in ESI − mode [89]. On the other hand, the atomic pressure chemical ionization (APCI) technique has been used by Tahboub for ionization of AT in positive mode [64]. Figures 2 and 3    AT, o-OH AT, and p-OH AT    AT and amlodipine  AT and metabolites  LOD and LOQ in the reported methods have been calculated based on their signal to noise ratio, as discussed earlier; nevertheless, in some developed methods, lowest concentration of their linear calibration curve has been chosen as LOQ for analysis. The bioanalytical method developed by Zhou et al. is most sensitive than all other reported LC-MS/MS methods with the lowest LOQ of 0.036 ng mL −1 and the calibration curve in the range of 0.03-25 ng mL −1 along with intra-day and inter-day RSD of 4.02% and 5.23% respectively [69]. In more than 50% of total reported LC-MS/MS methods, LOQ is equal to or less than 0.5 ng mL −1 which indicates the developed LC-MS/MS methods are highly sensitive. Importantly, Novakova and co-workers have demonstrated the role of the two variables, i.e., concentration and pH of buffer solution in the mobile phase that affects the signal to noise ratio of MS/MS detector, and observed that an increase in a concentration above 5 mM significantly reduces the response of mass spectrometer and the concentrations lower than 0.5 mM were not enough to keep buffering capacity and thus had a negative impact to the response of the MS/MS detector; thus, the best response of AT and simvastatin was observed at 0.5 mM concentration of buffers. Also, AA gave better peak shapes as compared to ammonium formate (AF), and finally, AA buffer (pH 4.0; 0.5 mM) was selected as an optimized buffer for mobile phase composition [81].

Analysis of metabolites and co-administered drugs
Most of the references cited in Tables 1 and 2 also describe the simultaneous determination of AT with certain other statins [13, 20, 24, 27-29, 50, 51, 54, 56, 61, 65, 81, 87] and other co-administered drugs such as amlodipine [35,40,59,64,65,69,70,77,78,95], ezetimibe [42,66,71], gemfibrozil [13,30,51], clopidogrel [14,15,41,93], fenofibrate [26], aspirin [15,23,72], ramipril [64,78], pioglitazone [17,22], glimepiride [64,68,73], olmesartan [62,98], and many other. Various methods have also been developed and validated using LC-MS/MS to quantify the metabolites of AT and are enlisted in Table 2   plasma using phenyl column [75]. Interestingly, Van pelt et al. modified the conventional LC-MS system by incorporating three valves and four columns into it for the analysis of AT and acid and lactone metabolites in human plasma to reduce the overall run time from 4.5 to 1.65 min [96]. Courlet et al. developed a UHPLC-MS/MS method for the simultaneous quantification of amlodipine, metoprolol, pravastatin, rosuvastatin, AT, o-OH AT, and p-OH AT in human plasma using their stable isotopically labeled analogs as IS. All the compounds were analyzed using ESI + mode, except for pravastatin which was detected using ESI − mode [65]. Vlčková et al. optimized a UHPLC-MS/MS method for selective quantification of seven statins along with their inter-converted products and metabolites in biological samples by comparing the three MS/MS detector with different specifications [55]. Likewise, Wagmann et al. proposed a highly selective method to detect seven statins and their metabolites in human blood plasma using LC-high-resolution MS/MS and applied in the therapeutic drug monitoring of these statins on 14 human plasma samples [53]. Also in 2014, eight different cardiovascular drugs were simultaneously quantified in human plasma using an APCI LC-MS/MS method by Tahboub [64]. However, an RP-HPLC-UV method was designed by Bhatia et al. for the simultaneous estimation of four cardiovascular drugs, i.e., AT, aspirin, atenolol, and losartan in tablet dosage form and plasma [23].

