The oral vitamin K antagonist (VKA) warfarin, a Coumadin derivative, is a very effective anticoagulant drug in preventing thromboembolic events. Its mechanism of action is inhibition of vitamin K-dependent coagulation factors II, VII, IX, and X, which are produced by the liver [2].

The pharmacokinetics of warfarin is complex and therefore requires periodical monitoring, laboratory testing, and adjustment of the dose. The levels of anticoagulation desired depend on the medical condition and are assessed by the international normalized ratio (INR).

Several factors interfere with the response to warfarin treatment, such as genetic and environmental factors. Some mutations in the gene coding for the hepatic enzyme cytochrome P450 might provoke a higher incidence of bleeding [3]. Among the environmental factors, warfarin can interact with other medications (and herbs) that the patient might be concomitantly taking and with the patient’s diet. Therefore, due to the narrow therapeutic index range and the diverse factors that might alter its response [4], warfarin therapy carries a higher risk of hemorrhage, ranging in magnitude from 1.0% to 7.4% per year; this risk is increased in the older population, those with a history of stroke, and concomitant use of other agents that increase bleeding risk [5]. Thus, the decision of whether or not to place a patient on warfarin therapy must involve carefully balancing antithromboembolic benefit versus risk of bleeding [6].

Several management strategies have been described for the operative patient with AF on warfarin [7]. One of the key factors to be taken into consideration is the risk of thromboembolism. The guidelines of the American College of Chest Physicians (ACCP) classify the risk for thromboembolism according to the medical condition that mandated the use of anticoagulant treatment, as well as the coexistence of other comorbidities [8]. These guidelines consider three groups: high risk, moderate risk, or low risk according to the indication for antithrombotic therapy.

High risk for arterial thromboembolism includes one of the following: CHADS₂ score of >5, a recent (within 3 months) stroke or transient ischemic attack, or rheumatic valvular heart disease. Patients at moderate risk include those with a CHADS₂ 2–4. Patients with low risk for thromboembolism include those with a CHADS₂ score of <2 and without history of prior stroke or transient ischemic attack. The overall risk of a thromboembolic event in the low-risk patient population is <5% per year.

According to the guidelines established by the ACCP, patients at high risk of developing a thromboembolic event and receiving chronic VKA therapy should be bridged with therapeutic dosages of low molecular weight heparins (LMWH). The authors of these guidelines prefer bridging with LMWH as opposed to unfractionated heparin (UFH). Ten cohort studies assessing bridging anticoagulation therapy in approximately 1,400 patients with chronic atrial fibrillation were reviewed. They concluded that the overall risk for perioperative arterial thromboembolism was 0.57% when bridging anticoagulant therapy was used. Warfarin should be stopped 5 days preoperatively and bridging anticoagulation initiated. Several therapeutic dose regimens have been studied with similar results; these include dalteparin 200 IU/kg daily, enoxaparin 1.5 mg/kg daily, tinzaparin 175 IU/kg daily, dalteparin 100 IU/kg twice per day, and enoxaparin 1 mg/kg twice per day. The last dose of therapeutic subcutaneous LMWH should be 24 h prior to surgery or procedure start time. It is also recommended that the last dose should consist of only half the recommended dose. For those patients receiving UFH, it is recommended to stop UFH 4 h before surgery or procedure start time.

If neuraxial anesthesia is indicated, the recommendation is to postpone needle placement for at least 10–12 h after prophylactic LMWH dose. In patients receiving therapeutic doses of LMWH, such as dalteparin 120 U/kg BID, dalteparin 200 U/kg QD, enoxaparin 1.5 mg/kg BID, or tinzaparin 175 U/kg QD, the recommendation is to delay needle placement for a minimum period of 24 h. If LMWH was administered 2 h preoperatively, neuraxial anesthesia should be avoided. There are no contraindications to neuraxial anesthesia for patients receiving UFH 5,000 U BID. However, in those patients receiving more than BID dosing or greater than 10,000 U of UFH daily, we recommend frequent neurologic exams if neuraxial technique has been performed. UFH should be administered 1 h after needle placement and the indwelling catheter removed 2–4 h after the last dose [9].

