Prevention of Venous Thromboembolic Disease

Menno V. Huisman

Sharif Pasha

Patrick Mismetti

Ajay Kakkar

Venous thromboembolic disease frequently complicates the course of patients undergoing a variety of surgical interventions. Indeed, it was in the determination of the frequency of thrombosis in general surgical patients using I¹²⁵ labeled fibrinogen leg scanning that Kakkar et al. in 1969¹ were able to describe the natural history of postoperative deep vein thrombosis (DVT). The understanding of the natural history of this disease, and, subsequently, risk factors associated with its occurrence in surgical patients and its wider epidemiology have allowed the development of strategies to both identify patient populations at risk for the development of venous thromboembolism (VTE) in the perioperative period, quantify that risk, and thus provide methods for prophylaxis against VTE.

Beyond recognising that thrombosis commonly occurs in surgical populations, the establishment of evidence that demonstrates the adverse consequences if thromboembolic disease is not prevented has been important. Not only would patients suffer symptomatic DVT or pulmonary embolism (PE), but they may also experience, as a first manifestation of their thrombosis, a fatal PE.² Approximately 80% of patients who develop PE will have no evidence of peripheral venous thrombosis before their presentation with PE.³ Because PE is frequently fatal, even if diagnosed and treated in hospital,⁴ strategies based on a policy of treating massive PE or DVT as a way of preventing death from fatal PE expose surgical patients to an unacceptable risk of fatal thromboembolic complications. Beyond the consequences of the acute thrombotic episode are the long-term implications. In the intermediate term there is the need for prolonged anticoagulant therapy for a thrombotic episode. This treatment is associated with serious, if not fatal, bleeding complications.⁵ Furthermore, patients with cancer or impaired inhibition of coagulation are at greater risk for the development of recurrent thromboembolism, which in some cases may be fatal.⁶ Beyond this intermediate risk, patients who have suffered either symptomatic or asymptomatic DVT or PE are at greater risk for the development of the postphlebitic syndrome, the symptoms of which include leg swelling, varicose veins, venous ulceration, and other trophic skin changes, which are disabling and have impact on the patients quality of life.⁷

RISK OF VENOUS THROMBOEMBOLIC DISEASE IN GENERAL SURGICAL PROCEDURES

More than three decades of research have identified risk factors frequently associated with the development of postoperative VTE in general surgical patients. Most important amongst these are advancing age, a history of VTE, long periods of immobility, the presence of malignancy, obesity, the use of estrogen-containing oral contraception or hormone replacement therapy (HRT), and the type of surgical intervention.⁸^,⁹ The way that these risk factors interact in an individual patient to create a specific risk for the development of thrombosis is unpredictable. Therefore, although it would be possible to assign an individual surgical patient a score with regard to their risk for the development of thromboembolic episodes in the perioperative period, the use of such models has yet to be fully validated. A second approach on which decisions about thromboprophylaxis can be made is to define specific surgical risk groups in whom decisions about the use of thromboprophylaxis can be made¹⁰ (see Table 86.1).

CLINICAL TRIALS IN VENOUS THROMBOEMBOLISM—CONSIDERATIONS ABOUT EFFICACY AND SAFETY

The adoption of any specific measure for the prevention of VTE in surgical patients must be based upon evidence for both the efficacy of a particular intervention with regard to preventing thrombosis in patients at differing risks, and the safety of the intervention, in particular for pharmacologic prophylaxis, this being the rate of the bleeding complications associated with anticoagulant therapy.

In terms of assessing the efficacy of any given method of thromboprophylaxis, a number of different efficacy endpoints have been identified. Clearly, the most important endpoint is that of the prevention of fatal PE.¹¹ Although this complication is rare, and trials designed with this endpoint are the most challenging, landmark studies in the field of thrombosis research have used autopsy-proven fatal PE as a primary clinical trial endpoint.¹¹ Demonstration that an individual method of prophylaxis is able to reduce the frequency of PE provides strong confirmation of its utility and relevance in routine clinical practice for the given patient population.

Beyond the endpoint of fatal PE, symptomatic VTE, whether it is nonfatal PE or DVT, represents an important clinical complication that occurs more frequently than that of fatal PE. The use of those clinical endpoints, objectively confirmed using appropriate diagnostic tests, is crucial in trials assessing the efficacy of various thromboprophylactic strategies.¹²^,¹³^,¹⁴

Most cases of thromboembolic disease are, however, asymptomatic. Therefore clinical trials comparing different methods of thromboprophylaxis that use asymptomatic DVT as an endpoint require fewer patients. Studies using either fibrinogen leg scanning or more recently venography for the detection of asymptomatic DVT have provided the foundations of our understanding of the efficacy of various methods for preventing
DVT in surgical patients. These studies form the basis of many of the guidelines on the use of antithrombotic therapy in the perioperative period.¹⁰^,¹⁵^,¹⁶^,¹⁷ It is essential, however, when using surrogate efficacy endpoints such as venographically proven DVT, that the consistency of the natural history of the thrombotic episode between the asymptomatic DVT and the potential fatal PE has been defined for the population in question. It is also essential that the study design and the antithrombotic intervention does not change that natural history, thereby potentially providing an erroneous result with the use of this surrogate endpoint.

