Journal of Pharmaceutical and Biomedical Analysis 96 (2014) 241–248
Contents lists available at ScienceDirect
Journal of Pharmaceutical and Biomedical Analysis journal homepage: www.elsevier.com/locate/jpba
Simultaneous determination of oral antidiabetic drugs in human plasma using microextraction by packed sorbent and high-performance liquid chromatography Iara Maíra de Oliveira Viana, Paula de Paula Rosa Lima, Cristina Duarte Vianna Soares, Christian Fernandes ∗ Departamento de Produtos Farmacêuticos, Faculdade de Farmácia, Universidade Federal de Minas Gerais, Av. Pres. Antônio Carlos 6627, 31270-901 Belo Horizonte, MG, Brazil
a r t i c l e
i n f o
Article history: Received 23 December 2013 Received in revised form 10 March 2014 Accepted 25 March 2014 Available online 2 April 2014 Keywords: Oral antidiabetic drugs Microextraction by packed sorbent Experimental design Fused-cored column Human plasma
a b s t r a c t In this study, a simple method using microextraction by packed sorbent and high-performance liquid chromatography with ultraviolet detection for simultaneous determination of chlorpropamide, gliclazide and glimepiride in human plasma was developed and validated. A fractional factorial design and a complete factorial design were applied to evaluate the parameters which could affect the extraction and desorption steps, respectively. All parameters in the extraction step (pH, sample volume, sample dilution and number of aspiration/ejection cycles) and in the desorption step (percentage of acetonitrile in the elution solvent and number of aspirations of elution solvent through the device) were statistically significant (p > 0.05) when recovery was used as response. The developed method allowed the use of small volumes of sample and solvents and rapid separation by using a fused core column (only 2.2 min were needed). This method was fully validated showing selectivity, precision, accuracy and linearity over the range 1.0–50.0 g mL−1 for chlorpropamide, 1.0–10.0 g mL−1 for gliclazide and 0.1–1.0 g mL−1 for glimepiride. Finally, the validated method was applied in the analysis of samples from volunteers containing the three tested analytes. © 2014 Elsevier B.V. All rights reserved.
1. 1 Introduction Diabetes mellitus is a chronic disease that occurs when pancreas does not produce enough insulin, or when cells do not respond to the insulin that is produced. Hyperglycemia, which could cause serious damage to many body’s systems, is a common effect of uncontrolled diabetes. According to World Health Organization, approximately 347 million people have diabetes mellitus worldwide and it will probably be the seventh leading cause of death in 2030 [1]. The non-pharmacological diabetes treatment – healthy diet and regular physical activity – can prevent or delay the onset of type 2 diabetes mellitus (T2DM) [1]. However, lifestyle modification alone is usually not able to achieve appropriate glycemic levels. Patients
∗ Corresponding author at: Departamento de Produtos Farmacêuticos, Faculdade de Farmácia, Universidade Federal de Minas Gerais, Av. Presidente Antônio Carlos, 6627 Campus Pampulha, 31270-901 Belo Horizonte, MG, Brazil. Tel.: +55 31 34096957; fax: +55 31 34096976. E-mail address: [email protected] (C. Fernandes). http://dx.doi.org/10.1016/j.jpba.2014.03.042 0731-7085/© 2014 Elsevier B.V. All rights reserved.
with T2DM are usually treated with oral medication or even insulin. Often the use of a single drug is unlikely to result in long-term glycemic control and patients require multiple agents with different mechanisms of action [2]. The sulfonylureas are insulin secretagog agents which are commonly prescribed with metformin when the first-line treatment (metformin monotherapy) failed [2,3]. Beyond that, the sulfonylureas become the first line option for people with normal weight or when metformin should not be taken [4]. Also, they must be the first choice when the patient has weight loss or too high blood glucose [3]. There are two groups of sulfonylureas, the first and second generation, but all drugs consist of substituted arylsulfonylureas (Fig. 1). The first generation group includes chlorpropamide (CL), tolazamide, acetohexamide and tolbutamide; while the second one contains glibenclamide (GB) (also known as glyburide), gliclazide (GZ), glimepiride (GM) and glipizide [5]. Determination of drugs in biological fluid is essential in therapeutical drug monitoring, pharmaceutical research, analytical and toxicological studies. However, conventional sample preparation methods are usually laborious, time-consuming and are
242
I.M.d.O. Viana et al. / Journal of Pharmaceutical and Biomedical Analysis 96 (2014) 241–248
Fig. 1. Chemical structures of the sulfolnylureas (A – chlorpropamide; B – gliclazide; C – glimepiride; D – glibenclamide).
responsible for at least one-third of the error generated during the performance of an analytical method. Moreover, they employ large sample volumes and toxic solvents [6–8]. Until now, several analytical methods have been described for the determination of CL, GZ or GM in human plasma samples. However, from our knowledge, in all of them conventional techniques, such as protein precipitation (PP) [9–13], liquid–liquid extraction (LLE) [14–18] and solid-phase extraction (SPE) [19,20], were employed. Microextraction by packed sorbent (MEPS) is a new miniaturized solid-phase extraction technique that can reduce the volume of solvent and sample needed. Furthermore, it is easy to use and rapid. The sorbent (1–4 mg) is inserted into the syringe (100–250 L) barrel as a plug or between the needle and the barrel as a cartridge [21]. It is a very versatile technique, since this cartridge can be packed with any solid-phase material such as silica-based (C2, C8, C18), restricted access material (RAM) or molecularly imprinted polymers (MIPs) and can handle small sample volumes (10 L of plasma, urine or water) as well as large volumes (1000 L). Compared with PP, LLE and SPE, MEPS can reduce sample preparation time and organic solvent consumption [22], as have already been reported for determination of other drugs in biological samples [23–29]. One of the most advantages of this technique is that the sorbent could be used several times; more than 100 times with plasma or urine samples, while conventional SPE cartridges is used only once. The washing step after extraction allows and increases reuse. Moreover, using appropriate pretreatment of the sample before the MEPS method, blockage and coagulation are avoidable. Commonly, pretreatment includes centrifugation to remove suspended materials, pH adjustment, hydrolysis or precipitation [30]. In this context, the aim of this study was to develop a modern and fast method to simultaneously determine CL, GZ and GM in human plasma. Despite the fact that these three drugs are not usually co-administered in the clinical practice, the proposed method allows the determination of CL, GZ and GM together or alone, avoiding the need for developing three independent methods. MEPS was used as sample preparation technique. Experimental design was employed to optimize the parameters that affect the extraction. Liquid chromatography was performed by using a fused core column, which allowed rapid and efficient separation. The developed
method was validated according to Brazilian legislation and applied to the analysis of samples from volunteers. 2. 2 Materials and methods 2.1. Chemicals, reagents and materials CL batch 1019, GZ batch 1057 and GB batch 1018 (used as internal standard – IS) analytical standards were purchased from Instituto Nacional de Controle de Qualidade em Saúde (Rio de Janeiro, Brazil). GM batch G0K135 analytical standard was obtained from United States Pharmacopeia (Rockville, USA). GB by Cadila Pharmaceuticals (Ahmedabad, Gujarat, India), GM by Mantena Laboratories (Nalgonda Dist, Andhra Pradesh, India), GZ by Shandong Keyuan Pharmaceutical (Jinan, Shandong, China) and CL by Kothari Phytochemicals International (Nagari, India) active pharmaceutical ingredients were used in the optimization step. Three different drug products were obtained in the local market: tablets containing 250 mg of chlorpropamide, 30 mg of gliclazide and 4 mg of glimepiride. Acetonitrile and methanol HPLC grade and sodium citrate dehydrate analytical grade were obtained from J.T. Baker (Xalostoc, Mexico). Phosphoric acid 85% (w/w) and sodium hydroxide were acquired from Merck (Darmstadt, Germany). Potassium phosphate monobasic was supplied by Sigma-Aldrich (São Paulo, Brazil); citric acid monohydrate and sodium carbonate anhydrous by Labsynth (Diadema, Brazil); and sodium bicarbonate by Vetec (Rio de Janeiro, Brazil). The water used for preparing all solutions and samples was purified in a Direct-Q 3 System from Millipore (Bedford, MA, USA). Extraction was performed using the MEPS 250 L syringe and the MEPS BIN (barrel insert and needle) containing 4 mg of silica C18 sorbent (SGE Analytical Science; Melbourne, Australia). 2.2. Sample collection Blood samples were collected from volunteers at Laboratório de Hematologia da Faculdade de Farmácia da Universidade Federal de Minas Gerais (UFMG; Belo Horizonte, Brazil). This study was approved by UFMG Ethics Committee. The blood samples were usually drawn in the morning from fasting volunteers and stored in glass tubes containing heparin as the anticoagulant until been
I.M.d.O. Viana et al. / Journal of Pharmaceutical and Biomedical Analysis 96 (2014) 241–248
centrifuged at 480 × g for 10 min at 4 ◦ C. Then, the supernatant (plasma) was stored at −80 ◦ C. Drug-free plasma samples (blank plasma) were spiked with analytes before being used for MEPS optimization and method validation. After, the validated method was applied in samples from nine different volunteers (three volunteers for each drug) who received single dose of antidiabetics orally. Only one drug was administered to each volunteer. Samples were obtained 2, 6 and 3 h after administration of 250 mg of chlorpropamide, 90 mg of gliclazide and 4 mg of glimepiride, respectively. These analyses were performed in duplicate. 2.3. Stock solution Stock solutions of CL, GZ, GB and GM at the concentration of 0.5 mg mL−1 were individually prepared by dissolving 12.5 mg of
243
each active pharmaceutical ingredient in acetonitrile and water 4:1 (v/v). Then, 1 mL of each stock solution was combined and diluted in acetonitrile and water 1:1 (v/v) to obtain a working solution at 100 g mL−1 , which was used to spike blank human plasma at the concentration of 4 g mL−1 . All stock solutions were stored at 4 ◦ C for one month. At the validation procedure, stock solutions of CL, GZ, GB (IS) and GM, at the concentrations of 5 mg mL−1 , 3.2 mg mL−1 , 1.25 mg mL−1 and 0.5 mg mL−1 respectively, were prepared by dissolving appropriate amounts of each analytical standard in acetonitrile and water 4:1 (v/v). Appropriate volumes of stock solutions were individually diluted in acetonitrile and water 1:1 (v/v) to get intermediate solutions: 1000 g mL−1 of CL, 320 g mL−1 of GZ and 41.8 g mL−1 of GM. Stock and intermediate solutions were used to prepare working solutions. Stock and intermediate solutions were stored for at least 40 days at 4 ◦ C. The IS intermediated solution at 225 g mL−1
Fig. 2. Pareto chart acquired from FFD 24−1 plan. (A) Sample volume, (B) Sample dilution, (C) pH of buffer used at sample dilution, (D) Number of aspiration/ejection cycles through the device.
