An efficient hydrophilic interaction liquid chromatographic method for the simultaneous determination of metformin and pioglitazone using high-purity silica column

Journal of Chromatography B, 997 (2015) 16–22

Contents lists available at ScienceDirect

Journal of Chromatography B journal homepage: www.elsevier.com/locate/chromb

An efficient hydrophilic interaction liquid chromatographic method for the simultaneous determination of metformin and pioglitazone using high-purity silica column Abdel-Maaboud Ismail Mohamed a , Fardous Abdel-Fattah Mohamed a , Sameh Ahmed b,a,∗ , Yahya Abduh Salim Mohamed a a b

Department of Pharmaceutical Analytical Chemistry, Faculty of Pharmacy, Assiut University, Assiut 71526, Egypt Pharmacognosy and Pharmaceutical Chemistry Department, College of Pharmacy, Taibah University, Al Madinah AlMunawarah 30001, Saudi Arabia

a r t i c l e

i n f o

Article history: Received 5 February 2015 Received in revised form 21 May 2015 Accepted 25 May 2015 Available online 30 May 2015 Keywords: Pioglitazone HCl Metformin HCl Oral hypoglycemic Hydrophilic interaction liquid chromatography Binary mixture Pharmacokinetics

a b s t r a c t Hydrophilic interaction liquid chromatography (HILIC) provides a feasible approach to effectively separate polar compounds in complex matrices. Herein, a simple, reproducible and efficient HILIC method was developed for the simultaneous determination of pioglitazone. HCl (PIO) and metformin HCl (MET) in rabbit plasma. High-purity silica column was used for rapid and efficient separation of these co-administered drugs. The chromatographic parameters were optimized for best separation. The proposed HILIC system provides high separation efficiency with good peak shape compared to reversed phase (RP) chromatography. Additionally, a simple isocratic elution mode with a mobile phase composed of a mixture of methanol and 10 mM phosphate buffer (pH 3.0) (94:6, v/v) was used and the effluent was monitored at 230 nm. The method was validated in accordance with the requirements of US-FDA guidelines and was found to behave efficiently for the intended purpose. The correlation coefficient of 0.9992 was obtained in the concentration ranges of 0.5–100 ␮g mL−1 . The limits of detection (S/N = 3) and quantification (S/N = 10) were 0.16 and 0.5 ng mL−1 , respectively. The retention times were 3.4 and 5.0 min for PIO and MET, respectively. Plasma levels were successfully determined in rabbit with satisfactory precision and accuracy. In addition, the stability tests in rabbit plasma proved reliable stability under the experimental conditions. The developed HILIC method was applied successfully to study the pharmacokinetic behaviors of the studied analytes in rabbit plasma after a single oral dose containing PIO and MET. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Type 2 diabetes is characterized by defects in both insulin secretion and insulin sensitivity in adipose tissue, live and skeletal muscles [1,2]. Approximately, 50% of patients treated by monotherapy requires a combination therapy with two oral agents to achieve the target blood glucose level [3]. The combination of metformin HCl (MET) and insulin-sensitizing agent is commonly used in clinical practice. Recently, a new class of insulin-sensitizing agents, the thiazolidinediones, is introduced for the treatment of type 2 diabetic patients [4]. MET is chemically designated as N,N-dimethylimidodicarbonimidic diamide hydrochloride (Fig. 1). It is the most well known member of the biguanide group, regarded as the main compound in

∗ Corresponding author. Tel.: +966 54 3110057; fax: +966 4 8475027. E-mail address: [email protected] (S. Ahmed). http://dx.doi.org/10.1016/j.jchromb.2015.05.032 1570-0232/© 2015 Elsevier B.V. All rights reserved.

mixed therapies, and is always used in high doses of about 500–1000 mg. Pioglitazone HCl (PIO) is chemically designated as 5-[[4-[2- (5-Ethyl-2-pyridinyl) ethoxy]phenyl]methyl]-2,4thiazolidinedione hydrochloride (Fig. 1). It is a member of the thiazolidinedione group and is used along with MET in a dose of 15 mg per tablet [5]. The combination therapy of MET and PIO had beneficial effects on lipid levels, improved insulin sensitivity and also improved insulin secretion [6]. To study efficacy and safety of the studied oral hypoglycemic combination during the treatment regime, it is essential to employ a selective, efficient and fully validated methodology to suite the plasma determination of these oral hypoglycemic combinations in complex matrices. Although several analytical methods were reported for analysis of each component individually, few methods were reported for the simultaneous determination of MET and PIO in combination dosage forms. These methods include; spectroscopic and chemometric methods [7–8], thin layer chromatographic (TLC) methods [9] and high performance liquid