Table 86.1 Risk stratification for postoperative VTE and frequency of VTE without prophylaxis

	Venographic DVT		Pulmonary Embolism
	Calf(%)	Proximal (%)	Symptomatic (%)	Fatal (%)
Low risk: <40 y and uncomplicated surgery and no additional risk factors	2	0.4	0.2	<0.01
Moderate risk: >40 y or prolonged or complicated surgery or additional “minor” risk factors	20	5	2	0.5
High risk: major surgery for malignancy or previous VTE or knee/hip surgery or heparinincluded thrombocytopenia	50	15	5	2
DVT, deep-vein thrombosis; VTE, venous thromboembolism.

The balance between the efficacy in preventing a thromboembolic episode versus perioperative bleeding complications must be addressed in the use of any antithrombotic agent. This balance must be considered in the specific clinical setting in which the patients find themselves The potential risk for development of thrombosis or bleeding is dependent upon the individual clinical scenario (type and site of operation, patient comorbidities, and concomitant medication).

Studies over the past three decades have established that pharmacologic prophylaxis is not only effective in preventing thrombosis but is associated with an increase in perioperative bleeding complications.¹⁸ Thus the balance between efficacy and safety for any recommended pharmacologic thromboprophylaxis should demonstrate a desirable benefit/risk ratio and have the additional benefit of proven cost efficacy, which provides an economic rationale for routine protection against thromboembolic disease in moderate- and high-risk surgical populations.

METHODS OF PROPHYLAXIS

Mechanical Methods of Prophylaxis

Mechanical methods of prophylaxis are designed either to eliminate stasis or to interrupt thrombus progression from the deep venous system to the pulmonary circulation.

None of the currently available mechanical methods of thromboprophylaxis have been as extensively investigated as the pharmacologic methods.¹⁰ Of those available methods, one of the most popular is graduated compression stockings. These stockings apply a decreasing pressure from ankle to thigh across the lower limb, reducing the venous cross-sectional area and thus enhancing venous flow, and reducing stasis. Other mechanical methods increase pulsatile flow such as pneumatic calf compression, electrical calf stimulation, and venous foot pumps. All of these methods enhance venous flow and can be used intraoperatively and in the postoperative period, although after surgery they are cumbersome to use to maximum effect. Another approach to the prevention of PE is with either venous ligation¹⁹ or inferior vena cava (IVC) filters, which interrupt the transition of thrombus from venous system to pulmonary circulation. No randomized trials have been done to evaluate IVC filters for primary prophylaxis in surgical patients. They have been evaluated in patients with an established DVT for the prevention of recurrent PE, where they have been shown to be effective in reducing the frequency of PE initially, but in the longer term having been associated with an excess of VTE.²⁰

Pharmacologic Methods of Prophylaxis

A more popular approach to the prevention of VTE in the perioperative period has been the use of pharmacologic methods.

Aspirin

Aspirin and other antiplatelet drugs have become the mainstay of antithrombotic therapy for patients with atherosclerosis. Aspirin is highly effective in the prevention of arterial thrombotic episodes in patients with established vascular disease. The mechanism of action is through the inhibition of prostaglandin H synthetase 1 and thus a resultant loss of platelet cyclooxygenase activity.²¹

Interestingly, although aspirin is effective in the prevention of arterial thromboembolic events, its role in the prevention of venous thrombosis remains controversial. Despite the fact that aspirin is delivered orally, studies that evaluated its use in the prevention of VTE have often been of limited size and attended by important methodologic flaws that make their interpretation very difficult.²²^,²³ When compared with studies using other methods of thromboprophylaxis, the studies with aspirin do not justify its routine use, as a sole agent, for VTE prevention in medium- or high-risk surgical patients. Aspirin alone is therefore not recommended as thromboprophylaxis in any patients at risk of VTE.¹⁰

Dextrans

The polysaccharide dextrans have been evaluated as potential thromboprophylactic agents.²⁴ Their mode of action appears to be that of hemodilution, combined with an inhibitory effect on fibrin polymerization and an enhancement of fibrinolysis with some antiplatelet effect. They must be given intravenously and run the risk of severe anaphylactic reaction and volume overload. They are therefore not routinely used for thromboprophylaxis in most centers.

Unfractionated Heparin

Low-dose unfractionated heparin (LDUH) has remained the mainstay of thromboprophylaxis in general surgical patients.⁹^,²⁵ Indeed, it is in the evaluation of this agent that so much of our understanding of the natural history of VTE in surgical patients has been generated. Unfractionated heparin (UFH) binds and potentiates the activity of circulating antithrombin with the resultant inhibition of activated clotting factors X and II (thrombin). The agent can be given in a fixed dose of 5,000 anti-Xa units subcutaneously twice daily, commenced preoperatively, and continued in the postoperative period without need for monitoring of anticoagulant activity.