244
I.M.d.O. Viana et al. / Journal of Pharmaceutical and Biomedical Analysis 96 (2014) 241–248
(obtained from the IS stock solution) was stored at 4 ◦ C for 12 days. Working solutions were always prepared in the same day of the analysis.
Table 1 LLOQ, low QC, medium QC, high QC and dilution QC concentration for the three tested analytes. Level
2.4. Apparatus and chromatographic conditions Chromatography was carried out using an HPLC system (Agilent 1200 Infinity Quaternary LC system) coupled with a diode-array detector (Agilent 1200 Infinity, G DAD). The chromatographic separation of the antidiabetic drugs was accomplished in a Kinetex C18 fused-core column (100 mm × 4.6 mm, 2.6 m) connected to a KrudKatcher ULTRA HPLC in-line filter (0.5 m depth filter × 0.004 in ID) from Phenomenex (Torrance, USA) kindly donated by Allchrom (São Paulo, Brazil). The experiments were performed at room temperature with a mobile phase consisting of acetonitrile: potassium phosphate solution 10 mM pH 3.0 (60:40) at a flow-rate of 1.2 mL min−1 . The injection volume was 20 L and detection was at 230 nm. The mobile phase was filtered at 0.45 m membrane and degassed prior to use. 2.5. Sample preparation Before being used for the first time, the MEPS sorbent was manually conditioned with 250 L of methanol followed by 250 L of water. After that, the spiked plasma samples (containing 4 g mL−1 of each oral antidiabetic drug) were diluted. Dilution conditions (type and pH buffer, and plasma-buffer rate) were defined after experimental design, which will be described hereafter. Then, 100 L of this sample was manually drawn through the sorbent and ejected in the same vial. The sorbent was washed four times with 100 L of water to remove interferences. The analyte was then desorbed with the eluent. A chemometric approach was employed for optimization of MEPS parameters. The MEPS extraction and desorption conditions were optimized separately. Firstly, a fractional factorial design (FFD) 24 − 1 was used for optimizing the variables which could affect the extraction step. Different plasma volumes (100, 250 and 400 L), sample dilution (1:1, 1:2 and 1:3), buffer solutions pH (citrate buffer solution pH 2.5, citrate buffer solution pH 6.0 and carbonate/bicarbonate buffer solution pH 9.5) and number of aspiration/ejection cycles through the device (5, 10 and 15) were evaluated. The FFD involved eight experiments without replication and three experiments at the central point, carried out randomly. Afterwards, desorption variables were optimized using a complete factorial design (CFD). The evaluated factors and levels were percentage of acetonitrile in the elution solvent (50 and 100% of acetonitrile) and number of aspirations of elution solvent through the device (2 and 4). The CFD was performed randomly and with one replication of each experiment. Additionally, different plasma samples volumes (100, 250 and 400 L) were assessed in order to improve the method sensitivity. 2.6. Method validation The developed method was validated according to Bioanalytical Validation Guidance RDC 27/2012 from Brazilian Health Surveillance Agency [31]. All experiments were carried out with drug-free plasma samples spiked with standard solutions of the drugs containing concentrations that included the therapeutic plasma levels. The selectivity was ensured by individual analyses of blank matrix obtained from six sources (four normal, one hemolyzed and one lipemic plasma samples) and comparing these results with the lower limit of quantification (LLOQ) in the same matrix. Carry-over was assessed by three injections of the same blank sample. One injection was made before and two injections were made after the injection of an upper limit of quantification (ULOQ).
LLOQ Low QC Medium QC High QC Dilution QCa
Analyte concentration (g mL−1 ) CL
GZ
GM
1.00 3.00 21.25 42.50 75.00
1.00 3.00 4.89 8.50 15.00
0.10 0.30 0.49 0.85 1.50
a Dilution QC was defined as 1.5 times higher than upper limit of quantification (ULOQ). After dilution with blank plasma its concentration was the same of ULOQ.
Linearity was evaluated using three independent calibration curves prepared in different days. Calibration curves consisted of a blank sample (matrix sample processed without internal standard), a zero sample (matrix sample processed with internal standard) and 6 concentration levels (n = 3) of plasma matrix spiked with antidiabetic drugs. Calibration curves were constructed in the following concentrations: 1.0, 2.0, 10.0, 20.0, 35.0 and 50.0 g mL−1 for CL, 1.0, 2.0, 4.0, 6.0, 8.0 and 10.0 g mL−1 for GZ and 0.1, 0.2, 0.4, 0.6, 0.8 and 1.0 g mL−1 for GM. GB (IS) was used at the concentration of 10.7 g mL−1 . LLOQ was defined as the lowest analyte concentration that could be measured with a signal-to-noise ratio of 10. Intra and inter-day precision and accuracy were determined by analysis of samples (n = 5) containing known amounts of the drugs at 5 concentration levels (LLOQ, low quality control – QC, medium QC, high QC and dilution QC) in three days, according to Table 1. The concentration of dilution QC sample was defined so that after its dilution (with blank plasma) the obtained concentration was the same of the ULOQ (50.0 g mL−1 for CL, 10.0 g mL−1 for GZ and 1.0 g mL−1 for GM). The recovery of the analytes and IS was performed by comparing the analytical results for extracted samples at three concentrations (low, medium and high) with extracted blank matrix spiked with standards at the same concentration levels that represents 100% recovery. Stability of CL, GZ and GM in human plasma was evaluated under different circumstances: simulating the handling, after long-term storage at −80 ◦ C and short-term storage at room temperature, and after freeze and thaw 12 hours’ cycles. Low and high QC samples were analyzed in replicate (n = 3). Stability of CL, GZ, GM and IS stock solutions were evaluated by comparing the results between a new solution and a solution stored by 40 days (12 days for IS) at 4 ◦ C. 3. 3 Results and discussion 3.1. Chromatographic conditions A fused-core column, which could provide an efficient and fast separation, was employed in the proposed method. This column is more sensitive to band broadening mainly when conventional HPLC instruments are employed [32]. Therefore, some chromatographic parameters (PEEK tubing diameter and length, injection volume, data acquisition frequency and response time of the detector) were optimized to reduce peak broadening and increase efficiency. As expected, PEEK tubing with smaller diameter and length (extra-column volume of 43.98 L before and 26.06 L after optimization), as well as reduced injection volume (5 L instead of 10 L), decreased extra-column dispersion. Also, higher frequency of acquisition and lower response time (10 Hz/0.5 s) reduced band broadening. After instrumental optimization extracolumn variance changed from 86.6 L2 to 22.1 L2 by using uracil
I.M.d.O. Viana et al. / Journal of Pharmaceutical and Biomedical Analysis 96 (2014) 241–248
245
at 5 g mL−1 , acetonitrile and phosphate 10 mM pH 3.0 solution (55:45) as mobile phase, flow-rate of 0.1 mL min−1 and detector set at 254 nm. Then, for antidiabetic drugs separation, additional tests were performed. Different ratios of acetonitrile and phosphate 10 mM pH 3.0 (this pH was chosen since most of the analytes were in the nonionized form) solution were tested and better resolution was found when 60:40 was employed. Flow-rate was 1.2 mL min−1 . Detector was set at 230 nm, since it is the wavelength where the analytes have the maximum absorbance. The injection volume was set at 20 L to improve sensitivity. After optimization, the chromatographic run was only 2.2 min with adequate resolution between all the analytes.
3.2. MEPS optimization
Fig. 3. Fitted surface acquired from CFD 22 plan for chlorpropamide, which is representative of the other analytes.
