In this work, SALLME was tried for alogliptin extraction from plasma. Different factors were studied to achieve the maximum enrichment. Different solvents, volumes of extractant, anions, cations, amounts of salt and pH values were investigated. Optimization of these parameters was performed using one- factor- at-a- time (OFAT). The peak area was the parameter to evaluate the effect of each factor on extraction.
Organic Solvent Optimization
Preliminary experiments were done to investigate the optimum extractant using different solvents including methanol, propylene glycol, glycerol, tetrahydrofuran and acetonitrile. A 2000 µL of each solvent was added to 1 mL of aqueous solution containing 250 mg of sodium chloride followed by vortexing for 2 min, then the tube was centrifuged for 5 min at 4000 rpm. The results showed that there is no phase separation with methanol, propylene glycol, and glycerol, which could be due to the high polarity of the hydroxyl groups and the multiple H-bonds formed between these solvents and water. On the other hand, both tetrahydrofuran and acetonitrile could induce phase separation, but the background noise in tetrahydrofuran after injection into HPLC/UV was significantly higher than acetonitrile. For this reason, acetonitrile was chosen as the optimum extractant in the following SALLME procedures.
Acetonitrile Volume Optimization
The extractant volume is the most important factor that could affect the sample enrichment in SALLME. Generally, analyte pre-concentration is inversely proportional to the volume of extractant [30]. Different volumes of acetonitrile were investigated in the range of 100 to 2000 µL. The results showed that 450 µL was the least volume of acetonitrile that could be used in SALLME. Using volumes of acetonitrile lower than 450 µL could not induce phase separation. As shown in Fig. 3, the highest response was observed using 500 µL acetonitrile, thus it was designated as the optimum acetonitrile volume in the following procedures.
Anion- type Optimization
Different anions were investigated to select the optimum anion that could achieve the best extraction efficiency. All anions were sodium salts of monovalent (Chloride and acetate), divalent (Sulfate, thiosulfate, carbonate) and trivalent (Phosphate) anions. Fig.4 shows that the highest extraction efficiency was achieved with chloride. The mechanism of salting-out depends on hydrophobic effect and electrostatic repulsion [31]. In this aspect, ions with high charge density (charge/size) are expected to interact strongly with water and induce more electrostatic repulsion [32]. The small size of chloride compared with other anions may explain the observed high efficiency. It is here worth mentioning that the volume retrieved of acetonitrile after adding chloride was small compared with other anions which helped make the analyte more concentrated in the separated layer of acetonitrile. Further optimization was performed by using a binary mixture of chloride/carbonate, the two anions that could attain the best results, but in different ratios. Adding carbonate to chloride resulted in higher phase ratios (Retrieved volume of ACN/Aqueous volume). However, using chloride alone was better than its mixtures with carbonate. Thus, chloride was selected as the optimum anion in this step.
Cation- type Optimization
The type of salt cation has a role in salting-out phenomena, but to less extent than the anion component. Seven cations were tested, all in chloride salts including monovalent (Soidum and potassium), divalent (Copper, cobalt, calcium and magnesium) and trivalent (Ferric) cations. Transition metals (Copper, cobalt and ferric) were found not suitable due to the observed color, which could complicate the sample matrix. No phase separation was observed in case of potassium and magnesium, while sodium and calcium could successfully induce salting out of the acetonitrile layer. Compared with calcium, sodium could achieve higher extraction efficiency due to the small size of sodium, which could enhance charge density and salting out capabilities [31]. Therefore, sodium chloride was selected as an optimum salt for SALLME of alogliptin from plasma, because it was cheap, safe, available and more efficient.
pH Optimization
In extraction methods, pH plays a major role due to its effect on solubility and ionization of drugs. Different pH values were investigated in the range of 8.3 to 13.9, adjusted using the appropriate concentration of sodium hydroxide to span the pKa value of alogliptin (pKa = 9.47). Fig. 5 shows low peak areas for alogliptin at pH values lower than 9.47 due to the predominance of the ionized form. Further increases in pH were associated with a corresponding pronounced increase in response up to pH 11.5. Increasing pH above 11.5 did not significantly affect the obtained response. The highest extraction efficiency was achieved with pH = 12 followed by a steady state in the range of 13 to 13.9. At pH = 12, alogliptin will be in the non-ionized form, leading to better extraction efficiency. Therefore, the selected optimum pH value was 12.
