Old and new anticoagulant drugs: A minireview

Abstract

The limits of traditional anticoagulants, such as heparin and warfarin, have prompted the search for new agents for prophylaxis and treatment of arterial and venous thromboembolism, including factor Xa and thrombin inhibitors. These agents can be given orally, and their most significant advantage is that no laboratory monitoring is needed. The anti-Xa inhibitor rivaroxaban and the direct thrombin inhibitor dabigatran etexilate are licensed for prophylaxis of venous thromboembolism (VTE) in high-risk orthopedic surgery. They are at least as safe and effective as heparins but much more expensive. Dabigatran, rivaroxaban, and other agents currently in the pipeline of clinical development have the potential to replace warfarin in the two most frequent indications for anticoagulation, i.e. secondary prophylaxis of VTE and atrial fibrillation. Prevention and treatment of coronary artery thrombosis in patients with ischemic heart disease is another area of investigation for the role of new anticoagulants. These drugs have the potential to meet some currently unmet needs of traditional anticoagulants, but available clinical data warrant confirmation and expansion. Lack of specific antidotes for anticoagulation reversal and the high cost are important limitations of their use.

Key words::

• Cardiovascular disease
• dabigatran
• factor Xa inhibitors
• fondaparinux
• low-molecular-weight heparins
• rivaroxaban
• thrombin inhibitors
• thromboembolism
• unfractionated heparin
• warfarin

Key messages

• Traditional anticoagulants must be administered parenterally and/or need laboratory monitoring for dose adjustment.

• Newly developed anticoagulants, such as factor Xa and thrombin inhibitors, can be given orally at fixed dosage.

• Currently licensed for the primary prophylaxis of venous thromboembolism in high-risk orthopedic surgery, they are in the advanced phases of licensing in atrial fibrillation and secondary prophylaxis of venous thromboembolism.

• Possible limitations of these drugs are lack of specific antidotes and the very high cost.

Introduction

Arterial and venous thromboembolism is still the most frequent cause of morbidity and mortality in high-and middle-income countries, as well as in emerging economies. For instance, in a European country such as Italy with a total population of nearly 60 million, at least 350,000 people die or become severely disabled each year owing to acute coronary syndromes, ischemic stroke, or venous thromboembolism. The burden of these scourges is also increasing in low-income countries, as shown by the INTERHEART and INTERSTROKE population studies (1,2).

The magnitude of the burden of thromboembolic diseases commands primary prevention as the most important community goal, through lifestyle interventions on such modifiable risk factors as smoking, hypertension, abdominal obesity, physical inactivity, and inadequate dietary styles. Despite the positive outcome of the implementation of these measures, many people still develop clinical manifestations of thromboembolism, so that a pharmacological approach to prevention and treatment is necessary to reduce morbidity and mortality. Even though formation of thrombi in blood vessels is a complex process that involves a multiplicity of interacting mechanisms (such as abnormal changes in the vessel wall and blood flow), blood hypercoagulability plays a pivotal role in thrombogenesis. Efficacious anticoagulant drugs such as heparin and vitamin K antagonists have been available for many years, but they have some limitations, and there are a number of unmet clinical needs. The past decade has witnessed major advances in the development of new anticoagulant drugs and the preliminary clinical evidence of their efficacy and potential advantages over traditional drugs (3). This narrative article, prompted by the most recent data reported at the 21st International Congress on Thrombosis (4), will briefly review the state of the art of traditional anticoagulants, the currently unmet clinical needs, and what the near future will look like provided newer anticoagulants are going to fulfill their current promises.

Where are we now?

The most widely available, licensed, and currently used anticoagulant drugs are unfractionated heparin (UH), low-molecular-weight heparins (LMWHs), the synthetic pentasaccharide fondaparinux, and vitamin K antagonists such as warfarin and related drugs (5–8). The direct thrombin inhibitors bivalirudin and lepirudin are only licensed for restricted indications, such as acute coronary syndromes and the thrombotic complications of heparin-induced thrombocytopenia (9).

UH has been available for nearly a century, this animal-derived extractive drug having been discovered in 1916 by Jay McLean at John Hopkins in Baltimore. In 1940, Karl Paul Link discovered coumarin as the substance contained in sweet clover hay that killed cattle by bleeding (10). Evolution of UH towards the chemically fractionated low-molecular-weight heparins (LMWHs), and then towards the synthesis of fondaparinux that contains the basic pentasaccharide anticoagulant structure shared by both UH and LMWHs, took place in the 1980s and in the late 1990s (6,10).