Recovery was the variable used as response in the experimental design. All tested parameters were statistically significant (with confidence limit of 95%), but pH and sample volume were by far the most important parameters (as observed in the Pareto chart at Fig. 2). The results demonstrated that there is a tendency of higher recovery when pH was 2.5, number of aspiration/ejection cycles was 15, sample dilution was 1:1 and sample volume was 100 L. As the tested analytes have pKa around 5.0 (CL 5.0, GZ 5.8, GB 5.3 [33], GM 5.1 [34]) almost all molecules are in the non-ionized form at pH 2.5. The best number of cycles was 15 since the more often the analytes come in contact with the sorbent the higher is the extraction. A lower dilution (1:1) means higher concentration which leads to increased recovery. Recovery was better when sample volume was lower, possibly because in this case the total volume will pass more times through the sorbent. So, these levels were fixed, except dilution and number of cycles. Sample dilution was changed to 1:2 because 1:1 generated a high pressure in the MEPS device, which can reduce its life cycle. Also, in order to simplify the extraction process 10 cycles of aspiration/ejection was chosen instead of 15 cycles. The experimental design for the desorption step showed better recoveries at higher percentage of acetonitrile as elution solvent and higher number of aspirations of elution solvent through
the device. However, according to the surface plot for chlorpropamide (Fig. 3), with 70% of acetonitrile and three cycles of aspiration/ejection of 50 L a high recovery is also observed. The same behavior was found for gliclazide, glimepiride and glibenclamide. The small desorption volume (150 L instead of 200 L) allowed to concentrate the analytes on samples and increase the response. Thus, these conditions were chosen for further tests. Based on the experimental design, 100 L of plasma sample was the condition that showed higher recovery. However, since the objective of this method is to quantify CL, GZ and GM in the levels usually found in patients’ samples (maximum concentration of CL, GZ and GM is 30–250 g mL−1 , 0.7–4.9 g mL−1 and 0.308 g mL−1 , respectively [33]), 400 L of plasma sample had to be used in order to achieve the required sensitivity. Thus, the final extraction method was as follows: human plasma samples spiked with the IS working solution (400:20 v/v) were vortexed-mixed for 30 s and centrifuged at 480 × g for 10 min at 4 ◦ C. The supernatants were filtered at 0.45 m membrane and transferred to a clean tube. Then, 420 L of plasma containing IS was pipetted and diluted with 780 L of pH 2.5 citrate buffer (1:2). The sample with buffer was vortexed for 1 min before the extraction step. After that, the
Fig. 4. Chromatogram of blank plasma and spiked plasma sample at LLOQ showing method selectivity.
246
I.M.d.O. Viana et al. / Journal of Pharmaceutical and Biomedical Analysis 96 (2014) 241–248
Fig. 5. Ultraviolet spectra of spiked plasma samples containing CL, GZ, GM and GB at LLOQ and standard solutions at CQB.
Table 2 Parameters used to evaluate linearity. Analyte
Day of analysisa
Linearity range (g mL−1 )
Slope (a)
Intercept (b)
r2
CL
First Second Third First Second Third First Second Third
1.0–50.0
0.0675 0.0726 0.0740 0.0647 0.0670 0.0667 0.0811 0.0832 0.0838
0.0119 −0.0128 0.0193 0.0056 0.0052 −0.0010 0.0044 −0.0024 −0.0031
0.9994 0.9980 0.9973 0.9973 0.9936 0.9985 0.9957 0.9903 0.9953
GZ
GM
a
1.0–10.0
0.1–1.0
One independent curve was prepared each day of analysis.
Fig. 6. Chromatograms of samples from three volunteers who received single dose of chlorpropamide 250 mg, gliclazide 90 mg or glimepiride 4 mg.
I.M.d.O. Viana et al. / Journal of Pharmaceutical and Biomedical Analysis 96 (2014) 241–248
sample was manually drawn in the MEPS sorbent (10 cycles of 100 L). Then, the sorbent was washed with water (4 × 100 L) to remove interferences and the analytes were desorbed with 150 L of 70% acetonitrile in water (3 × 50 L). Before being re-used, the sorbent was washed/reconditioned with 4 × 100 L of pure acetonitrile followed by 4 × 100 L of purified water.
Table 3 Precision and accuracy assays. Analyte
3.3.2. Carry-over All injections of the blank plasma, before and after the injection of spiked plasma samples at ULOQ, were similar. The responses of interfering peaks were lower than 20% of the responses for spiked samples at LLOQ for CL, GZ and GM, and lower than 5% for the IS. 3.3.3. Linearity All calibration curves (Y = aX + b) obtained in human plasma for CL, GZ and GM by plotting concentration (X, g mL−1 ) versus drug-internal standard peak area ratio (Y), were linear over the range tested for each analyte (Table 2), once the determination coefficients (r2 ) were higher than 0.9903. According to the current legislation, calibration standards are accepted when the deviation related to nominal concentration are up to 20% for LLOQ and 15% for the other levels. At least 75% of the calibration standards should follow these criteria. The results obtained for the analytes were satisfactory with the proposed method. 3.3.4. Precision and accuracy Table 3 shows the precision and accuracy results. All values were lower than the limit of 20% for LLOQ and 15% for QC levels. 3.3.5. Recovery The recovery results are described in Table 4. Kim et al. compared different sample preparation techniques to analyze glimepiride in plasma samples. According to these authors recovery at 200 g mL−1 was 77.1% and 85.4% when LLE and SPE were used, respectively. The obtained result for glimepiride at 300 g mL−1 (low QC) described in this study was 71.9%, close to that for LLE and slightly lower than that for SPE. In other studies, recoveries found for gliclazide and chlorpropamide were 86.5% and 95.8%, respectively [35,36]. Although the results found in this study were lower than those reported in other studies, the recovery was satisfactory for the purposes of this method. It should be highlighted that higher recovery was obtained when 100 L, instead of 400 L, was used, as already mentioned in the MEPS optimization. Fig. 3 shows that recoveries close to 100% were observed. However, in order to improve sensitivity, 400 L was chosen. 3.3.6. Stability All results obtained for low QC and high QC were within the limits; therefore, analytes were stable in the tested conditions. Further,
Precision (%RSD)a
Level
c
Accuracy (%RSE)b d
Intra-day
Inter-day
Intra-dayc
Inter-dayd
CL
LLOQ Low QC Medium QC High QC Dilution QC
3.8 5.5 1.0 1.0 4.9
11.3 9.4 5.6 3.8 8.4
−17.8 4.5 −6.1 −6.3 −9.1
8.5 −9.5 3.4 6.3 1.2
GZ
LLOQ Low QC Medium QC High QC Dilution QC
1.9 1.4 0.8 1.2 7.7
10.0 6.3 8.5 4.4 9.0
−5.5 −14.4 −2.7 −2.8 −0.4
0.8 6.9 0.6 4.5 −4.4
GM
LLOQ Low QC Medium QC High QC Dilution QC
5.1 2.2 1.8 1.0 3.0
7.6 5.5 8.2 5.4 6.7
−16.1 −8.3 −8.4 −3.3 10.2
12.3 5.7 3.2 4.2 −5.9
3.3. Method validation 3.3.1. Selectivity Interfering peaks at the same retention times of those of the compounds of interest (CL, GZ and GM) showed response (peak area) lower than 20.0%, when compared with results of the analytes at LLOQ. Also, interfering peaks next to the IS retention time showed response lower than 5%. Fig. 4 shows typical chromatograms obtained from blank and spiked plasma samples. Aiming to identify the peaks of CL, GZ, GM and GB, spectra in the UV region of the spiked sample at LLOQ were obtained and compared with spectra of standard solutions at CQB (Fig. 5). The similarity between sample and standard spectra allowed drugs identification and confirmed method selectivity.
247
a
RSD = relative standard deviation. RSE = relative standard error. c Each value is the mean of five independent assays. d Inter-day precision and accuracy were obtained by combining data from three different days (n = 15). b
Table 4 Recovery obtained for CL, GZ, GM and GB. Analyte
Amount added (g mL−1 )
Extraction yield (%)a
RSD (%)
CL
Low QC Medium QC High QC
45.8 37.2 46.7
0.02 0.07 0.44
GZ
Low QC Medium QC High QC
59.4 50.2 52.6
0.31 0.18 0.59
GM
Low QC Medium QC High QC –
71.9 50.6 49.3 50.4
3.25 0.15 0.56 0.33
GB (IS) a
Each value is the mean of three injections of the same sample.