A.-M.I. Mohamed et al. / J. Chromatogr. B 997 (2015) 16–22

17

2. Experimental 2.1. Chemicals and reagents Pioglitazone HCl (PIO) was obtained from Dr. Reddy”s; European Egyptian Pharmaceuticals. Metformin HCl (MET) was obtained from EI-Nasr Pharmaceutical Chemical Co. (Abu-Zaabal, Egypt). ATVC was from (Eipico Pharmaceuticals Co., 10th of Ramadan City, Cairo, Egypt). It was used without further purification and certified to contain 98% for PIO and 99.96% for MET dry weight basis. Sodium dihydrogen orthophosphate monohydrate was obtained from EINasr Pharmaceutical Chemical Co. (Abu-Zaabal, Egypt). Sodium dihydrogen phosphate, sodium hydroxide, and phosphoric acid for pH adjustment were from (Cairo Pharmaceuticals Co., Cairo, Egypt). 10 mM Phosphate buffer pH 3 [26] was prepared by dissolving 0.3 g of sodium dihydrogen orthophosphate monohydrate in 250 mL of distilled water and adjusted to pH 3 with diluted phosphoric acid (10–12 drops of 85% orthophosphoric acid in 100 mL distilled water). Double distilled water was obtained through WSC-4D water purification system (Hamilton Laboratory Milton Glass Ltd., Kent, USA). All chemicals were of analytical or HPLC grade and were supplied from Sigma Aldrich (Seelze, Germany. Pharmaceutical formulations were purchased from the local market; Pioglumet® tablets (Elrazy Pharmaceuticals Co., Cairo, Egypt) labeled to contain 850 mg MET and 15 mg PIO. Blank rabbit plasma used in this study was supplied from Animal house, Assiut University (Assiut, Egypt), and they were stored in deep-freezer at −30 ◦ C until analysis. 2.2. Instrumentations and chromatographic conditions Fig. 1. Chemical structures of pioglitazone HCl (PIO), metformin HCl (MET), and atorvastatin calcium (IS).

chromatographic (HPLC) method [10–16]. On the other hand, the only report for their simultaneous determination in plasma sample was using LC–MS technique [17]. LC–MS technique is a relatively expensive instrument and requires technical experiences. Additionally, the bad resolutions obtained for some chromatographic methods and laborious extraction procedures limit the use of the others. The name of hydrophilic interaction liquid chromatography (HILIC) was suggested by Dr. Andrew Alpert in his 1990 paper [18]. Recently, there has been a significant increase of interest for HILIC methods. This is due to the growing need for the analysis of polar compounds that do not bind to reversed-phase (RP) materials and the constantly increasing complexity of samples. It was suggested that the mechanism of HILIC involves partitioning of separated substances between a hydrophobic mobile phase and a partially immobilized layer of water on the stationary phase [19]. Several recent review articles have been published on the topic of HILIC, all of which concluded that the mechanism is in fact multidimensional including additional components such as adsorption and electrostatic interactions [20–24]. MET is a highly polar compound while PIO has a low polarity. Using HILIC technique, the two compounds could be separated with good resolution maintaining good peak shape. The objective of this research was to develop a new, simple, reliable and efficient HILIC method for the simultaneous determination of PIO and MET in rabbit plasma samples to be convenient for pharmacokinetic studies. The proposed method used high-purity silica column to provide efficient separation. The method was optimized and validated in accordance with US-FDA guidelines [25]. Finally, this method was applied for the determination of PIO and MET in rabbit plasma using atorvastatin calcium (ATVC) as internal standard in a pharmacokinetic study and all the pharmacokinetic parameters were assessed.