These time-honored anticoagulants have been employed in an array of clinical conditions caused by arterial and venous thromboembolism. Among them, the most important and clinically relevant are primary prophylaxis of venous thromboembolism (deep-vein thrombosis and pulmonary embolism) in high-risk orthopedic surgery and in medical conditions, treatment and secondary prophylaxis of acute venous thromboembolism, prevention of cardioembolic stroke in non-valvular atrial fibrillation, and treatment of acute coronary syndromes (11). To briefly summarize a long story of therapeutic success, UH and LMWHs do reduce by approximately 60% venous thromboembolic complications in hip and knee arthroplasties and in the frame of high-risk medical conditions (heart failure, acute inflammatory diseases, prolonged immobilization in bed). They are also needed, in addition to dual antiplatelet therapy with aspirin and clopidogrel, in acute coronary syndromes, whether or not managed with reperfusion techniques. Vitamin K antagonists reduce by more than 90% recurrence of venous thromboembolism, and by approximately 60% cardioembolic stroke in non-valvular atrial fibrillation (12). Fondaparinux is at least as effective (and equally safe) as heparin and LMWH for the initial treatment of venous thromboembolism, employing a standard dosing regimen that offers a uniform approach of patient management (once daily subcutaneous injection with no body-weight tailoring) (13).

Limitations of traditional anticoagulants

In spite of the excellent clinical results obtained with traditional anticoagulants, there is much space for improvement in clinical practice in terms of their clinical applicability, safety, and efficacy. The comparative advantages and disadvantages of traditional anticoagulants, and the corresponding clinical consequences, are summarized in Table I.

Table I. Comparative advantages and disadvantages of traditional anticoagulant and corresponding clinical consequences.

Comparative advantage or disadvantage	Consequence
LMWH versus UH:
Better bioavailability	Subcutaneous administration
Less binding to endothelium	Longer half-life, which permits once-daily administration
Less binding to plasma proteins	More predictable anticoagulant response, which obviates need for monitoring
Less binding to platelet factor 4	Lower risk of thrombocytopenia
Less binding to bone cells	Lower risk of osteoporosis
Protamine sulphate less effective as antidote	Less rapid reversal in case of overdose and/or bleeding complications
Fondaparinux versus LMWH :
Improved safety	No risk of biological contamination; lower risk of thrombocytopenia and osteoporosis
Defined target	Single target (factor Xa) versus multiple targets
Less potent inhibition of factor Xa than heparin	Risk of catheter thrombosis in acute coronary syndromes, unless supplemental heparin is administered
Vitamin K antagonists versus LMWH and fondaparinux:
Oral administration	Easier outpatient treatment
Interaction with food, drugs and genetic polymorphisms	Need of regular laboratory control for dose adjustment
Longer plasma half-life	Less rapid reversal in case of overdose and/or bleeding complication

LMWH = low-molecular-weight heparins; UH = unfractionated heparin.

The advantages of LMWH over UH are higher availability and longer half-life, which allow once-daily subcutaneous administration, more predictable anticoagulant responses that obviate the need of laboratory monitoring (for dosage tailoring), and less binding to platelet factor 4 and bone cells that results in a lower risk for thrombocytopenia and osteoporosis. The additional potential advantages of subcutaneous fondaparinux over UH and LMWHs are improved safety due to the synthetic source of this drug and a lower risk of thrombocytopenia and osteoporosis (11). Fondaparinux has activated coagulation factor Xa (FXa) as a single inhibition target, whereas UH and LMWH act upon both FXa and thrombin (13). Because all these drugs carry a risk of bleeding complications, the anticoagulant activity of UH can be quickly reversed by an effective antidote such as protamine sulphate, only partially effective for LMWH neutralization and not at all for fondaparinux (13).

