INTRODUCTION

The potent anticancer drug abiraterone acetate {[(3S,8R,9S,10R,13S,14S)-10,13-dimethyl-17-pyridin-3-yl-2,3,4,7,8,9,11,12,14,15-decahydro-1H-cyclopenta[a]phenanthren-3-yl]acetate} is an acetyl ester and a significant prodrug of its active metabolite abiraterone.^1,2,3 It is a well-built receptor blocker of androgen with potent bioavailability, especially in oral administration.⁴ Abiraterone acetate is found to be more intense and effective than ketoconazole and liarozole in CYP17A1 inhibition; which is a rate-limiting enzyme for the biosynthesis of androgens.^4,5 Principally, abiraterone acetate is specified for use in combination with prednisone for treatment of men with metastatic castration-resistant prostate cancer who have already received prior chemotherapy comprising docetaxel.⁵ Abiraterone acetate is poorly soluble over an extensive pH range. However, it is faintly soluble in HCl as well as in organic solvents (Figure 1). Abiraterone acetate can be detected by liquid chromatography-tandem mass spectrometry,^6,7,8 spectrofluorimetric for measuring fluorescence emission, and absorption spectroscopy in cancer patients.^9,10,11 Furthermore, the drug has been thoroughly recognized by other chromatographic techniques for the quantification of metabolites in biological samples bioanalytically.^8,9,12 However, there is scant research using QbD Paradigm, MODR Concepts, design space, and systematic development by chemometrics-assisted statistical optimization of critical method parameters (CMPs) and critical method attributes.¹³ These optimization techniques with analytical quality by design (AQbD) are efficient, cost-effective, and innovative methodologies, which signify the “finest solution” to a meticulous “complexity” or problems raised in method development but deliver consistent quality output where other general High-performance liquid chromatography (HPLC) methods have been unable to achieve.^13,14,15,16 Notably, more relatively efficient optimization techniques such as Box-Behnken design (BBD) than other designs (central composite, d-optimal, and mixture design) have been applied for analytical method development and optimization of pharmaceutical drugs. BBD is a second-order three-level factorial design with advantages of user-friendly, economical, simple, and limited experimental runs.^17,18 Apart from that, such BBD optimization is a sequential design empowered with a feasible design matrix showing the quadratic model with a straightforward assessment of CMPs or critical material attributes. It also requires limited variables or responses, mainly three coded levels such as [low (-1), middle (0), and high (+1)], for the experimental design to fit the statistics or lack of fit values of the distinctive quadratic model.^17,18,19 Therefore, the developed method of validation, optimization, and stability-indicating reversed-phase (RP)-HPLC by employing the AQbD approach for the quantification of the bulk drug and its pharmaceutical formulations of abiraterone acetate is more economical, robust, and less expensive and reduces the heat and trial methods of optimization as well as the revalidation process as compared to one factor-at-a-time approach.

MATERIALS AND METHODS

A pure standard of the drug abiraterone acetate was obtained from Sun Pharmaceutical Industries Ltd., Halol, Gujarat, India. Abiraterone acetate was obtained commercially [XIBRA (Cipla Ltd, India), ZYTIGA (Janssen-Cilag Ltd., India), and ABIRAPRO (Glenmark Pharmaceuticals, India)] with a labeled claim of 250 mg. Phosphate buffer and acetonitrile (HPLC grade) were procured from Merck Laboratories, Mumbai.

Instrumentation

A Shimadzu HPLC system (LC-2010C HT) with a ultraviolet (UV)-visible detector, ultra-sonicator from Remi Instruments, Mumbai, and nylon filter (0.45 µm) from Millipore, Mumbai, India, was used.

Chemicals and reagents

Orthophosphoric acid (OPA) (AR grade) and acetonitrile (HPLC grade) were obtained from Merck Laboratories Pvt. Ltd., Mumbai. Potassium dihydrogen phosphate (KH₂PO₄) was obtained from Fischer Scientific, Mumbai, India.