Analysis in biological fluids other than blood
Instead of analysis in common biological fluids such as plasma and serum from humans or animals, certain researchers have successfully developed liquid chromatography-based bioanalytical procedures for the quantification of AT in various other biological fluids. Silva et al. determined AT and curcumin in mouse plasma and also in the lung, brain, liver, and spleen homogenate supernatants using an HPLC-DAD assay [12]. Identically, AlKhani et al. designed a precise HPLC-UV method to quantify AT in both rat plasma and rat intestinal perfusion solution for its further application in pharmacokinetic studies [39]. The development of analytical methods for the detection of drugs and its metabolites in the urine samples generally helps to understand in vivo metabolism and clearance process, so, various methods have been developed to analyze AT in the urine samples [13,43,44,47,92]. In the year 2017, Beydokhti et al. proposed a method to simultaneously determine AT, rosuvastatin, and gemfibrozil in deionized water, wastewater, human plasma, and urine using HPLC-UV method and urine samples were prepared for analysis by diluting with 10 mL of deionized water [13]. Likewise, eight statins were simultaneous determined in human urine using a highly selective and sensitive LC-MS/MS method developed by Jang et al. in which dilute and shot approach was used for urine sample preparation [92]. Sample preparation or extraction of drug from the biological matrix Traditionally, sample preparation is carried out by diverse techniques such as precipitation of the plasma proteins (PP), liquid-liquid extraction (LLE), and solid-phase extraction (SPE), before the final bioanalysis using liquid chromatography interfaced with UV or MS or MS/MS or capillary gas chromatography to prevent the interference of endogenous substances such as lipids, proteins, salts, acids, bases, and cells present in biological fluid with the chromatographic detection. Out of the total methods cited in Tables 1 and 2, 38% used PP for sample preparation, 33% applied LLE techniques, and 12% employed SPE technique. Protein precipitation (PP) is the most commonly applied sample preparation technique used for the extraction of AT from biological matrix because of its rapidity, simplicity, and suitability to both hydrophilic and lipophilic analytes. PP process generally involves the addition of ACN or MeOH as precipitating agents followed by subsequent vortex and centrifugation. The % mean recoveries obtain from PP are quite significant. Acidification of the solvent has been generally employed in many PP methodologies to extract AT because the addition of acids such as trichloroacetic acid (TCA) or formic acid (FA) or AcA in the organic solvent enhanced the precipitation process due to changes in the pH [15,25,33,35,40,65,75,77,93]. Also, ice-cold extraction solvents have been used in many procedures to prevent degradation and achieve high purification of samples [15,25,34,40,46,99]. Mathur and Devi used 10% perchloric acid as a precipitating agent along with a mixture of K 2 CO 3 (2.5 M) and KOH (6 M) to maintain pH between 6.0 and 6.5 [48].
Various bioanalytical methods used LLE strategy for the sample pre-treatment of AT before the final analysis by using ethyl acetate, methyl-tertiary butyl ether, dichloromethane, or diethyl ether as organic extraction solvents. Instead of the single solvent procedure, various methods have also used a combination of different solvents [17,27,53,54,69,85,86] or various buffering agents such as FA [17,60]. AA [20] and phosphate buffer [33,[70][71][72]85] for the extraction of AT using LLE method. Also, the addition of buffering agents to the extraction solvent has improved the recovery of analyte from biological matrices. Bullen et al. used NaOH as a buffering agent to boost the extraction of AT from the rat, dog, and human plasma [90]. Despite its few disadvantages such as time-consuming, use of toxic, expensive, and environmentally unfriendly organic solvents [18] and less selectivity toward hydrophilic drugs and metabolites [103], LLE is still extensively employed as a sample preparation technique.
Because of its laborious multistep procedure including column conditioning, sample loading, washing, and elution, followed by evaporation of the eluent, only 12% of the total reported methods adopted SPE approaches for the extraction of AT from biological matrices. Different types of reverse phase SPE cartridges or columns such as HLB or C18 SPE cartridges have been used in SPE as a stationary phase. Moreover, ACN and MeOH have been commonly employed as a solvent for column conditioning, washing, and elution along with some acid, base, or buffers to facilitate elution. SPE isolates analyte more effectively than LLE but not widely employed. Apart from its tedious procedures, cartridges used in SPE are also too expensive, and the plugging of the cartridge is also a major concern in this extraction procedure. Interestingly, Partani et al. carried out their whole SPE sample preparation process in an ice-cold water bath (excluding vortex mixing, SPE, and drying) and under low light conditions to prevent the degradation of the analyte by the temperature and light [67].