In high-risk patients requiring therapeutic LMWH post-procedure, the risks of achieving inadequate hemostasis must be weighed carefully against the risk of developing a perioperative thromboembolic event. For minor surgical or invasive procedures where the risk of hemorrhage is remote, anticoagulation with either therapeutic LMWH or UFH should be resumed 24 h after the surgery or procedure. In those undergoing major surgery or procedures with a high risk of bleeding, therapeutic LMWH or UFH should be initiated 48–72 h after surgery or when surgical hemostasis has been achieved [7]. An alternative for patients undergoing procedures with a high risk of bleeding is to start intravenous UFH as soon as hemostasis has been confirmed with the goal of maintaining a partial thromboplastin time two times the normal range, to allow for tighter control of postoperative anticoagulation since the intravenous administration of heparin can easily be stopped and reversed with protamine. Warfarin is restarted prior to discharge [10]. Perioperative anticoagulation therapy should be continued until the warfarin level is therapeutic as defined by the indicated patient-specific goal (INR).

Much controversy exists surrounding the management of intermediate- and low-risk patients on chronic VKA therapy regarding the necessity of bridging strategies, especially in patients with no preexisting rheumatic disease and atrial fibrillation. Patients with atrial fibrillation with moderate risk of perioperative arterial thromboembolism, history of a prior venous thromboembolism (VTE), or with a mechanical heart valve should receive therapeutic LMWH, therapeutic UFH, or low-dose LMWH as opposed to no bridging anticoagulation during temporary cessation of chronic VKA regimen (Grade 2C recommendation) [7].

Contrary to the recommendations of the aforementioned ACCP guidelines, there are two uncontrolled observational studies that investigated the effects of not “bridging” patients after discontinuation of OAC. A prospective observational study by Garcia et al. investigated the occurrence of thromboembolic events 30 days postoperatively in a cohort of 1,024 patients whose chronic VKA therapy was interrupted. Five hundred and fifty of these patients had atrial fibrillation, making atrial fibrillation the largest subgroup within this study population. Of these 550 patients, 535 did not receive any bridging anticoagulation therapy. The authors reported that four patients developed either a stroke or a systemic embolic event (0.7%, 95% CI, 0.2–1.9%). In this cohort subset, 90 patients had >3 risk factors for developing arterial thromboembolism; therefore, the majority of patients were within the intermediate- to low-risk categories as defined by the authors. However, it must be noted that risk stratification for arterial thromboembolism was not based on the CHADS₂ classification, and two of the four reported cases of arterial thromboembolism were in those patients with a distant history of stroke. Interestingly, 23 out of the 1,024 patients experienced a major bleeding episode, and of those 23, fourteen received bridging anticoagulation therapy. The authors concluded that for patients receiving long-term VKA therapy undergoing minor outpatient procedures with low-to-intermediate risk of thromboembolism, perioperative cessation of VKA therapy without a bridging anticoagulation regimen is associated with a low risk of arterial thromboembolism [11]. Wyskonski and colleagues reached similar conclusions in their prospective cohort study examining the 3-month cumulative incidence of thromboembolism, bleeding, and death among consecutive patients with non-valvular atrial fibrillation referred to the Thrombophilia Center at the Mayo Clinic over a 7-year period. Of the 345 patients followed, 271 (79%) had persistent AF, and 118 (34%) reported prior thromboembolic events (stroke, TIA, peripheral artery embolus, or left atrial thrombus). The CHADS₂ classification system stratified high- and low-risk patients. The most common procedures were orthopedic, gastrointestinal, and urologic. The authors reported four patients with six thromboembolic events after four procedures. The 3-month cumulative thromboembolic incidence was 1.1% (95% CI, 0–2%) from the 345 patients followed. Bridging LMWH therapy was implemented in two of the four patients. Interestingly, nine of the 345 patients who underwent ten procedures in total experienced ten major bleeding events (hemoglobin decrease of 2 g/dL or transfusion of two units PRBC). Five of the nine patients received bridging LMWH, and the most common bleeding incident was from a gastrointestinal source. Eleven patients experienced 11 minor bleeding events after 11 procedures of which 10 received bridging LMWH. In their study, warfarin was stopped 4–5 days before surgery, and those indentified to be at high risk of stroke were either treated with intravenous unfractionated heparin or bridged with LMWH. The regimen for bridging therapy was as follows: ardeparin sodium 130 IU SC q 12 h, dalteparin sodium 100 IU SC q12 h or 200 IU/kg, and enoxaparin sodium 1 mg/kg SC q 12 h or 1.5 mg/kg SC q 24 h. The last LMWH injection occurred 24 h before surgery at a dose 50% of the calculated daily dose. (Table 13.3) Intravenous UFH was stopped 6–8 h before surgery. Warfarin was restarted immediately after the procedures, and warfarin and LMWH therapy overlapped for at least 5 days until the INR exceeded 2. The authors concluded that LMWH bridging therapy should only be initiated in patients at the highest risk for thromboembolism (prior stroke; CHADS₂ score > or equal to 4) while considering the procedure-associated risk of bleeding [12]. In patients with lower CHADS scores, routine perioperative DVT prophylaxis should be employed.