Methods

Method optimization was performed with Box-Behnken statistical design comprising the CMPs, which include three significant factors (the composition of the determined mobile phase, flow rate, and pH of buffer), encompassing three levels. Seventeen experimental runs were established with five center points. The flow rate was tested at 1.0, 1.2, and 1.5 mL/min; the pH was measured at 4, 5, and 6; and the concentration of the mobile phase was monitored at 20%, 50%, and 80%. The responses considered were retention time (Rt) and peak area, which were designated as critical analytical attributes (CAAs). The data were analyzed, and the model was validated with Design-Expert software. The quadratic model revealed a virtuous correlation with the experimental data executed for design space navigation. The 2D & 3D response surface techniques and perturbation sequential plots were scrutinized to assess the indicative critical factors’ impact upon the observed responses or CAAs found within the predicted range. The predicted values from the practical responses were found to be satisfactory, and it was confirmed that it was acquired within the design space of the optimized results.^19,20

Statistical analysis

Design-Expert (Version 12), Stat-Ease Inc., Minneapolis, MN, USA, was utilized for method optimization and estimation of its CMPs and randomization of the runs.¹⁴ Microsoft Excel 2007 (Microsoft, USA) was used for the remaining calculations.

Preparation of solvent

Phosphate buffer (pH 4): Disodium hydrogen phosphate (5.04 g) and KH₂PO₄ (3.01 g) were dissolved in water to a volume of 1000 mL. The pH was adjusted to 4.0 with OPA. The resulting solution was passed through (0.45 µm) filter paper, filtered by vacuum filtration, and sonicated for about 15 min.¹⁵ While preparing the mobile phase, only the phosphate buffer was filtered using (0.45 µm) nylon membrane filters. Acetonitrile was not filtered and was used as provided by the supplier.

Procedure for the stock standard solution

Standard stock solution of the drug was made by accurately weighing 10 mg of the drug and mixed with 10 mL of acetonitrile to acquire a concentration of 1000 µg mL^-1. Serial dilutions of the stock solution were prepared by diluting with the mobile phase to obtain concentrations of 1 to 100 µg mL^-1 and filtered through a 0.45 µm syringe filter before chromatography. Calibration curves were constructed as depicted in Figures 2a-c. Here, a mixture of acetonitrile and phosphate buffer (10 mM KH₂PO₄) (20:80, %v/v) was selected as a mobile phase with a flow rate of 1 mL min^-1 and UV detection at 235 nm. The limit of quantification (LOQ) and limit of detection (LOD) were estimated based on the linearity plot (concentration vs. peak area).

Chromatographic conditions

A Merck-Hibar Purospher STAR Analytical column (C18, 250 mm × 4.6 mm × 5 µm) was used for chromatographic separation due to its advantage of having the highest carbon loading for better separation of the desired analyte, minimizing the variabilities of the mode of selection. A mobile phase of acetonitrile and phosphate buffer (10 mM KH₂PO₄) (20:80 %v/v) was used. The flow rate was 1 mL min^-1 with an injection volume of 10 µL. The oven temperature of the chromatographic column was 30°C, and the sample temperature was ambient.

Optimization of the chromatographic method using the analytical QbD approach

Initially, trial and error practices were executed to obtain rigorous data for the selected chromatographic method’s performance and its finding of vital independent variables with its considerable effect upon the dependent variables. The utmost significance of developing the RP-HPLC method mostly separates the drug from its excipients and the degradants. Statistical analysis was accomplished by implementing a suitable experimental design by response surface methodology (RSM) through BBD principles. The statistical design intensifies ANOVA principles for establishing the optimized experimental conditions of the method.^16,21,22 A simple, accurate, and AQbD-based RP-HPLC method was developed, which is also stability indicating. This was further subsequently validated as per the International Conference on Harmonization (ICH) recommended stability guidelines for quantification of abiraterone acetate in its various pharmaceutical formulations (tablets). A mixture of phosphate buffer (10 mM KH₂PO₄ and acetonitrile) (20:80, %v/v) was used as the mobile phase with a 1 mL min^-1 flow rate.

Optimized chromatographic conditions

The optimized conditions are as follows: Optimized trail concentration: 10 µg mL^-1; flow rate: 1mL min^-1; mobile phase: Acetonitrile/phosphate buffer [10 mM KH₂PO₄, (20:80) %v/v]. The optimized chromatograms of the blank, standard, and mixture of excipients (10 µg mL^-1) of the developed analytical method are shown in Figure 2a-c, respectively.