Low- and Ultralow-Molecular-Weight Heparins

Low-molecular-weight heparins (LMWHs) are derived from UFH through depolymerization.²⁶ A variety of different LMWHs are available for clinical practice, each of which has a different molecular weight and a different ratio of anti-Xa to anti-IIa inhibitory activity. The mechanism of action of LMWHs is the same as that of UFH, that is, the potentiation of antithrombin activity and the ability to release tissue factor (TF) pathway inhibitor and therefore inhibiting TF:VIIa activity. LMWHs achieve a lesser degree of inhibition of thrombin activity compared with UFH but a greater inhibition factor Xa activity. Advantages of the LMWHs include a longer plasma half life, the ability to dose once or twice daily subcutaneously without the need for dose monitoring, a limited effect on platelet activation, and a reduced incidence of heparin-induced thrombocytopenia²⁷ and osteoporosis.

A new generation of LMWHs and ultra-LMWHs is currently being investigated for the prevention of VTE.²⁸ Among these are bemiparin²⁹^,³⁰ and semuloparin,²⁸^,³¹ both of which are being evaluated in major general and orthopedic surgery. These drugs have higher anti-factor Xa and only minimal anti-factor IIa activities compared with first-generation LMWHs, and may in the future offer an alternative to currently approved drugs.

Indirect Factor Xa Inhibitors

The pentasaccharide fondaparinux—the first in a new class of indirect-acting selective inhibitors of activated factor X— became available in the early 2000s. Fondaparinux has a short 5-saccharide sequence, and can bind antithrombin and potentiate its inhibition of activated factor X. The molecular length is not sufficiently long to bind and inhibit thrombin. This drug has been evaluated primarily in the prevention of thromboembolic disease in orthopedic patients undergoing high-risk hip or knee replacement procedures³²^,³³^,³⁴^,³⁵ and in the treatment of DVT and PE.³⁶

A single study has been undertaken to evaluate its use in the prevention of thromboembolic disease in general surgical patients.³⁷ This study compared fondaparinux at a dose of 2.5 mg once daily with the LMWH dalteparin at a dose of 5,000 IU once daily. There were no significant differences in efficacy or safety between fondaparinux and the LMWH.

Direct Thrombin Inhibitors

Another interesting class of antithrombotic agents is the direct thrombin inhibitors. These agents, either naturally occurring or synthetic, are able to bind directly to and inhibit thrombin. They include hirudin, bivalirudin, desirudin, argatroban, melagatran-ximelagatran, and dabigatran etexilate. Most of these agents require parental administration, apart from ximelagatran and dabigatran etexilate, which are prodrugs and orally bioavailable. Development of ximelagatran ceased in 2006 following reports of hepatotoxicity.

These direct thrombin inhibitors are primarily being evaluated in either prophylaxis against thromboembolic disease in high-risk orthopedic patients undergoing joint arthroplasty,³⁸^,³⁹^,⁴⁰ or as a rescue antithrombotic strategy in patients with heparin-induced thrombocytopenia.⁴¹ Desirudin has been approved in the United States for VTE prophylaxis in patients undergoing elective hip replacement surgery.⁴²

New Oral Antithrombotic Agents

The limitations of UFH, LMWH as well as of vitamin K antagonists (VKAs) have led to the development of new oral antithrombotic drugs, of which specific and direct inhibitors of either thrombin or FXa are the most advanced evaluated. In contrast to LWMH and UFH, direct thrombin inhibitors inactivate both fibrin-bound thrombin and free thrombin, while direct FXa inhibitors inhibit FXa within the prothrombinase complex. In addition, both classes of new drugs have a rapid onset and offset of action, a low propensity for food and drug interactions and produce very predictable anticoagulant responses after fixed-dose administration so that routine coagulation monitoring is unnecessary.

The direct thrombin inhibitor dabigatran has been approved in many countries and is administered as the prodrug dabigatran etexilate. It has a longer half-life than the anti-Xa inhibitors (14 to 17 hours) and is eliminated to 80% via renal excretion. Two FXa inhibitors, rivaroxaban and apixaban, have been approved. They are given as active drugs and have multiple elimination pathways. All three drugs reach maximum plasma concentration after 2 to 4 hours. They are reviewed in detail in Chapter 110.

INTERPRETING DATA FROM CLINICAL TRIALS THAT HAVE ASSESSED ANTITHROMBOTIC AGENTS FOR THROMBOPROPHYLAXIS IN SURGICAL PATIENTS

Thromboprophylaxis in surgical patients presents a major challenge. Surgical trauma makes patients prone to perioperative bleeding complications. The delicate balance between efficacy (in terms of preventing VTE) and safety (in terms of not inducing excessive bleeding complications) of thromboprophylaxis has been discussed above.

