Heparins are large water soluble polysaccharides which mainly act through blocking the activated intrinsic pathway. Heparin is also referred to as unfractionated heparin (UFH) to differentiate it from other lower molecular weight derivatives discussed below. It acts by catalyzing the binding of antithrombin III (a serine protease inhibitor) to serine protease coagulation factors (IIa, IXa, Xa, XIa, and XIIa) which results in immediate inactivation of coagulation cascade either in vivo or in vitro [2, 3]. Heparin can be administered intravenously (IV) or subcutaneously (SC). The half-life is approximately 2 h and the metabolism is through hepatic or reticuloendothelial system . The half-life is dose dependent with higher doses having longer half-life. At higher doses, the reticuloendothelial system binding sites become saturated, and clearance by hepatic metabolism is typically slower. Its use is safe without dose adjustment in patients with renal failure.

The effect is monitored through partial thromboplastin time (PTT) or activated clotting time (ACT) measurement [4, 5]. The major advantage of UFH is its short half-life, which is especially useful in patients in whom hour-to-hour control of the intensity of anticoagulation is desired. UFH is dosed to achieve a target activated PTT of 60–80 s. The most popular nomogram uses an initial bolus of 80 U/kg, followed by an initial infusion rate of 18 U/kg/h. Heparins’ specific structure inhibits them from placental exchange and makes them drugs of choice in pregnant patients who need anticoagulation.

Heparin is commonly used for inpatient treatment of venous and arterial thromboembolism. The main side effects are bleeding and heparin-induced thrombocytopenia (HIT) , but osteoporosis and hypersensitivity have been also reported [6]. There are some important limitations in UFH use. First, it has a narrow therapeutic range window, with the risk of either bleeding or inadequate anticoagulation. This limitation was illustrated in the Enoxaparin and Thrombolysis in Reperfusion for Acute Myocardial Infarction Treatment Thrombolysis in Myocardial Infarction (ExTRACT TIMI) 25 trial [7, 8]. This randomized clinical trial enrolled 20,506 patients with myocardial infarction and showed significant decrease in the composite of death, nonfatal reinfarction, or urgent revascularization in patients treated with enoxaparin compared to UFH . This difference was thought to be due partly to low predictability of dose-response and difficulty to stay in the therapeutic zone with UFH. Guervil et al. have demonstrated that aPTT monitoring was outside the therapeutic range in 60% of the time in patients who were receiving the UFH infusion [9].

Heparin is not suited for outpatient therapy because of the absence of an oral formulation. There is also a risk for thrombus extension during heparin therapy due to low effect on activated thrombin which has bound to factor Xa or fibrin in the formed thrombus. UFH anticoagulant effects can be reversed by protamine sulfate. Usually 1 mg of protamine sulfate is used to neutralize up to 100 units of heparin. Neutralization occurs in less than 5 min after IV administration through complexing with heparin and its inactivation [10].

Low molecular weight heparins (LMWHs) are fragments of UFH that exhibit less binding to plasma proteins and endothelial cells. Therefore, LMWHs have a greater bioavailability, a more predictable dose-response especially with weight-based dosing, and a longer half-life. No monitoring or dose adjustment is needed unless the patient is markedly obese or has chronic kidney disease. Due to their size, they are less likely but still have a significant chance of HIT. This group has more activity against factor Xa, and their levels can be monitored more efficiently by anti-factor Xa assay rather than PTT. Protamine sulfate has limited antidote effect against LMWH. Currently available LMWHs in the USA are enoxaparin, dalteparin, and tinzaparin [11]. They are different in molecular weight and manufacturing methods. Both enoxaparin and dalteparin have been used widely in VTE and share similarity in efficacy and safety profile; however drug administration can be incommodious. After initial drug selection, it is preferable not to try using different LMWHs interchangeably, since they have a different tendency to affect factor Xa versus factor IIa (enoxaparin has approximately 50% more anti-Xa effect compared to dalteparin with similar anti-IIa effect) [12].