Method validation

Method validation was performed to substantiate that the analytical procedure employed for a definite experiment is appropriate for use.¹⁹ The outcomes of method validation parameters can be highly requisite to judge the reliability, quality, and steadiness of analytical results. Validation of a method for parameters linearity, accuracy, precision, and robustness was done according to recommendations of ICH guidelines (ICH Q2) (R1).^23,24,25

Linearity

The linearity plot of the proposed method was constructed as per the stated ICH guidelines.^26,27,28 The linearity chart of abiraterone acetate was found to be within the concentration range of 2-100 µg mL^-1; further, the calibration plot was constructed using the peak area versus the concentration.

Accuracy

A series of solutions were spiked with known standard concentrations of abiraterone acetate of 50%, 100%, and 150% (5, 10, and 15 µg/mL) in triplicate, performed as per recommendations of ICH guidelines.^26,27,28

Precision

The precision of the developed technique, expressed in % relative standard deviation (RSD), was calculated by performing repeatability and intermediate precision studies.²¹ The developed analytical QbD-based method was validated by the precision studies (both intraday and interday).

LOD and LOQ

The LOD and LOQ of the current investigation were evaluated from the baseline noise of abiraterone acetate through comparisons of measured signals of samples with known analyte concentrations with that of the blank by (signal-to-noise) S/N ratios of 3:1 (LOD) & 10:1 (LOQ) as per guidance and protocol recommended by ICHQ2B.²⁹

Robustness

A robustness study was performed to recognize and evaluate the toughness of an analytical method.³⁰ To check the ability of the projected method, different factors were deliberately altered such as alternation in mobile phase composition, change in flow rate, etc.

Specificity

Specificity was carried out to evaluate the analyte noticeably in the occurrence of components that may be expected to be present during development. Specificity was established by representing no interference from the excipients and the degradation products.^28,29

Forced degradation studies

Forced degradation generally includes exposure of drug substances to a range of stress conditions to demonstrate the stability profile and possible degradants of developed analytical methods.^26,27,28 Acid degradation was performed by adding 1 mL of 0.1 N HCl, then heating at 60°C for 30 min, cooling to room temperature, and neutralizing. Alkali degradation was performed by adding 1 mL of 0.1 N NaOH, heating at 60°C for 30 min, cooling to room temperature, and neutralizing. The degradative process through oxidation was performed by exposing the drug to 1 mL of 3% H₂O₂ and heating at 60°C for 30 min. The degradation study was achieved by heating the drug content solutions at 105°C on a thermostatically regulated water bath for half an hour. Photolytic degradation was done by subjecting the drug solutions to UV light in a UV chamber at 365 nm for 3 h. The degradation samples were accurately prepared through appropriate aliquots of the drug and in the solution form of their drug products, instructed by the stress testing protocol.^26,29 After a definite time, the treated solutions were adjusted with the mobile phase.

Stability of analytical solution

The solution stability study was effectively performed by observing the standard and sample solutions to determine the stability potential of the drug substance. This factor was analyzed by injecting the drug sample and standard solutions at distinctive intervals. This evaluation parameter is intended for the evaluation of the chemical stability of the drug and sample solution whether any significant changes occurred at varied time intervals.^29,31,32

Assay of formulations

Twenty tablets of commercial brands of abiraterone acetate were chosen arbitrarily, and their average weight was determined.³¹ The tablets were crushed to fine powder, 250 mg was weighed, and the powder was dissolved in 200 mL of acetonitrile (in a 250 mL volumetric flask). Then, it was shaken for 20 min and ultrasonicated for 20 min. After that, it was allowed to cool at room temperature, and the solution was diluted up to mark with diluents (1000 mcg mL^-1). The final obtained solution was then diluted to 10 mcg mL^-1 with acetonitrile/phosphate buffer (20:80 %v/v) and subsequently injected to an HPLC system for estimation in triplicate.