Chloride Amount Optimization
To study the effect of salt amount on SALLME performance, different amounts of sodium chloride were tried in the range of 100 mg to 600 mg. The results showed that the salt amount had a small effect on extraction efficiency. The peak areas of alogliptin benzoate were comparable regardless of the amount of sodium chloride. As shown in Fig. 6, 250 mg of NaCl resulted in slightly higher responses, thus it was selected as the optimum amount in the following procedures.
Method Validation
The chromatographic method was validated according to the International Conference of Harmonization (ICH) guidelines M10 [29]. The following parameters were investigated: linearity, precision, accuracy, limit of detection (LOD) and limit of quantitation (LOQ). The method was found linear in the range of 0.1 to 50 µg/mL Accuracy and precision were determined by % recovery and % RSD, respectively. The results were satisfactory for bioanalysis application.
Linearity
The method linearity was investigated in the concentration range of 0.1 to 50 µg/mL. The calibration curve was constructed by plotting the response ratio (ratio between peak area of alogliptin and peak area of sitagliptin) on the y-axis and alogliptin concentration on the x-axis (in µg/mL). The calibration curve indicated a linear relationship between response ratio and alogliptin concentration with an acceptable correlation coefficient and regression parameters as summarized in Table 1.
Table 1 Quantitative analysis and regression line
Compound
|
Slope
|
Intercept
|
r
|
Range
|
LOD
|
LOQ
|
Alogliptin
|
0.0272
|
0.0012
|
0.997
|
0.1-50 µg/ml
|
0.019 µg/ml
|
0.06 µg/ml
|
R2: Regression coefficient, LOD: Limit of detection, LOQ: Limit of quantitation
Accuracy and Precision
Method accuracy was investigated by analyzing plasma samples, spiked with alogliptin at three concentrations: 5, 20 and 40 µg/mL, each prepared in triplicate. Table 2 indicates that the % recovery was in the range of 98 to 101, with an average %recovery of 99.76. The precision of the analytical method was investigated by analyzing plasma samples, spiked with alogliptin at the same three concentrations, each prepared in triplicate. The concentrations were analyzed in the same day to test repeatability and in 3 consecutive days to evaluate intermediate precision. Based on the %RSD shown in Table 2, the method intraday and interday precision were acceptable for bioanalysis application.
Table 2 Intra-day and inter-day precision and accuracy
Parameter
|
Accuracy & precision
|
Intraday
|
Interday
|
Added
(µg/mL)
|
Found
(µg/mL)
|
Found
(%)
|
Added
(µg/mL)
|
Found
(µg/mL)
|
Found
(%)
|
|
5
|
5.05
|
101.00
|
5
|
5.02
|
100.44
|
Alogliptin
|
20
|
19.97
|
99.85
|
20
|
19.99
|
99.93
|
|
40
|
39.38
|
98.43
|
40
|
39.60
|
99.01
|
Mean
|
|
|
99.76
|
|
|
99.79
|
% RSD
|
|
|
1.56
|
|
|
2.57
|
RSD: Relative standard deviation
Limit of detection and Limit of quantitation
Limit of detection (LOD) and limit of quantitation (LOQ) were calculated according to the following equations:
Where SD is the standard deviation of 10 blank injections. LOD was 0.019 µg/mL and LOQ was 0.06 µg/mL. SALLME was found sensitive for determination alogliptin in biological samples obtained from alogliptin clinical study.
