Beyond the standard robust testing, method validation according to the ICH criteria does not give much reliability in terms of mitigating method variability. (Sandhu et al. 2016) In recent times, pharmaceutical companies are adopting QbD in analytics for trouble free compilation with FDA and ICH guidelines. (Peraman et al. 2015a) AQbD delves into the scientific understanding of technique factors and their interconnections, resulting in a region that is both robust and cost-effective. When an AQbD methodology is used in the development stage, an analytical method’s flexibility is granted without the need for revalidation or regulatory assessment. (Vogt and Kord 2011; Peraman et al. 2015b; Das and Maity 2017) As a result, using Quality by Design concepts towards the construction of analytical methods has become rather prevalent in order to achieve high robustness and superior method performance. (Monks et al. 2012) In recent times, many analytical methods for single drug estimation (Alruwaili 2021; Babar and Padwal 2021; Jena et al. 2021; Patel et al. 2021; Srujani et al. 2021; Wadhwa et al. 2021) and simultaneous quantification (Gundala et al. 2019; Babar et al. 2021; Saurabh Chaudhari et al. 2022) have been developed using AQbD method. Analytical quality by design refers to the systematic and scientific creation and optimization of analytical procedures. (Palakurthi et al. 2020) The AQbD approach allows a scientific and risk assessed understanding of the major causes of variability, then the identification of CMPs through risk assessment and factor screening studies to identify the significant variables that have a massive effect on analytical performance, and then optimization of those variables through appropriate experimental designs to broaden method performance. (Rozet et al. 2012) Quality by design is well emphasized in the design and manufacture of pharmaceutical drug substance and drug product processes, as specified in ICH Q8, Q9, and Q11 (Palakurthi et al. 2020).

When using a QbD strategy, it is vital that sufficient consideration be paid to the method’s intended usage so that method’s objectives are articulated. The Analytical Target Profile is reflected by all this. In order to meet the ATP’s standards, an acceptable technique and procedure conditions must be chosen. An exercise oriented towards understanding the method can be done based on a risk assessment (i.e., the method complexity and the possibility for robustness and ruggedness issues) to have a better understanding of the impact that key input factors may have on the method’s operational qualities. As a result, a set of operational method variables can be identified. Experimentation can thus establish a functional relationship between method input variables and method performance attributes. Insights obtained during the method’s creation and early implementation is fed into risk assessment using tools like the Fishbone diagram that may be used to pick out the factors need to be studied and which need to be regulated. (Jadhav and Tambe 2013) Critical Quality Attributes (CQAs) are measurable chromatogram attributes that must be within a certain limit (or) range to ensure the method’s targeted quality. CQAs are used in chromatographic techniques as resolution, theoretical plate number, tailing, and analytical peak’s retention time. Critical Method Parameters are variables identified through a procedure called quality risk assessment that do have an impact on Critical Quality Attributes. Following this, based on developmental information and experiments, critical method attributes of this experiment are designated as flow rate, mobile phase ratio, and column temperature.

Candesartan Cilexetil and Pioglitazone were obtained as gift samples from Madras Pharmaceuticals, Chennai, Tamil Nadu. Dosage form to be analysed claims 8 mg of Candesartan Cilexetil and 15 mg Pioglitazone per tablet. Ortho phosphoric acid, Potassium dihydrogen orthophosphate, sodium chloride, hydrochloric acid, hydrogen peroxide, acetonitrile (HPLC grade) were purchased from sigma-Aldrich, Chennai, India.

The technique was devised using Agilent technologies 1220 infinity II HPLC system fitted with Kromasil C₁₈ (150 × 4.6 mm, 5 µm) column. Digital pH meter (MP -1PLUS) was used for pH measurements and Professional Ultra-Sonicator (ANTECH, GT SONIC) was employed to degas the mobile phase and filtered through 0.45 µm Millipore filter attached to a vacuum pump. Chromatographic separation was achieved using C ₁₈ column and mobile phase consisting of Buffer and acetonitrile at 60:40 v/v ratios. 20 mM Potassium dihydrogen orthophosphate is used as buffer with pH 3.5 adjusted with ortho phosphoric acid. The method development was performed in HPLC at isocratic system. 10 µL was the injection volume with a flow rate of 0.9 mL /min. UV detection was set at 220 nm for elution. The % organic phase, flow rate, and column temperature were optimized using Central Composite Design through Design Expert Trail version 11 Stat-Ease, U.S.A.

