Background

Fixed-dose combination (FDC) therapy of lamivudine (LAM) and tenofovir disoproxil fumarate (TDF) is part of the three antiretroviral drugs recommended by WHO—and it is among the preferred first- and second-line regimens for adolescents, adults, children and infants [1]—the third being efavirenz, a generation non-nucleoside reverse transcriptase inhibitor (NNRTI). This therapy is also referred to as highly active antiretroviral therapy (HAART), and it is believed to be the most effective treatment in slowing HIV-1 infection progression and retarding the emergence of resistant mutants [2, 3]. Generally, its use has improved and alleviated the challenges faced in the management and treatment of people living with AIDS [4, 5].

Lamivudine, 2′,3′-dideoxy-3′-thiacytidine-4-amino-1-[(2R,5S)-2-(hydroxymethyl)-1,3-oxathiolan-5-yl]-1,2-dihydropyrimidin-2-one (Fig. 1a)—a nucleoside reverse transcriptase inhibitor (NRTI) prodrug analogue of dideoxycytidine—is known to be active against human immune deficiency virus (HIV) [6]. It requires three phosphorylation steps intracellularly, to elicit its pharmacological active anabolite, lamivudine triphosphate [7, 8]. The nucleoside analogue is infused into viral DNA by HIV reverse transcriptase and HBV polymerase leading to the DNA chain termination. Tenofovir disoproxil fumarate (TDF) (Fig. 1b) is an acyclic nucleoside phosphonate diester analogue of deoxyadenosine 5′-monophosphate, chemically named 9—[(R)-2-[[bis [[isoproxy carbonyl) oxy] methoxy] phospinyl] methoxy propyl], that belongs to the class of antiretrovirals—nucleotide reverse transcriptase inhibitors (NtRTIs). It acts by blocking the enzyme reverse transcriptase that is pivotal to viral production in persons infected by HIV [9, 10]. The active form of TDF is the diphosphate metabolite—tenofovir-diphosphate (TFV-DP) that arises from its inhibition of reverse transcriptase, by competing with the natural substrate deoxyadenosine 5′-triphosphate through intracellular phosphorylations [11], leading to the termination of the DNA chain by its incorporation [12].

Fig. 1
figure 1

Chemical structure of a lamivudine and b tenofovir disoproxil fumarate

Full size image

LAM and TDF, like most anti-retrovirals, do exhibit similar solubility in many solvent systems (such as ethanol, methanol, dilute mineral acids) and have maximum absorptions (λmax) between 250 and 280 in these solvents [5, 13, 14]. The spectra of these drugs are either over-laying or overlapping, hence the difficulty in separating multicomponent FDCs into their components [5]. The aforementioned difficulties associated in assaying multicomponent antiretrovirals have been resolved with the application of HPLC techniques as recommended by various pharmacopeias [15, 16]. However, in developing countries, access to functional HPLC is limited, because of its astronomical cost and the availability of associated accessories and consumables [5]. The UV–visible spectrophotometry is relatively cheaper and easier instrument for assaying pharmaceuticals. However, to overcome the occurrence of overlain spectra, the use of derivative UV-spectrophotometry has been applied for the evaluation of different drug compounds [17,18,19]. Other reported methods are simultaneous UV spectrophotometric method [13, 14, 20] and TLC-UV-spectrophotometric method [5]. Derivative spectrophotometry (DS) offers a range of applications, which are more reliable with respect to the normal spectrophotometry. Other spectra derivative techniques such as ratio derivative (RDS), difference derivative (DD) and compensation method (CM) have been found very usefulness in the assay of pharmaceuticals in binary mixtures [21]—these techniques are computer oriented. It has been used to resolve overlaying or overlapping spectra of multicomponent FDCs simultaneously [17, 22], determination of trace analytes in various matrices, amino acids and protein assay, environmental analysis, identification of organic and inorganic substances [23]. In addition, it has been widely applied for quantitative analysis, characterization and quality assurance in the agro and pharmaceutical industries and in biomedical-related disciplines [17, 22]. These outstanding features are mainly due to its enhanced sensitivity, selectivity, specificity and the elimination of spectral interference [24, 25]. It is also characterized by simplicity, rapidity and reproducibility. The aforementioned advantages are as a result of its spectral differential and resolution enhancement, quantitative and qualitative methods that distinguishes small variation between almost similar spectra [22]. The versatility of derivative spectroscopy (DS) is hinged on the associated data processing techniques—which comprise of zero-crossing, least-square deconvolution, Fourier transforms, etc.

