Introduction

Hepatitis C virus (HCV) is a well-known and dangerous virus that can cause a life-threatening disease. In the United States (US), HCV affects approximately three to five million people, while approximately 170 million people are infected globally. In its early stages, the disease caused by HCV has no symptoms; however, when it becomes chronic, it can lead to serious fatal complications, including hepatocellular carcinoma and death [1]. Velpatasvir (VPS), a direct-acting antiviral, is an effective agent that can be administered to HCV patients. VPS acts as a defective substrate for nonstructural protein 5A (NS5A), which plays a key role in the replication of HCV ribonucleic acid (RNA). Hence, VPS helps to prevent HCV-RNA replication [2]. Sofosbuvir (SFV), a nucleotide analog NS5B polymerase inhibitor that acts as a prodrug, is also used to treat HCV patients [2].

In 2016, US Food and Drug Administration (FDA) approved the combination of VPS and SFV for treating chronic HCV infection (genotypes 1, 2, 3, 4, 5, or 6) in adult patients [3]. VPS has been reported as practically insoluble in aqueous buffers having pH ˃5.0, slightly soluble in aqueous buffers of pH 2.0 and soluble in aqueous buffers of pH 1.2. However, SFV has been reported as slightly soluble in water and aqueous buffers of pH 2.0–7.7 [3, 4]. The bioavailability of VPS (25–30%) has been reported as low while it has been reported higher for SFV (92%) after oral administration of tablets [3]. These two drugs have been investigated using different chromatographic and spectrophotometric techniques [4,5,6,7,8]. The forced degradation of SFV was studied by liquid chromatography–mass spectrometry (LC–MS) technique [9]. The concurrent analysis of VPS and SFV in commercial tablets was studied extensively using reversed-phase high-performance liquid chromatography (HPLC) techniques [10,11,12,13,14,15,16,17,18,19,20,21]. HPLC technique was also applied for the concurrent analysis of VPS and SFV in human plasma [14]. Some LC–MS techniques have also been reported for the concurrent analysis of VPS and SFV in human plasma [22, 23]. Few ultra-performance liquid chromatography techniques were also applied for the concurrent analysis of VPS and SFV in tablet dosage forms [7, 24]. Capillary electrophoresis technique has also been reported for the concurrent analysis of VPS and SFV in formulations [25]. A single spectrodensitometric technique has been reported for the concurrent determination of VPS and SFV in tablets [26]. However, the most reported methods of analysis are based on chromatographic techniques that are more expensive and complicated than spectrophotometric and spectrodensitometric techniques. Spectrodensitometric techniques offer several advantages over other analytical techniques such as HPLC and these advantages include cost, economy, sample preparation and sample analysis etc. [26,27,28]. Thus, there remains an urgent need for accurate, simple and cost-effective methods for the simultaneous estimation of VPS and SFV. Hence, the goal of the present study was to establish accurate, simple and cost-effective analytical methodologies for the concurrent quantification of VPS and SFV either in their pure forms or tablets.

Experimental

Chemicals and materials

Chromatography grade water was utilized in all experiments and it was obtained from Milli-Q Water Purification Unit. VPS (purity of 100.23%) was obtained from BioVision (Milpitas, CA, USA). SFV (PSI-7977; purity of 99.98%) was purchased from Cayman Chemical Company (Ann Arbor, MI, USA). Methanol was procured from Sigma-Aldrich (St. Louis, MO, USA). All other reagents or chemicals used were of analytical/pharmaceutical grades.

Instrumentation

A double-beam spectrophotometer (JASCO, Kyoto, Japan) containing matched quartz cuvettes with path lengths of 1 cm was used. The spectrophotometer was attached with an IBM-compatible computer with an HP 680 inkjet printer (Hewlett Packard, USA). A CAMAG thin-layer chromatography (TLC) scanner (3 S/N 130319) operated with winCATS software and a CAMAG Linomat 5 Autosampler (Muttenz, Switzerland) with a CAMAG microsyringe (100 μL) were used.

Pharmaceutical formulations

The pharmaceutical formulations were Epclusa® extended-release tablets (Gilead Sciences International, Cambridge, UK). The claimed amounts of VPS and SFV in Epclusa® tablet were 100 mg and 400 mg/tablet, respectively.