8 mg of Candesartan Cilexetil, 15 mg of pioglitazone working standards were weighed and added to 50 mL dry volumetric flasks and are then diluted with diluent and sonicated. The flask was vigorously shaken before being filled with diluent solution to the desired volume. To obtain the needed standards, 1 mL of each of the two stock solutions was placed in a 10 mL volumetric flask and diluted to 10 mL.

10 tablets each containing 8 mg of Candesartan Cilexetil and 15 mg of pioglitazone was weighed, powdered and quantity equivalent to label claim, transferred to a 100 mL volumetric flask and dissolved with diluent. The solution was filtered and 2 mL of each of the filtrate i.e., 16 µg/m L of Candesartan Cilexetil and 30 µg/m L of Pioglitazone was diluted to 10 mL with diluent and analysed.

UV spectra of Candesartan Cilexetil and pioglitazone were acquired using a UV-VIS spectrophotometer in order to identify the analytical wavelength. Both stock solutions were scanned against a blank in the UV region between 200 and 400 nm.

Selecting the right solvents was a major hurdle for method development. The mobile phase is generally alluded to as the lifeline of the HPLC system. It is critical in transporting the analyte through the separation column and subsequently to the detector for identifying the separated components. Mobile phase is almost never a single solvent. It entails mixing water with organic solvents, aqueous buffers with polar solvents, or organic solvent mixtures at desired quantities. It is important to use different solvent mixtures in order to achieve desired polarity in the mobile phase for completely solubilize the sample, as well as to control interaction between analyte and stationary phase to attain the ideal degree of separation for separated components in the quickest possible time. Majority of RP-HPLC are performed with buffered aqueous solutions, such as a polar mobile phase, or with other polar solvents, such as methanol or acetonitrile. Polarity governs the sequence of solute elution in HPLC. More polar solutes elute first in RP-HPLC. Both the drugs were soluble in a bunch of solvents majorly used for reverse phase liquid chromatography such as acetone, dimethyl sulfoxide, acetonitrile, ethanol, and methanol and insoluble in water. Methanol and water are used in combination as a mobile phase applied in the initial trail of method development studies. Here, Candesartan Cilexetil was not eluted. Also, when pH of mobile phase was not adjusted, peak tailing and peak broadening occurred. Hence, in order to increase the retention of both the compounds onto the stationary phase, the amount of organic content in mobile phase was adjusted and water is substituted with phosphate buffer. Ortho phosphoric acid was used to deliver the pH to 3.5. Temperature affects the solubility of compounds in the mobile phase in HPLC. The retention time decreases as the column temperature goes up. Each 1 °C increase in temperature reduces retention time by 1–2%. The chromatographic separation process becomes faster and more efficient as the column temperature rises. As a result, column was kept at 30 °C throughout the study to avoid any disruptions with elution.

Candesartan Cilexetil showed maximum absorbance at 222 nm and Pioglitazone at 253 nm. An isosbestic point at 220 nm was observed which was taken as the detection wavelength. Also, the run was extended for 10 minutes to confirm that all traces of components had been removed and the column had been re-equilibrated. A chromatogram created over 7 minutes can be shown in Fig. 3, which displays the peak concentrations of both Pioglitazone and Candesartan Cilexetil.

Figure 3. 

Chromatogram showing peaks of Pioglitazone and Candesartan Cilexetil.

The definition of the Analytical Target Profile is a critical component of the AQbD paradigm. It provides the sum of all performance criteria needed to effectively characterize what a technique must analyze. Determining Pioglitazone and Candesartan Cilexetil simultaneously as API and in a tablet dosage form is the target and the liquid chromatographic assay method is the target method. A RP-HPLC with UV detection is selected as the analytical method. The % organic phase, flow rate of mobile phase and column temperature were chosen as the Critical Method Attributes.