stock and intermediate solutions showed to be stable for at least 40 days at 4 ◦ C. 3.4. Analysis of samples from volunteers After being validated, the method was applied for determination of CL, GZ and GM in samples from nine volunteers, in duplicate. The mean plasma concentration and their respective standard deviations are shown on Table 5. The mean plasmatic concentration obtained with the three volunteers for chlorpropamide, gliclazide and glimepiride were 27.289 g mL−1 (from 22.141 to 30.850 g mL−1 ), 3.846 g mL−1 (from 2.675 to 6.093 g mL−1 ) and Table 5 Plasma concentration from volunteers obtained with the developed method. Drug
Volunteer
Mean plasma concentration (n = 2) (g mL−1 )
Standard deviation between replicates
Chlorpropamide
1 2 3
30.850 22.141 28.874
1.548 0.035 0.119
Gliclazide
4 5 6
2.675 6.093 2.770
0.107 0.443 0.002
Glimepiride
7 8 9
0.331 0.332 0.425
0.003 0.019 0.101
248
I.M.d.O. Viana et al. / Journal of Pharmaceutical and Biomedical Analysis 96 (2014) 241–248
0.363 g mL−1 (from 0.331 to 0.425 g mL−1 ), respectively. The mean plasma levels are in accordance with those already described in the literature [17,37,38]. Also, the appropriate standard deviation observed between replicates confirms the method applicability. Fig. 6 shows three chromatograms obtained with samples from volunteers containing CL, GZ and GM. 4. 4 Conclusions The experimental design used in the extraction and desorption step allowed to rationally choose the better levels for each parameter and evaluate their possible interactions. MEPS showed to be an adequate technique for sample preparation of antidiabetic drugs. Compared to other sample preparation techniques, small volumes of samples and organic solvents can be employed. Moreover, the developed method enabled the concentration of the antidiabetic drugs in the sample matrix without the need of an additional step, such as solvent evaporation, which is commonly used. Also, MEPS conjugates extraction and clean-up in only one step. The use of a fused core column allowed the development of a fast chromatographic run with adequate separation of the four tested drugs in less than 2.2 min. The developed method was fully validated and showed to be suitable to be applied in plasma samples from volunteers. Acknowledgments The authors are grateful to Pró-Reitoria de Pesquisa da UFMG and CAPES for the financial support and Dra. Luci Maria Sant’ Ana Dusse (Departamento de Análises Clínicas da UFMG) for collaboration in the collection of plasma samples. Also, the authors would like to thank Farmácia Artesanal, Laboratório Cifarma and Fundac¸ão Ezequiel Dias for donating the raw materials. References [1] World Health Organization, Fact sheet n. 312, 2013. http://www.who.int/mediacentre/factsheets/fs312/en/index.html (Accessed: 12/10/2013). [2] H.E. Lebovitz, Type 2 diabetes mellitus-current therapies and the emergence of surgical options, Nat. Rev. Endocrinol. 7 (2011) 408–419. [3] Brazil, Ministry of Health, Cadernos de Atenc¸ão Básica no. 36: Estratégias para o cuidado da pessoa com doenc¸a crônica – Diabetes mellitus, Brasília, 2013. [4] ALAD Guide diagnosis, control and treatment of type 2 diabetes mellitus, Pan American Health Organization, 2008. [5] J.S.L. Laurence, L. Brunton, Keith L. Parker, Goodman & Gilman’s the pharmacological basis of therapeutics, eleventh ed., McGraw-Hill, New York, 2006. [6] R.E. Majors, An overview of sample preparation, LC-GC 9 (1991) 16–19. [7] C. Fernandes, A.J. Santos-Neto, J.C. Rodrigues, C. Alves, F.M. Lancas, Solid-phase microextraction-liquid chromatography (SPME-LC) determination of fluoxetine and norfluoxetine in plasma using a heated liquid flow through interface, J. Chromatogr. B 847 (2007) 217–223. [8] P.L. Kole, G. Venkatesh, J. Kotecha, R. Sheshala, Recent advances in sample preparation techniques for effective bioanalytical methods, Biomed. Chromatogr. 25 (2011) 199–217. [9] N.M. El-Enany, A.A. Abdelal, F.F. Belal, Y.I. Itoh, M.N. Nakamura, Development and validation of a repharsed phase- HPLC method for simultaneous determination of rosiglitazone and glimepiride in combined dosage forms and human plasma, Chem. Cent. J. 6 (2012) 9. [10] U. Danlami, M.T. Odunola, G. Magaji, S.A. Thomas, The effect of chloroquine on the pharmacokinetics of chlorpropamide in human volunteers, Afr. J. Pharm. Pharmacol. 5 (2011) 1682–1686. [11] K.S. Lakshmi, T. Rajesh, Development and validation of rp-hplc method for simultaneous determination of glipizide, rosiglitazone, pioglitazone, glibenclamide and glimepiride in pharmaceutical dosage forms and human plasma, J. Iran. Chem. Soc. 8 (2011) 31–37. [12] X. Hu, Y. Zheng, J. Sun, L. Shang, G. Wang, H. Zhang, Simultaneous quantification of benazepril, gliclazide and valsartan in human plasma by LC-MS-MS and application for rapidly measuring protein binding interaction between rhein and these three drugs, Chromatographia 69 (2009) 843–852. [13] S.M. Foroutan, A. Zarghi, A. Shafaati, A. Khoddam, Application of monolithic column in quantification of gliclazide in human plasma by liquid chromatography, J. Pharm. Biomed. Anal. 42 (2006) 513–516.
[14] P. Sengupta, U. Bhaumik, A. Ghosh, A.K. Sarkar, B. Chatterjee, A. Bose, T.K. Pal, LC–MS-MS development and validation for simultaneous quantitation of metformin, glimepiride and pioglitazone in human plasma and its application to a bioequivalence study, Chromatographia 69 (2009) 1243–1250. [15] M.T. Bakare-Odunola, I. Enemali, M. Garba, O.O. Obodozie, K.B. Mustapha, The influence of lamivudine, stavudine and nevirapine on the pharmacokinetics of chlorpropamide in human subjects, Eur. J. Drug Metab. Pharmacokinet. 33 (2008) 165–171. [16] L. Chakradhar, R. Kallem, A. Karthik, B.T. Sundari, S. Ramesh, R. Mullangi, N.R. Srinivas, A rapid and highly sensitive method for the determination of glimepiride in human plasma by liquid chromatography-electrospray ionization tandem mass spectrometry: application to a pre-clinical pharmacokinetic study, Biomed. Chromatogr. 22 (2008) 58–63. [17] M.T.B. Odunola, I.S. Enemali, M. Garba, O.O. Obodozie, Rapid high performance liquid chromatographic determination of chlorpropamide in human plasma, Afr. J. Biotechnol. 6 (2007) 1378–1381. [18] G.X. Ling, J. Sun, J.L. Tang, X.D. Xu, Y.H. Sun, Z.G. He, Liquid chromatographyelectrospray ionization mass spectrometric method for determination of gliclazide in human plasma, Anal. Lett. 39 (2006) 1381–1391. [19] H. Kim, K.Y. Chang, C.H. Park, M.S. Jang, J.A. Lee, H.J. Lee, K.R. Lee, Determination of glimepiride in human plasma by LC-MS–MS and comparison of sample preparation methods for glimepiride, Chromatographia 60 (2004) 93–98. [20] M.L. Kundlik, B.H. Zaware, S.R. Kuchekar, Rapid, specific approach for direct measurement of glimepiride in human plasma by LC-ESI-MS–MS employing automated 96 well format: application to a bioequivalence study, J. Chromatogr. Sci. 50 (2012) 64–70. [21] Z. Altun, L.G. Blomberg, E. Jagerdeo, M. Abdel-Rehim, Drug screening using microextraction in a packed syringe (MEPS)/mass spectrometry utilizing monolithic-, polymer-, and silica-based sorbents, J.Liq. Chromatogr. Relat. Technol. 29 (2006) 829–840. [22] M. Abdel-Rehim, New trend in sample preparation: on-line microextraction in packed syringe for liquid and gas chromatography applications – I. Determination of local anaesthetics in human plasma samples using gas chromatography-mass spectrometry, J. Chromatogr. B 801 (2004) 317–321. [23] L. Somaini, M.A. Saracino, C. Marcheselli, S. Zanchini, G. Gerra, M.A. Raggi, Combined liquid chromatography-coulometric detection and microextraction by packed sorbent for the plasma analysis of long acting opioids in heroin addicted patients, Anal. Chim. Acta 702 (2011) 280–287. [24] H. Vlckova, M. Rabatinova, A. Miksova, G. Kolouchova, S. Micuda, P. Solich, L. Novakova, Determination of pravastatin and pravastatin lactone in rat plasma and urine using UHPLC-MS/MS and microextraction by packed sorbent, Talanta 90 (2012) 22–29. [25] M. Rodrigues, G. Alves, M. Rocha, J. Queiroz, A. Falcao, First liquid chromatographic method for the simultaneous determination of amiodarone and desethylamiodarone in human plasma using microextraction by packed sorbent (MEPS) as sample preparation procedure, J. Chromatogr. B 913 (2013) 90–97. [26] R. Mandrioli, L. Mercolini, D. Lateana, G. Boncompagni, M.A. Raggi, Analysis of risperidone and 9-hydroxyrisperidone in human plasma, urine and saliva by MEPS-LC-UV, J. Chromatogr. B 879 (2011) 167–173. [27] R. Said, Z. Hassan, M. Hassan, M. Abdel-Rehim, Rapid and sensitive method for determination of cyclophosphamide in patients plasma samples utilizing microextraction by packed sorbent online with liquid chromatography-tandem mass spectrometry (MEPS-LC–MS/MS), J. Liq. Chromatogr. Relat. Technol. 31 (2008) 683–694. [28] A.R. Chaves, F.Z. Leandro, J.A. Carris, M.E.C. Queiroz, Microextraction in packed sorbent for analysis of antidepressants in human plasma by liquid chromatography and spectrophotometric detection, J. Chromatogr. B 878 (2010) 2123–2129. [29] M.A. Saracino, K. Tallarico, M.A. Raggi, Liquid chromatographic analysis of oxcarbazepine and its metabolites in plasma and saliva after a novel microextraction by packed sorbent procedure, Anal. Chim. Acta 661 (2010) 222–228. [30] M. Abdel-Rehim, Recent advances in microextraction by packed sorbent for bioanalysis, J. Chromatogr. A 1217 (2010) 2569–2580. [31] BRAZIL, Brazilian Health Surveillance Agency, Bioanalytical Validation Guidance, RDC 27/2012, Brasília, 2012. [32] J.J. Salisbury, Fused-core particles: a practical alternative to sub-2 micron particles, J. Chromatogr. Sci. 46 (2008) 883–886. [33] A.C. Moffat, M.D. Osselton, B. Widdop, Clarke’s Analysis of Drugs and Poisons, third ed., Pharmaceutical Press, London, 2004. [34] American Chemical Society, SciFinder, 2014. https://scifinder.cas.org (Accessed: 01/14/2014). [35] J.Y. Park, K.A. Kim, S.L. Kim, P.W. Park, Quantification of gliclazide by semi-micro high-performance liquid chromatography: application to a bioequivalence study of two formulations in healthy subjects, J. Pharm Biomed Anal. 35 (2004) 943–949. [36] M.T. Bakare, A. Mustapha, I. Abduaguye, An improved high-performance liquid chromatographic determination of chlorpropamide in human plasma, Chromatographia 39 (1994) 107–109. [37] J.K. McGavin, C.M. Perry, K.L. Goa, Gliclazide modified release, Drugs 62 (2002) 1357–1364. [38] C. Pistos, M. Koutsopoulou, I. Panderi, Improved liquid chromatographic tandem mass spectrometric determination and pharmacokinetic study of glimepiride in human plasma, Biomed. Chromatogr. 19 (2005) 394–401.
