Warfarin, also known as Coumadin, is a vitamin K antagonist inhibiting the synthesis of clotting factor II, VII, IX, and X. Vitamin K is an essential component for hepatic synthesis of clotting factor II, VII, IX, and X. Therefore, use of warfarin leads to deficiency of the vitamin K-dependent clotting factors and anticoagulation effect. The effective half-life of warfarin ranges from 20 to 60 h, and the anticoagulation effect lasts for two to 5 days. For decades, warfarin has been the most commonly used anticoagulants in the world. However, proper use of warfarin is frequently restricted by the narrow therapeutic range and need for permanent monitoring of anticoagulant activity by international normalized ratio (INR). Even in patients with well-controlled INR (2–3) on warfarin, INR should be monitored at least every 12 weeks, and out of therapeutic range is common. Warfarin also has multiple interactions with drugs and foods which interrupt maintenance within the narrow therapeutic range. The safety and effectiveness of warfarin are closely associated with the quality of anticoagulation. Low INR is ineffective for ischemic stroke, and bleeding risk significantly increases proportionally with international normalized ratio. In patients with AF taking warfarin, time in therapeutic range of INR (TTR) is commonly suboptimal and varied from 30 to 80% [2]. Both a highly variable INR and low TTR of INR are independent predictors of bleeding and thromboembolic complications. Mean TTR below 60% indicates that warfarin is inefficient, and consideration for a switch to one of the NOACs might be reasonable. Although warfarin has many clinical limitations, adjusted-dose warfarin significantly and efficiently reduces risk of ischemic stroke and systemic embolism in patients with atrial fibrillation (AF) (about 50–60% risk reduction compared to placebo). Compared to antiplatelet, warfarin reduces risk of ischemic stroke and systemic embolism to about one third. In clinical trials with warfarin and NOAC in non-valvular AF, subgroup patients with well-controlled warfarin therapy (TTR > 66%) were not significantly inferior for the prevention of thromboembolism to NOACs [3]. To improve INR control and therapeutic potential of warfarin treatment, validated decision support tools (paper nomogram or computerized dosing program) and patient self-test of INR using point-of-care system might be effective. One of the common misconceptions is that persons on warfarin should avoid taking foods containing vitamin K. However, warfarin therapy needs stable therapeutic level of INR derived by maintaining consistent dietary habit for vitamin K-containing foods which is balanced with dosage of warfarin intake. Avoidance of whole food containing vitamin K is impossible and not necessary.

To overcome the clinical limitations of vitamin K antagonist, a number of NOAC were developed. Currently four NOACs (dabigatran, rivaroxaban, apixaban, and edoxaban) got an approval from the European Medicines Agency (EMA) and US Food and Drug Administration (FDA) after clinical trials for the prevention of thromboembolism in patients with non-valvular AF. These NOACs have no interaction with food and have stable therapeutic effect with fixed dosage, and there is no need for routine monitoring of coagulation assay. Dabigatran is a selective inhibitor of factor IIa, and rivaroxaban, apixaban, and edoxaban are selective inhibitors of factor Xa. Because the underlying characteristics of study participants varied on each clinical trials with NOACs, direct compartment among the NOACs is impossible. Generally, NOACs for non-valvular AF have comparable preventive effect for ischemic stroke or systemic thromboembolism compared to conventional warfarin treatment and have relatively low bleeding complications especially in hemorrhagic stroke (Table 29.1). On review of the individual trials, apixaban and dabigatran 150 mg are superior to warfarin for the prevention of stroke or systemic thromboembolism. Dabigatran 150 mg bid is the only NOAC which showed a significant reduction of risk of ischemic stroke compared to warfarin (RR 0.76, 95% CI 0.60–0.98). All-cause mortality is significantly reduced with apixaban (HR 0.89, 95% CI 0.80–0.99) and a low dose of edoxaban (HR 0.87, 95% CI 0.79–0.96). However, a low dose of edoxaban is inferior to warfarin therapy in preventing ischemic stroke. Major bleeding risk with NOACs is significantly low or comparable to the risk with warfarin, but there is concern for increase of GI bleeding with a high dose of edoxaban, dabigatran 150 mg, and rivaroxaban compared to warfarin. Because kidney function can influence drug metabolism and excretion of NOACs, kidney function should be considered before prescription of NOACs. Currently, all NOACs are contraindicated for patients with end-stage renal disease. For patients with reduced kidney function, dose adjustment is needed. Table 29.2 summarizes the recommended regimen of NOACs according to renal function. In subgroup analysis from the ENGAGE AF trial, the risk of ischemic stroke is higher than in warfarin in patients with creatinine clearance >95 mL/min; therefore, the FDA does not recommend edoxaban for patients with higher creatinine clearance.