RESULTS AND DISCUSSION

Method development and optimization Using Box-Behnken experimental design

In the current investigation, trials were proposed and conducted using Box-Behnken experimental design.²¹ The proposed BBD desires 17 experimental runs of examination to acquire data and model the response surface for a reliable analysis. A 3² level with 17 experimental runs was executed to identify the optimized design space for detection of the predicted response. The 3² BBD with RSM was executed.^16,21 The statistical design with optimization methodology and its data analysis through BBD were statistically evaluated using Design-Expert-Ver.12 software.^21,22 The three most influential factors (CMPs), the composition of the mobile phase (X1), flow rate (X2), and pH (X3), were selected; the peak area (Y1) and Rt (Y2) were used as observed responses. The design matrix of the statistical BBD and experimental runs are shown in Table 1.²¹ The contour plots illustrate that the effect of both the responses stand in need about factors X1 (acetonitrile and phosphate buffer %) and X2 (flow rate), while the X3 (pH) is found to produce a nullified effect upon the obtained responses. The outline of 3² factors and ANOVA results, through its calculated mean and standard deviation values, are summarized in Table 2.²² The statistically constructed model was suitably validated by interaction studies using the effect of various factors upon the found responses. The 2D counter plot analysis of peak area and Rt with the observed responses (R1) & (R2) are depicted in Figure 3a, b, respectively. Similarly, the 3D counter plot analysis of peak area (R1) response and Rt (R2) response is depicted in Figure 3c, d, respectively. The statistical model signifies that predicted values for both the responses (Rt and peak area) are a bit closer to the actual values representing superior accuracy and precision values for the obtained responses. The 2D and 3D counter plots of predicted versus actual values for peak area (R1) and Rt (R2) are depicted in Figure 4a, b, respectively. The perturbation plot displays the impact of all the influential factors (mobile phase and flow rate) at a particular point within the design space for the selected responses peak area and Rt. The representative plots of perturbation analysis for the observed responses peak area (R1) and Rt (R2) are depicted in Figure 4c, d, respectively. Here, for 2D and 3D surface numerical optimization, the Rt (R2) and peak area (R1) are depicted in Figure 5a-d, respectively. Figure 6 elucidates the parameters intended for numerical optimization for desirability and optimized data for factors indorsed by design. Finally, the optimized apparent chromatographic conditions can be well predicted from the arithmetical model, and it has strongly been recommended for the developed analytical method (Table 3).

Results of validation studies

The optimized feasible chromatographic conditions were aggrandized and effectively implemented to validate the QbD-based method by a range of validation parameters, linearity, accuracy, precision, LOD, LOQ, robustness, etc., as per recommendations of ICH guidelines.

Linearity

Six concentrations of the abiraterone acetate working standard ranging from 2 to 100 ppm were prepared and analyzed for the linearity study. The calibration plot was constructed by plotting the chromatographic peak areas versus the known predicted concentrations in µg mL^-1, and values of observed concentrations were determined. The regression equation was found to be Y= 73399x - 71154, and the regression coefficient r² is 0.998. The detail of the calibration plot and its residual plot data analysis are depicted in Figure 7a, b, respectively.

Accuracy

The accuracy of the developed method was studied to make sure the agreement between the true and reference values in three significant levels of 50%, 100%, and 150%. The percentage recoveries of three various concentrations were found to be within the range of 98% to 102%, and % RSD was obtained within the acceptance limit, that is, NMT 2.0%. The data of all the recovery studies are shown in Table 4.

Precision

Precision studies of the developed method were carried out by the system, method, and intermediate precision studies. Precision of the system suitability test for six replicate standard injections was calculated, within the acceptance criteria, that is, NMT 2%, and both intraday and interday precisions through three replicate injections of standards were calculated. Data of % RSD are reported in Table 5.

LOD and LOQ

The detection and quantification limits (LOD & LOQ) of the current investigation were actively quantified as per the recommendation of ICHQ2B guidelines for validation of analytical methodology. The detection limit was derived as a signal-to-noise (S/N) ratio of 3:1, whereas the quantification limit was indicated as a S/N ratio of 10:1, as an effect of the response by the detector. The estimated value for LOD was found to be 0.45 ppm, and the estimated value for LOQ was 1.35 ppm. The calculated values are shown in Table 3.²³

Robustness

Robustness is the capacity to replicate the analytical method in diverse labs or under varied conditions without the manifestation of unanticipated differences in the obtained results. The estimated values of mobile phase composition and flow rates were taken, and changes in flow rate and mobile phase data are summarized in Table 6. The results show the calculated % RSD within the acceptance limit, that is, less than 2%.