Contents lists available at ScienceDirect
Journal of Pharmaceutical and Biomedical Analysis journal homepage: www.elsevier.com/locate/jpba
Simultaneous determination of oral antidiabetic drugs in human plasma using microextraction by packed sorbent and high-performance liquid chromatography Iara Maíra de Oliveira Viana, Paula de Paula Rosa Lima, Cristina Duarte Vianna Soares, Christian Fernandes ∗ Departamento de Produtos Farmacêuticos, Faculdade de Farmácia, Universidade Federal de Minas Gerais, Av. Pres. Antônio Carlos 6627, 31270-901 Belo Horizonte, MG, Brazil
a r t i c l e
i n f o
Article history: Received 23 December 2013 Received in revised form 10 March 2014 Accepted 25 March 2014 Available online 2 April 2014 Keywords: Oral antidiabetic drugs Microextraction by packed sorbent Experimental design Fused-cored column Human plasma
a b s t r a c t In this study, a simple method using microextraction by packed sorbent and high-performance liquid chromatography with ultraviolet detection for simultaneous determination of chlorpropamide, gliclazide and glimepiride in human plasma was developed and validated. A fractional factorial design and a complete factorial design were applied to evaluate the parameters which could affect the extraction and desorption steps, respectively. All parameters in the extraction step (pH, sample volume, sample dilution and number of aspiration/ejection cycles) and in the desorption step (percentage of acetonitrile in the elution solvent and number of aspirations of elution solvent through the device) were statistically significant (p > 0.05) when recovery was used as response. The developed method allowed the use of small volumes of sample and solvents and rapid separation by using a fused core column (only 2.2 min were needed). This method was fully validated showing selectivity, precision, accuracy and linearity over the range 1.0–50.0 g mL−1 for chlorpropamide, 1.0–10.0 g mL−1 for gliclazide and 0.1–1.0 g mL−1 for glimepiride. Finally, the validated method was applied in the analysis of samples from volunteers containing the three tested analytes. © 2014 Elsevier B.V. All rights reserved.
1. 1 Introduction Diabetes mellitus is a chronic disease that occurs when pancreas does not produce enough insulin, or when cells do not respond to the insulin that is produced. Hyperglycemia, which could cause serious damage to many body’s systems, is a common effect of uncontrolled diabetes. According to World Health Organization, approximately 347 million people have diabetes mellitus worldwide and it will probably be the seventh leading cause of death in 2030 [1]. The non-pharmacological diabetes treatment – healthy diet and regular physical activity – can prevent or delay the onset of type 2 diabetes mellitus (T2DM) [1]. However, lifestyle modification alone is usually not able to achieve appropriate glycemic levels. Patients
∗ Corresponding author at: Departamento de Produtos Farmacêuticos, Faculdade de Farmácia, Universidade Federal de Minas Gerais, Av. Presidente Antônio Carlos, 6627 Campus Pampulha, 31270-901 Belo Horizonte, MG, Brazil. Tel.: +55 31 34096957; fax: +55 31 34096976. E-mail address: [email protected] (C. Fernandes). http://dx.doi.org/10.1016/j.jpba.2014.03.042 0731-7085/© 2014 Elsevier B.V. All rights reserved.
with T2DM are usually treated with oral medication or even insulin. Often the use of a single drug is unlikely to result in long-term glycemic control and patients require multiple agents with different mechanisms of action [2]. The sulfonylureas are insulin secretagog agents which are commonly prescribed with metformin when the first-line treatment (metformin monotherapy) failed [2,3]. Beyond that, the sulfonylureas become the first line option for people with normal weight or when metformin should not be taken [4]. Also, they must be the first choice when the patient has weight loss or too high blood glucose [3]. There are two groups of sulfonylureas, the first and second generation, but all drugs consist of substituted arylsulfonylureas (Fig. 1). The first generation group includes chlorpropamide (CL), tolazamide, acetohexamide and tolbutamide; while the second one contains glibenclamide (GB) (also known as glyburide), gliclazide (GZ), glimepiride (GM) and glipizide [5]. Determination of drugs in biological fluid is essential in therapeutical drug monitoring, pharmaceutical research, analytical and toxicological studies. However, conventional sample preparation methods are usually laborious, time-consuming and are
242
I.M.d.O. Viana et al. / Journal of Pharmaceutical and Biomedical Analysis 96 (2014) 241–248
Fig. 1. Chemical structures of the sulfolnylureas (A – chlorpropamide; B – gliclazide; C – glimepiride; D – glibenclamide).
responsible for at least one-third of the error generated during the performance of an analytical method. Moreover, they employ large sample volumes and toxic solvents [6–8]. Until now, several analytical methods have been described for the determination of CL, GZ or GM in human plasma samples. However, from our knowledge, in all of them conventional techniques, such as protein precipitation (PP) [9–13], liquid–liquid extraction (LLE) [14–18] and solid-phase extraction (SPE) [19,20], were employed. Microextraction by packed sorbent (MEPS) is a new miniaturized solid-phase extraction technique that can reduce the volume of solvent and sample needed. Furthermore, it is easy to use and rapid. The sorbent (1–4 mg) is inserted into the syringe (100–250 L) barrel as a plug or between the needle and the barrel as a cartridge [21]. It is a very versatile technique, since this cartridge can be packed with any solid-phase material such as silica-based (C2, C8, C18), restricted access material (RAM) or molecularly imprinted polymers (MIPs) and can handle small sample volumes (10 L of plasma, urine or water) as well as large volumes (1000 L). Compared with PP, LLE and SPE, MEPS can reduce sample preparation time and organic solvent consumption [22], as have already been reported for determination of other drugs in biological samples [23–29]. One of the most advantages of this technique is that the sorbent could be used several times; more than 100 times with plasma or urine samples, while conventional SPE cartridges is used only once. The washing step after extraction allows and increases reuse. Moreover, using appropriate pretreatment of the sample before the MEPS method, blockage and coagulation are avoidable. Commonly, pretreatment includes centrifugation to remove suspended materials, pH adjustment, hydrolysis or precipitation [30]. In this context, the aim of this study was to develop a modern and fast method to simultaneously determine CL, GZ and GM in human plasma. Despite the fact that these three drugs are not usually co-administered in the clinical practice, the proposed method allows the determination of CL, GZ and GM together or alone, avoiding the need for developing three independent methods. MEPS was used as sample preparation technique. Experimental design was employed to optimize the parameters that affect the extraction. Liquid chromatography was performed by using a fused core column, which allowed rapid and efficient separation. The developed
method was validated according to Brazilian legislation and applied to the analysis of samples from volunteers. 2. 2 Materials and methods 2.1. Chemicals, reagents and materials CL batch 1019, GZ batch 1057 and GB batch 1018 (used as internal standard – IS) analytical standards were purchased from Instituto Nacional de Controle de Qualidade em Saúde (Rio de Janeiro, Brazil). GM batch G0K135 analytical standard was obtained from United States Pharmacopeia (Rockville, USA). GB by Cadila Pharmaceuticals (Ahmedabad, Gujarat, India), GM by Mantena Laboratories (Nalgonda Dist, Andhra Pradesh, India), GZ by Shandong Keyuan Pharmaceutical (Jinan, Shandong, China) and CL by Kothari Phytochemicals International (Nagari, India) active pharmaceutical ingredients were used in the optimization step. Three different drug products were obtained in the local market: tablets containing 250 mg of chlorpropamide, 30 mg of gliclazide and 4 mg of glimepiride. Acetonitrile and methanol HPLC grade and sodium citrate dehydrate analytical grade were obtained from J.T. Baker (Xalostoc, Mexico). Phosphoric acid 85% (w/w) and sodium hydroxide were acquired from Merck (Darmstadt, Germany). Potassium phosphate monobasic was supplied by Sigma-Aldrich (São Paulo, Brazil); citric acid monohydrate and sodium carbonate anhydrous by Labsynth (Diadema, Brazil); and sodium bicarbonate by Vetec (Rio de Janeiro, Brazil). The water used for preparing all solutions and samples was purified in a Direct-Q 3 System from Millipore (Bedford, MA, USA). Extraction was performed using the MEPS 250 L syringe and the MEPS BIN (barrel insert and needle) containing 4 mg of silica C18 sorbent (SGE Analytical Science; Melbourne, Australia). 2.2. Sample collection Blood samples were collected from volunteers at Laboratório de Hematologia da Faculdade de Farmácia da Universidade Federal de Minas Gerais (UFMG; Belo Horizonte, Brazil). This study was approved by UFMG Ethics Committee. The blood samples were usually drawn in the morning from fasting volunteers and stored in glass tubes containing heparin as the anticoagulant until been
I.M.d.O. Viana et al. / Journal of Pharmaceutical and Biomedical Analysis 96 (2014) 241–248
centrifuged at 480 × g for 10 min at 4 ◦ C. Then, the supernatant (plasma) was stored at −80 ◦ C. Drug-free plasma samples (blank plasma) were spiked with analytes before being used for MEPS optimization and method validation. After, the validated method was applied in samples from nine different volunteers (three volunteers for each drug) who received single dose of antidiabetics orally. Only one drug was administered to each volunteer. Samples were obtained 2, 6 and 3 h after administration of 250 mg of chlorpropamide, 90 mg of gliclazide and 4 mg of glimepiride, respectively. These analyses were performed in duplicate. 2.3. Stock solution Stock solutions of CL, GZ, GB and GM at the concentration of 0.5 mg mL−1 were individually prepared by dissolving 12.5 mg of
243
each active pharmaceutical ingredient in acetonitrile and water 4:1 (v/v). Then, 1 mL of each stock solution was combined and diluted in acetonitrile and water 1:1 (v/v) to obtain a working solution at 100 g mL−1 , which was used to spike blank human plasma at the concentration of 4 g mL−1 . All stock solutions were stored at 4 ◦ C for one month. At the validation procedure, stock solutions of CL, GZ, GB (IS) and GM, at the concentrations of 5 mg mL−1 , 3.2 mg mL−1 , 1.25 mg mL−1 and 0.5 mg mL−1 respectively, were prepared by dissolving appropriate amounts of each analytical standard in acetonitrile and water 4:1 (v/v). Appropriate volumes of stock solutions were individually diluted in acetonitrile and water 1:1 (v/v) to get intermediate solutions: 1000 g mL−1 of CL, 320 g mL−1 of GZ and 41.8 g mL−1 of GM. Stock and intermediate solutions were used to prepare working solutions. Stock and intermediate solutions were stored for at least 40 days at 4 ◦ C. The IS intermediated solution at 225 g mL−1
Fig. 2. Pareto chart acquired from FFD 24−1 plan. (A) Sample volume, (B) Sample dilution, (C) pH of buffer used at sample dilution, (D) Number of aspiration/ejection cycles through the device.