This study is aimed at evaluating the application of derivative spectrophotometric method—by referencing it with the pharmacopeia recommended HPLC–UV technique for the estimation of LAM and TDF in FDC formulations.

Methods

Apparatus

Ultraviolet–Visible spectrophotometer—JENWAY 6305 model, with 1.0-cm quartz cells, was used for all spectral measurements. The analytical weighing balance (Sartorius MSU66S Model) and Eppendorf micropipettes used were previously calibrated. Other instruments/equipments used were Hp Probook 6550b laptop, Microsoft excel 2007 and OriginLab, 2019 software.

Agilent HPLC, model 1200 series was used for the quantification of LAM and TDF. The instrument was inter-phased to a UV-detector and quaternary pump, using a RP18, ODS, octadecyl column (5 μm, 150 × 4.6 mm, ZORBAX Eclipse XDB-C18), for chromatographic separation. Elutions were performed using mobile phase made up of methanol (70%, v/v) and 10 mM KH2PO4 (30%, v/v), with a flow rate of 1 mL min−1 at ambient temperature.

Materials and reagents

Chemicals used were of HPLC and spectroscopic grade. The potassium dihydrogetn phosphate and methanol were manufactured by SIGMA-ALDRICH GmbH, Germany, while the concentrated HCl was manufactured by Merck, Darmstadt, Germany. Lamivudine (LAM) and tenofovir disoprixil fumarate (TFD) standards were gifted from NAFDAC, Yaba, Lagos, and made by European Directorate of Quality Medicine (EDQM). The FDC drug samples—lamivudine/tenofovir disoprixil fumarate (300/300 mg), were gifted by the Federal Medical Center, Yenagoa, Nigeria, and manufactured by Mylan Laboratories Ltd., Hyderabad, Telangana, India, and Hetero Laboratories Ltd Telangana, India.

Preparation of reagents

  1. (i)

    Hydrochloric acid (0.1 M) reagent: Transfer 8.5 mL concentrated HCl acid to 100 mL of distilled water in 1-L volumetric flask, and make to mark with distilled water.

  2. (ii)

    Lamivudine and tenofovir disoprixil fumarate standard solutions: Weigh 50 mg of LAM and TDF standards into separate 50-mL volumetric flasks, dissolve with 5 mL methanol, and make to volume with 0.1 M HCl to obtain a stock solutions of 1000 µg mL−1 each.

Analytical techniques

Zero-order derivative spectra and determination of maximum wavelength (λ max )

Procedure by Vaikosen et al. [5] was adopted. Separate solutions of LAM and TDF reference standards and combined standards (LAM/TDF) as formulated in binary FDC were prepared from stock solutions in 0.1 M HCl. These solutions were scanned in the UV region (200 to 350 nm) to obtain individual and combined drug spectra. The maximum absorptions (λmax) of each drug in 0.1 M HCl were then obtained from the spectra.

Derivative of spectra

To resolve LAM and TDF spectra overlap, their derivatives (1st–4th) were calculated and corresponding spectra plots were done using OriginLab and Microsoft Excel 2007 softwares. The most appropriate of the four derivatives was chosen.

HPLC–UV method

Aliquots of clear drug solutions were diluted with the mobile phase mixture to obtain appropriate concentrations, and 10 μL was injected into instrument. Three injections per FDC brand were made, while peak areas of each drug were computed. The amounts of the anti-retrovirial drugs—LAM and TDF—were determined from their calibration curves.

Method validation

The analytical performances of methods were assessed in accordance with ICH guidelines [26] and in addition, by applying the proposed methods to formulated FDC drugs. Under the ICH guidelines, the following parameters, specificity, interference, precision, accuracy, linearity, sensitivity, ruggedness, and robustness, were studied. The linearity and sensitivity of the methods were established by carrying out a five point calibration curve for standards. The least squares method was used to obtain the regression equations and other parameters. The limits of detection (LOD) and quantification (LOQ)—which depicted sensitivity of methods—were evaluated using the expressions, LOD = 3.3 Sd/x; LOQ = 10 Sd/x (where Sd is the standard deviation of the intercept of regression line and “x” is the slope of the regression line) [27, 28]. The ruggedness methods were assessed by applying both to assay brands of FDC antiretroviral drugs—this measured the reliability of methods for routine laboratory quality assessment. The recovery studies were done by spiking drug samples containing 1.0, 1.5, 2.0 and 2.5 mg of LAM, with drug standard at concentrations 2.0, 1.5, 1.0 and 2.5 mg, respectively, while for TDF, sample containing 10.0, 12.5, 15.0 and 10.0 mg was spiked with 2.0, 2.5, 5.0 and 10.0 mg pure drug standard. The intra-day and inter-day precision was determined by replicate analysis at four concentration levels—4, 6, 8, 10 μg mL−1 for LAM and 8, 10, 15 and 20 μg mL−1 for TDF, while for HPLC–UV, levels were at 0.5, 1.0, 2.5 and 4.0 μg mL−1; these were spiked with 1, 2.5, 5, 2.5 and 1.0 μg mL−1, respectively, for recovery studies. The concentrations for intra-day and inter-day studies were 0.5, 1.0, 2.5, and 5 μg mL−1. The drugs were replicated thrice on the same day, while the inter-day assay was done on 3 days—every other day, within a week using the same concentrations and two brands of the FDC.