Standard and working solutions

Spectrophotometric method

Stock standard solution of VPS was obtained by accurately weighing 100 mg of pure VPS into a 100-mL volumetric flask. The powder was dissolved, and the volume was brought to 100 mL using methanol. Stock standard solution of SFV (2.5 mg/mL) was prepared in the same manner but with 25 mg of pure SFV in a 25-mL volumetric flask.

Working standard solutions for the studied drugs were prepared by diluting 5 mL of the VPS or SFV stock solution in a 50-mL volumetric flask and diluting with methanol to obtain standard working solutions with concentrations 100 and 250 μg/mL for VPS and SFV, respectively.

TLC–densitometric method

Standard stock solutions of VPS and SFV were obtained by dissolving 25 mg of VPS or SFV into 20 mL methanol. Both the drugs were dissolved separately under vortex mixing. The final volume of each drug was maintained with methanol in order to obtain stock solution of 1.0 mg/mL for both drugs.

Validation of the proposed methods

The above methods were validated according to ICH-Q2B guidelines [29].

Linearity

First derivative of ratio spectra (1DD) method

VPS (10–500 μg/mL) and SFV (50–800 μg/mL) working standards were prepared separately using methanol as a solvent. The 1DD technique was applied to determine VPS (1–50 μg/mL) by measuring the peak amplitudes at 290 nm using SFV (80 μg/mL) as a divisor. The proposed 1DD technique was also applied to quantify SFV (5–80 μg/mL) by obtaining the peak amplitudes at 243 nm using VPS (50 μg/mL) as a divisor.

TLC–densitometric method

Aliquiots of VPS (5–50 μL) and SFV (10–70 μL) stock standards (1 mg/mL) were applied to glass-coated TLC plates (silica gel 60 F254; 20 cm × 20 cm, 0.25 mm) in the form of bands using an |autosampler (CAMAG Linomat 5) and a 100-μL microsyringe (Hamilton) to obtain the concentrations in the range of 5–50 μg/band for VPS and 10–70 μg/band for SFV. The bands were applied at spacings of 10 mm. The air-dried plates were developed and pre-saturated for at least 1 h using mobile phase [chloroform:methanol, 9.5:0.5 (% v/v)] at ambient temperature. The developed plates were air-dried and the spots were identified using a deuterium lamp at λmax of 265 nm in reflectance/absorbance mode. p-Anisaldehyde was used as a visializing agent to identify the spots. Calibration curves for VPS and SFV were prepared separately by plotting the measured peak areas and concentration. Linear regression analysis was performed to generate regression equations and various stattistical parameters.

Accuracy

Accuracy was evaluated as the percent of recovery (% recovery) [29]. Accuracy was evaluated at three different concentrations of VPS (i.e., 5, 10 and 40 μg/mL for the spectrophotometric technique and 10, 20 and 30 μg/band for the TLC–densitometric technique) and SFV (5, 20 and 40 μg/mL for the spectrophotometric technique and 20, 30 and 40 μg/band for the TLC–densitometric technique).

Precision

Precision of both analytical methodologies were estimated as the repeatability and intermediate variation [29]. Three different concentrations of VPS (i.e., 5, 10 and 40 μg/mL for the spectrophotometric technique and 10, 20 and 30 μg/band for the TLC–densitometric technique) and SFV (5, 20 and 40 μg/mL for the spectrophotometric technique and 20, 30 and 40 μg/band for the TLC–densitometric technique) were analyzed for repeatability (intra-day precision) and intermediate precision (inter-day variation) using the two proposed analytical methods. The results were documented as the percent of the coefficient of variation (%CV).

Specificity and selectivity

Specificity was determined by comparing the ultraviolet, 1DD spectra and TLC–densitometric results recorded for VPS, SFV and commonly used excipients with the blank (a solution of excipients in methanol without drugs) [29]. The selected excipients were those indicated in the manufacturer’s monograph of the Epclusa® extended-release tablets. Selectivity was verified by the quantification of mixtures prepared in the laboratory containing different amounts of VPS and SFV.