244
I.M.d.O. Viana et al. / Journal of Pharmaceutical and Biomedical Analysis 96 (2014) 241–248
(obtained from the IS stock solution) was stored at 4 ◦ C for 12 days. Working solutions were always prepared in the same day of the analysis.
Table 1 LLOQ, low QC, medium QC, high QC and dilution QC concentration for the three tested analytes. Level
2.4. Apparatus and chromatographic conditions Chromatography was carried out using an HPLC system (Agilent 1200 Infinity Quaternary LC system) coupled with a diode-array detector (Agilent 1200 Infinity, G DAD). The chromatographic separation of the antidiabetic drugs was accomplished in a Kinetex C18 fused-core column (100 mm × 4.6 mm, 2.6 m) connected to a KrudKatcher ULTRA HPLC in-line filter (0.5 m depth filter × 0.004 in ID) from Phenomenex (Torrance, USA) kindly donated by Allchrom (São Paulo, Brazil). The experiments were performed at room temperature with a mobile phase consisting of acetonitrile: potassium phosphate solution 10 mM pH 3.0 (60:40) at a flow-rate of 1.2 mL min−1 . The injection volume was 20 L and detection was at 230 nm. The mobile phase was filtered at 0.45 m membrane and degassed prior to use. 2.5. Sample preparation Before being used for the first time, the MEPS sorbent was manually conditioned with 250 L of methanol followed by 250 L of water. After that, the spiked plasma samples (containing 4 g mL−1 of each oral antidiabetic drug) were diluted. Dilution conditions (type and pH buffer, and plasma-buffer rate) were defined after experimental design, which will be described hereafter. Then, 100 L of this sample was manually drawn through the sorbent and ejected in the same vial. The sorbent was washed four times with 100 L of water to remove interferences. The analyte was then desorbed with the eluent. A chemometric approach was employed for optimization of MEPS parameters. The MEPS extraction and desorption conditions were optimized separately. Firstly, a fractional factorial design (FFD) 24 − 1 was used for optimizing the variables which could affect the extraction step. Different plasma volumes (100, 250 and 400 L), sample dilution (1:1, 1:2 and 1:3), buffer solutions pH (citrate buffer solution pH 2.5, citrate buffer solution pH 6.0 and carbonate/bicarbonate buffer solution pH 9.5) and number of aspiration/ejection cycles through the device (5, 10 and 15) were evaluated. The FFD involved eight experiments without replication and three experiments at the central point, carried out randomly. Afterwards, desorption variables were optimized using a complete factorial design (CFD). The evaluated factors and levels were percentage of acetonitrile in the elution solvent (50 and 100% of acetonitrile) and number of aspirations of elution solvent through the device (2 and 4). The CFD was performed randomly and with one replication of each experiment. Additionally, different plasma samples volumes (100, 250 and 400 L) were assessed in order to improve the method sensitivity. 2.6. Method validation The developed method was validated according to Bioanalytical Validation Guidance RDC 27/2012 from Brazilian Health Surveillance Agency [31]. All experiments were carried out with drug-free plasma samples spiked with standard solutions of the drugs containing concentrations that included the therapeutic plasma levels. The selectivity was ensured by individual analyses of blank matrix obtained from six sources (four normal, one hemolyzed and one lipemic plasma samples) and comparing these results with the lower limit of quantification (LLOQ) in the same matrix. Carry-over was assessed by three injections of the same blank sample. One injection was made before and two injections were made after the injection of an upper limit of quantification (ULOQ).
LLOQ Low QC Medium QC High QC Dilution QCa
Analyte concentration (g mL−1 ) CL
GZ
GM
1.00 3.00 21.25 42.50 75.00
1.00 3.00 4.89 8.50 15.00
0.10 0.30 0.49 0.85 1.50
a Dilution QC was defined as 1.5 times higher than upper limit of quantification (ULOQ). After dilution with blank plasma its concentration was the same of ULOQ.
Linearity was evaluated using three independent calibration curves prepared in different days. Calibration curves consisted of a blank sample (matrix sample processed without internal standard), a zero sample (matrix sample processed with internal standard) and 6 concentration levels (n = 3) of plasma matrix spiked with antidiabetic drugs. Calibration curves were constructed in the following concentrations: 1.0, 2.0, 10.0, 20.0, 35.0 and 50.0 g mL−1 for CL, 1.0, 2.0, 4.0, 6.0, 8.0 and 10.0 g mL−1 for GZ and 0.1, 0.2, 0.4, 0.6, 0.8 and 1.0 g mL−1 for GM. GB (IS) was used at the concentration of 10.7 g mL−1 . LLOQ was defined as the lowest analyte concentration that could be measured with a signal-to-noise ratio of 10. Intra and inter-day precision and accuracy were determined by analysis of samples (n = 5) containing known amounts of the drugs at 5 concentration levels (LLOQ, low quality control – QC, medium QC, high QC and dilution QC) in three days, according to Table 1. The concentration of dilution QC sample was defined so that after its dilution (with blank plasma) the obtained concentration was the same of the ULOQ (50.0 g mL−1 for CL, 10.0 g mL−1 for GZ and 1.0 g mL−1 for GM). The recovery of the analytes and IS was performed by comparing the analytical results for extracted samples at three concentrations (low, medium and high) with extracted blank matrix spiked with standards at the same concentration levels that represents 100% recovery. Stability of CL, GZ and GM in human plasma was evaluated under different circumstances: simulating the handling, after long-term storage at −80 ◦ C and short-term storage at room temperature, and after freeze and thaw 12 hours’ cycles. Low and high QC samples were analyzed in replicate (n = 3). Stability of CL, GZ, GM and IS stock solutions were evaluated by comparing the results between a new solution and a solution stored by 40 days (12 days for IS) at 4 ◦ C. 3. 3 Results and discussion 3.1. Chromatographic conditions A fused-core column, which could provide an efficient and fast separation, was employed in the proposed method. This column is more sensitive to band broadening mainly when conventional HPLC instruments are employed [32]. Therefore, some chromatographic parameters (PEEK tubing diameter and length, injection volume, data acquisition frequency and response time of the detector) were optimized to reduce peak broadening and increase efficiency. As expected, PEEK tubing with smaller diameter and length (extra-column volume of 43.98 L before and 26.06 L after optimization), as well as reduced injection volume (5 L instead of 10 L), decreased extra-column dispersion. Also, higher frequency of acquisition and lower response time (10 Hz/0.5 s) reduced band broadening. After instrumental optimization extracolumn variance changed from 86.6 L2 to 22.1 L2 by using uracil
I.M.d.O. Viana et al. / Journal of Pharmaceutical and Biomedical Analysis 96 (2014) 241–248
245
at 5 g mL−1 , acetonitrile and phosphate 10 mM pH 3.0 solution (55:45) as mobile phase, flow-rate of 0.1 mL min−1 and detector set at 254 nm. Then, for antidiabetic drugs separation, additional tests were performed. Different ratios of acetonitrile and phosphate 10 mM pH 3.0 (this pH was chosen since most of the analytes were in the nonionized form) solution were tested and better resolution was found when 60:40 was employed. Flow-rate was 1.2 mL min−1 . Detector was set at 230 nm, since it is the wavelength where the analytes have the maximum absorbance. The injection volume was set at 20 L to improve sensitivity. After optimization, the chromatographic run was only 2.2 min with adequate resolution between all the analytes.
3.2. MEPS optimization
Fig. 3. Fitted surface acquired from CFD 22 plan for chlorpropamide, which is representative of the other analytes.
