All chemicals and reagents used for method development were HPLC grade. The analytical standards of RDV, DXM, MTF, STG, and GLM were purchased from Sigma-Aldrich Co. Acetonitrile, methanol, and potassium dihydrogen phosphate used for mobile phase and stock solutions preparation were of HPLC grade. A blank human plasma standard (Sigma-Aldrich Co.) was used for calibration curve construction. All additional reagents needed to develop the analytical method were suitable for HPLC analysis.

The stock solutions of RDV, DXM, MTF, STG (1 mg/ml), and GLM (0.2 mg/ml) were prepared by dissolving the required amount of each substance in methanol and diluting it to 20.0 ml with the same solvent. A mixed stock solution (solution A, 40 µg/ml) was prepared by transferring suitable aliquots of each stock solution to one and the same 50.0 ml volumetric flask and further diluted with methanol. For plasma studies stock solutions of RDV, DXM (4 mg/ml), and GLM (0.2 mg/ml) were prepared in methanol. A mixed stock solution containing RDV, DXM, and GLM (solution B, 160 µg/ml) was prepared by transferring the required aliquots of the RDV, DXM, and GLM stock solutions to a 20.0 ml volumetric flask and further diluted with methanol.

Six working solutions (RDMSG -SM1, RDMSG -SM2, RDMSG -SM3, RDMSG -SM4, RDMSG -SM5 and RDMSG -SM6) in the concentration range of 1–24 µg/ml and three working quality control solutions (RDMSG -QC1, RDMSG -QC2, and RDMSG -QC3) with concentrations 1.6, 6, and 20 µg/ml were prepared by appropriately diluting aliquots from the solution A with methanol. For plasma studies six working solutions (RDG-P1, RDG-P2, RDG-P3, RDG-P4, RDG-P5, and RDG-P6) in the concentration range of 5–120 µg/ml and three working quality control solutions (RDG-QC1, RDG-QC2, and RDG-QC3) with concentrations 8, 30, and 100 µg/ml were prepared by appropriately diluting aliquots from the solution B with methanol. All solutions were stored at 2–4 °C temperature before analysis.

For plasma analysis, STG was chosen as an internal standard (IS). The solution was prepared by dissolving 40 mg of the substance in methanol and diluting it to 50.0 ml with the same solvent. The obtained concentration was 0.8 mg/ml.

Calibration and quality control solutions were prepared in standard human plasma. To 400 µl blank plasma a 100 µl of RDG-P1, RDG-P2, RDG-P3, RDG-P4, RDG-P5, RDG-P6, RDG-QC1, RDG-QC2 and RDG-QC3 were added, respectively. The resulting samples were vortexed for 2 min. The obtained concentrations were 1, 2, 4, 8, 16, and 24 µg/ml for the calibration standards and 1.6, 6, and 20 µg/ml for the quality control samples.

Plasma protein precipitation was performed according to the previously published method (Smerikarova et al. 2023) with some modifications and adjustments. A plasma aliquot of 150 µl was mixed with 50 µl of the IS solution, then the sample was vortexed for 5 min. Further on, 600 µl of acetonitrile was added as a precipitating agent. After homogenization on a vortex mixer for 10 min, 15 min of sonication in an ultrasonic bath and 15 minutes of shaking at 500 rpm were applied. The supernatant was separated after 10 min of centrifugation at 13 000 rpm and 600 µl were filtered through a 0.45 µm syringe filter.

A SHIMADZU Corporation chromatographic system consisting of online degassing unit DGU-20A5, solvent delivery unit LC-20AD, autosampler Sil-20A, column oven CTO-20A, and UV-VIS detector SPD-20A was. The recording and processing of the results were done with Lab Solution Software. An RP-18 (250 × 4.6 mm, 5 µm) chromatographic column, equipped with a suitable guard column, was applied as a stationary phase. Isocratic elution was performed at a flow rate of 1.0 ml/min, and at ambient temperature. A mixture of acetonitrile and potassium dihydrogen phosphate buffer (pH 3) in a ratio of 45:55 v/v was used as а mobile phase. The mobile phase was filtrated and degassed in an ultrasonic bath before use. UV – VIS detector was set at 250 nm and the injection volume was 20 µl.

During the development of the chromatographic method, columns with different hydrocarbon chain lengths were tested, namely – octadecylsilane (C18) and octylsilane (C8). Considering the very short MTF retention time on the C8 column as well as the better separation on the C18 column, the C18 column was chosen for further studies. The effect of column length was investigated (two columns of different lengths -150 × 4.6 mm and 250 × 4.6 mm were used) and the longer column was preferred as it showed some additional increase in MTF retention time.

To find a suitable mobile phase, experiments were carried out with different compositions and ratios of solutions. Acetonitrile was preferred as the organic solvent because the use of methanol resulted in broad, diffuse, and asymmetric peaks. However, the application of mixtures of acetonitrile and water in different ratios did not produce the desired result. As the percentage of acetonitrile increased, overlapping peaks were observed and conversely, as the acetonitrile content decreased, the peaks broadened and along with this, the retention time of the last eluting peak increased too much. The introduction of a phosphate buffer in the composition of the mobile phase led to an improvement in the results. The effect of pH in the range of 3 to 6 was investigated, with the best results obtained at pH 3. The effect of buffer concentration (concentrations from 0.5 mM to 20 mM) on the separation and peak shape was also investigated. A concentration of 20 mM was the most suitable. It should be noted that lowering the concentration of the buffer solution results in broader peaks especially for antidiabetic agents and also to an unacceptable increase in the retention time of the latter component. Finally, a mixture of acetonitrile and potassium dihydrogen phosphate buffer (pH 3) in a ratio of 45:55 v/v was used as а mobile phase. Varying the flow rate and temperature did not significantly improve the separation of the compounds, so 1 ml/min flow rate and room temperature were chosen. The absorption maxima of the investigated compounds vary in relatively wide ranges, approximately from 230 nm for GLM to 270 nm for STG. Therefore, considering the spectra of the compounds (Fig. 2), as a good compromise, the UV detection was performed at 250 nm.