Retention times of Pioglitazone and Candesartan Cilexetil, resolution, number of plates for Pioglitazone and Candesartan Cilexetil were chosen as Critical Method Parameters which were found to have effects on the chosen Critical Method Attributes. Thus, risk assessment was based on the fishbone diagram. Temperature of the column plays an important role on the analytical method and hence was chosen as essential parameter. The HPLC equipment and column require a specific flow rate. The analyte may not have enough time to interact with the stationary phase if the flow rate is higher than normal. A high flow rate reduces retention times, whereas lowering the flow rate increases resolution. This change is caused by the flow rate’s effect on the number of column plates, not by the relative peak spacing. Hence, flow rate was selected as a factor for the study. Quantity of organic phase plays a tremendous role in analyte retention in RP-HPLC. This is purely based on the polarity of the mobile phase and analytes to be separated. As the stationary phase is non-polar in a reverse phase chromatography, polar of the mobile phase has to be set carefully in order to elute the analytes from the column. Thus, this was also considered as an important factor which impacts the retention times and number of theoretical plates of an analyte.

Narrowing down of essential parameters through risk assessment identified risk factors. These were 0.9 mL/min flow rate, 40% organic phase and column temperature of 30 °C. Finally, the optimized mobile phase was 60% volume of 20 Mm Potassium dihydrogen orthophosphate and 40% of acetonitrile. Here Design of Experiment was incorporated and a flow rate of 0.8–1.0 mL/min, 35–45% organic phase and a column temperature of 27–33 °C were set as parameters. Retention times of Pioglitazone and Candesartan Cilexetil, resolution, number of plates for Pioglitazone and Candesartan Cilexetil were observed and optimized as factors. The critical method parameters were optimized using a Central Composite experimental design with three independent variables. Design Expert software was used to simulate the experimental conditions and runs were obtained. The chromatograms from all 20 trials were analysed for retention time, resolution, and count of theoretical plates for peaks of both Candesartan Cilexetil and Pioglitazone. The model prediction equations were estimated and ANOVA was used to statistically analyse the obtained results. Table 2 displays the ANOVA findings for all factors for the quadratic response surface model.

The prediction equations for all the responses are as follows:

• For RT1 : +2.505-0.276A-0.103B-0.211 C-0.126AB-0.115AC- 0.312BC +0.126 A2-0.381B2-0.291C2
• For RT2: +3.656-0.415A+0.231 B- 0.154 C-0.236 AB-0.117 AC-0.126 BC+ 0.143 A2 +0.176 B2+0.243 C2
• For RS : +8.171-0.164 A+0.903 B-0.255 C- 0.012 AB-0.137 AC-0.121 BC- 0.191 A2-0.126 B2-0.179 C2
• For TP1 : +7648.47-29.792A-457.889B+89.448C-53.5AB-195.25AC+61.75 BC-148.087 A2-786.074 B2-322.742 C2
• For TP2 : +9014.95-93.446 A-579.562 B+86.735C-68.125 AB-265.125 AC+43.125 BC-117.601 A2-841.502 B2-355.542 C2.

The co-efficient and statistical results have been tabulated below.

The DOE model was found to be statistically significant (P value < 0.0001) for predicting all responses using ANOVA. The statistical significance of the model can be inferred from its F-values. The F-value for the lack of fit for all components indicates that the lack of fit is insignificant in contrast to the pure error. This is clearly a nice fit for the model. The regression coefficients show that the variables chosen for the study had a considerable influence on the replies. The interaction terms show how the responses alter when variables are simultaneously changed. Using contour plots, we were able to better understand the relationship between variables through Fig. 4.

Figure 4. 

3-D contour plots for dependent variables in developing the optimized model. a) indicates interaction of flow rate and organic phase b) indicates interaction of flow rate and temperature c) indicates interaction of organic phase and temperature.

The design space or control space defined as the method operable design (MODR) is established by the specific CMA combinations. Fig. 5 displays the contour or desirability plots for various combinations of CMAs for the optimized method.