Recovery was the variable used as response in the experimental design. All tested parameters were statistically significant (with confidence limit of 95%), but pH and sample volume were by far the most important parameters (as observed in the Pareto chart at Fig. 2). The results demonstrated that there is a tendency of higher recovery when pH was 2.5, number of aspiration/ejection cycles was 15, sample dilution was 1:1 and sample volume was 100 L. As the tested analytes have pKa around 5.0 (CL 5.0, GZ 5.8, GB 5.3 [33], GM 5.1 [34]) almost all molecules are in the non-ionized form at pH 2.5. The best number of cycles was 15 since the more often the analytes come in contact with the sorbent the higher is the extraction. A lower dilution (1:1) means higher concentration which leads to increased recovery. Recovery was better when sample volume was lower, possibly because in this case the total volume will pass more times through the sorbent. So, these levels were fixed, except dilution and number of cycles. Sample dilution was changed to 1:2 because 1:1 generated a high pressure in the MEPS device, which can reduce its life cycle. Also, in order to simplify the extraction process 10 cycles of aspiration/ejection was chosen instead of 15 cycles. The experimental design for the desorption step showed better recoveries at higher percentage of acetonitrile as elution solvent and higher number of aspirations of elution solvent through
the device. However, according to the surface plot for chlorpropamide (Fig. 3), with 70% of acetonitrile and three cycles of aspiration/ejection of 50 L a high recovery is also observed. The same behavior was found for gliclazide, glimepiride and glibenclamide. The small desorption volume (150 L instead of 200 L) allowed to concentrate the analytes on samples and increase the response. Thus, these conditions were chosen for further tests. Based on the experimental design, 100 L of plasma sample was the condition that showed higher recovery. However, since the objective of this method is to quantify CL, GZ and GM in the levels usually found in patients’ samples (maximum concentration of CL, GZ and GM is 30–250 g mL−1 , 0.7–4.9 g mL−1 and 0.308 g mL−1 , respectively [33]), 400 L of plasma sample had to be used in order to achieve the required sensitivity. Thus, the final extraction method was as follows: human plasma samples spiked with the IS working solution (400:20 v/v) were vortexed-mixed for 30 s and centrifuged at 480 × g for 10 min at 4 ◦ C. The supernatants were filtered at 0.45 m membrane and transferred to a clean tube. Then, 420 L of plasma containing IS was pipetted and diluted with 780 L of pH 2.5 citrate buffer (1:2). The sample with buffer was vortexed for 1 min before the extraction step. After that, the
Fig. 4. Chromatogram of blank plasma and spiked plasma sample at LLOQ showing method selectivity.
246
I.M.d.O. Viana et al. / Journal of Pharmaceutical and Biomedical Analysis 96 (2014) 241–248
Fig. 5. Ultraviolet spectra of spiked plasma samples containing CL, GZ, GM and GB at LLOQ and standard solutions at CQB.
Table 2 Parameters used to evaluate linearity. Analyte
Day of analysisa
Linearity range (g mL−1 )
Slope (a)
Intercept (b)
r2
CL
First Second Third First Second Third First Second Third
1.0–50.0
0.0675 0.0726 0.0740 0.0647 0.0670 0.0667 0.0811 0.0832 0.0838
0.0119 −0.0128 0.0193 0.0056 0.0052 −0.0010 0.0044 −0.0024 −0.0031
0.9994 0.9980 0.9973 0.9973 0.9936 0.9985 0.9957 0.9903 0.9953
GZ
GM
a
1.0–10.0
0.1–1.0
One independent curve was prepared each day of analysis.
Fig. 6. Chromatograms of samples from three volunteers who received single dose of chlorpropamide 250 mg, gliclazide 90 mg or glimepiride 4 mg.
I.M.d.O. Viana et al. / Journal of Pharmaceutical and Biomedical Analysis 96 (2014) 241–248
sample was manually drawn in the MEPS sorbent (10 cycles of 100 L). Then, the sorbent was washed with water (4 × 100 L) to remove interferences and the analytes were desorbed with 150 L of 70% acetonitrile in water (3 × 50 L). Before being re-used, the sorbent was washed/reconditioned with 4 × 100 L of pure acetonitrile followed by 4 × 100 L of purified water.
Table 3 Precision and accuracy assays. Analyte
3.3.2. Carry-over All injections of the blank plasma, before and after the injection of spiked plasma samples at ULOQ, were similar. The responses of interfering peaks were lower than 20% of the responses for spiked samples at LLOQ for CL, GZ and GM, and lower than 5% for the IS. 3.3.3. Linearity All calibration curves (Y = aX + b) obtained in human plasma for CL, GZ and GM by plotting concentration (X, g mL−1 ) versus drug-internal standard peak area ratio (Y), were linear over the range tested for each analyte (Table 2), once the determination coefficients (r2 ) were higher than 0.9903. According to the current legislation, calibration standards are accepted when the deviation related to nominal concentration are up to 20% for LLOQ and 15% for the other levels. At least 75% of the calibration standards should follow these criteria. The results obtained for the analytes were satisfactory with the proposed method. 3.3.4. Precision and accuracy Table 3 shows the precision and accuracy results. All values were lower than the limit of 20% for LLOQ and 15% for QC levels. 3.3.5. Recovery The recovery results are described in Table 4. Kim et al. compared different sample preparation techniques to analyze glimepiride in plasma samples. According to these authors recovery at 200 g mL−1 was 77.1% and 85.4% when LLE and SPE were used, respectively. The obtained result for glimepiride at 300 g mL−1 (low QC) described in this study was 71.9%, close to that for LLE and slightly lower than that for SPE. In other studies, recoveries found for gliclazide and chlorpropamide were 86.5% and 95.8%, respectively [35,36]. Although the results found in this study were lower than those reported in other studies, the recovery was satisfactory for the purposes of this method. It should be highlighted that higher recovery was obtained when 100 L, instead of 400 L, was used, as already mentioned in the MEPS optimization. Fig. 3 shows that recoveries close to 100% were observed. However, in order to improve sensitivity, 400 L was chosen. 3.3.6. Stability All results obtained for low QC and high QC were within the limits; therefore, analytes were stable in the tested conditions. Further,
Precision (%RSD)a
Level
c
Accuracy (%RSE)b d
Intra-day
Inter-day
Intra-dayc
Inter-dayd
CL
LLOQ Low QC Medium QC High QC Dilution QC
3.8 5.5 1.0 1.0 4.9
11.3 9.4 5.6 3.8 8.4
−17.8 4.5 −6.1 −6.3 −9.1
8.5 −9.5 3.4 6.3 1.2
GZ
LLOQ Low QC Medium QC High QC Dilution QC
1.9 1.4 0.8 1.2 7.7
10.0 6.3 8.5 4.4 9.0
−5.5 −14.4 −2.7 −2.8 −0.4
0.8 6.9 0.6 4.5 −4.4
GM
LLOQ Low QC Medium QC High QC Dilution QC
5.1 2.2 1.8 1.0 3.0
7.6 5.5 8.2 5.4 6.7
−16.1 −8.3 −8.4 −3.3 10.2
12.3 5.7 3.2 4.2 −5.9
3.3. Method validation 3.3.1. Selectivity Interfering peaks at the same retention times of those of the compounds of interest (CL, GZ and GM) showed response (peak area) lower than 20.0%, when compared with results of the analytes at LLOQ. Also, interfering peaks next to the IS retention time showed response lower than 5%. Fig. 4 shows typical chromatograms obtained from blank and spiked plasma samples. Aiming to identify the peaks of CL, GZ, GM and GB, spectra in the UV region of the spiked sample at LLOQ were obtained and compared with spectra of standard solutions at CQB (Fig. 5). The similarity between sample and standard spectra allowed drugs identification and confirmed method selectivity.
247
a
RSD = relative standard deviation. RSE = relative standard error. c Each value is the mean of five independent assays. d Inter-day precision and accuracy were obtained by combining data from three different days (n = 15). b
Table 4 Recovery obtained for CL, GZ, GM and GB. Analyte
Amount added (g mL−1 )
Extraction yield (%)a
RSD (%)
CL
Low QC Medium QC High QC
45.8 37.2 46.7
0.02 0.07 0.44
GZ
Low QC Medium QC High QC
59.4 50.2 52.6
0.31 0.18 0.59
GM
Low QC Medium QC High QC –
71.9 50.6 49.3 50.4
3.25 0.15 0.56 0.33
GB (IS) a
Each value is the mean of three injections of the same sample.