Only a few antithrombotic agents have been assessed to the extent that there is a clear demonstration that their use is associated with a significant reduction in fatal PE, in well-designed randomized clinical trials.¹¹^,⁴³ However, in meta-analyses of methodologically sound clinical trials, where PE has been
included as one of the endpoints in the study, a number of agents have been shown to possess antithrombotic efficacy. With the use of symptomatic VTE as an endpoint, fewer patients, other than those with fatal PE, are required to demonstrate efficacy. Asymptomatic endpoints are only relevant when they predict both symptomatic DVT and potentially fatal PE.²⁵^,⁴⁴^,⁴⁵

Consistent assessment of bleeding in surgical patients is difficult. Bleeding in the perioperative period is determined not only by factors inherent to the patient, but to the nature and extent of operation, the experience of the surgeon, and any other concomitant drug therapy that the patient may be receiving. These factors interact in an unpredictable way with the administration of any particular antithrombotic therapy in the intra- and postoperative period. Since bleeding is an expected complication of operation, blood loss per se, both intraoperatively and in the early postoperative period, and consequent requirement for transfusion are difficult surrogate markers for assessment of bleeding risk for an antithrombotic agent in the perioperative setting. This situation is contrary to that seen, for instance, in patients who are receiving antithrombotic therapy for treatment of established thrombosis where bleeding would not be expected.⁴⁶ Differences between baseline risk and risk attributed to an antithrombotic agent are necessary to identify as their occurrence outside the controlled setting of a clinical trial may have important adverse consequences.

Few perioperative pharmacologic prophylaxis trials have used endpoints that assess the potential impact of excessive bleeding on surgical outcome.⁴⁷ However, such endpoints, which include need for reoperation secondary to bleeding, wound hematoma, and in particular, the need for surgical evacuation of wound hematomas, or consequences of bleeding, resulting in either anastomotic dehiscence, wound dehiscence, or arthroplasty joint sepsis, may represent more important endpoints, when considering the potential adverse consequences. However, when looking across the board at results of clinical trials that assessed antithrombotic agents in the perioperative period, it is clear that routine antithrombotic therapy with LDUH, LMWH, or fondaparinux is not associated with a significant adverse impact in terms of bleeding on surgical outcome in the general surgical population. With any drugs that are cleared by the kidneys, however, caution must be exercised in patients with reduced renal function, particularly in elderly patients, in those with diabetes mellitus, and in patients at high risk of bleeding.¹⁰

GENERAL SURGERY

Frequency of Thromboembolic Events

The rate of thrombosis reported in patients without prophylaxis undergoing general surgical procedures ranges from 15% to 30%, with rates of PE ranging between 0.2% and 5% and fatal PE < 0.012%¹^,⁹^,¹¹^,²⁵^,⁴⁸ (see Table 86.1). These historic rates of thrombosis may overestimate the magnitude of the current problem in surgical practice. Earlier mobilization, shorter length hospital stay, improved anesthetic techniques, greater attention to a reduction in perioperative sepsis through antibiotic prophylaxis, and greater use of thromboprophylaxis itself, have had an important impact on reducing the frequency of postoperative DVT in general surgical patients. However, against this is the clear trend for surgical procedures to be undertaken in far sicker patients, who spend longer time severely immobilized in the intensive care unit, and that procedures are now undertaken in older patients and far greater numbers of patients with malignant disease. Thus, in an era where studies using control or placebo against novel antithrombotic agents in surgical practice are no longer ethically justified, it is not possible to know the current rate of thromboembolic disease in patients who do not receive any form of thromboprophylaxis, but it would be wrong to assume that VTE is no longer a serious complication in contemporary surgical patients.⁴⁹ Indeed, the Million Women Study highlighted the ongoing risk of VTE among women undergoing inpatient surgery in the United Kingdom over the period 1996 to 2001, with an estimated 1 in 140 middle-aged women being admitted with VTE over the 3 months after surgery.⁵⁰ A U.S. study, based on data from the Nationwide Inpatient Sample, also reported a doubling between 1998 and 2005 in the number of hospitalized patients with a clinically recognized episode of acute PE among hospitalized surgical and medical patients. At the same time, the rates of bleeding and heparin-induced thrombocytopenia related to anticoagulant prophylaxis remained stable or declined.⁵¹

Methods for prophylaxis against VTE in general surgical patients include general measures and specific interventions with antithrombotic strategies. General measures include careful history taking to assess thrombosis risk and in particular to identify important risk factors such as increasing age, obesity, the presence of malignancy, and a history of VTE⁵²^,⁵³ (see Table 86.2). Careful attention to the prevention of perioperative sepsis, maintenance of fluid balance, and early mobilization also play an important role in the reduction of thrombosis risk.
Patients who are taking either oral contraceptives or HRT should have this therapy discontinued, ideally 1 month before surgery.

Table 86.2 Risk factors for venous thromboembolic disease

Factors	Severity
General Risk Factors
Age	Moderate
Obesity	Mild
Ethnicity	Mild
Immobilization	Moderate
Previous VTE	Severe
Estrogen therapy	Mild
Inherited thrombophilia	Severe
Malignancy	Moderate
Heparin-induced thrombocytopenia	Severe
Surgical Factors
Orthopedic surgery	Severe
Tourniquet application	Moderate
General anesthesia >30 min	Moderate
Venous catheterization	Moderate
Major trauma	Severe
Paralysis	Moderation
Cancer chemotherapy, antiphospholipid antibodies, and hyperhomocystinemia, all recognized risk factors for venous thromboembolism (VTE) in nonsurgical patients, are also likely to increase the risk of postoperative VTE.

The most popular methods for the prevention of thromboembolic disease in moderate- or high-risk general surgical patients are LDUH, LMWH, and fondaparinux. Indeed the highest quality of evidence in terms of both efficacy and safety with regard to thromboprophylaxis in the perioperative period is available with these agents.

