Association between plasma concentration of edoxaban determined by direct and indirect methods in Japanese patients with non-valvular atrial fibrillation (CVI ARO 7)

Direct oral anticoagulants, including edoxaban, primarily do not need routine monitoring of the anticoagulant effect. However, extremely high/low plasma concentrations of edoxaban (PC-Ed) should be properly evaluated, especially when patients under anticoagulation therapy are at an emergency state. For this purpose, PC-Ed determined by an anti-Xa assay (indirect PC-Ed) is more convenient and, therefore, more useful compared with PC-Ed determined by an LC–MS/MS (direct PC-Ed) in daily clinical practice. Consecutive 97 patients with non-valvular atrial fibrillation (NVAF) under edoxaban therapy were evaluated, in whom edoxaban 60/30 mg doses were prescribed for 48/49 patients, 71 (73.2%) were men, and the average age was 69 years. CHADS₂ score 0, 1, and ≥ 2 were 26.8%, 44.3%, and 28.9%, while CHA₂DS₂-VASc score 0, 1, and ≥ 2 were 14.4%, 16.5%, and 69.1%, respectively. Median values of direct and indirect PC-Ed by LC–MS/MS and anti-Xa assay were 187.1 and 176.1 ng/mL at peak (2–4 h post-dose) and 14.4 and 17.5 ng/mL at trough (pre-dose), respectively. The PC-Ed at peak and trough by two methods were significantly correlated, and the correlation coefficients were r = 0.973 and 0.963 (both, p < 0.0001), respectively. By a Bland–Altman plot, mean differences between the direct and indirect PC-Ed [lower to upper percent limit of agreement] were − 4.87 [− 46.71 to 36.98] and 4.66 [− 1.37 to 10.69] ng/mL at peak and trough, respectively. Moreover, mean % error for difference between the direct and indirect PC-Ed [lower to upper percent limit of agreement] was − 1.22 [− 20.59 to 18.14] and 31.75 [− 14.03 to 77.53] % at peak and trough, respectively, where the % error extremely increased around the lower limit of detection (LLOD) in the anti-Xa assay. Strong similarity was observed between the direct and indirect PC-Ed, especially at peak. The indirect PC-Ed was higher than the direct PC-Ed, especially around the LLOD, suggesting the need for caution when we use the anti-Xa assay for measurement of trough PC-Ed (UMIN 000032492).