A limitation common to UH, LMWHs, and fondaparinux is that they need parenteral administration (14), an obstacle particularly in the frame of outpatient management. This limitation is not shared by oral vitamin K antagonists like warfarin and related compounds. However, they need dosage tailoring by means of laboratory monitoring with the prothrombin time and related methods, because they interact with other drugs, food, and genetic polymorphisms (12). The development and adoption of a standardized method to express the prothrombin time results (the international standardized ratio (INR)) has been a major step forward in the optimal clinical use of warfarin and other vitamin K antagonists, with a substantial improvement in safety and efficacy. However, applicability of these drugs is still unsatisfactory, because many elderly patients with limited mobility and autonomy find it difficult to attend hospital-located monitoring facilities at the frequent and regular intervals needed for safe and effective treatment.

Pertaining to UH, the accuracy of its laboratory monitoring for dose adjustment is lagging behind that of vitamin K antagonists, because no standardization is currently available of the results obtained with different laboratory tests and reagents (activated partial thromboplastin time, activated clotting time). However, the need for laboratory monitoring of UH has been circumvented by LMWHs, which only need dosage tailoring based upon patient weight (15). Fondaparinux was another step forward, because this drug can be administered at a fixed dose, barring patients with abnormal renal function. Finally, non-anticoagulant side-effects of UH such as thrombocytopenia and osteoporosis have been substantially reduced with the advent of LMWHs and fondaparinux but are still a potential risk, because these drugs share with UH the pentasaccharide moiety (16).

Beside these limitations related to the structural, pharmacokinetic, and pharmacodynamic features of traditional anticoagulants, their efficacy is not optimal in some clinical conditions. For instance, in the frame of primary prophylaxis of venous thromboembolism in high-risk orthopedic surgery, LMWHs are definitely more efficacious than UH, but the incidence of venous thromboembolism is still unacceptably high (17). Subcutaneous fondaparinux (2.5 mg) started 6 hours after surgery (an administration timing often preferred by orthopedic surgeons, who are afraid of intraoperative bleeding) reduces venous thromboembolism by approximately half compared with the LMWH enoxiparin. However, prolonged continuation of prophylaxis with LMWHs or fondaparinux for 30–40 days after hospital discharge is sometimes difficult when nursing facilities are not available for subcutaneous administration at home.

Pertaining to the treatment of acute venous thromboembolism, most guidelines recommend 5–10 days of subcutaneous LMWHs, which are more convenient than UH in the frame of the highly recommended outpatient setting of management (18,19). The initial parenteral treatment with UH, LMWHs, or fondaparinux (needed to obtain an immediate anticoagulant protection) is followed by bridging to the oral administration of vitamin K antagonists: for 3 months if venous thrombosis is provoked by transient risk factors (such as surgery, pregnancy, or immobilization), for at least 6 months if thrombosis is unprovoked, and lifelong if thrombosis is recurrent (18,19). As mentioned above, this approach to therapeutic anticoagulation has dramatically reduced the rate of recurrence. However, a substantial further improvement would be made by the use of only one, rapidly acting oral drug able to avoid the initial use of LMWHs and laboratory monitoring of vitamin K antagonists, so that there would be no distinction between initial and long-term therapy (18,19).

Pertaining to cardiological atherothrombotic conditions, anticoagulants are being used for treatment of acute coronary syndromes, whether or not treated also with vascular reperfusion techniques such as pharmacological thrombolysis or percutaneous intervention. However, there is a need for improvement, particularly owing to the frequent occurrence of bleeding complications, a problem looming large in patients concomitantly treated with multiple drugs that impair hemostasis (UH, LMWHs, or fondaparinux, administered on top of double or triple antiplatelet agents such as aspirin, ADP-receptor, and platelet glycoprotein IIb/IIIa inhibitors) (20). Another unmet clinical need of traditional anticoagulants in cardiology is their use in re-infarction prevention. Vitamin-K antagonists are of proven efficacy and are superior to monotherapy with aspirin (20,21). However, they are rarely used by cardiologists, owing to the need for long-term laboratory monitoring and the perceived high risk of bleeding. An oral drug with no laboratory requirement would certainly increase the propensity of clinicians to choose anticoagulants for secondary prophylaxis of myocardial infarction, instead of the platelet function inhibitors that are currently preferred (21).

