Development, validation and application of a new HPLC-DAD method for simultaneous quantification of apixaban, dabigatran, edoxaban and rivaroxaban in human plasma.

Direct oral anticoagulants (DOACs) have been commonly used for the treatment of venous thromboembolism and for the prevention of stroke in patients with atrial fibrillation. Despite not being initially recommended, monitoring DOACs plasma concentrations is now recognized as essential in emergency situations and in special populations. Moreover, the inter-individual variability found in real studies as well as the high reported non-adherence are corroborating the importance of determining the individual relationship between administered doses, plasma concentrations and pharmacological effects. Therefore, accurate but user-friendly bioanalytical techniques are required to monitor DOACs plasma concentrations in routine clinical practice and phase IV clinical trials. Herein, a fast and simple high performance liquid chromatography (HPLC) method coupled to diode array detection (DAD) was developed, validated and applied to quantify the four currently marketed DOACs (apixaban, edoxaban, dabigatran and rivaroxaban). Sample preparation was performed by solid phase extraction followed by evaporation and concentration of the analytes. Chromatographic separation was accomplished within 6 min on a reversed-phase column (octadecyl-silica packing material; 55 mm × 4 mm, 3 μm particle size), applying a mobile phase composed of an aqueous solution of formic acid (0.1 %, v/v) and acetonitrile, pumped with a gradient elution at 30 °C. The proposed method was linear (r² ≥ 0.993) within the concentration ranges of 0.017-5.28 μg mL^-1, 0.066-5.28 μg mL^-1, 0.033-5.28 μg mL^-1 and 0.017-5.28 μg mL^-1 for apixaban, dabigatran, edoxaban and rivaroxaban, respectively, all of them including the expected range of therapeutic concentrations. Overall, intra- and inter-day trueness of quality control samples, including at the lower limit of quantification (LLOQ), varied between -12.98 to 5.79 %, while imprecision was lower than 16.43 %, supporting that the method is accurate and precise in accordance to international guidelines. Recovery and stability were also assessed and allowed the method to be applied in clinical practice, during therapeutic drug monitoring.