The International Multicentre Trial published in 1975 randomized 4,121 patients aged over 40 years undergoing major surgical intervention to a control group or to receive LDUH in a dose of 5,000 U started 1 to 2 hours before surgery and continued three times daily in the postoperative period until the patient was fully mobile.¹¹ The primary endpoint in this study was autopsy-proven fatal PE. Nearly 70% of patients in this study who died underwent autopsy. The trial demonstrated a significant reduction in the frequency of fatal PE (16 patients in the control group vs. two in the lose-dose heparin group; P < 0.005) and provided the first evidence that heparin-based thromboprophylaxis not only prevented DVT, but was able to reduce the frequency of fatal PE and to have a significant impact on improving surgical outcome. The International Multicentre Trial hailed the modern era of thromboprophylaxis and has formed the basis of recommendations for thromboprophylaxis adopted throughout the world. The only safety concern was an increase in wound hematoma rates, associated with low-dose prophylactic heparin therapy (P < 0.01), also seen in other studies.

Numerous studies have since been completed that have assessed the use of LDUH with endpoints including PE, symptomatic VTE, and asymptomatic DVT.²⁵ These studies all confirm that when compared with control groups, LDUH therapy is associated with a >60% reduction in the frequency of both DVT and symptomatic thromboembolic disease.²⁵

The striking feature of LDUH prophylaxis is the associated reduction in perioperative mortality, with the frequency of PE-associated mortality falling from eight to one per 1,000 operated patients, thus profoundly impacting on surgical outcome. However, concerns about potential bleeding complications as manifest by wound hematoma formation drove the quest for the development of safer antithrombotic agents in surgical practice.¹¹^,²⁵

As a result, LMWH has been investigated extensively⁴⁷^,⁴⁸^,⁵⁴^,⁵⁵^,⁵⁶^,⁵⁷^,⁵⁸ since the first report of its use for the thromboprophylaxis in man in general surgical patients by Kakkar et al. in 1982.⁵⁹ In this study, patients were randomized to receive one of two doses of a LMWH. The study confirmed that LMWH given in a fixed dose once daily was capable of providing thromboprophylatic efficacy and was safe.

A recent meta-analysis has reviewed studies that randomized over 44,000 general surgical patients to trials comparing LMWH against LDUH.⁴⁸ In this meta-analysis, both agents provided equal efficacy and safety. The advantage of LMWH was that it may in general be administered once daily versus twice or three times daily with LDUH to achieve antithrombotic efficacy in surgical patients. LMWHs have broadly replaced LDUH for thromboprophylaxis in general surgical patients.

A recent multinational randomized study—SAVE-ABDO— compared enoxaparin 40 mg/d started preoperatively with semuloparin, an ultra-LMWH, 20 mg/d started postoperatively for the prevention of VTE in 4,413 patients with risk factors for VTE who were undergoing major abdominal surgery.³¹ While semuloparin failed to achieve noninferiority over enoxaparin for the prevention of VTE and all-cause mortality, the rates of major bleeding (odds ratio [OR] 0.63; 95% confidence interval [CI], 0.46 to 0.87) and the composite of clinically relevant nonmajor and major bleeding (OR 0.71; 95% CI, 0.54 to 0.93) were lower with semuloparin. This potential benefit in terms of safety is worthy of further investigation.

The selective inhibitor of factor Xa, fondaparinux, has been evaluated in patients undergoing high-risk abdominal surgery.³⁷ In a trial of some 3,000 general surgical patients, most of whom underwent operation for malignant disease, there was no significant difference in either efficacy with regard to thromboprophylaxis, major bleeding complications, or death in patients who received fondaparinux commenced postoperatively with LMWH (dalteparin) commenced preoperatively. In a study involving 1,309 patients undergoing major abdominal surgery, all of whom had intermittent pneumatic compression (IPC), postoperative fondaparinux reduced the VTE rate by 70% versus mechanical prophylaxis alone, at a low bleeding risk.⁶⁰

Mechanical Methods of Prophylaxis in General Surgery

Mechanical methods of prophylaxis have been assessed in general surgical patients.²⁴ Although they have no risk of bleeding complications, few trials of methodologic quality have evaluated their efficacy. In particular, no trials have demonstrated that mechanical methods of thromboprophylaxis, alone, are able to reduce the frequency of fatal PE in general surgical patients.

The use of graduated compression stockings alone is not advocated, with the exception of patients at high risk for bleeding, although recent analyses suggested a 50% reduction in the rate of DVT with their use compared with control.⁶¹^,⁶² However, when used in combination with LDUH, they appear beneficial. Studies have demonstrated a 75% reduction in the rate of DVT identified for patients using LDUH and graduated compression stockings compared with those who received LDUH alone.⁶² There is no evidence that this method of prophylaxis alone can reduce the frequency of PE. In addition, graduated compression stockings that are used inappropriately may cause a tourniquet effect on the lower limb, resulting in a potential enhancement of the risk of thromboembolic disease.⁶³