Finally, chronic atrial fibrillation is the cardiological condition in which the use of anticoagulant therapy is at the moment definitely more suboptimal. Atrial fibrillation is very frequent in the elderly, as 1% of people over the age of 60 years develop this arrhythmia, and the rate rises to 10% at the age of 80. Oral anticoagulant therapy with warfarin or related drugs is highly effective and reduces the rate of ischemic stroke by more than 60%. Also aspirin is effective, but 3-fold less than vitamin K antagonists with no clear advantage in term of reduction of bleeding complications (22). Although the efficacy and cost-effectiveness of anticoagulant prophylaxis is clearly demonstrated particularly in the elderly, the adherence to guidelines is still very poor in most health care settings (community, outpatient clinics, nursing homes, and academic hospitals) (23). There is also an evident trend for a decreasing use of antithrombotic prophylaxis with increasing age. In particular, warfarin and related drugs are under-used irrespective of the levels of cardioembolic risk, with a clear association between their under-use and poorer outcome. The sub-optimal use of this highly efficacious treatment has multiple reasons. One is the concern of physicians, who tend to give more weight to the age-related high risk of bleeding than to the similarly high age-related risk of cardioembolism. Another important concern is for poor patient compliance in terms of regular drug intake and laboratory monitoring, particularly for elderly frail people who have some degree of cognitive impairment. Perhaps these limitations could be to some degree circumvented if new drugs with less risk of bleeding and no need of laboratory monitoring were available.

State-of-the-art of new anticoagulant drugs

Led by the aforementioned unmet clinical needs, in the last few years the pharmaceutical industry has devoted considerable efforts to developing new anticoagulants, one of the fields in clinical pharmacology where more progress has taken place. Of the new drugs in clinical development, approximately two-thirds are FXa inhibitors and one-third direct thrombin inhibitors, with little effort at the moment to develop new heparins and vitamin K antagonists (source: www.citoline.com/tiraltrove.htlm). The strategies of the industry were mainly concentrated on agents suitable for oral administration at a fixed dose, with no need of laboratory monitoring. Figure 1 and Table II report the mechanisms of action and pharmacological characteristics of new anticoagulant drugs, compared with those of traditional agents.

Figure 1. Mechanism of interaction between coagulation factors and corresponding active enzymes (solid lines) and drugs inhibiting enzymatic activity (broken lines).

Table II. Pharmacokinetic and pharmacodynamic characteristics of new and traditional anticoagulants.

	UH	Warfarin	LMWHs	Fondaparinux	Rivaroxaban	Dabigatran
Target	Factor Xa and thrombin	All vitamin-K-dependent coagulation factors and inhibitors	Factor Xa and thrombin	Factor Xa	Factor Xa	Thrombin
Administration	Subcutaneous	Oral	Subcutaneous	Subcutaneous	Oral	Oral
Antidote	Yes	Yes	Partial	No	No	No
Source	Biological	Synthetic	Biological	Synthetic	Synthetic	Synthetic
Bioavailability (%)	30	90	90	100	60–80	4–10
Degree of renal excretion	Low	Low	High	High	High	High
Half-life (hours)	1	15–44	4	17	6–9	14–17
Rate of thrombocytopenia	<5%	No	<1%	?	No	No

At the time of writing, only two anticoagulant drugs are licensed for the indication of prevention of venous thromboembolism after high-risk orthopedic surgery: the direct thrombin inhibitor dabigatran etexilate and the anti-Xa inhibitor rivaroxaban (24–27). However, licensing of other drugs and other indications in clinical conditions such as the treatment of venous thromboembolism and the prevention of cardioembolism in chronic atrial fibrillation is approaching (27). Clinical trials in patients with acute coronary syndromes are also on-going, but licensing for this indication is predicted to occur later than for atrial fibrillation and treatment of venous thromboembolism (4,9).

The results of fully published, phase III clinical trials can be briefly summarized as follows. Dabigatran etexilate was evaluated (in doses of 150 or 220 mg given orally once daily) in patients undergoing hip or knee arthroplasty, in comparison with a control group treated subcutaneously with the LMWH enoxiparin (28–30). Either daily dose of this drug was at least as effective as enoxiparin in thromboprophylaxis (administered at doses of 40 mg once daily starting preoperatively or 30 mg twice daily starting postoperatively), with no excess of bleeding (Table III). In the same prophylactic indication, the FXa inhibitor rivaroxaban (20 mg/day orally once daily) was superior to enoxiparin using as outcome all clinical manifestations of venous thromboembolism and all-cause mortality, with no excess of bleeding (31–34) (Table IV). On the basis of these results, both drugs are now licensed in Europe for the prophylaxis of venous thromboembolism in high-risk orthopedic surgery.

