Quality by Design for the optimization of a fast HPLC-DAD method for the analysis of cannabinoids in magistral preparations

Cannabis sativa contains a large variety of different chemical compounds, among which cannabinoids are generally accepted as the main medicinal ingredients. Already more than 70 different cannabinoids have been identified so far, anyway, among the major cannabinoids present in Cannabis sativa Cannabidiol (CBD) and Δ9-tetrahydrocannabinol (THC) are the compounds that possess the most relevant psychoactive properties. A fast and simple HPLC method for the quantitation of these two cannabinoids in magistral oil preparations has been developed and validated following the Quality by Design (QbD) principles. Risk assessment and Design of Experiments represent fundamental tools of this new systematic approach. QbD leads to the establishment of the Design Space (DS), namely the multidimensional space of combination and interaction of variables that have been demonstrated to provide quality. A scouting system was set-up to test different types of organic solvents. A mixture of acetonitrile and phosphate buffer led to good selectivity and fast time of analysis. The Critical Quality Attributes (CQAs), representative of the quality of the chromatogram, were resolution and analysis time. The Critical Process Parameters (CPPs) were selected using a fishbone diagram and were related to the characteristics of both the mobile phase and the column: flow, temperature, pH of the mobile phase. In the Response Surface Methodology, a quadratic polynomial model was postulated to link CQAs to CPPs. The models were calculated by means of Doehlert Design and were graphically represented by contour plots. The search for the global optimum zone was performed by the sweet spot plots, which were analysed in order to identify the zones where the multidimensional combinations of the CPPs values allowed the desired values for both the CQAs to be obtained (resolution value ≥0.85, analysis time ≤6 min). The DS was identified by a risk of failure map and was defined on the basis of the calculated models and on the basis of Monte Carlo simulations. The final conditions selected for the analysis were (with the related DS range): chromatographic column, Poroshell® 120 SB-C18 (150 mm x 2.1 mm i.d., particle diameter 2.7 µm, Agilent Technologies); type of organic eluent, acetonitrile/phosphate buffer (75:25 v/v); flow, 0.38 mL min-1 (0.31-0.39 mL min-1); temperature, 53 °C (44-57 °C); pH of the mobile phase, 3.45 (3.10-3.50). The baseline separation of the analytes was obtained in less than 5 min. In order to validate the DS, additional points at the extremities of the DS were tested, selected by a Plackett-Burman matrix, verifying the agreement between measured and predicted values for the CQAs. The same type of matrix was successively used for method robustness testing. Finally, the method was applied to a real sample of magistral oil preparation.