Unfractionated heparin has a binding capacity to antithrombin III (ATIII) and induces conformational change of ATIII. The heparin-enhanced ATIII leads to inactivation of thrombin and factor Xa, which induces anticoagulative effect. Unfractionated heparin also has some anticoagulation effect through inhibition of factor IXa, XIa, and XIIa. Unfractionated heparin has a half-life of 1–2 h and can reach therapeutic range immediately upon intravenous administration. Although unfractionated heparin is the most commonly used parenteral anticoagulant, it has a number of major limitations including a narrow therapeutic window and highly variable dose-response relation that requires laboratory monitoring of activated partial thromboplastin time and dose adjustment. Low-molecular-weight heparins (LMWH) also bind to ATIII, but have more selective inhibitory effect on factor Xa than on IIa. With subcutaneous administration of LMWH, the peak level is reached at 2–4 h and the half-life is 3–4 h. LMWH has many merits compared to UFH including consistent dose-related anticoagulant response which leads to low bleeding complication and usage at fixed dose based on body weight; therefore, dose adjustment or monitoring of LMWH is not necessary except in patients with renal insufficiency. For many years, parenteral anticoagulation using heparin or low-molecular-weight heparin is commonly administrated for acute ischemic stroke patients for the prevention of secondary ischemic events. However, there is lacking evidence supporting the early use of parenteral anticoagulation even in patients with cardioembolic source. Some clinical data suggest parenteral anticoagulation may reduce early recurrent ischemic stroke, but there was significantly increased risk for symptomatic hemorrhage resulting in no net benefit [4]. Therefore, guidelines by the American Heart Association (AHA) and European Stroke Organization do not recommend the use of early parenteral anticoagulation for the treatment of patients with acute ischemic stroke. Even it is not for the prevention of stroke, LMWH can be recommended for prophylaxis of deep vein thrombosis to the immobilized patients with acute stroke.

Bivalirudin (commercial name: Angiomax) is a highly specific direct thrombin inhibitor which binds to both free and clot-bound thrombin resulting in inhibition of thrombin-mediated fibrin formation and thrombin-mediated platelet activation. Bivalirudin is primarily indicated for anticoagulants during percutaneous coronary intervention in place of heparin. By intravenous administration, bivalirudin has immediate anticoagulation effect with predictable dose response and short half-life time (about 30 min in normal renal function) without binding to plasma proteins. Therefore, bivalirudin has more predictable anticoagulative effect than unfractionated heparin and has low rate of bleeding complication in trials with percutaneous coronary intervention. Bivalirudin can be used for patients who were contraindicated with heparin due to heparin-induced thrombocytopenia or heparin-induced thrombosis–thrombocytopenia syndrome. Considering the more frequent intracranial hemorrhage during neuroendovascular procedure than percutaneous coronary intervention, bivalirudin might be an ideal alternative anticoagulant to heparin. However, there is insufficient data for the efficacy and safety of bivalirudin during the neuroendovascular procedures.