Detection and quantification limits

The limit of detection (LOD) is the lowest detectable amount of drug which could be distinguished from the background, while the limit of quantification (LOQ) is the lowest accurate and precise amount of the target drug [29]. LOD and LOQ of the proposed analytical methodologies were obtained using standard deviation (σ) technique by applying Eqs. (1) and (2), respectively:

$$ \mathrm{LOD}=3.3\times \frac{\upsigma}{\mathrm{S}} $$
(1)
$$ \mathrm{LOD}=10\times \frac{\upsigma}{\mathrm{S}} $$
(2)

Where, S is the slope of the calibration curve.

Robustness

Robustness can be checked by evaluating the effect of minute changes in assay conditions on method validity. For the 1DD method, robustness was evaluated by making minute changes in the degree of smoothing of the derivative ratio spectra. For the TLC–densitometric method, robustness was checked by making minute changes to the solvent mixture used in densitometric separation.

Analysis of VPS and SFV in Epclusa® extended-release tablets

Ten Epclusa® extended-release tablets were weighed and the average weight was obtained. The tablets were then crushed into a fine powder and mixed well. Tablet powder equivalent to 100 mg of VPS and 400 mg of SFV was then transferred to a 250-mL beaker and an amount of 40 mL of methanol was added followed by stirring for approximately 20 min. The solution was then filtered into a 100-mL volumetric flask. The residue was washed twice with approximately 10 mL methanol and the volume was brought to the mark using methanol followed by extensive mixing. The necessary dilutions were performed for both methods. The contents of VPS and SFV were determined using the proposed analytical methods. The concentration of each drug was calculated from the calibration curve of each drug.

Standard addition was performed by spiking various known amounts of VPS and SFV in the tablets. The samples were then re-analyzed using the proposed analytical methods.

Comparison with reference published methods

The results obtained using the proposed analytical techniques for quantifying VPS and SFV were statistically compared with those obtained by reported reference methods in order to evaluate the accuracy and precision of the proposed techniques [9].

Results and discussion

Spectrophotometric method

Molecular absorption spectroscopy was used to quantify the active pharmaceutical ingredients (APIs). However, the use of molecular absorption spectra in drug analysis by the aid of Beer–Lambert’s law is hindered especially when the studied drugs have severe overlap in the zero order curves as in the case of the present study. Therefore, different alternatives were used like the use of first derivative of ratio spectra method in this study. The DD1 absorption spectra of VPS and SFV recorded in this work are presented in Fig. 1. The overlap between the VPS and SFV spectra was addressed by applying the 1DD method (Figs. 2 and 3). This technique depends on the derivatization of the ratio spectra to remove the interference between the components of the binary mixture. The most important merit of this method is the presence of many maxima and minima, which allow us to determine the concentrations of the target drugs in the presence of excipients and other interfering materials [30].

Fig. 1
figure 1

Absorption spectra of SFV (20 μg/mL) and VPS (5 μg/mL)

Full size image
Fig. 2
figure 2

DD1 spectra of VPS in the concentration range of 1–50 μg/mL

Full size image
Fig. 3
figure 3

DD1 spectra of SFV in the concentration range of 5–80 μg/mL

Full size image

TLC–densitometric method

Without prior separation, the simultaneous determination of VPS and SFV was performed using a coupled TLC–densitometric method. Different solvent systems were used to separate the two drugs. Satisfactory results were obtained by applying the experimental conditions mentioned previously with chloroform:methanol (9.5:0.5; % v/v) as the mobile phase. The values of retention factor (Rf) for SFV and VPS were obtained as 0.19 and 0.29, respectively (Fig. 4).

Fig. 4
figure 4

TLC–densitometric separation patterns of VPS and SFV

Full size image

Method validation

Linearity

Spectrophotometric method

To determine the VPS concentration in the range of 1–50 μg/mL, a linear standard curve was plotted between the peak heights for the 1DD spectra at 290 nm using SFV (80 μg/mL) as a divisor (Fig. 2). To determine SFV concentration in the range of 5–80 μg/mL, the peak heights in the 1DD-spectra at 243 nm were used with VPS (50 μg/mL) as the divisor (Fig. 3). The linear regression equations for VPS and SFV were expressed as Eqs. (3) and (4), respectively:

$$ {}^1{\mathrm{DD}}_{\mathrm{SFV}}=0.0344C+0.1099\ \left({r}^2=0.9990\right) $$
(3)
$$ {}^1{\mathrm{DD}}_{\mathrm{SFV}}=0.0045C+0.005\ \left({r}^2=0.9999\right) $$
(4)

Where, 1DD is the peak height, C is the corresponding concentration and r2 is the correlation coefficient.