The HPLC system used in this study was a Younglin Autochro3000HPLC system (Younglin, Korea) with UV detector, a Rheodyne injection valve with a 20 ␮L loop was used. Data acquisition and peak areas calculations were performed on YoungLin Autochro3000 chromatography software. Compounds were separated on Kromosil 100 5 SIL column (Kromosil® Sweden) (250mm × 4.6 mm, 5 ␮m i.d.) with the isocratic mobile phase of methanol and phosphate buffer (10 mM, pH 3.0) (94:6; v/v). The flow-rate was 1 mL min−1 and detection was adjusted at  = 230 nm. The column was maintained at ambient temperature (25 ± 2 ◦ C). The mobile phase was filtered and degassed by sonication before use. Kromasil 100 5C18 column (250mm × 4.6 mm, 5 ␮m i.d.) was used in comparison study. In addition, ultrasonic cleaner (Cole-Parmer, Chicago, USA), pH meter, model 3305 (Jenway, London, UK), Sartorius handy balance H51 (Hanover, Germany) and oil-less vacuum pump (Rocker, Tiwan) and laboratory centrifuge model 800 (Republic of China were used in this study. 2.3. Preparation of standard solutions Stock solutions of and metformin HCl (MET) and pioglitazone HCl (PIO) reference standards (250 ␮g mL−1 ) was prepared in methanol. The working standard solutions were prepared by further dilution of the stock solution with the same solvent to obtain concentrations ranging from 0.5 to 100 ␮g mL−1 . The stock and working standard solutions were kept in refrigerator at ± 4 ◦ C in light protected flasks. 2.4. Preparation of calibration and quality control samples Seven calibration samples were prepared by spiking appropriate amounts of the working solution of MET and PIO with concentration range of (0.5–100 ␮g mL−1 ) in blank plasma obtained from rabbits and spiked with amount of ATVC (I.S.) equivalent to 50 ␮g mL−1 from I.S. The concentration of the calibration samples in plasma were 0.5, 1, 2, 10, 25, 50 and 100 ␮g mL−1 . Quality control (QC)

18

A.-M.I. Mohamed et al. / J. Chromatogr. B 997 (2015) 16–22

samples in plasma were prepared by further dilution of spiked plasma samples with blank rabbit plasma to acquire plasma concentration of 0.5 ␮g mL−1 (LQC), 10 ␮g mL−1 (MQC), and 100 ␮g mL−1 (HQC) for MET and PIO. The spiked plasma samples and quality control samples were stored at −30 ◦ C in well-closed, light resistant containers prior to analysis. All plasma samples were thawed at room temperature immediately before assayed. 2.5. Sample preparation Into 2 mL eppendorf tubes, an aliquot of 200 ␮l blank rabbit plasma was spiked with of 200 ␮L of the standard PIO and MET in the concentration range (0.5–100 ␮g mL−1 ), 100 ␮L of internal standard ATVC (50 ␮g mL−1 ) then 500 ␮L of methanol was added. The contents of the tube were vortexed for 2.0 min and then centrifuged at 4000 × g r.p.m for 15.0 min at room temperature. The clear supernatant was taken and filtered using filter disc (0.22 ␮m) and 20 ␮L of the filtrate was injected onto HPLC system. The blank plasma sample was also analyzed in each run to confirm absence of interferences. The extraction residue was reconstituted in 200 ␮l of the mobile phase, then, injected into the HPLC system. 2.6. Validation of analytical method The developed HILIC method in plasma was validated in accordance with US-Food and drug administration (FDA) guidance for bioanalytical method validation [25]. For the statistical analysis, Excel 2003 (Microsoft Office) was used. The studied validation parameters were; selectivity, linearity, range, limit of detection (LOD), limit of quantitation (LOQ), precision, accuracy, robustness and stability study. 2.7. Application to pharmacokinetic study The pharmacokinetic study was performed in healthy female rabbits and the study proposal was approved by the Committee of Animal Care, Assiut University, Assiut, Egypt. A total number of five white female rabbits (Boskat breed) obtained from the animal house of Assiut University weighing from 1.5 to 2.5 kg were used as a model animal for determining the bioavailability. Animals were kept individually in cages with wire-mesh bases constructed of galvanized steel for two weeks before treatment for acclimatization in a room with controlled lighting (14 h/day), constant temperature (16–20 ◦ C) and relative humidity (55–65%). Sick, injured and abnormal animals were eliminated. Rabbits were fed a commercial pellet diet and water. Single oral doses of pioglumet® tablets were administered for each of 5 female rabbits labeled to contain 15 mg PIO and 850 mg MET. The blood samples were collected into sodium heparin-containing tubes predosing and at 0.5, 1.0–4.0 h postdosing. The samples were collected from retro-orbital plexuses according to the method of Shermer [27]; Transferred to heparinized tubes and centrifuged for 10 min at 4 ◦ C, 10,000 × g rpm, then the plasma was separated and stored at −30 ◦ C in dark until analysis. The pharmacokinetic parameters of PIO and MET were calculated by non-compartmental analysis model using the Moment analysis, Microsoft Excel 2003. The investigated pharmacokinetic parameters included; the maximum plasma concentration (Cmax ) and the maximum plasma concentration time (tmax ) were directly obtained from the raw data. The area under the plasma concentration-time curve up to the last measurable time point, AUC0−t and to infinity AUC0-∞) were obtained by the linear trapezoidal method. The AUC0−t was extrapolated to infinity (i.e. AUC0-∞ ) by adding the quotient of Clast /Kel , where Clast represents the last measurable concentration at last time of study (in this study 4 h) and Kel represents the apparent terminal rate constant. Kel was calculated by the linear regression of the