Intermittent pneumatic calf compression has been assessed for the prevention of DVT in general surgical patients.⁶⁴^,⁶⁵ It is not possible, however, on the basis of these small studies to make recommendations about its routine use. The results suggest either a similar efficacy to LDUH or no definite thromboprophylactic effect for the use of IPC alone. In a randomized trial of 2,551 cardiac surgical patients, the combination of pneumatic calf compression with LDUH reduced the rate of symptomatic PE to 1.5% compared to 4.0% for LDUH alone {(P < 0.001).⁶⁶ In general surgical patients with multiple risk factors and at very high risk for VTE, a combination of pharmacologic and mechanical prophylaxis with graduated compression stockings and/or IPC is advocated.¹⁰

Laparoscopic Surgery

Laparoscopic surgery is a major advance in general surgical practice, providing opportunities for reducing surgical trauma, patient discomfort associated with intra-abdominal and pelvic
surgical procedures, and hospital length of stay. The problem of VTE has not been investigated extensively in patients undergoing minimal access surgical procedures.⁶⁷ From the point of view of the pathogenesis of VTE, minimal-access surgical procedures such as cholecystectomy are associated with varying reports of activation of blood coagulation, ranging from only minimal⁶⁸^,⁶⁹^,⁷⁰ to the same degree of coagulopathy as that observed in the open cholecystectomy procedure.⁷¹^,⁷² Often, laparoscopic procedures may take longer than the open surgical procedure, and during operation both the pneumoperitoneum and the reverse Trendelenburg position used in operations such as cholecystectomy or fundoplication result in profound venous stasis of the lower limbs.⁷³ Laparoscopic surgical procedures appear associated with a shorter hospital stay and therefore a potentially more rapid postoperative mobilization. The outcome of patients once discharged from hospital in terms of their mobility after laparoscopic procedures, compared to those who remain in hospital, has not been investigated extensively.⁶⁷

A number of surveys have determined current surgical practice of thromboprophylaxis after laparoscopic cholecystectomy and reported rates of thromboembolic complications. In a UK study of 417 surgeons, 91% reported never having encountered a thromboembolic episode after laparoscopic cholecystectomy.⁷⁴ Most surgeons use LDUH heparin in these patients. In a similar study undertaken in Denmark, 80% of surveyed surgical departments reported no problems with postoperative VTE in patients undergoing laparoscopic cholecystectomy.⁷⁵

Beyond these surveys, there have been a few studies using objective screening techniques that have attempted to determine the rates of VTE associated with various laparoscopic procedures. In a small study, with the screening endpoint of contrast venography 6 to 10 days after operation, no case of DVT was identified in 25 patients who underwent laparoscopic cholecystectomy without thromboprophylaxis.⁷⁶ Similarly, low rates in the absence of thromboprophylaxis have been identified in a series of laparoscopic surgical patients screened with lower limb Duplex ultrasonography.⁷⁷^,⁷⁸

Registries of surgical outcome also indicate a low frequency for thromboembolic complications in laparoscopic procedures. In a North American analysis of more than 100,000 laparoscopic cholecystectomies, the rate of symptomatic VTE was 0.2% up to 3 months after operation.⁷⁸ Similar findings were identified in a further literature review including over 150,000 laparoscopic cholecystectomies where various thromboprophylactic strategies were used and where the rates of DVT, PE, and fatal PE were 0.03%, 0.06%, and 0.02%, respectively.⁷⁹ Finally, the Swedish registry of laparoscopic cholecystectomy reported a rate of VTE of 0.2% in more than 11,000 cases.⁸⁰

As a consequence of these very low reported rates of VTE, there have been few randomized trials to determine the potential benefit of routine thromboprophylaxis in minimal-access surgical patients. In one study, patients were randomized to placebo or to the LMWH dalteparin in a dose of 2,500 IU once daily for up to 10 days after operation. There was no evidence of DVT screened venographically in either group of patients.⁷⁶ In a second trial, graduated compression stockings alone or combined with the LMWH reviparin were compared. The rate of DVT, as screened with ultrasound 5 to 7 days after operation, was below 1%.⁷⁷

In conclusion, for simple laparoscopic procedures such as cholecystectomy or tubal ligation, VTE prophylaxis is not generally recommended. However, for patients with a history of VTE or active cancer or with more complicated procedures such as colectomy, it is indicated.

Cancer Surgery

Cancer is an important risk factor for VTE.⁹^,⁴⁸ The frequency of thromboembolic complications in patients undergoing major cancer surgery is thought to be about twice that of patients undergoing the equivalent operation without malignant disease.⁹^,⁸¹^,⁸² In general, LDUH¹¹ and LMWH⁴⁸ have been validated for the prevention of thromboembolic disease in patients undergoing major abdominal or pelvic procedures for cancer. In a recent subgroup analysis of more than 6,000 patients with malignant disease receiving perioperative LDUH or LMWH, compared with 17,000 patients without malignancy, fatal PE was three times as common in cancer than in noncancer surgical patients despite use of prophylaxis.⁸²