Specificity

The specificity of the method has been studied to determine the interference from the degradation products as per ICH through a forced degradation study. The results denote no sign of peak formation at the Rt of abiraterone acetate and the case of degradation products since the peak purity passed. The data reveal that the purity angle was less than the purity threshold, presenting no specific interference (Table 7).

Forced degradation studies

Forced degradation studies provide knowledge on practical degradation pathways and degradation products of the active ingredients and help expound the configuration of the degradants as per ICH recommendations.²⁷ The following conditions were applied for degradation: Acid (1 mL of 0.1 N HCl, 60°C for 30 min), alkali (1 mL of 0.1 N NaOH, 60°C for 30 min), peroxide (1 mL of 3% v/v H₂O₂), thermal (105°C), and photolytic degradations at 231 nm, for determining the steady nature of drugs. Percent drug degradation values during acidic and alkaline degradation were observed to be 26.5% and 12.3%, respectively. In contrast, less than 2% degradation was reported in the case of oxidative, thermal, and photolytic degradations, which indicate the drug’s completely resistant behaviors to the above stress conditions. The detailed descriptions of forced degradation activities are listed in Table 7. The representative chromatograms of the sample in various stress conditions by incorporating a mixture of excipients (a), acid (b), alkali (c), oxidative (d), photolytic (e), and thermal degradation (f) studies are depicted in Figure 8a-f respectively.

Stability of analytical solution

The solution stability study was carried out by observance of sample and standard solutions at 25±2°C for 40 h. After analysis, it has been concluded that the drug standard and sample were stable for up to 40 h (Table 7). The anticipated method was effectively validated and met the necessities as per the recommended stability guidelines of ICH.

Assay of formulations

The chromatogram of the marketed sample is depicted in Figure 9. The calculated values of the percentage of the assay of various marketed formulations are exhibited in Table 8. The results show that all the values of the formulations are within the acceptance limit, that is, 98-102%.

CONCLUSION

Chemometrics-assisted method development affords regulatory flexibility, the formation of homogeneous or robust finished products of assured quality characteristics as per Food and Drug Administration (FDA) concerns, and ICH stability stressed conditions. In addition to this, the AQbD method minimizes the process of revalidation, lessens solvent consumption and development time, and enhances optimum robust analytics. By implementing the DoE, the Box-Behnken statistical design can evaluate the independent variables (CMPs) concurrently with adding common interactions among the critical factors to optimize the tentative conditions. It is explicit that the application of this BBD with RSM is an adaptable practice to reduce the total experimental runs to obtain sustainable, robust analytics, reducing the chance of revalidation, requisite for analytical development. The optimization of the RP-HPLC method for abiraterone acetate can produce the highest intense data and boost efficiency within a short period as per the ICH Q8 (R2) and FDA perspectives. The proposed method was found to be linear, with concentrations of 2-100 ppm having r²: 0.998. The remarkable % recovery (within 98-102%) of the drug reflects that the excipients existing in the tablet formulation have no impediment in the quantitation of the drug. The optimized conditions by AQbD of the anticipated method established that the proposed study was cost-effective, extremely robust, and stability indicating. Therefore, the developed stability-indicating method was economical, accurate, and precise. It can be successfully implemented for the routine analysis of abiraterone acetate in its bulk and pharmaceutical formulations.

ACKNOWLEDGMENTS

The authors are thankful to the Board of Management, School of Pharmacy and Life Sciences, Centurion University, Bhubaneswar, India, and the Principal of KL College Pharmacy, KL University (Deemed to be University), Vaddeswaram, Guntur, India, for providing the suitable instrumental facilities to carry out the research activities.

Conflict of interest: No conflict of interest was declared by the authors. The authors are solely responsible for the content and writing of this paper.