Fig. 2. Chromatograms for separation of binary mixture of PIO and MET using (A) HILIC and (B) RP-C18 chromatography.

log-transformed concentrations of the drug in the terminal phase. MRT represented the maximum residual time (the tendency of drugs and metabolites to remain in the body) and calculated using the formula MRT = AUMC/AUC. The half-time of drug elimination (t1/2 ) was calculated as 0.693/Kel . The clearance (CL) was calculated as dose/AUC. 3. Results and discussion 3.1. Optimization of chromatographic conditions The most important aspect of method development in liquid chromatography is the achievement of sufficient resolution within a reasonable analysis time. This goal can be achieved by adjusting different chromatographic factors to give a desired separation. The main analytical parameters to be optimized are stationary phase and mobile phase compositions (organic solvent, buffer concentration, buffer volume and pH). 3.1.1. Stationary phase The choice of the separation method depends on several factors such as the nature of the analytes, the complexity of the sample, and the intended use. It is important to study and understand the performance of the high-purity silica column. Its performance was compared with traditional C18 column for separation of the tested mixture (Fig. 2). The chromatographic parameters including; retention factors (k ), number of theoretical plates (N), peak resolution (Rs), and the peak asymmetry factor at 10% peak height (As0.1 ), were calculated for both HILIC and RP-C18 techniques for the separation of the studied drugs. It is clear from the results presented in Table 1 that the resolution, efficiency represented by number of theoretical plates (N) and peak symmetry of tested mixture using HILIC method is far better than traditional RP-C18 technique the tested antidiabetic drugs. Therefore, HILIC technique was selected in the proposed method as it guarantees efficient separation maintaining good peak shape. 3.1.2. Effect of organic solvent type in the mobile phase and its ratio The effect of the organic solvent in the mobile phase was tested by using acetonitrile, ethanol and methanol (Table 2). The best performance and efficiency of the column was obtained when using methanol. In addition, methanol gave the best separation and peak symmetry. Therefore; methanol was selected as the ideal organic solvent for the investigated oral hypoglycemic drugs. Moreover, the effect of different proportions of methanol on the chromatographic behavior of the eluted peaks was studied. It was observed that with small amounts of methanol in the mobile phase, the reten-

A.-M.I. Mohamed et al. / J. Chromatogr. B 997 (2015) 16–22

19

Table 1 Performances of HILIC and RP-C18 chromatographic methods for the separation of PIO and MET. Column

HILIC RP-C18

Rs

K

–

PIO

MET

PIO

MET

PIO

MET

PIO

MET

3.3 4.6

0.89 0.58

2.7 1.0

5476 160

5595 3136

45.6 1562.5

44.68 79.72

1.0 1.2

1.0 1.0

N

HETP

As0.1

Rs = Resolution, K  = Capacity factor, N = Number of theoretical plates, HETP = Height equivalent theoretical plate, As0.1 = Asymmetry factor at 10% peak height. Table 2 Effects of alterations of the organic solvents in the mobile phase on the efficiency and performances of HILIC method. Organic solvent

Ethanol Methanol Acetonitrile

Rs

K

–

PIO

MET

PIO

MET

PIO

MET

3.6 3.3 8.8

0.20 0.89 0.20

0.7 2.7 2.1

1310 5476 2153

218 5595 3418

1.2 1.0 1.3

1.2 1.0 1.5

N

As0.1

Fig. 3. Van Deemter plot for separation of tested oral hypoglycemic mixture by HILIC method. Conditions; Kromasil 100 5 SIL column, methanol: phosphate buffer pH 3.0 (94:6; v/v), detection at  = 230 nm.

tion values were extended and peak resolution was decreased for the investigated drugs. Since a short retention time is desirable provided that good resolution and peak symmetry was obtained, the proportion of 94% (v/v) of methanol was selected for the subsequent work. 3.1.3. Effect of pH variation and phosphate buffer concentration The effect of pH variation on the separation of PIO and MET was studied in the range from pH 2.5 to pH 4.0. The retention factors (k ) increased with increasing the buffer pH for the studied drugs. However, good retention values with acceptable peak shapes and high numbers of theoretical plates were obtained at buffer pH of 3.0; hence it was selected as the optimal pH. The effect of buffer concentration also has been tested in the range 5–20 mM. Results of N, Peak area and As0.1 were presented in Table 3. Phosphate buffer with concentration of 10 mM was chosen for further experiments to avoid peak asymmetry of the interested peaks in the presence of significant amounts of organic solvent and minimize the abrasive effect on pump seals [28]. 3.1.4. The effect of the flow rate on the column efficiency The influence of the mobile phase flow rate on the performance of the HILIC for the separation of tested mixture was investigated. The flow rate was varied in the range 0.6–1.4 mL min−1 . The Van Deemter plot (height equivalent to a theoretical plate (HETP) versus flow rate) of the tested mixture by HILIC method was shown in Fig. 3. As can be seen, the efficiency of the normal naked silica col-