Fondaparinux, a selective antithrombin III-mediated anti-Xa pentasaccharide, is administered as a weight-based once-daily subcutaneous injection in a prefilled syringe. No laboratory monitoring is required routinely. In certain populations (renal impairment, pregnancy, obesity, children), monitoring can be achieved using anti-factor Xa assay which is currently the gold standard for monitoring LMWH and fondaparinux therapy [13]. Target anti-factor Xa levels may be different in each laboratory due to variability in the type of used assays and titration curves [10]. Target level of 0.25–0.35 units/mL was recommended for anti-Xa based on the study in patients with renal impairment due to bleeding complications associated with anti-Xa between 0.45 and 0.8 units/mL [14, 15]. Fondaparinux is synthesized in a laboratory and, unlike LMWH or UFH, is not derived from animal products . It does not cause heparin-induced thrombocytopenia.

UFH has the largest size compared to other heparins and hence most likely to cause HIT. It is a serious antibody-mediated reaction resulting from irreversible aggregation of platelets. Currently, the term HIT is used without any dividing to describe only the immune-mediated reaction [16]. It can occur in up to 5% of patients who use UFH or LMWH. Patients who develop HIT typically are noted to have more than 50% decrease in platelet count within 5–10 days after initiation of heparin without an alternative cause. They are at risk of developing a new thrombus (heparin-induced thrombocytopenia and thrombosis [HITT] ). The assumption of diagnosis of HIT requires immediate discontinuation of any heparin product and initiation of alternative anticoagulants, typically a parenteral direct thrombin inhibitor [17]. Confirmation can be reached by HIT antibody testing immunoassay (ELISA) or functional assay like serotonin release assay (SRA) [18]. HIT or HITT may be delayed and can occur up to several weeks after discontinuation of heparin [19, 20].

Parenteral direct thrombin inhibitors (DTIs) block thrombin and prevent its cleaving effect on fibrinogen [21]. The drugs developed in this group are bivalirudin, argatroban, and lepirudin. Lepirudin was taken off the market in 2012. This section will focus on argatroban and bivalirudin. DTIs have no interaction with heparin platelet factor 4 (PF4) antibodies. These drugs are useful for the treatment of venous thromboembolism in the setting of HIT [22, 23]. Despite their benefits, development of this class of drugs was slower than many other classes probably due to lack of investment and research funding. One of the reasons of this disinterest is their short-term use and route of administration which results in narrow target population . On the other hand, the popularity of heparins and their high safety and efficacy profile limited the development of alternative parenteral DTIs [24]. The two main disadvantages are lack of antidote and difficulty in monitoring [25, 26]. Parenteral DTIs are also expensive and are usually reserved for patients who cannot receive heparin and LMWHs. Oral DTIs are discussed separately in this chapter.

Argatroban is commonly used for treatment of VTE in patients with HIT, whereas bivalirudin is used as an alternative to heparin in patients with acute coronary syndrome who undergo angiography and intervention [27]. Both drugs, unlike heparin, can inhibit clot-bound thrombin in addition to soluble thrombin. The relatively short half-life of parenteral DTIs necessitates precise monitoring of their level to assure effectiveness and inhibition of side effects. Bivalirudin is now commonly used in the USA as an alternative to heparin alone or heparin and GPIIb/GPIIIa inhibitor in patients with acute coronary syndrome. Compared to heparin, it has demonstrated non-inferiority effect and less major bleeding and no risk of thrombocytopenia [27].

Until recently, the vitamin K antagonists (VKAs) were the only oral anticoagulant agents available, and warfarin remains the most commonly prescribed oral anticoagulant worldwide. The main indications of warfarin use include primary and secondary prevention of VTE, prevention of systemic embolism, and stroke in patients with mechanical heart valves and atrial fibrillation (AF) [28]. Warfarin affects coagulation cascade by interfering with γ-carboxylation of vitamin K-dependent clotting factors II, VII, IX, and X. Besides anticoagulation effect , warfarin also plays a role as a procoagulant as it acts against proteins C and S. It is readily absorbed after oral administration, reaching peak concentration within 4 h. It however has significant variability in dose-response across individuals and requires maintenance in a narrow therapeutic range of international normalized ratio (INR) of 2–3 for most indications [29, 30].