TLC–densitometric method

The TLC–separation technique was also used to determine VPS and SFV concentrations. Linearity was established by plotting the measured peak area versus the corresponding concentrations at 265 nm over the concentration ranges of 5–50 and 10–70 μg/band for VPS and SFV, respectively (Fig. 5). The respective regression equations for VPS and SFV were expressed as Eqs. (5) and (6), respectively:

$$ {\mathrm{PA}}_{\mathrm{VPS}}=1.4514C+2330.9\kern1.25em \left({r}^2=0.9994\right) $$
(5)
$$ {\mathrm{PA}}_{\mathrm{SFV}}=3.888C+446.29\kern1.75em \left({r}^2=0.9995\right) $$
(6)
Fig. 5
figure 5

Three-dimensional separation patterns for VPS and SFV in their linearity ranges

Full size image

Where, PA is the area under the peak.

Accuracy

Accuracy was obtained by the evaluation of various quality control (QC) samples at three different concentrations of VPS and SFV within the specified linearity ranges for each analyte. The recovery and %CV were calculated and results are tabulated in Table 1. The results of recovery and %CV were found to be satisfactory for both analytical methods. These results suggested that both analytical methods were accurate for concurrent analysis of VPS and SFV.

Table 1 Validation parameters for the determination of VPS and SFV by the proposed methods
Full size table

Precision

To determine the repeatability and intermediate precisions, three different concentrations of VPS or SFV were analyzed on the same day and three successive days, respectively. The %CV values for the repeatability and intermediate precisions were recorded as less than 2% for both the analytical methods. These results suggested the precision of the proposed analytical methods (Table 1).

Specificity and selectivity

None of the examined excipients interfered with the analytes, indicating that the proposed methods are specific. Hence, the adopted methods are suitable for application in the assays for the two investigated HCV agents in commercial products since no detectable interference from the formulation matrix was found.

To check selectivity, the proposed analytical methods were applied in the quantification of VPS and SFV in the mixtures prepared in the laboratory. Good results with acceptable %CV values were obtained, confirming the selectivity of the proposed analytical methods (Table 2).

Table 2 Application of the proposed methods to the determination of VPS and SFV in laboratory-prepared mixtures
Full size table

LOD and LOQ

The acceptability of the proposed methods was evaluated based on the LOD and LOQ values. The LOD and LOQ values confirmed that the proposed methods are sufficient and acceptable (Table 1).

Robustness

Slight changes in the determination conditions resulted in effects smaller than 1%CV, confirming the robustness of the both analytical methods (Table 1).

Quantification of VPS and SFV in Epclusa® extended-release tablets

VPS and SFV in Epclusa® extended-release tablets were quantified to ensure content uniformity using the proposed methods (Table 3). The marketed combination showed acceptable recoveries within the acceptable content uniformity limits. The standard addition technique was also applied by spiking various known amounts of VPS and SFV into the tablets in order to ensure the reliability and applicability of both analytical techniques. The results indicated satisfying recoveries of spiked drugs using the proposed methods (Table 3).

Table 3 Determination of VPS and SFV in Epclusa® extended-release tablets and standard addition recovery evaluation using the proposed methods
Full size table

Comparison with reference method

The developed analytical techniques of the present study were compared statistically with previously reported technique [9]. The results of this comparison are tabulated in Table 4. The results indicated insignificant differences between the proposed methods and the reference method, confirming the accuracy and precision of the developed methods.

Table 4 Comparison of the proposed methods and a reference method for the determination of VPS and SFV in bulk samples
Full size table

Conclusions

Based on the results presented herein, we can conclude that the proposed methods are accurate, selective, precise and simple approaches for simultaneously quantifying VPS and SFV over the specified linearity ranges in either bulk or tablet form. The developed methods can easily be adopted in laboratories without HPLC capabilities. The proposed analytical methods could be successfully applied for the routine analysis of all pharmaceutical products containing VPS and SFV as the active ingredients.