Table III. Dabigatran etexilate for thromboprophylaxis in major orthopedic surgery: results of the phase III trials.

				Primary efficacy outcome^a			Primary safety outcome^b
Study (ref.)	Patients	Surgery	Study arms	Enoxaparin	Dabigatran 150 mg (P)	Dabigatran 220 mg (P)	Enoxaparin	Dabigatran 150 mg (P)	Dabigatran 220 mg (P)
RE-NOVATE (28)	3494	Hip arthroplasty	Enoxaparin 40 mg o.d., 28–35 days	6.7%	8.6%	6.0%	1.6%	1.3%	2.0%
			Dabigatran etexilate 150 mg o.d., 28–35 days		(<0.0001)^c	(<0.0001)^c		(0.60)	(0.44)
			Dabigatran etexilate 220 mg o.d., 28–35 days
RE-MODEL (29)	2076	Knee arthroplasty	Enoxaparin 40 mg o.d., 6–10 days	37.7%	40.5%	36.4%	1.3%	1.3%	1.5%
			Dabigatran etexilate 150 mg o.d., 6–10 days		(0.017)^c	(0.0003)^c		(–)	(0.82)
			Dabigatran etexilate 220 mg o.d., 6–10 days
RE-MOBILIZE (30)	2715	Knee arthroplasty	Enoxaparin 30 mg b.i.d., 12–15 days	25.3%	33.7%	31.7%	1.4%	0.6%	0.6%
			Dabigatran etexilate 150 mg b.i.d., 12–15 days		(0.0009)^d	(0.02)^d		(NA)	(NA)
			Dabigatran etexilate 220 mg b.i.d., 12–15 days

o.d. = once daily; b.i.d. = twice daily; NA = not available.

^aComposite of total (venographic and symptomatic) VTE and death from all causes.

^bIncidence of major bleeding on treatment.

^cNon-inferior to enoxaparin.

^dInferior to enoxaparin.

Table IV. Rivaroxaban for thromboprophylaxis in major orthopedic surgery: results of the phase III trials.

				Primary efficacy outcome^a			Primary safety outcome^b
Study (ref.)	Patients (n)	Surgery	Study arms	Rivaroxaban	Enoxaparin	P	Rivaroxaban	Enoxaparin	P
RECORD 1 (31)	4541	Hip arthroplasty	Rivaroxaban 10 mg o.d. 5 weeks	1.1%	3.7%	<0.001	0.3%	0.1%	0.18
			Enoxaparin 40 mg o.d. 5 weeks
RECORD 2 (32)	2509	Hip arthroplasty	Rivaroxaban 10 mg o.d. 5 weeks	2.0%	9.3%	<0.001	0.1%	0.1%	–
			Enoxaparin 40 mg o.d. 10–14 days
RECORD 3 (33)	2531	Knee arthroplasty	Rivaroxaban 10 mg o.d. 10–14 days	9.6%	18.9	<0.001	0.6%	0.5%	0.77
			Enoxaparin 40 mg o.d. 10–14 days
RECORD 4 (34)	2300	Knee arthroplasty	Rivaroxaban 10 mg o.d. 10–14 days	6.9%	10.1%	0.012	0.7%	0.3%	0.11
			Enoxaparin 30 mg b.i.d. 10–14 days

o.d. = once daily; b.i.d. = twice daily.

^aComposite of total VTE (any DVT, non-fatal PE) and death from all causes up to the end of treatment period.

^bIncidence of major bleeding on treatment.

Phase III trials designed to assess the role of dabigatran in patients with atrial fibrillation and acute venous thromboembolism have been completed and published. In the non-inferiority RE-LY, a total of 18,113 patients with atrial fibrillation at risk of embolic stroke were randomized to receive dabigatran or warfarin (35). The rate of the primary outcome (systemic embolism) was lower with dabigatran at a dosage of 150 mg twice daily (1.11% per year) than with dabigatran at a dosage of 110 mg twice daily (1.53%) or warfarin (1.69%). The rate of major bleeding in warfarin-treated patients was higher than with low-dose dabigatran (3.36% versus 2.71%), but not when warfarin was compared with the higher dosage of dabigatran (3.11%). Importantly, a 70% reduction in intracerebral bleeding was seen in patients receiving both doses of dabigatran compared to warfarin-treated patients. The findings that the 150 mg dosage is more efficacious and that the 110 mg dosage is safer than warfarin and yet equally efficacious are important and suggest that dosing of this anticoagulant drug might be personalized, with the lower dosage preferably chosen in patients at higher risk of bleeding, and the higher dosage in those at increased risk of cardioembolic stroke (36).