The anticoagulant effect of warfarin takes several days to take action; hence concomitant bridging with other anticoagulants such as heparin or LMWH is usually done until the INR reaches therapeutic range. Then warfarin oral administration alone can be continued. Initiation with low-dose warfarin at 5mg has been found to be superior to 10 mg in patients treated with warfarin after heart valve replacement [31, 32]. Conversely, in patients with acute venous thromboembolism, initial administration of warfarin at 10 mg allowed more rapid achievement of therapeutic INR [33]. The current guidelines however suggest initiating VKA therapy with 10 mg daily for the first 2 days followed by dosing based on INR values (Grade 2C) [30].

The patient response to warfarin varies significantly because of genetic factors, comorbidities, the use of concomitant medications, as well as dietary factors, each of which can affect its metabolism. Clinical outcome and efficacy of warfarin are highly associated with patient adherence and the time for which patient’s PT-INR values are maintained within therapeutic range [34]. According to a study by De Caterina et al., patients with an average time >70% in therapeutic range are considered to be at a low risk of a major hemorrhagic or thrombotic event [35]. Patients with a low body weight [36, 37], significant congestive heart failure, liver disease, or concurrent medications with interactions may require lower doses. However, renal clearance does not play a significant role in warfarin elimination [38].

Warfarin is often initiated in the evening, so that an INR can be obtained with morning laboratory testing, allowing time in the afternoon to obtain the results and determine the next dose. This may not be however sufficient time to determine the effect of the drug. Gage suggests that INR testing needs to be done 15 or more hours after first administration [39]. It may be thus better to administer warfarin in the afternoon, if INR testing is done the following morning. The optimal frequency of INR testing to maintain patients within therapeutic range is still unclear, as patients exhibit fluctuations of INR with diet, medications, and adherence. When therapy is just initiated, INR monitoring is done every few days until it is therapeutic. The INR is then usually obtained weekly for 1–2 weeks, to verify dosing by stability of the INR within range. Commonly, testing is then obtained biweekly for one to two times. If the INR remains stable within the therapeutic range and all else remains constant, the duration between tests can be extended to 3 monthly visits. Current guidelines recommend that the frequency of testing should be scheduled to every 12 weeks (Grade 2B) [30].

The major limitation of warfarin is the interaction with drugs and dietary restrictions. Patient education and identification of factors which may lead to over or under-anticoagulation are thus necessary [40]. When combined with low-dose aspirin , nonsteroidal anti-inflammatory drugs (NSAIDs) , or clopidogrel , the risk of bleeding is significantly increased [41, 42]. It is recommended to avoid use of these drugs except when benefit outweighs risk of bleeding [30]. Most drug interactions affecting warfarin involve inhibition of the CYP450 enzymes . Broad-spectrum antibiotics also affect vitamin K production by intestinal flora in malnourished patients with limited stores. Chronic alcohol use can decrease activity of warfarin by increasing clearance, but the presence of concomitant liver disease can potentiate its effect. Patient education is essential to develop a diet with a relatively constant level of vitamin K and to identify foods rich in vitamin K [28].

Intracranial hemorrhage is the most feared bleeding associated with warfarin therapy. In the SPORTIF III and V clinical trials, warfarin was used for prevention of stroke in non-valvular atrial fibrillation (AF) patients and showed a rate of intracranial and subdural hematomas at 0.4%, and the rate of major bleeding was 2.5% per year [43, 44]. The DURAC trial study group in 1994 for patients with VTE concluded that 75% of INR values of their study cohort (1124 patients) were ≥2.0, and 58% were in the therapeutic range. There were eight patients with recurrent VTE (1.3 in 100 patient years; 95% CI, 0.2–1.2). Seventeen hemorrhagic events were recorded (2.8/100 patient years; 95% CI, 0.8–2.2), among which two were fatal (0.3/100 patient years; 95% CI, 0–0.4) [45]. When comparing low-dose warfarin to conventional dose of warfarin, there was no significant difference in frequency of overall bleeding between the two groups (OR, 1.3; 95% CI, 0.8–2.1), and the frequency of recurrent VTE was higher in low-dose warfarin group (OR, 2.8; 95% CI, 1.1–7) [46]. However, after conventional full-dose anticoagulation therapy for 6 months, when patients were followed with low-dose warfarin compared to placebo (PREVENT), 48% risk reduction of recurrent VTE was observed (OR, 0.36; 95% CI, 0.19–0.67) with no significant difference in major bleeding episodes [47]. In a recent meta-analysis of 11 clinical trials (41,015 patients treated for both VTE and AF), safety of warfarin in “high-risk” populations was compared. A significant association with all-cause bleeding and age>75 years (RR,1.62; 95%CI, 1.28–2.05), low body weight (RR, 1.2; 95%CI, 1.03–1.4), and those with impaired renal function (RR, 1.59; 95%CI, 1.3–1.94) was noted [48]. However, its use in patients with end-stage renal disease in a meta-analysis (56,146 patients with ESRD and AF) showed no association with major bleeding (OR, 1.18; 95% CI, 0.82–1.69) or gastrointestinal bleeding (OR, 1.19; 95% CI, 0.81–1.76) [49].