Figure 2. 

UV spectra of RDV, DXM, MTF, STG and GLM.

The developed method was validated according to the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH) guidelines (International Council for Harmonisation 2022) by various parameters including selectivity, linearity, accuracy, precision, system suitability, limit of detection and limit of quantification.

A method is considered selective when it can distinguish the target compounds from other components in the sample. The selectivity of the proposed method was proven by its capability to determine RDV, DXM, MTF, STG, and GLM in their synthetic mixture without interference. As can be seen from the representative chromatograms (provided in Fig. 3A, B) there are no interferences from mobile phase components during the analysis of RDV, DXM, MTF, STG, and GLM in their synthetic mixture. The peaks of all five analyzed drug substances are well separated.

Figure 3. 

Representative chromatograms obtained from A. The mobile phase and B. Synthetic mixture. Legend: Peak numbers: MTF-1, DXM-2, STG-3, RDV-4, and GLM-5.

System suitability tests were performed in order to evaluate the performance of the chromatographic system. Parameters like retention time, column efficiency (column theoretical plates), capacity factor, selectivity, resolution, and tailing factor were determined by a sixfold analysis of the test sample at the concentration of 8 µg/ml. The requirements of the European Pharmacopoeia were fulfilled, as can be seen from the detailed results, expressed as mean values in Table 1.

Linearity studies were performed to track the dependence of the analytical signal on the concentration of the analytes in the samples. Six standard solutions containing the analyzed substances were used in a concentration range from 1 to 24 µg/ml. For calibration curve construction, peak areas against solution concentrations were plotted. The achieved high correlation coefficients (R2>0.999) showed good linearity of the proposed method. Detailed results are listed in Table 2.

Quality control samples at three concentration levels within the range of the calibration curve were used for accuracy and precision studies. Three independently prepared samples were triplicate analyzed in one day or on three consecutive days, to assess the accuracy and precision within a run and between runs. Reproducibility and accuracy were assessed by a coefficient of variance (CV%) and by the percentage deviation of the average concentration compared to the weighted one (d%). The coefficients of variance were below 2.5% and d% ranged from -3.5 to 6.2%. Detailed results are listed in Table 3.

Limits of detection and limits of quantification were determined experimentally by the signal-to-noise ratio. The LODs achieved were 0.05 µg/ml for MTF, DXM, and RDV. For STG LOD was 0.5 µg/ml and for GLM 0.3 µg/ml. The limits of quantification were 0.2 µg/ml for MTF, DXM, and RDV, and 1.0 µg/ml for STG and GLM.

A protein precipitation procedure was used as a simple and fast sample preparation method. Unfortunately, under specific chromatographic conditions, MTF cannot be determined in plasma samples due to its very short retention time. An overlap of the plasma and MTF peaks was observed. However, the proposed method allows the simultaneous determination of RDV and DXM in the presence of the GLM and IS. All drug substances were well separated and no interferences from the biological matrix components were observed (Fig. 4A, B).

Figure 4. 

Representative chromatograms obtained from A. The blank and B. Spiked plasma samples. Legend: Peak numbers: DXM-2, STG-3, RDV-4 and GLM-5.

The area ratio of the chromatographic peaks of each analyte and the IS was plotted against concentration for the calibration curve in plasma. Linear relationship over the studied concentration range (from 1 to 24 µg/ml) and satisfactory correlation coefficients (R2>0.997) were observed. The details are summarized in Table 4.

Three QC samples at different concentration levels were analyzed for within-run and between-run accuracy and precision studies. The coefficient of variance (CV%) and the percentage deviation of the average concentration compared to the weighted one (d%) were used for reproducibility and accuracy assessment. The coefficients of variance were below 3.7% and d% ranged from -7.3 to 8.5%. The results summarized in Table 5 confirmed the method as accurate and reproducible.

The recovery studies were performed by blank plasma samples which were spiked with a laboratory-prepared mixture (RVD, DXM, and GLM) and IS at three different concentration levels. The obtained recovery percentages, ranging from 85.1% to 108.5% are listed in Table 6.

Table 7 presents comparative data for the proposed method and some previously published reports. It can be seen that in terms of limits of detection and quantification, as well as linearity, the results obtained are better in some cases. It can be said that the obtained retention time results occupy an intermediate position compared to the published ones. An exception is the retention time of GLM, which is higher than the published ones.

Regarding the recovery rate (Table 6) for RDV, DXM, and GLM from the biological samples, the obtained results can be considered similar to the published reports. In the available literature, the recovery rates for RDV range from 97% to 103% (Kishore et al. 2021; Emam et al. 2022), for DXM from (77%–82%) (Kumar et al. 2006) to (97%–103%) (Emam et al. 2022), and for GLM from 80% to 98%) (El-Enany 2012; Sebaiy et al. 2020).

However, the obtained results can hardly be compared due to the lack of published methods dealing with the simultaneous separation and quantification of the studied medicinal substances.