Fig. 4. A typical chromatogram for the extract of pioglitazone PIO and MET spiked with ATVC as I.S in rabbit plasma analyzed by the developed HILIC method.

umn in HILIC method was better and low pack pressure at flow rate of 1.0 mL min−1 , so it was chosen for further experiments. 3.2. Internal standard selection Choosing a suitable internal standard (I.S) is an important aspect to obtain good accuracy as the interfering plasma components may lead to inaccuracy of analytical results. Several compounds such as gliclazide, glipizide, glibenclamide and ATVC were investigated to find a suitable internal standard. ATVC was selected as the internal standard in our study because it gives sharp and well resolved peak from the studied binary mixture as shown in Fig. 4 under the optimum conditions. 3.3. Extraction efficiency MET has a high polarity, so it was difficult to extract it from plasma using liquid–liquid extraction method. The solid phase extraction was proved to be feasible for both MET and PIO from plasma samples but the extraction procedures were tedious and time-consuming. To obtain high extraction efficiency, plasma protein precipitation was more suitable for the studied oral hypoglycemic and IS. Three different protein precipitation agents, acetonitrile, methanol and ammonium sulfate (salting out), were investigated. It was found that good resolution and separation when using methanol, so it was selected as a precipitating agent.

Table 3 Effect of alterations of buffer concentrations on the efficiency and performance of the HILIC method. Phosphate buffer strength

5 mM 10 mM 15 mM 20 mM

N

Peak area

As0.1

PIO

MET

PIO

MET

PIO

MET

3552 5476 1246 259

2172 5595 1590 309

2559.04 3765.35 2793.07 2184.36

2679.93 4820.79 4743.3 4299.74

1.0 1.0 1.0 1.5

1.2 1.0 1.0 1.1

20

A.-M.I. Mohamed et al. / J. Chromatogr. B 997 (2015) 16–22

Table 4 parameters and statistical Quantitative proposed HILIC method for the simultaneous determination of PIO and MET in rabbit plasma.

data

of

the

Drug

TR

Rangea (␮g mL−1 )

Slopeb ± SD

Interceptb ± SD

r

LOD (␮g mL−1 )

LOQ (␮g mL−1 )

PIO MET

3.4 5.0

0.5–100 0.5–100

52.63 ± 3.05 90.98 ± 6.92

122.49 ± 2.52 27.40 ± 3.11

0.9992 0.9999

0.16 0.16

0.5 0.5

a b

Peak area ratio of analyte/IS versus plasma concentration (␮g mL−1 ). Data presented as mean ± SD of three experiments.

Fig. 5. HPLC chromatograms of rabbit plasma collected predose (blank plasma) (A), after 1 h postdose (B) from administration of pioglumet® tablet (15 mg PIO and 850 mg MET) and spiked with ATVC (IS). Chromatographic conditions were the optimum conditions.

When a simple single-step methanol protein precipitation was adopted, high extraction efficiencies with good peak shapes were achieved for PIO, MET and ATVC (IS) without visible endogenous interference at the retention times. 3.4. Validation of the developed HILIC method In order to ensure that the developed HILIC method is valid for the determination of PIO and MET in plasma sample, validation

studies were carried out in accordance with US-FDA guidelines for bioanalytical method validation [25]. The selectivity of the proposed method was assessed by testing blank rabbit plasma from five different sources for the interferences in the retention times of PIO and MET. Chromatograms obtained from blank plasma demonstrated that there were no interferences from endogenous plasma components that emphasize the high selectivity of the proposed method.

A.-M.I. Mohamed et al. / J. Chromatogr. B 997 (2015) 16–22

21

Table 5 Intra-day and inter-day precision of the proposed HILIC method for simultaneous determination of PIO and MET in spiked rabbit plasma. Drug

Concentration (␮g mL−1 )

0.5 10 100 0.5 10 100

PIO

MET

a

Intra-day precision (n = 5) %Recoverya ± SD

RSD%

%Recoverya ± SD

RSD%

102.41 ± 3.65 100.24 ± 2.29 99.59 ± 1.89 97.36 ± 3.77 100.65 ± 1.69 100.61 ± 0.88

3.56 2.28 1.90 3.87 1.68 0.87

100.27 ± 3.28 99.96 ± 1.90 97.77 ± 1.35 98.66 ± 2.58 102.18 ± 2.63 99.60 ± 1.45

3.28 1.9 1.38 2.61 2.57 1.46

Average of five determinations.