AF is the most common cause of cardioembolic stroke and well-established indications which need long-term anticoagulation. In the USA, approximately 2% of people under the age of sixty-five have AF, and 9% of people over or equal to the age of sixty-five have AF [5]. As the prevalence of AF dramatically increases with age, the burden of AF and the AF-related stroke consistently increases. Without proper anticoagulation, AF increases the risk of stroke to about four to five times. If mitral valve stenosis is coexistent with AF, the risk increases up to 15-fold. AF is subdivided into paroxysmal or persistent by the duration of AF, but the stroke risk is considered similar between them. In the prior randomized trials, long-term oral anticoagulation with warfarin has been established as the effective management for the prevention of ischemic stroke rather than placebo (OR 0.34) or antiplatelet (OR 0.53) [6]. However, use of warfarin significantly increases systemic bleeding complication even in well-controlled therapeutic range. In AF patients, the absolute risk reduction with anticoagulation is dependent on the underlying thromboembolic risk; those with large risk have more absolute risk reduction with anticoagulation. For patients with low thromboembolic risk, routine use of anticoagulation is not recommended due to the low absolute risk reduction and expected bleeding complication. Therefore, risk stratification is essential for identification of candidates who get preventive benefit with long-term anticoagulation. CHADS₂ score is an established prediction tool for risk stratification in patients with non-valvular AF. The CHADS₂ scheme has a 0 to 6 score (each 1 point for congestive heart failure, hypertension, age ≥ 75 years, and diabetes mellitus and 2 points for prior stroke or transient ischemic attack), and there was approximately 2.0% increase of stroke rate for each 1 point increment of CHADS₂ score (Table 29.3) [5]. Based on the prior studies, non-valvular AF patients with CHADS₂ score ≥ 2 is indicated for long-term oral anticoagulation with the significant net benefit with prevention of stroke. All AF patients with history of ischemic stroke or transient ischemic attack (CHADS₂ score ≥ 2) should get long-term anticoagulation in the absence of major contraindication such as bleeding complication. CHADS₂ score is validated and a widely used risk assessment tool for the need of anticoagulation in non-valvular AF. However, CHADS₂ score has some limitations. Substantial portion of patients were classified as moderate risk (CHADS₂ score of 1) and it is uncertain whether anticoagulation or antiplatelet is better for them. CHADS₂ score has wide confidence intervals for each score and inadequate discrimination ability identifying truly low-risk patients who do not need neither anticoagulation nor antiplatelet. CHA₂DS₂-VASc score is a modified scoring system based on CHADS₂ score incorporating other common risk factors which provide better discrimination of truly low-risk and high-risk patients who need anticoagulation. The 2016 European Society of Cardiology (ESC) and 2014 American College of Cardiology (ACC)/AHA guidelines recommended the use of CHA₂DS₂-VASc score as the risk assessment tool in non-valvular AF [5, 7]. With non-valvular AF with CHA₂DS₂-VASc score ≥ 2, oral anticoagulation is recommended in the absence of major contraindications. If patients with AF and CHA₂DS₂-VASc score ≥ 2 are not suitable for anticoagulation, dual antiplatelet (aspirin plus clopidogrel) is recommended for the prevention of stroke. For non-valvular AF patients with CHA₂DS₂-VASc score of 1, optimal antithrombotic strategy is controversial, and recommendations varied between several guidelines for non-valvular AF (Table 29.4). The 2016 ESC guideline for AF management recommended oral anticoagulation for the patients with CHA₂DS₂-VASc score of 1. On the other hand, 2014 AHA/ACC guidelines for AF stated that no antithrombotic therapy or treatment with oral anticoagulation or aspirin may be considered for those with CHA₂DS₂-VASc score of 1 [5]. For the truly low-risk group with CHA₂DS₂-VASc score of 0, no antithrombotic therapy is more reasonable than the use of antiplatelet. The 2016 ESC guideline considered females with CHA₂DS₂-VASc score of 1 (no risk factor except female) as low-risk group which do not need antithrombotic therapy [7].