Effects on Bone Metabolism Markers and Arterial Stiffness by Switching to Rivaroxaban from Warfarin in Patients with Atrial Fibrillation

In recent years, direct oral anticoagulants (DOACs) of dabigatran, rivaroxaban, apixaban, edoxaban, which are all alternatives to warfarin, have been released. The use of DOACs is becoming more widespread in the clinical management of thrombotic stroke risk in patients with atrial fibrillation (AF). In large-scale clinical trials of each drug, DOACs were reported to inhibit intracranial hemorrhage, stroke, and death compared to warfarin. Warfarin is an endogenous vitamin K antagonist; therefore, patients who are taking warfarin must be prohibited from taking vitamin K. Vitamin K is an essential cofactor required for the ɤ-carboxylation of vitamin K-dependent proteins including coagulation factors, osteocalcin (OC), matrix Gla protein (MGP), and the growth arrest-specific 6 (GAS6). OC is a key factor for bone matrix formation. MGP is a local inhibitor of soft tissue calcification in the vessel wall. GAS6 prevents the apoptosis of vascular smooth muscle cells. Therefore, decrease of blood vitamin K levels may cause osteoporosis, vascular calcification, and the inhibition of vessels angiogenesis. This study aimed to evaluate the effects of changing from warfarin to rivaroxaban on bone mineral metabolism, vascular calcification, and vascular endothelial dysfunction. We studied 21 consecutive patients with persistent or chronic AF, who were treated with warfarin at least for 12 months. Warfarin administration was changed to rivaroxaban (10 or 15 mg/day) in all patients. Osteopontin (OPN), bone alkaline phosphatase (BAP), and under-carboxylated osteocalcin (ucOC) were measured. Pulse wave velocity (PWV) and augmentation index (AI) were also measured as atherosclerosis assessments. All measurements were done before and six months after the rivaroxaban treatment. There was a significant increase in serum level of BAP compared to baseline (12.5 ± 4.6 to 13.4 ± 4.1 U/L, P < 0.01). In contrast, there was a significant decrease in the serum level of ucOC (9.5 ± 5.0 to 2.7 ± 1.3 ng/ml, P < 0.01). Also, in the ucOC levels, there was a significant negative correlation between baseline values and baseline to 6-months changes in high ucOC group (r = -0.97, P < 0.01). The atherosclerosis- and osteoporosis-related biomarker, serum level of OPN were significantly decreased compared to baseline (268.3 ± 46.8 to 253.4 ± 47.1 ng/ml, P < 0.01). AI and PWV were significantly decreased after 6 months of treatment with rivaroxaban (33.9 ± 18.4 to 24.7 ± 18.4%, P = 0.04; 1638.8 ± 223.0 to 1613.0 ± 250.1 m/s, P = 0.03, respectively). Switching to rivaroxaban from warfarin in patients with atrial fibrillation was associated with an increase of bone formation markers and a decrease of bone resorption markers, and also improvements of PWV and AI.