In the double-blind, non-inferiority RE-COVER trial (37), a total of 2,564 patients with acute venous thromboembolism were randomized to dabigatran (150 mg twice daily) or warfarin (dosage adjusted to an INR of 2 to 3) for 6 months, the latter oral drug being started after initial treatment with a parenteral anticoagulant. Rates of the primary efficacy combined events (recurrent symptomatic venous thromboembolism and related death) were 2.4% and 2.1% in the dabigatran-and warfarin-treated groups (P < 0.001 for non-inferiority). Rates of major bleeding were 1.6% for dabigatran and 1.9% for warfarin, and those of any bleeding were 16.1% (dabigatran) and 21.9% (warfarin). The results of this trial indicate that, for the treatment of acute venous thromboembolism, a fixed dosage of dabigatran alone has an efficacy and safety profile at least similar to that of warfarin preceded by heparins. However, dabigatran offers the advantages over warfarin of minimal interactions with foods or drugs, needs no routine monitoring of the prothrombin time INR, and permits the use of only one oral drug throughout the entire acute phase of venous thromboembolism (38).

What will the future look like?

As outlined above, several new oral anticoagulants are under development, and the outcomes of the few fully published phase III trials on dabigatran and rivaroxaban are so far quite consistent with the expectations. Many other drugs (particularly FXa but also thrombin inhibitors) are in clinical development (phases II and III), and preliminary results are available (4). Hence the armamentarium of anticoagulants is likely to expand in the near future, providing new opportunities (38).

On the basis of the published clinical data, it seems that the advantages originally promised by these drugs have so far generally materialized. Among these advantages are: the predictable dose-response with no need for laboratory monitoring, the relatively high efficacy-to-safety index, and the rapid onset of action with minimal interactions with other drugs. On the other hand, some limitations are apparent. The cost of these drugs is much higher than that of traditional anticoagulants, and it remains to be demonstrated that some cost-savings offered by them (no laboratory monitoring, more clinical events spared) do compensate for the higher costs. Cost-effectiveness studies are urgently needed to fill this gap of knowledge. A few limitations of these drugs are related to patient management, such as the lack of methods to assess compliance, the risk of self-adjustment of dosage, as well as the unjustified perception by patients and caregivers of a therapy made more trivial by less need of close surveillance. Other limitations are the disputable criteria for non-inferiority adopted in some phase III trials, and a lack of suitable antidotes for anticoagulant reversal in case of emergency bleeding. Moreover, the fixed dosage may be too high and hence risky in patients with chronic renal dysfunction (most of these drugs are eliminated through the kidney) and inadequate in weight-extreme patients. Other limitations are related to health care structures, such as the perception of a diminished role for the anticoagulation clinics that have contributed so much in the last 20–30 years to improving the efficacy and safety of warfarin and related vitamin K antagonists. It is likely that anticoagulation clinics will continue to have an important role in the future, because patients treated with the new drugs will still need care to assess compliance, monitor clinical progress, and determine the optimal duration of anticoagulation. Moreover, clinics would be necessary to monitor potential side-effects, such as the unpredicted increase of liver enzymes that led to the withdrawal from the market of ximelagatran, the first licensed oral thrombin inhibitor.

It is too early to make comments on how to choose among the different new anticoagulants, not only because no head-to-head comparative study has been yet performed, but also because the currently available data are insufficient to make a meaningful choice possible. It is unlikely that an individual drug will be effective and suitable for all the potential therapeutic indications. It is perhaps more likely that for each indication (and perhaps for each patient category) an individual agent may prove more suitable than others.

Declaration of interest: P. M. Mannucci has received honoraria from Bayer Schering and Boehringer Ingelheim for acting as a speaker at meetings. M. Franchini declares no conflicts of interest.