Calibration graphs of D3 spectra for drug standards

A five-point calibration curve for the third–order derivative spectra was prepared by carrying out serial dilutions from stock solutions of reference standards. The spectra were measured at two wavelengths with respect to the order of the derivative, where zero crossing was observed for TDF and maximum for LAM (λmax for LAM) and conversely zero crossing for LAM and a maximum for TDF (λmax for TDF). The values for D3 amplitudes were obtained for concentration ranges of 8–24 µg mL−1 and 2–10 µg mL−1 for TDF and LAM, respectively. The absorbance values were plotted against the concentrations of the solutions to obtain a straight line calibration curve, while amount of drugs in FDC was deduced for test samples.

The standard solutions for the calibration of LAM and TDF for the HPLC method were prepared from stock solutions to obtain co-mixed standards in the mobile phase. The working concentrations for a five-point calibration curve and drug quantification in brands ranged from 0.25 to 5.0 μg mL−1 for both drugs.

Procedure for simultaneous extraction and application of methods to drugs

Derivative spectroscopic method

An equivalent of 50 mg each of LAM and TDF in pulverized FDC tablets was weighed and transferred into a 25-ml calibrated volumetric flask, shaken gently with 10 ml of methanol for about 2–3 min and then made to volume with methanol. This solution was filtered into a clean volumetric flask, with the first 5 mL of the filtrate discarded. Appropriate dilutions were made to obtain concentrations within the working range for each analyte using 0.1 M HCl. The absorbance of the solution was measured at two wavelengths from the derivative spectra, where zero crossings were observed for TDF (λmax for LAM) and LAM (λmax for TDF).

HPLC–UV method

For the high-performance liquid chromatographic, suitable aliquots of clear drug extract were diluted with the mobile phase and 10 μl was injected into instrument. Three runs were made for brand and the peak areas of the drugs were computed, while the quantities of each in FDC tablets were determined from the regression equations obtained.

Statistical analysis

The statistical analyses were carried out using OriginLab80 (Origin, China, 2019 version) and Microsoft Office Excel 2010.

Result

Zero- and higher-order derivative spectra of TDF and LAM

Figure 2 shows the zero-order derivative spectrum of TDF, LAM and overlay of co–mixed standards (TDF/LAM as found in FDC tablets). The maximum absorptions for LAM and TDF were observed at 270 and 260 nm, respectively, and there was significant spectra overlap between LAM and TDF reference standards, while the co–mixed showed a single maximum absorption at 260 nm. Figure 3 represents four different orders of derivative UV spectra for LAM and TDF standards. The third-order (D3) and fourth-order derivative spectra (D4) were found satisfactory; however, the D3 spectrum was found to be the more appropriate, with the overlapping of both drugs properly resolved (Fig. 3c). The amplitudes at 240 nm and 262.5 nm showed maxima for TDF and LAM, respectively—these also corresponded to the zero-crossing points for LAM and TDF. The observed maximum amplitude for LAM in this study is close to the amplitude of 265.6 nm reported by Uslu and Özkan [17] in the first derivative spectra of lamivudine. Figure 4 shows the linear response of both drugs at three levels of concentration for the third-order derivative spectra with alternate maxima and zero-crossing points for LAM and TDF to enable their simultaneous quantification.