Besides conventional sample preparation approaches, many researchers used modified and novel extraction procedures to isolate AT from biological matrices. Both Hassan et al. and Yang et al. employed a novel Saltingout assisted liquid-liquid extraction (SALLE) technique for the extraction of AT from human serum and human plasma respectively using ACN salted with AA and magnesium chloride as a water-miscible solvent [18,62]. Interestingly, this fast sample pre-treatment procedure shortens the chromatographic run time of AT. Martins et al. utilized dispersive liquid-liquid micro-extraction (DLLME) as a novel extraction method for the isolation of AT from biological matrices and compare it with the conventional SPE method [24]. A further modification in the DLLME technique with solidification in the prepared floating organic droplets has been made by Taheri et al. using a low-density solvent (Undecanol), which floats and solidifies at the top of the extraction vial during the extracted phase and separates the analyte easily [16]. Furthermore, Farajzadeh et al. used sugar cubes as a disperser base instead of a disperser solvent in the DLLME procedure to extract AT from human serum [38]. Notably, Beydokiti et al. resolved the problem of lack of sample clean-up by using tandem air-agitated LLE based on solidification of floating organic droplets (TAALLME-SFO) technique for the extraction of AT from a biological matrix. However, the use of chlorinated solvents is also a major drawback of this method [50]. Interestingly, instead of two-phase liquid microextraction, in 2009, Farahani et al. designed a novel three-phase liquid microextraction process (LLLME) in droplet-based mode to separate the analyte from complex biological matrix before HPLC assay [31].
Rukthong et al. utilized a simple, sensitive, and rapid solid-phase microextraction (SPME) technique to extract AT from human plasma using an RP-C18 SPE sorbent material which is packed in a 100-μL pipette tip to eliminates the clogging problem [97]. Because of comparative high sorption capacity and high selectivity for analytes, many researchers employed magnetic nanoparticles (MNPs) as sorbents instead of conventional SPE sorbent in the extraction of AT from biological matrices and the extraction technique involving MNP is termed as magnetic solid-phase extraction (MSPE). Tekkeli et al. used graphene nanosheets decorated with Fe 3 O 4 nanoparticles as a magnetic sorbent for the extraction of AT from human plasma along with ACN as an extraction solvent [37]. Similarly in another study, Khoshhesab et al. used an ultrasound-assisted MSPE technique using sodium dodecyl sulfate-coated Fe 3 O 4 MNP as a sorbent for the extraction and found it comparatively easy and fast while comparing with conventional LLE and LLLME methods [44]. Dastkhoon et al. in 2017 developed dispersive micro solid-phase extraction (Dμ-SPE) approach as a novel modification in the SPE technique for extraction of AT from human plasma and urine using Cu@SnS/SnO 2 nanoparticles loaded on activated carbon as a sorbent for separation of AT from matrices [43]. Interestingly, a combined magnetic-dispersive SPE (MD-SPE) approach was designed by Gholami and Ahmadi for the extraction of AT from human serum using polypyrrole-Fe 3 O 4 MNP as sorbent [49]. Notably, Azadi and Ahmadi further used magnetic dispersive micro solid-phase extraction (MDμ-SPE) technique for the extraction of AT and valsartan from human serum and urine using CMCcoated Fe 3 O 4 MNP [47]. In the same way, Beydokhti et al. developed a novel extraction technique by the combining magnetic dispersive micro SPE process and supra-molecular solvent-based microextraction process (Mdμ-SPE-SSME) to extract AT, rosuvastatin, and gemfibrozil from human plasma and urine sample with the help of layered double hydroxide-coated Fe 3 O 4 . Also, in comparison with the conventional SPE technique, Mdμ-SPE-SSME reduces the overall extraction time and eliminates the elution step [13]. Interestingly, Vlčková et al. developed microextraction by packed sorbent (MEPS) technique to extract AT and its metabolites from biological samples using a gas-tight syringe as an extraction device [56,74]. Briefly, MEPS is a novel and miniature version SPE technique, and in comparison with the conventional SPE methods, the MEPS technique is quite simpler and time-saving and requires a lesser amount of sample. Moreover, the MEPS technique has good recovery in a short time with a lesser volume of samples.

Conclusion
Analysis of AT in the biological fluid is quite useful for pharmacokinetic studies, therapeutic drug monitoring, and drug-drug interaction studies. Numerous developed bio-analytical methods that are useful for quantitation of AT either alone or with its metabolites or coadministered drugs in plasma, serum, or urine are herein summarized and discussed. Due to advancement in technology, many new LC methods have been developed in the last decade with high sensitivity and accuracy. However, LC-MS/MS technique is more sensitive, selective, and precise as compared to the HPLC-UV technique for the quantification of AT. Also, various extraction techniques used to extract AT from biological matrices are discussed herein. Indeed, PP is the most commonly applied technique for sample preparation, followed by LLE and SPE, yet numerous advance methods have been developed such as SALLE, DLLME, and MEPS for the extraction of AT from biological fluids, to get high recovery. Thus, this review can be used as an informative tool to develop and optimize certain new LC methods for quantification of AT in biological matrices.