Freeze and thaw stability study
The freeze and thaw stability study was performed by spiking human plasma with alogliptin at three quality control concentration levels (5, 20and 40 µg/mL). The plasma samples were frozen for 12 hours, thawed for three cycles, and the concentration of alogliptin was determined after each cycle to be compared with the zero cycle. Table 3 indicates that alogliptin was stable for three freeze/thaw cycles.
Table 3 Freeze and thaw stability study
|
Added (µg/mL)
|
Found (µg/mL)
|
Recovery (%)
|
RSD (%)
|
RE (%)
|
1st cycle
|
5.00
|
4.98
|
99.63
|
4.28
|
0.37
|
20.00
|
19.98
|
99.91
|
3.39
|
0.09
|
40.00
|
41.03
|
102.58
|
3.26
|
2.58
|
2nd cycle
|
5.00
|
5.21
|
104.23
|
1.94
|
4.23
|
20.00
|
19.88
|
99.38
|
1.54
|
0.62
|
40.00
|
39.34
|
98.36
|
1.49
|
1.64
|
3rd cycle
|
5.00
|
5.09
|
101.73
|
3.66
|
1.73
|
20.00
|
20.57
|
102.83
|
0.98
|
2.83
|
40.00
|
38.89
|
97.23
|
4.92
|
2.77
|
Comparison with other reported methods
Due to the low plasma concentrations of alogliptin, most analytical methods used LC-MS/MS for quantitation after sample preparation using protein precipitation. While protein precipitation is a simple and fast method for sample treatment, its efficiency to remove interference and protect analytical instruments is less than perfect. Moreover, the dilution effect of the added precipitating agent compromises method sensitivity, which could be compensated by the inherent high sensitivity of mass detection. For sake of comparison, the developed SALLME method was compared with the reported protein precipitation procedure [27] under the same chromatographic conditions. As shown in Fig. 7, a huge plasma peak appeared in the beginning of the chromatogram (retention time = 1.55 min) compared with a very small peak at the same retention time in SALLME. This could be due to the dual function of acetonitrile to precipitate protein and to extract the drug after salting out. Moreover, the peak area of alogliptin treated with SALLME was more than seven times higher than that in the protein precipitation method. These results were achieved while only one third of the acetonitrile volume was consumed in the SALLME method which decreases organic solvent consumption, reduces organic waste, protects the operator and the environment and bestows green characteristics on the developed method.
Besides protein precipitation, conventional LLE [33] was also used for alogliptin determination in plasma by UPLC/DAD using diethyl ether as an extractant and a mobile phase consisting of acetonitrile:phosphate buffer (50:50, v/v) for reconstitution after evaporation. The results of the developed SALLME method were compared with the results of the reported LLE method using Student t-test and F-test at 95% confidence level (α=0.05) as Table 4 shows. The statistical analysis indicated that there were no significant differences between the two methods. Although these results showed comparable accuracy and precision, the procedures in LLE were time consuming, non-ecofriendly, more expensive and required vacuum for solvent evaporation. On the other side, SALLME was simpler, cheaper, eco-friendlier and could achieve lower LOQ than the reported LLE method.
Table 4: Comparison between the developed SALLME method and the reference LLE method [33].
Parameter
|
Comparison with reference method
|
The SALLME method
|
Reference Method
|
Added
(µg/mL)
|
Found*
(µg/mL)
|
Found
(%)
|
Added
(µg/mL)
|
Found
(µg/mL)
|
Found
(%)
|
|
5.00
|
4.96
|
99.15
|
5.00
|
4.96
|
99.29
|
|
10.00
|
10.08
|
100.75
|
10.00
|
10.23
|
102.26
|
|
20.00
|
20.02
|
100.08
|
20.00
|
20.20
|
101.01
|
Mean
|
|
|
99.99
|
|
|
100.85
|
% RSD
|
|
|
0.81
|
|
|
1.48
|
t-test**
|
|
|
0.429
|
|
|
|
F-test**
|
|
|
0.450
|
|
|
|
*Average of five determinations for each concentration
** p-values at 95% confidence level (α=0.05)