Fig. 2
figure 2

Zero-order derivative spectra for TDF, LAM, and mixture of TDF + LAM standards

Full size image
Fig. 3
figure 3

Overlay spectra of a first (dA/dλ)-, b second (d2A/dλ2)-, c third (d3A/dλ3)- and d fourth (d4A/dλ4)-order derivatives of TDF and LAM

Full size image
Fig. 4
figure 4

Third-order derivative spectra of lamivudine (1) 2 µg mL−1, (2) 4 µg mL−1, (3) 6 µg mL−1 and tenofovir disproxal fumarate (1) 16 µg mL−1, (2) 20 µg mL−1, (3) 24 µg mL.−1

Full size image

HPLC assay of FDC TDF and LAM

Figure 5 shows the HPLC chromatogram for LAM and TDF standards, with retention times at 2.316 and 3.577 min, respectively. Both drugs were detected at an optimum wavelength of 254 nm, and their peaks were well resolved.

Fig. 5
figure 5

Chromatogram of mixed standards of lamivudine and tenofovir disoproxil fumarate

Full size image

Analytical performance of methods: derivative spectroscopic and HPLC methods

The results obtained for the measurement of performance of methods are as enumerated in the sections below.

Linearity range and sensitivity

The calibration graphs for the D3 spectrophotometry using Beer’s law plot (n = 5) for LAM and TDF showed good linearity at concentration ranges of 2–10 µg mL−1 and 8–24 µg mL−1, respectively (Table 1). The regression equations were obtained using the least squares method, with very small intercepts (≤ 1.62 × 10−4) and slopes (≤ 3.58 × 10−5). Calculated standard errors were ≤ 8.04 × 10−5, with correlation coefficient (R2) of 0.998 for both drugs. These values were considered satisfactory and indicated good sensitivity of the proposed derivative method. Similarly, the HPLC curves were linear, with correlation coefficients for LAM and TDF being 0.999. Also, the intercepts and slopes were 175.05 and 1214.50, respectively, for LAM, with corresponding values of 124.74 and 2040.60 for TDF. These values depicted good sensitivity and accuracy of both methods. The standard errors of the intercept and slope were ≤ 2.32 and ≤ 20.35, respectively.

Table 1 Optimum conditions for drug assay by proposed methods
Full size table

The LOD for the proposed D3 spectrophotometric method was 0.46 and 2.61 μg mL−1 for LAM and TDF, respectively, with corresponding LOQ values as 1.40 and 7.90 μg mL−1. These values confirmed the reliability and repeatability of the D3 method.

Precision and accuracy

The results for the intra-day and inter-day studies are presented in Tables 2 and 3 for D3 and HPLC methods, respectively. The computed relative standard deviation (RSD) for intra-day and inter-day assays for D3 method ranged from 0.16 to 1.90% for both drugs, while for HPLC it was from 0.16 to 1.99%. The RSD was found to be less than 2%—this shows that the methods are precise and accurate [29]. The standard errors (SEs) were ≤ 0.08 and ≤ 0.02 for all runs in the D3 and HPLC methods, respectively. These values depicted high reproducibility, good precision and accuracy of both methods [30].

Table 2 Summary of precision and accuracy studies for third derivative spectroscopy method
Full size table
Table 3 Summary of precision and accuracy studies for HPLC method
Full size table

Ruggedness, robustness and recovery studies

The ruggedness of the third derivative and HPLC methods was assessed by applying methods to assay two FDC brands of LAM/TDF—thus evaluating the reliability of both methods for routine laboratory quantification. Results obtained in varying some experimental conditions/parameters—such as brands, standard addition, varying of drug concentrations and comparative studies with established pharmacopeia HPLC methods [15, 16]—were useful indices for the evaluation of the reliability of the D3 method (Table 4). The percentage recovery for the proposed method ranged from 94.80–100.13% to 96.63–99.93% for LAM and TDF, respectively, while values obtained for HPLC method were from 96.00–100.33% and 96.00–100.13%. Observed variations in results were insignificant; hence, the D3 method is considered reliable, rugged and robust.

Table 4 Recovery studies for lamivudine and tenofovir disoproxil fumarate in FDC
Full size table

Application of analytical methods to dosage form

Table 5 shows the results obtained for the successful application of the proposed D3 spectroscopic and the HPLC methods for the assay of two FDC brands containing LAM/TDF. The amounts of drugs found in the formulations were within the BP and USP specifications [15, 16] and also agree with the label claim. The content of LAM and TDF in brand I was 100.19 ± 0.59% and 100.89 ± 0.38%, respectively, with corresponding values for brand II being 99.59 ± 1.19% and 99.39 ± 0.63% for the proposed D3 spectroscopic method. For the HPLC method, the amounts of LAM in brands I and II were 99.86 ± 0.50 and 99.87 ± 0.71, respectively, while those for TDF percentage content were 99.87 ± 0.32% and 100.06 ± 0.35%. The Student-t test for accuracy with respect to the amount of drugs in formulations for both methods was between 0.075 and 1.016 for 5 replicates—this implies that no significant difference between the claims on brands and values obtained in evaluating the both methods at 95% confidence level [31].