Despite wide usage and sufficient data from clinical trials demonstrating efficacy for a variety of thrombotic and thromboembolic conditions , warfarin is becoming underutilized because its management is cumbersome for both patients and physicians. Besides hemorrhage, warfarin skin necrosis is the most serious adverse effect and is induced by a transient hypercoagulable state. It occurs with intake of warfarin in individuals with congenital or acquired protein C or S deficiency [50, 51]. Affected individuals develop skin lesions which begin as an erythematous macule and if left untreated progress to an indurated lesion before becoming necrotic.

The current guidelines for reversal established by the Ninth American College of Chest Physicians Conference on antithrombotic and thrombolytic therapy recommend [30]:

Oral vitamin K can also be used for reversal, as it was found to lower the INR rapidly in asymptomatic patients who have INR above therapeutic range [52, 53].

Several direct oral anticoagulants (DOACs) have been developed to overcome the limitations of heparin and coumadin . Among them, the oral direct thrombin inhibitor dabigatran etexilate (Pradaxa^®) and the oral direct activated factor Xa inhibitors rivaroxaban (Xarelto^®), apixaban (Eliquis^®), and edoxaban (Lixiana^®/Savaysa^®) are currently approved for anticoagulation as prophylaxis and treatment of VTE, as well as prevention of stroke and embolic events in non-valvular atrial fibrillation.

Unlike VKA which has an indirect role in inhibition of factor synthesis, DOACs directly inhibit either thrombin or factor Xa . DOACs thus have a rapid onset and offset of action. All the DOAC agents are rapidly absorbed following oral administration and have a relatively short half-life (5–17 h). Table 21.2 illustrates the pharmacokinetics of each of the DOAC agents. The disadvantages of dabigatran pharmacodynamics include low bioavailability (6%) despite the presence of prodrug etexilate to enhance its intestinal absorption, high renal clearance (85% of absorbed dose), and low metabolism (<10%, by glucuronidation). On the contrary, factor Xa inhibitors have a good oral bioavailability (>50%); lower renal clearance than dabigatran , although still significant (54–73% of absorbed dose); and high metabolism rate by mainly CYP3A4 activity (rivaroxaban), CYP3A4/CYP3A5 (apixaban), and hydrolysis (edoxaban ) [54].

Since the recent shift to DOACs, four randomized clinical trials have been conducted to study the safety and efficacy of each of these agents compared with VKA therapy [55–59]. The DOACs have proven to be non-inferior to VKA in patients with venous thromboembolism (VTE) [60]. The risk of bleeding with DOACS, and particularly clinically relevant bleeding including intracranial bleeding, is less with DOACs than with VKA therapy. Dabigatran , rivaroxaban , and edoxaban tend to have higher GI bleeding compared to VKA for treatment of atrial fibrillation [61–63]. However this has not been the case in patients treated for VTE [64]. Also, the risk of bleeding may be lower with apixaban as compared with the other DOACs [65, 66]. On the other hand, the risk that a major bleeding will be fatal appears to be no higher than VKA therapy [60, 66]. Thus, based on its less bleeding tendency and greater patient convenience while providing similar or superior efficacy, DOACs are currently recommended as first choice of drugs in patients for initial and long-term treatment of VTE in patients without cancer [67] (Table 21.3).