Table 6 The accuracy of the proposed HILIC method for determination of PIO and MET in spiked rabbit plasma. Drug

Concentration (␮g mL−1 )

Recoverya % ± SD

RSD%

PIO

0.5 10 100 0.5 10 100

97.33 ± 3.08 99.25 ± 2.65 100.12 ± 1.96 98.56 ± 3.41 101.95 ± 2.77 99.95 ± 1.23

3.16 2.67 1.96 3.46 2.72 1.23

MET

a

Inter-day precision (n = 5)

Table 7 Robustness of the proposed HILIC method for simultaneous determination of PIO and MET. Parameters variations

Proposed procedures 1-pH (±3.0) 2.8 3.2 2-Methanol (94%) 93% 95% 3-Buffer conc.(10 mM) 7.5 mM 12.5 mM 4-Flow rate(1 mL min−1 ) 0.9 mL min−1 1.1 mL min−1 5-Detection wavelength (230 nm) 227 nm 233 nm

Average of five determinations.

Under the optimum chromatographic conditions for HILIC, the relationship between the plasma concentrations and peak areas ratio for the drug and ATVC (IS) were linear in the range (0.5–100 ␮g mL−1 ) for MET and PIO. Quantitative parameters and statistical data of the proposed HILIC method for quantitation of PIO and MET in rabbit plasma were summarized in Table 4. The correlation coefficients (r) for calibration curves were ranged from (0.9992–0.9999). The slopes of the calibration curves reflect the sensitivity of the developed HILIC method. Limits of detection (LOD) were 0.16 ␮g mL−1 and limits of quantitation (LOQ) were 0.5 for PIO and MET, respectively. Although the sensitivity was lower than the only report for their simultaneous determination in plasma sample by LC–MS [17], it satisfies the quantitation of the target analyte in plasma samples. Intra-day precision of the developed HILIC method was assessed by assay (n = 5) at low (LQC), medium (MQC), and high (HQC) concentration levels (0.5, 10 and 100 ␮g mL−1 ) for PIO and MET (Table 5). The inter-day precision was conducted by repeating the analysis over five working days. The overall precision of the method was expressed as relative standard deviations (RSD). RSD% values were ranged from 1.38 to 3.56% for PIO and 0.87–3.87 for MET indicating good repeatability and precision. The obtained precisions were satisfactory for quality control measurements. The accuracy of the proposed HILIC method in rabbit plasma was determined by investigating the recovery percentages of the studied drugs at three QC levels (0.5, 10, and 100 ␮g mL−1 ) (three replicates of each concentration). The results were shown in Table 6. The accuracy data revealed good accuracy and recovery percent-

a

%Recoverya ± SD PIO

MET

99.69 ± 1.73 101.70 ± 0.15 98.85 ± 1.24

101.58 ± 0.20 100.28 ± 0.75 100.64 ± 0.49

99.14 ± 0.01 99.89 ± 0.54

100.93 ± 0.29 100.56 ± 0.75

101.25 ± 0.30 101.37 ± 0.86

100.90 ± 1.12 99.24 ± 0.46

102.45 ± 0.91 101.51 ± 1.19

100.25 ± 1.15 99.65 ± 0.54

100.13 ± 0.96 99.33 ± 0.16

99.23 ± 0.88 101.54 ± 1.12

Average of five determinations.

ages ranging from 97.3 to 100.12% with RSD% 1.96–3.08 and from 98.56 to 101.95% with RSD% 1.23–3.41 for PIO and MET; respectively. Dilution integrity was conducted by spiking the blank rabbit plasma with PIo and MET (1 mg mL−1 ) and diluting this sample with blank plasma to 0.5, 10, and 100 ␮gmL−1 (n = 5). The% RSD for the obtained data was less than 4.9 and the% recoveries were 92.56–102.33%. The accuracy and precision of the developed IPC method was superior than most of the reported methods proving method reproducibility. The robustness is an important validation parameter as it measures method reliability. For the determination of the HILIC method’s robustness, several parameters were studied such as pH, mobile phase composition, methanol percentage, phosphate buffer concentration, detection and flow rate. Results were shown in Table 7. It was found that slight variation of these variables didn’t significantly affect the performance of the proposed HILIC method. So the proposed method could be considered robust and reliable during the normal usage.

Table 8 Stability of PIO and MET in spiked rabbit plasma analyzed by the developed HILIC method. Condition

Three times freeze thaw (−30 ◦ C) Room temperature (24 h) Refrigeration overnight(4 ◦ C) Freezer at −30 ◦ C for 1 month Data presented as%recovery ± SD (n = 5).