The majority of patients undergoing elective surgery are at low risk for thromboembolism. If VKA is interrupted and bridging therapy initiated, the authors of this chapter recommend that warfarin be stopped 4–5 days preoperatively, so that there is enough time for the INR to normalize prior to the procedure. An alternative is to interrupt the treatment with warfarin 2 days before surgery and reverse the effect of VKA with vitamin K. By doing so, the time period during which the patient is at increased risk for thrombosis is reduced. Additionally, to further decrease the risk of thromboembolism after reversal of warfarin, prophylactic heparin (5,000 U) can be given subcutaneously every 12 h. Then during the postoperative period, once the risk of bleeding has been determined to be minimal, prophylactic doses of heparin can be restarted along with warfarin, monitoring the level of anticoagulation until the INR reaches a therapeutic level. In cases where the risk of thromboembolism is very high, we recommend the administration of therapeutic LMWHs. The heparin administration must be held 24 h before surgery. Patients in whom anticoagulation is mandatory (e.g., warfarin treatment for anticoagulation in patients with mechanical heart valves), a continuous intravenous UFH infusion is advised in an inpatient setting, and the patient must therefore be hospitalized a few days before the procedure. The intravenous UFH infusion should be stopped 6–8 h prior to the procedure.

Atrial fibrillation is the most common cause of arterial thromboembolism resulting in a five- to sixfold increased risk of stroke. Although many newer pharmacologic oral anticoagulants have been implemented to reduce this risk, VKA therapy remains the cornerstone of treatment and is associated with a stroke risk reduction of 68% (range 45–82%). When these patients present for urologic procedures, a careful evaluation of risk for developing arterial thromboembolic event versus the risk of surgical bleeding must be evaluated when developing a perioperative anticoagulation management protocol. In patients at high risk for arterial thromboembolism as defined by the CHADS₂ classification system, the need to prevent a thromboembolic event will dominate management irrespective of bleeding risk. For those considered at moderate risk, no single perioperative strategy has currently been shown to be superior, and management will depend on the individual patient risk assessment. For patients at low risk, bridging therapies can be avoided. In addition to the CHADS₂ classification system, other factors must be considered including preexisting atrial and ventricular dysfunction, chronicity of anticoagulant therapy, and other concomitant medical conditions predisposing to a hypercoagulable state. Therefore, devising a perioperative anticoagulation strategy should not be the sole responsibility of the proceduralist, but rather a multidisciplinary effort between the internist, cardiologist, anesthesiologist, as well as the urologist.

The U.S. Food and Drug Administration (FDA) approved dabigatran etexilate (Pradaxa) in October 2012 as an alternative to warfarin in reducing the risk of stroke associated with atrial fibrillation. It is an oral reversible thrombin inhibitor with numerous pharmacokinetic and pharmacodynamic advantages over oral warfarin therapy. Since its activation and metabolism is independent of CYP450 system, dabigatran has less drug and dietary interactions. However, there are a few notable interactions that the perioperative clinician should know, serum levels of dabigatran are reduced when taking rifampin concurrently; additionally, serum levels are increased in the presence of dronedarone. Furthermore, the coadministration of proton pump inhibitors reduces the absorption of dabigatran by approximately 25%.

Dabigatran is predominantly excreted renally with the majority of the drug emerging unchanged in the urine. Its oral bioavailability is 6.5%, and its half-life is 14–17 h in those devoid of renal insufficiency. However, in those with renal dysfunction, the half-life can be prolonged up to 28 h. After oral ingestion, the time to peak anticoagulant activity is 2–3 h resulting in a dose-dependent increase in PT, aPTT, and TT. Furthermore, it does not demonstrate any inhibitory effects on other platelet-stimulating agents [20, 21].