In a study of more than 2,000 patients undergoing laparotomy for cancer, the hypothesis that a higher dose of LMWH would be associated with a low incidence of postoperative thromboembolic complications in cancer patients was tested. In this trial, in the two-thirds of patients undergoing operation for malignant disease, increasing the dose of the LMWH dalteparin from 2,500 to 5,000 IU once daily was associated with a reduction in the frequency of postoperative DVT from 14.9% to 8.5%. Among the patients with cancer in this trial, there was no significant increase in bleeding complications associated with increasing the dose of LMWH. For the one-third of patients without malignant disease, although there was a reduction in the frequency of postoperative DVT with the higher dose of LMWH, this was associated with a significant increase in perioperative bleeding complications.⁸³

Beyond the intensity of perioperative antithrombotic therapy has been the suggestion that prolonging the duration of prophylaxis into the postdischarge period might be associated with a lower frequency of late thromboembolic complications in patients with malignant disease. This hypothesis has been tested in two recent randomized clinical trials involving LMWH.⁸⁴^,⁸⁵ Patients undergoing laparotomy for cancer were randomized to either 1 week of in-hospital prophylaxis or 4 weeks of in-hospital and postdischarge prophylaxis with the LMWH enoxaparin 40 mg once daily. Prolonged prophylaxis was associated with a reduction in the frequency of postoperative DVT as screened by venography at the end of the 4-week treatment period.⁸⁴ A similar study including patients with or without malignant disease demonstrated the benefit for prolonged thromboprophylaxis in cancer surgical patients randomized to in-hospital or in-hospital and postdischarge thromboprophylaxis with the LMWH dalteparin 5,000 IU once daily for up to 4 weeks.⁸⁵ In another trial of patients undergoing major general surgery, some of whom had a history of cancer, 4 weeks of therapy with dalteparin 5,000 IU once daily compared with 1 week of therapy reduced the risk of VTE with no increase in risk of bleeding.⁸⁶

Studies involving the LMWH bemiparin have been conducted recently in patients with cancer, with the focus on identifying the optimal dosing and duration of thromboprophylaxis.²⁹^,³⁰ In a Spanish study, Balibrea et al.²⁹ reported that patients with cancer were still inadequately assessed for risk of VTE, and that bemiparin 3,500 IU/d was more effective than 2,500 IU/d, with no increase in risk of complications. Kakkar et al.³⁰ compared
the efficacy and safety of 1 versus 4 weeks of treatment with bemiparin 3,500 IU in patients undergoing abdominal or pelvic cancer surgery; while extended treatment did not reduce significantly or change the incidence of the composite of DVT, nonfatal PE, and all-cause mortality, the incidence of major VTE (proximal DVT, nonfatal PE, and VTE-related deaths) was reduced (relative risk reduction [RRR] 82.4%; 95% CI, 21.5 to 96.1), with no increase in the rate of hemorrhagic events. These data suggest that for certain cancer surgical patients at high risk for the development of postoperative DVT, prolonged thromboprophylaxis may be indicated, although no definite recommendations about routine use for postdischarge prophylaxis can be made on the basis of currently available studies.

ORTHOPEDIC SURGERY

Patients undergoing orthopedic surgery are at an increased risk of developing VTE,¹⁰^,⁸⁷ especially in case of major procedures like hip and knee replacement or hip fracture surgery.

The risk for any DVT, proximal DVT, any PE or fatal PE after hip or knee replacement or hip fracture surgery in absence of prophylaxis is summarized in Table 86.3.¹⁰ PE was the cause of death in 14% of the patients who died after hip fracture surgery.⁸⁸ The standardized and routine use of thromboprophylaxis has rendered fatal PE after orthopedic surgery an uncommon event. However, symptomatic VTE remains present in 1.3% to 10% of the patients.¹⁰

The pathologic basis for postdischarge thrombosis is still unclear. The thrombosis could be asymptomatic on discharge and become clinically overt later or could be newly developed after discharge. A study in patients undergoing elective hip replacement with a negative bilateral venography at discharge, showed a remarkable high incidence of thrombosis over the succeeding 3 weeks.⁸⁹^,⁹⁰ This suggests a late development of thrombosis after hospital discharge.

Table 86.3 Prevalence of VTE after major orthopedic surgery without prophylaxis

	DVT (%)		PE (%)
Procedure	Total	Proximal	Total	Fatal
Hip replacement	42-57	18-36	0.9-28	0.1-2.0
Knee replacement	41-85	5-22	1.5-10	0.1-1.7
Hip fracture surgery	46-60	23-30	3-11	0.3-7.5

The use of antithrombotics for the prevention of VTE has been assessed in multiple randomized controlled trials in patients undergoing elective knee or hip replacement therapy and to a lesser extent in patients after hip fracture therapy. However, data on arthroscopic knee surgery, elective spine surgery, and injury of the lower extremities remain scarce. In addition, issues including the optimal timing and prolongation of prophylaxis remain debatable.