PIO

MET

LQC (0.5 ␮g mL−1 )

MQC (10 ␮g mL−1 )

HQC (100 ␮g mL−1 )

LQC (0.5 ␮g mL−1 )

MQC (10 ␮g mL−1 )

HQC (100 ␮g mL−1 )

92.3 ± 2.62 89.3 ± 3.25

94.6 ± 2.29 90.4 ± 2.12

95.6 ± 2.91 91.2 ± 3.12

93.5 ± 2.49 89.7 ± 2.75

94.9 ± 2.89 91.2 ± 2.72

94.2 ± 1.81 90.6 ± 2.42

99.6 ± 2.21 93.6 ± 2.98

98.9 ± 1.91 97.2 ± 2.38

98.5 ± 2.56 95.2 ± 2.88

99.1 ± 2.75 93.5 ± 2.18

97.9 ± 1.96 98.9 ± 2.73

99.5 ± 2.76 93.9 ± 1.54

22

A.-M.I. Mohamed et al. / J. Chromatogr. B 997 (2015) 16–22

Table 9 Pharmaco kinetic parameters of PIO and MET in the female rabbits after a single oral administration of pioglumet® tablet(15 PIO and 850 mg MET). Pharmacokinetic parameters

PIO

MET

Cmax (␮g mL−1 ) Tmax (h) t1/2 (h) MRT (h) Auc0→4 (␮g h. mL−1 ) Auc0→∞ (␮g h. mL−1 ) Kel CL(mL h−1 )

4.77 1.00 4.97 7.57 15.43 39.10 0.14 0.97

9.41 1.00 1.75 2.17 23.49 26.62 0.397 36.2

Fig. 6. Plasma concentration–time profiles of PIO and MET after oral administration of one pioglumet® tablet (15 mg PIO and 850 mg MET). Data are represented as mean ± S.D. (n = 5).

The stabilities of PIO and MET in spiked rabbit plasma were investigated at three different concentrations; low (LQC), medium (MQC), and high (HQC) concentration levels (0.5, 10 and 100 ␮g mL−1 ). The results of stability tests of RAN in plasma were summarized in Table 8. The obtained stabilities for PIO and MET were were ranged from 89.3 to 99.6%. The results showed that PIO and MET in QC plasma samples proved to be stable in sample preparation and after storage at room temperature for 24 h. The results of process stability of QC plasma QC demonstrated that the post extraction solution were stable at 4 ◦ C for 24 h. PIO and MET was also mild stable after three freeze–thaw cycles and after storage in freezer at −30 ◦ C for 1 month. The conducted stability tests indicated reliable stability under the tested conditions. 3.5. Application to pharmacokinetic studies In order to test the applicability of the developed HILIC method to measure PIO and MET in the course of pharmacokinetic studies, the proposed HILIC method was used for simultaneous determination of PIO and MET in rabbit plasma after a single administration of pioglumet® tablet containing a binary mixture (15 mg PIO and 850 mg MET). The smaller amount of plasma required for the determination enabled many sampling points. PIO and MET were determined in rabbit plasma predose and postdose up to 240 min. Typical HPLC chromatograms of rabbit plasma collected predose (blank plasma) and after 1 h (postdose) from administration of pioglumet® tablet were shown in Fig. 5. The pharmacokinetic parameters of PIO and MET were calculated by non-compartmental model using the Moment analysis, Microsoft Excel 2003. The investigated pharmacokinetic parameters were Cmax and Tmax that were directly obtained from the raw data. In addition, AUC0−t , AUC0−∞ , Kel, t1/2 and CL were calculated. The pharmacokinetic parameters of PIO and MET in the female rabbits after a single oral administration of pioglumet® tablet were presented in Table 9. Plasma concentration–time profiles of PIO and MET after oral administration of one pioglumet® tablet (15 mg PIO and 850 mg MET) (n = 5) were presented in Fig. 6. The obtained results were consistent with previous studies for the pharmacokinetic studies of PIO and MET [17,29]. These results proved the applicability of the developed HILIC method to determine simultaneously the investigated binary mixture in pharmacokinetic studies.

