Use of chronic vitamin K antagonist (VKA) induces a long-term deficiency of vitamin K, which may cause arterial stiffness and bone-related disease. Switching from VKA to rivaroxaban could induce rapid sufficiency of vitamin K and improvement of arterial stiffness. The K2 SUMMIT-3 study is a multicenter, open-label, prospective, and randomized design. Patients with atrial fibrillation who have been taking VKA for more than 6 months but less than 10 years were randomly assigned to two groups; those switching from VKA to rivaroxaban and those continuing with VKA medication. The primary endpoint was the percentage difference of brachial-ankle pulse wave velocity (baPWV) in 3 months. A total of 77 patients were randomly assigned to receive rivaroxaban (n = 38) or VKA (n = 39). The average age was 74 ± 9 years. The duration for which VKA was prescribed prior to randomization was 90 ± 87 months.

Abnormally high levels of Des-gamma carboxyprothrombin (PIVKA-II) or uncarboxylated osteocalcin (ucOC) indicating vitamin K insufficiency were observed in 100% or 82% of the patients at baseline but it reduced to 2% (p < 0.0001) or 55% (p = 0.01) at 3 months in the rivaroxaban group. To the contrary, theses data had no changes in the VKA group. The percentage difference in baPWV was − 1.4 ± 10.0% vs. 3.5 ± 14.7% in the rivaroxaban and the VKA groups, respectively. (p = 0.02). Switching from VKA to rivaroxaban resulted in rapid sufficiency of vitamin K and reduction of arterial stiffness in 3 months.