Figure 5. 

Contour plots for optimized method for simultaneous estimation of Pioglitazone and candesartan Cilexetil: i) Interaction between flow rate and organic phase. ii) Interaction between flow rate and temperature. iii) Interaction between organic phase and temperature.

In this design space, the selected point or a desirability region is a combination of mobile phase consisting of acetonitrile in organic phase (40%) at 30 °C column temperature with 0.9 mL/min flow rate. At this point, the predicted experimental conditions were RT1 at 2.562, RT2 at 3.635, RS at 3.521, TP1 at 6697 and TP2 at 7613. The predicted experimental conditions were confirmed experimentally with RT1 at 2.462, RT2 at 3.598, RS at 3.529, TP1 at 6458 and TP2 at 7365.The design space depicts this method operable zone where alterations will have no effect on the analysis quality. Design space is an essential component of QbD. It was established by creating a plot by overlaying plots of independent variables with their ranges to meet the required response while keeping one variable constant. The yellow color in this plot denotes the region when all response variable requirements are met and no revalidation is necessary.

Figure 6. 

Overlay plots for: i) Interaction of flow rate and organic phase. ii) Interaction of flow rate and temperature. iii) Interaction of organic phase and temperature.

The validation parameters such as accuracy, precision, linearity and robustness of the optimized method were evaluated at a selected working point. The linearity equation is determined and co-efficients are shown in Table 3 analysed from Fig. 7.

Figure 7. 

Linearity overlay chromatogram for Pioglitazone and Candesartan Cilexetil.

The r² values derived by linear regression analysis were 0.9999 and 0.9997 respectively for Pioglitazone and Candesartan Cilexetil demonstrating the linearity between peak area and the drug concentration. The calculated LOD, LOQ values mentioned in Table 3 assured that the technique is sufficiently sensitive. Through accuracy and precision investigation, % recoveries ranged from 98 to 99 for Pioglitazone and from 98 to 100 for Candesartan Cilexetil. Precision in terms of % RSD were < 1% as mentioned in Table 4.

The percentage recovery of both drugs when tested simultaneously was found to be within limits. Both Pioglitazone and Candesartan Cilexetil had percentage recoveries that were statistically acceptable. It was also revealed that the drug peaks did not interfere with one another. As a result, the approach can be said to be specific to the two drugs in combination. The method‘s ruggedness/robustness was evaluated by purposefully modifying chromatographic parameters like % mobile phase, flow rate, and column temperature. Changes in experimental conditions by increasing and decreasing the flow rate by 0.1 m L/ min, by expanding and diminishing the temperature of the column by 5 degrees and by altering the organic phase by 5% had little effect on chromatogram quality parameters as retention time, resolution, and theoretical number of plates as mentioned in Table 4. % RSD was found to be within limits. As a result, the proposed method can be used for simultaneous, quantitative analysis for the combination of Pioglitazone and Candesartan Cilexetil.

The outcomes of the Candesartan Cilexetil and pioglitazone assays were compared to the labelled quantities. The chromatogram obtained for analysis of the prepared formulation, which shows that there are no additional peaks, indicating that the formulation excipients employed in the tablet do not interact. Candesartan Cilexetil and pioglitazone were found to have a % assay of 100.04% and 99.50%, respectively. These values were within the % limit.

Sample chromatograms were exposed to a variety of forced degradation conditions. To induce the stress conditions, 0.1M HCL, 0.1M NaOH, H₂O₂, UV light and high temperatures were used. The solutions underwent stress conditions from 6–7 hrs. This showed an additional peak along with analytes depicted in Fig. 8 at the end of stress duration. However, chromatograms in neutral circumstances did not exhibit distinct peaks for the degradation products. When exposed to acidic deterioration, both Candesartan Cilexetil and Pioglitazone degraded significantly more than under any other stress state. Summary of degradation studies was reported in Table 5.

Figure 8. 

chromatograms of analytes subjected to varying levels of forced degradation: A) acidic degradation, B) alkali degradation, C) neutral, D) thermal degradation, E) peroxide degradation, F) UV-light degradation.