4. Conclusion Herein, we developed a simple, reliable and efficient HILIC method suitable for the simultaneous determination of PIO and MET in rabbit plasma samples. High-purity silica column was used for rapid and efficient separation of PIO and MET. Applications of HILIC separation technology provided improved peak resolution with good peak shape in short analysis time and augmented method selectivity compared with the frequently used RP-C18 methods. The method has been validated and proved convenient reliability and reproducibility for pharmaceutical industry and quality control laboratories. Furthermore, sample processing procedures involved a single protein precipitation step with a sample size of 200 ␮l which enabled many sampling points. Finally, this method was applied for the determination of PIO and MET in rabbit plasma using ATVC as internal standard in a pharmacokinetic study. The proposed HILIC method provided a simple and reliable alternative for the reported LC-MS method convenient for their determination in plasma samples. Acknowledgments Thanks for Assiut University Animal care center for their help and support in this study. References [1] R.L. DeFronzo, Diabetes 37 (1988) 667–687. [2] H.P. Rang, M.M. Dale, J.M. Ritter, P.K. Moore, Pharmacology, 5 edn., Churchill Livingstone, London, 2003, pp. 380. [3] R.C. Turner, C.A. Cull, V. Frighi, R.R. Holman, JAMA 281 (1999) 2005–2012. [4] D.G. Grahame-Smith, J.K. Aronson, Oxford Textbook Of Clinical Pharmacology And Drug Therapy, 3 edn, Oxford University Press, New York, 2002, pp. 324. [5] A.G. Gilman, Goodman And Gilman’s, The Pharmacological Basis Of Therapeutics, 10 edn, McGraw-Hill inc, New York, 2001, pp. 1701. [6] A. Ramachandran, C. Snehalatha, J. Salini, V. Vijay, J. Assoc. Physicians India 52 (2004) 459–463. [7] K. Sujana, G.S. Rani, M.B. Prasad, M.S. Reddy, J. Biomed. Sci. Res. 2 (2010) 110–115. [8] M.A. Hegazy, M.R. El-Ghobashy, Ali M. Yehia, A.A. Mostafa, Drug Test. Anal. 1 (2009) 339–349. [9] D. Kale, R. Kakde, J. Planar Chromatogr. Mod. TLC 24 (2011) 331–336. [10] N.M. Mansoory, A. Jain, Int. J. Pharm. Pharm. Sci 4 (2012) 72–76. [11] B.L. Kolte, B.B. Raut, A.A. Deo, M.A. Bagool, D.B. Shinde, J. Chromatogr. Sci. 42 (2004) 27–31. [12] K.S. Lakshmi, T. Rajesh, S. Sharma, Int. J. Pharm. Pharm. Sci. 1 (2009) 162–166. [13] P. Srinivas, K. Venkataramana, R.J. Srinivasa, R.N. Srinivasa, Int. J. Pharm. Chem. Biol. Sci. 2 (2012) 104–109. [14] J. Swapna, C. Madhu, M. Srivani, M. Sumalatha, Y. Nehalatha, Y. Anusha, Asian J. Pharm. Anal. 2 (2012) 85–89. [15] S. Alexandar, R. Diwedi, M.J. Chandrasekar, Res. J. Pharm. Chem. Biol. Sci. 1 (2010) 858–866. [16] F.H. Havaldar, D.L. Vairal, Int. J. Appl. Biol. Pharm. Tech. 1 (2010) 1000–1005. [17] P. Sengupta, U. Bhaumik, A. Ghosh, A.K. Sarkar, B. Chatterjee, A. Bose, T.K. Pal, Chromatographia 69 (2009) 1243–1250. [18] A.J. Alpert, J. Chromatogr. A 499 (1990) 177–196. [19] W. Naidong, J. Chromatogr. B 796 (2003) 209–224. [20] P. Hemstrom, K. Irgum, J. Sep. Sci. 29 (2006) 1784–1821. [21] P. Jandera, J. Sep. Sci. 31 (2008) 1421–1437. [22] Z. Hao, B. Xiao, N. Weng, J. Sep. Sci. 31 (2008) 1449–1464. [23] Y. Hsieh, J. Sep. Sci. 31 (2008) 1481–1491. [24] T. Ikegami, K. Tomomatsu, H. Takubo, N. Horie, K. Tanaka, J. Chromatogr. A 1184 (2008) 474–503. [25] US-Food and drug administration, guidance for industry: bioanalytical method validation, center for drug evaluation and research, rockville, MD, 2001, http://www.fda.gov/downloads/Drugs/Guidances/ucm070107 [26] The European Pharmacopoeia, The European Pharmacopoeia, 5th edition, Buffer solution, 2005, 2015, pp. 430. [27] S. Shermer, The Blood Morphology Of Laboratory Animals, PBI FA Davis Co, Philadelphia, USA, 1968, pp. 42. [28] B. Kaushal, K. Srivastava, J. Chem. Pharm. Res. 2 (2010) 519–545. [29] S. Gananadhamu, V. Laxmikanth, S. Shantikumar, V. Sridhar, C. Geetha, C. Sandhya, Am. J. Anal. Chem. 3 (2012) 849–858.