Prevention of Venous Thromboembolism in Elective Hip Replacement

Prevention of VTE is recommended in all patients undergoing elective hip replacement¹⁰ due to the high risk for proximal DVT, and nonfatal and fatal PE (Table 86.3).¹⁰ It could be argued that these numbers are an overestimation of current orthopedic surgery due to shorter hospitalization, earlier mobilization, improved anesthetic techniques, and the standardized use of antithrombotic agents. However, surgical procedures are undertaken in older patients, elevating the risk for VTE. Other risk factors are obesity, the presence of malignancies, and a history of VTE.⁹¹

Mechanical prophylaxis methods that have been studied in elective hip replacement surgery are graduated compression stockings,⁹²^,⁹³^,⁹⁴^,⁹⁵^,⁹⁶^,⁹⁷ IPC⁹⁸^,⁹⁹^,¹⁰⁰^,¹⁰¹^,¹⁰²^,¹⁰³ and a venous foot pump.⁹⁶^,¹⁰⁴ These non-pharmacologic methods of prophylaxis give a RRR of 20% to 70% for the development of DVT. However, mechanical prophylaxis is less effective in primary prevention than currently available anticoagulant drugs.⁹⁷^,⁹⁹^,¹⁰³ Therefore, primary prevention by mechanical prophylaxis is only recommended in patients with a high risk of bleeding.¹⁰

In a meta-analysis, aspirin gives a reduction in fatal and nonfatal PE of 25% compared to placebo and no prophylaxis.¹⁰⁵ Since alternative prophylactic therapies are more effective,²³^,¹⁰⁶^,¹⁰⁷ the use of aspirin as thromboprophylaxis in elective hip replacement is not recommended.¹⁰

LDUH is more effective than placebo or no prophylaxis²⁵ and is considered safe in the preoperative and postoperative setting. However, it is considered to be less effective than other regimens available in routine clinical practice.¹⁰⁷^,¹⁰⁸^,¹⁰⁹

The most commonly used thromboprophylaxis for total hip replacement in North America is adjusted-dose oral VKAs like warfarin.¹¹⁰ The primary advantage of VKA is the delayed onset of action, hereby allowing surgical hemostasis. However, VKAs are hardly used in Europe due to the variable response between patients, the lower efficacy compared to LMWH,¹⁰⁸^,¹¹¹ interaction with other drugs, and due to the need for frequent monitoring. The dosage of VKA is adjusted to adhere to the recommended International Normalized Ratio (INR) range of 2.0 to 3.0.¹⁰

In Europe, the most frequently used thromboprophylaxis is LMWH, which proved to be highly effective and safe for the primary prevention of VTE. LMWH is more effective than LDUH¹²^,¹¹²^,¹¹³^,¹¹⁴^,¹¹⁵^,¹¹⁶ and when compared to adjusted-dose warfarin no difference in the incidence of total and proximal DVT was found.¹¹⁷^,¹¹⁸^,¹¹⁹ However, in one study¹²⁰ LMWH prophylaxis started preoperatively was found to be more effective than warfarin for the prevention of thrombosis but the use of LMWH was associated with a significantly increased incidence of bleeding at the operative site and of blood product transfusion. Another study¹²¹ found a significant reduction in total, proximal, and objectively confirmed DVT in patients using
LMWH < 2 hours preoperatively or at least 4 hours after surgery, compared to warfarin started postoperatively (2.2% vs. 4.4%). Combining the results of the five studies in which LMWHs were compared with adjusted dose warfarin,¹¹⁷^,¹¹⁸^,¹¹⁹^,¹²⁰^,¹²¹ the rates of DVT were significantly lower in the LMWH group (13.7 vs. 20.7%, P ≤ 0.01). However, the proximal DVT rates were 4.8% and 3.4% (P = 0.08), respectively. Major bleeding occurred more frequently in the LMWH-treated patients (5.3% and 3.3%, P < 0.01). In a meta-analysis¹²² in patients undergoing hip or knee replacement surgery, VKAs were less effective in the prevention of total and proximal DVT compared to LMWHs (P < 0.01). Subgroup analysis for patients undergoing elective hip surgery did not significantly modify the results (P = 0.02). In conclusion, LMWH are more effective in the prevention of DVT but cause more major bleeding than dose-adjusted VKAs.

The synthetic pentasaccharide fondaparinux has been proven to be highly effective in the prevention of DVT among patients with elective hip replacement surgery.³³^,¹²³ In one study,¹²³ fondaparinux 2.5 mg sc (subcutaneously) once daily 4 to 8 hours after surgery was compared to enoxaparin 40 mg sc once daily started 12 hours before surgery. The rates of VTE were 4% in the fondaparinux group compared to 9% in the enoxaparin group (P < 0.01). The rate of proximal DVT was lower in the patients randomized to fondaparinux (1%) compared to LMWH (2%; P < 0.01). In another study,³³ the identical fondaparinux regimen 2.5 mg once daily was compared to enoxaparin 30 mg bid (twice a day) 12 to 24 hours after surgery. There was no significant difference between the overall risk of VTE (6% in the fondaparinux group compared to 8% in LMWH patients), nor in the rate of proximal DVT (2% vs. 1%, respectively). In both trials fondaparinux was associated with a nonsignificant increase in bleeding (P = 0.11). These data indicate, by indirect comparison, that LMWH and probably fondaparinux are more effective than VKAs in the prevention of VTE in patients undergoing elective hip replacement. However, there is a slight increase in bleedings in the surgical site. The higher efficacy and higher bleeding risks could be attributable to the more rapid onset of anticoagulant effects compared to VKAs.

Only gold members can continue reading. Log In or Register to continue