For several decades, oral anticoagulants have been used in the treatment and prevention of venous thrombosis. Oral anticoagulants block the vitamin K-dependent liver production of the plasma clotting factors II (prothrombin), VII, IX and X. They have a relatively narrow therapeutic window that requires close international normalised ratio (INR) monitoring – overdosing may result in lifethreatening bleeding and underdosing inefficacy. Recently, some major improvements in the monitoring of oral anticoagulation have been made. The efficacy and safety of oral anticoagulation were found to be correlated with the reached INR values in trials in patients with AF, in those with artificial heart valves and in those after myocardial infarction (MI). Moreover, INR self-monitoring, which may even be more efficient than laboratory monitoring, has become a reality.

However, oral anticoagulation remains a laborious and poorly predictable therapy. Recently, oral direct thrombin inhibitors have been introduced. These agents lack the need for anticoagulant monitoring. In a large clinical trial on venous thromboprophylaxis, ximelagatran showed better efficacy than low molecular-weight heparin and in the large Efficacy and Safety of oral direct Thrombin inhibitor ximelagatran in patients with recent Myocardial damage (ESTEEM) study in coronary artery disease (CAD), ximelagatran plus aspirin showed superiority over aspirin alone. After a proper dose-finding study the drug was tested against warfarin in patients with AF in two large trials: Stroke Prevention using ORal Thrombin Inhibitor in atrial Fibrillation (SPORTIF)-III and SPORTIF-V.

In patients with NVAF, ximelagratran 36kg twicedaily in 3,407 patients in an open-label design and in 3,922 patients in a double blind set-up proved not inferior to warfarin (INR 2:3) in stroke prevention with similar major but less minor bleeding (see Table 1). Transient liver enzyme elevations were seen in up to 3% with 24mg, and 6% and 7% with 36mg twice-daily in the SPORTIF trials and ESTEEM, respectively, as was seen in previous trials. Recently, the new oral direct thrombin blocker dabigatran has been evaluated in a 12-week dose-finding warfarincontrolled study in 502 patients with AF. It shows an acceptable efficacy and safety profile, but liver enzyme elevation was only seen in less than 1% of patients being treated with dabigatran (see Table 2). These results are the basis for the very large phase- III trial of dabigatran versus warfarin (RELY).

Not only direct thrombin inhibitors have been tested for stroke prevention in AF. The novel once-weekly subcutaneous factor Xa-specific pentasaccharide idraparinux was compared with warfarin in the Prevention of Thromboembolic Events, Warfarin Versus Idraparinux (pentasaccharide) (AMADEUS) study of 5,700 patients. Unfortunately, this trial was prematurely terminated due to increased severe bleeding in idraparinux-treated patients. Possibly, the very-long-acting pentasaccharide cannot be adequately antagonised in case of bleeding. In the near future, oral facto