stock and intermediate solutions showed to be stable for at least 40 days at 4 ◦ C. 3.4. Analysis of samples from volunteers After being validated, the method was applied for determination of CL, GZ and GM in samples from nine volunteers, in duplicate. The mean plasma concentration and their respective standard deviations are shown on Table 5. The mean plasmatic concentration obtained with the three volunteers for chlorpropamide, gliclazide and glimepiride were 27.289 g mL−1 (from 22.141 to 30.850 g mL−1 ), 3.846 g mL−1 (from 2.675 to 6.093 g mL−1 ) and Table 5 Plasma concentration from volunteers obtained with the developed method. Drug
Volunteer
Mean plasma concentration (n = 2) (g mL−1 )
Standard deviation between replicates
Chlorpropamide
1 2 3
30.850 22.141 28.874
1.548 0.035 0.119
Gliclazide
4 5 6
2.675 6.093 2.770
0.107 0.443 0.002
Glimepiride
7 8 9
0.331 0.332 0.425
0.003 0.019 0.101
248
I.M.d.O. Viana et al. / Journal of Pharmaceutical and Biomedical Analysis 96 (2014) 241–248
0.363 g mL−1 (from 0.331 to 0.425 g mL−1 ), respectively. The mean plasma levels are in accordance with those already described in the literature [17,37,38]. Also, the appropriate standard deviation observed between replicates confirms the method applicability. Fig. 6 shows three chromatograms obtained with samples from volunteers containing CL, GZ and GM. 4. 4 Conclusions The experimental design used in the extraction and desorption step allowed to rationally choose the better levels for each parameter and evaluate their possible interactions. MEPS showed to be an adequate technique for sample preparation of antidiabetic drugs. Compared to other sample preparation techniques, small volumes of samples and organic solvents can be employed. Moreover, the developed method enabled the concentration of the antidiabetic drugs in the sample matrix without the need of an additional step, such as solvent evaporation, which is commonly used. Also, MEPS conjugates extraction and clean-up in only one step. The use of a fused core column allowed the development of a fast chromatographic run with adequate separation of the four tested drugs in less than 2.2 min. The developed method was fully validated and showed to be suitable to be applied in plasma samples from volunteers. Acknowledgments The authors are grateful to Pró-Reitoria de Pesquisa da UFMG and CAPES for the financial support and Dra. Luci Maria Sant’ Ana Dusse (Departamento de Análises Clínicas da UFMG) for collaboration in the collection of plasma samples. Also, the authors would like to thank Farmácia Artesanal, Laboratório Cifarma and Fundac¸ão Ezequiel Dias for donating the raw materials. References [1] World Health Organization, Fact sheet n. 312, 2013. http://www.who.int/mediacentre/factsheets/fs312/en/index.html (Accessed: 12/10/2013). [2] H.E. Lebovitz, Type 2 diabetes mellitus-current therapies and the emergence of surgical options, Nat. Rev. Endocrinol. 7 (2011) 408–419. [3] Brazil, Ministry of Health, Cadernos de Atenc¸ão Básica no. 36: Estratégias para o cuidado da pessoa com doenc¸a crônica – Diabetes mellitus, Brasília, 2013. [4] ALAD Guide diagnosis, control and treatment of type 2 diabetes mellitus, Pan American Health Organization, 2008. [5] J.S.L. Laurence, L. Brunton, Keith L. Parker, Goodman & Gilman’s the pharmacological basis of therapeutics, eleventh ed., McGraw-Hill, New York, 2006. [6] R.E. Majors, An overview of sample preparation, LC-GC 9 (1991) 16–19. [7] C. Fernandes, A.J. Santos-Neto, J.C. Rodrigues, C. Alves, F.M. Lancas, Solid-phase microextraction-liquid chromatography (SPME-LC) determination of fluoxetine and norfluoxetine in plasma using a heated liquid flow through interface, J. Chromatogr. B 847 (2007) 217–223. [8] P.L. Kole, G. Venkatesh, J. Kotecha, R. Sheshala, Recent advances in sample preparation techniques for effective bioanalytical methods, Biomed. Chromatogr. 25 (2011) 199–217. [9] N.M. El-Enany, A.A. Abdelal, F.F. Belal, Y.I. Itoh, M.N. Nakamura, Development and validation of a repharsed phase- HPLC method for simultaneous determination of rosiglitazone and glimepiride in combined dosage forms and human plasma, Chem. Cent. J. 6 (2012) 9. [10] U. Danlami, M.T. Odunola, G. Magaji, S.A. Thomas, The effect of chloroquine on the pharmacokinetics of chlorpropamide in human volunteers, Afr. J. Pharm. Pharmacol. 5 (2011) 1682–1686. [11] K.S. Lakshmi, T. Rajesh, Development and validation of rp-hplc method for simultaneous determination of glipizide, rosiglitazone, pioglitazone, glibenclamide and glimepiride in pharmaceutical dosage forms and human plasma, J. Iran. Chem. Soc. 8 (2011) 31–37. [12] X. Hu, Y. Zheng, J. Sun, L. Shang, G. Wang, H. Zhang, Simultaneous quantification of benazepril, gliclazide and valsartan in human plasma by LC-MS-MS and application for rapidly measuring protein binding interaction between rhein and these three drugs, Chromatographia 69 (2009) 843–852. [13] S.M. Foroutan, A. Zarghi, A. Shafaati, A. Khoddam, Application of monolithic column in quantification of gliclazide in human plasma by liquid chromatography, J. Pharm. Biomed. Anal. 42 (2006) 513–516.
[14] P. Sengupta, U. Bhaumik, A. Ghosh, A.K. Sarkar, B. Chatterjee, A. Bose, T.K. Pal, LC–MS-MS development and validation for simultaneous quantitation of metformin, glimepiride and pioglitazone in human plasma and its application to a bioequivalence study, Chromatographia 69 (2009) 1243–1250. [15] M.T. Bakare-Odunola, I. Enemali, M. Garba, O.O. Obodozie, K.B. Mustapha, The influence of lamivudine, stavudine and nevirapine on the pharmacokinetics of chlorpropamide in human subjects, Eur. J. Drug Metab. Pharmacokinet. 33 (2008) 165–171. [16] L. Chakradhar, R. Kallem, A. Karthik, B.T. Sundari, S. Ramesh, R. Mullangi, N.R. Srinivas, A rapid and highly sensitive method for the determination of glimepiride in human plasma by liquid chromatography-electrospray ionization tandem mass spectrometry: application to a pre-clinical pharmacokinetic study, Biomed. Chromatogr. 22 (2008) 58–63. [17] M.T.B. Odunola, I.S. Enemali, M. Garba, O.O. Obodozie, Rapid high performance liquid chromatographic determination of chlorpropamide in human plasma, Afr. J. Biotechnol. 6 (2007) 1378–1381. [18] G.X. Ling, J. Sun, J.L. Tang, X.D. Xu, Y.H. Sun, Z.G. He, Liquid chromatographyelectrospray ionization mass spectrometric method for determination of gliclazide in human plasma, Anal. Lett. 39 (2006) 1381–1391. [19] H. Kim, K.Y. Chang, C.H. Park, M.S. Jang, J.A. Lee, H.J. Lee, K.R. Lee, Determination of glimepiride in human plasma by LC-MS–MS and comparison of sample preparation methods for glimepiride, Chromatographia 60 (2004) 93–98. [20] M.L. Kundlik, B.H. Zaware, S.R. Kuchekar, Rapid, specific approach for direct measurement of glimepiride in human plasma by LC-ESI-MS–MS employing automated 96 well format: application to a bioequivalence study, J. Chromatogr. Sci. 50 (2012) 64–70. [21] Z. Altun, L.G. Blomberg, E. Jagerdeo, M. Abdel-Rehim, Drug screening using microextraction in a packed syringe (MEPS)/mass spectrometry utilizing monolithic-, polymer-, and silica-based sorbents, J.Liq. Chromatogr. Relat. Technol. 29 (2006) 829–840. [22] M. Abdel-Rehim, New trend in sample preparation: on-line microextraction in packed syringe for liquid and gas chromatography applications – I. Determination of local anaesthetics in human plasma samples using gas chromatography-mass spectrometry, J. Chromatogr. B 801 (2004) 317–321. [23] L. Somaini, M.A. Saracino, C. Marcheselli, S. Zanchini, G. Gerra, M.A. Raggi, Combined liquid chromatography-coulometric detection and microextraction by packed sorbent for the plasma analysis of long acting opioids in heroin addicted patients, Anal. Chim. Acta 702 (2011) 280–287. [24] H. Vlckova, M. Rabatinova, A. Miksova, G. Kolouchova, S. Micuda, P. Solich, L. Novakova, Determination of pravastatin and pravastatin lactone in rat plasma and urine using UHPLC-MS/MS and microextraction by packed sorbent, Talanta 90 (2012) 22–29. [25] M. Rodrigues, G. Alves, M. Rocha, J. Queiroz, A. Falcao, First liquid chromatographic method for the simultaneous determination of amiodarone and desethylamiodarone in human plasma using microextraction by packed sorbent (MEPS) as sample preparation procedure, J. Chromatogr. B 913 (2013) 90–97. [26] R. Mandrioli, L. Mercolini, D. Lateana, G. Boncompagni, M.A. Raggi, Analysis of risperidone and 9-hydroxyrisperidone in human plasma, urine and saliva by MEPS-LC-UV, J. Chromatogr. B 879 (2011) 167–173. [27] R. Said, Z. Hassan, M. Hassan, M. Abdel-Rehim, Rapid and sensitive method for determination of cyclophosphamide in patients plasma samples utilizing microextraction by packed sorbent online with liquid chromatography-tandem mass spectrometry (MEPS-LC–MS/MS), J. Liq. Chromatogr. Relat. Technol. 31 (2008) 683–694. [28] A.R. Chaves, F.Z. Leandro, J.A. Carris, M.E.C. Queiroz, Microextraction in packed sorbent for analysis of antidepressants in human plasma by liquid chromatography and spectrophotometric detection, J. Chromatogr. B 878 (2010) 2123–2129. [29] M.A. Saracino, K. Tallarico, M.A. Raggi, Liquid chromatographic analysis of oxcarbazepine and its metabolites in plasma and saliva after a novel microextraction by packed sorbent procedure, Anal. Chim. Acta 661 (2010) 222–228. [30] M. Abdel-Rehim, Recent advances in microextraction by packed sorbent for bioanalysis, J. Chromatogr. A 1217 (2010) 2569–2580. [31] BRAZIL, Brazilian Health Surveillance Agency, Bioanalytical Validation Guidance, RDC 27/2012, Brasília, 2012. [32] J.J. Salisbury, Fused-core particles: a practical alternative to sub-2 micron particles, J. Chromatogr. Sci. 46 (2008) 883–886. [33] A.C. Moffat, M.D. Osselton, B. Widdop, Clarke’s Analysis of Drugs and Poisons, third ed., Pharmaceutical Press, London, 2004. [34] American Chemical Society, SciFinder, 2014. https://scifinder.cas.org (Accessed: 01/14/2014). [35] J.Y. Park, K.A. Kim, S.L. Kim, P.W. Park, Quantification of gliclazide by semi-micro high-performance liquid chromatography: application to a bioequivalence study of two formulations in healthy subjects, J. Pharm Biomed Anal. 35 (2004) 943–949. [36] M.T. Bakare, A. Mustapha, I. Abduaguye, An improved high-performance liquid chromatographic determination of chlorpropamide in human plasma, Chromatographia 39 (1994) 107–109. [37] J.K. McGavin, C.M. Perry, K.L. Goa, Gliclazide modified release, Drugs 62 (2002) 1357–1364. [38] C. Pistos, M. Koutsopoulou, I. Panderi, Improved liquid chromatographic tandem mass spectrometric determination and pharmacokinetic study of glimepiride in human plasma, Biomed. Chromatogr. 19 (2005) 394–401.