Table 5 Application of methods to drug formulation
Full size table

Discussion

Maximum and zero-crossing amplitudes

The proposed D3 method has effectively shown the propensity to resolve overlaying/underlaying and overlapping problem observed with zero derivative spectra for LAM/TDF FDC and has been used for the quantification of both drugs simultaneously. The amplitudes at 240 nm and 262.5 nm showed maxima for TDF and LAM, respectively, with corresponding zero-crossing points for LAM and TDF, thus making the simultaneous determination feasible.

Specificity and interference

Interference of extraneous materials on the proposed method was negligible, while specificity was enhanced. Drug samples were in solid dosage form, and the quantification of analytes was evaluated in the UV-region; hence, there is no chromophore-bearing compound, as co-extractive that is likely to interfere in the methanol used for the extraction of LAM and TDF. In addition, solid dosage formulations are made up of pharmaceutical inorganic excipients or aids such as—magnesium stearate, sodium lauryl sulfate, starch sodium glycolate, carboxymethylcellulose (CMC), talc, lactose spray dried, titanium dioxide, microcrystalline cellulose, pre-gelanitizated starch and hydroxypropylcellulose [30, 32]. These substances are chromophore-free and insoluble in methanol and ethanol. Also, the presence of dyes or colored substances that are alcohol soluble would absorb in the visible region. The absence of interference implied that the proposed D3 spectra method is highly selective for FDC tablet formulations of LAM/TDF and could be used for routine laboratory quality control analysis of pure and solid dosage forms.

Comparison between D3 spectroscopy and HPLC methods

Both analytical techniques were compared using statistical analysis. The Student’s t-test values for both methods and analytes were ≤ 1.016 (Table 5), while the tabulated value for 5 replicates at 95% confidence is 2.776. This implies that there was no significant difference in the quantities as claimed by the manufacturers and the results obtained in applying both methods [33]. Table 6 shows the calculated paired t-test and variance ratio F-test values between both methods and the amount of actives found in the brands. The experimental t- and F-values ranged from 0.21 to 1.32 and 1.41 to 3.61, respectively—none of these values exceeded the stipulated critical values (t = 2.776, F = 6.39) for four degrees of freedom. The aforementioned statistical values suggest also that there was no significant difference between the proposed D3 spectroscopy and HPLC methods at 95% confidence level [34].

Table 6 Paired t-test/F-test for D3 spectroscopy and HPLC methods
Full size table

The LOD and LOQ values for both methods suggested that the HPLC method was more sensitive than D3 method. However, the D3 technique is simpler, more economic, more time saving, more robust and sufficient samples can be run within a day compared to the HPLC method (without automation devices). In addition, the D3 spectroscopy method has shown the inclination of being free from interferences associated with excipients such as starch, glucose, talc, lactose and/or from frequent degradation products compared to the HPLC method—where residual analytes and impurities build-up in columns and are likely to interfere with the assay [30, 31, 33]. For method validation with respect to precision and accuracy, the D3 method seemed better than the HPLC method. Tables 2 and 3 reaffirm D3 statistical preference—where the %RSD for two brands of LAM/TDF antiretrovirals (Heteros and Mylan brands) ranged from 0.08–0 1.86% to 0.16–1.99% for D3 spectrophotometric and HPLC methods, respectively. From the aforementioned, both assay techniques do not exhibit any significant difference and do resolve the overlaying challenges often encountered in FDC analyses using zero-order derivative spectrophotometric method.

Conclusion

The study presents a comparative use of third-order derivative spectroscopic method with the pharmacopoeia recommended HPLC method for the assay of antiretroviral FDC containing LAM and TDF. Statistically, both methods were adjudged to have no significant difference and do have the ability and capacity to resolve overlaying problems often associated with zero derivative spectrophotometry in the assay of FDCs. Although the D3 spectrophotometric technique is mathematical at deducing the wavelengths of interest, it was considered more time saving, cheaper, simpler, more environmental friendlier and more economical in terms of consumables and the generation of waste, than the HPLC method. The reliability, precision and accuracy of the proposed method are reflected in the validation parameters assessed and could be used for routine analysis of FDCs where HPLC is not available.