Optimization of a liquid chromatography method for the analysis of related substances in daclatasvir tablets using design of experiments integrated with the steepest ascent method and Monte Carlo simulation.

Analytical method for the determination of related substances (RS) in Daclatasvir tablets was optimised using quality by design (QbD) approach. Seven degradants (each more than 1.0%) generated during oxidation study, adversely affected the selectivity of the method. Coelution of the degradant peaks with API and known impurities, suggested failure in developing a stability indicating method. To overcome the shortcomings and develop a robust method, QbD principles were incorporated. Resolution was the critical quality attribute (CQA) and buffer pH, column oven temperature, gradient slope and flow rate were the critical method variables (CMVs) studied through design of experiments (DoE). Discovery of an unknown impurity (named as impurity D, about1.0%) was a key finding from this DoE study. The most crucial responses viz. Resolution between impurity D and the main peak and resolution between the main peak and impurity E demanded contradictory pH requirements. To select the right pH, responses were prioritised and eventually to attain the desired resolution between Daclatasvir and impurity E the value for pH was fixed to 3.0. Next, to improve resolution between impurity D and Daclatasvir, method of steepest ascent was applied to locate an apt value for column oven temperature. Accordingly, experiments were performed at different temperatures along the path of rapid increase in response. Finally, at 45 °C (pH :3.0), both the critical pairs were well resolved. The global optimum was determined through a Response surface methodology (RSM) design with pH and column oven temperature as critical factors. pH 3.0, column oven temperature 44 °C, % MP. B 45% and flow rate 1.0 mL min-1 was found to be the optimum condition. Further, the design space was complimented by establishment of a robust zone through Monte Carlo simulation and capability analysis. An analytical control strategy (ACS) was set up to ensure that the method repeatedly meets its acceptance criteria. The optimised method was successfully validated within the factor ranges mentioned in the ACS. Despite various intricacies, the QbD approach facilitated systematic optimisation of a stability indicating robust method.

A simultaneous determination of related substances by high performance liquid chromatography in a drug product using quality by design approach.

The combination of Abacavir, Lamivudine and Dolutegravir is an anti-retroviral formulation that displays high efficacy and superiority in comparison to other anti-retroviral combinations. Analysis of related substances in this combination drug product was very challenging due to the presence of nearly thirty peaks including the three active pharmaceutical ingredients (APIs), eleven known impurities and other pharmaceutical excipients. Objective of this study was to develop a single, selective, and robust high performance liquid chromatography method for the efficient separation of all peaks. Initially, one-factor-at-a-time (OFAT) approach was adopted to develop the method. But, it could not resolve all the critical peaks in such complex matrix. This led to the advent of two different HPLC methods for the determination of related substances, one for Abacavir and Lamivudine and the other for Dolutegravir. But, since analysis of a single sample using two methods instead of one is time and resource consuming and thus expensive, an attempt was made to develop a single and robust method by adopting quality by design (QbD) principles. Design of Experiments (DoE) was applied as a tool to achieve the optimum conditions through Response surface methodology with three method variables, pH, temperature, and mobile phase composition. As the study progressed, it was discovered that establishment of the design space was not viable due to the completely distant pH requirements of the two responses, i.e. (i) retention time for Lamivudine carboxylic acid and (ii) resolution between Abacavir impurity B and unknown impurity. Eventually, neglecting one of these two responses each time, two distinguished design spaces have been established and verified. Edge of failures at both design spaces indicate high probability of failure. It therefore, becomes very important to identify the most robust zone or normal operating range (NOR) within the design space with low risk of failure and high quality assurance. For NOR establishment, Monte Carlo simulation was performed on the basis of which process capability index (Cpk) was derived. Finally, the selectivity issue problem faced due to the pH dependency and the dissimilar pH needs of the two critical responses was resolved by introducing pH gradient into the program. This new ternary gradient program has provided a single robust method. Thus, two HPLC methods for the analysis of the combination drug product have been replaced with a selective, robust, and cost effective single method.