|
 |
CONTENTS |
|---|
TOP
CONTENTS
DEEP VENOUS THROMBOSIS
PULMONARY EMBOLISM
TREATMENT
COMPLICATIONS AND SPECIAL...
FUTURE DIRECTIONS
REFERENCES
|
|---|
Introduction
Deep Venous Thrombosis
Risk Factors and Prevalence
Prevention
Diagnosis
Natural History
Pulmonary Embolism
The Integrated Approach
Treatment
Complications and Special Circumstances
Hemorrhage
Venous Thromboembolism and Pregnancy
Heparin, Thrombocytopenia, and Thrombosis
Bone Loss with Heparin
Coumarin Skin Necrosis
Lupus Anticoagulant and Antiphospholipid Syndromes
Elective Surgery in Patients Receiving Warfarin
Venous Thromboembolism and Cancer
Future Directions
Venous thromboembolism (VTE) is a complex vascular disease with a multifactorial pathogenesis that results in two major clinical manifestations. The first and more common manifestation is deep venous thrombosis (DVT), which usually arises in the deep veins of the calf and spreads upward. Pulmonary embolism (PE), the second and more serious manifestation, occurs as a complication of DVT proximal to the deep calf veins. Symptomatic PE occurs in approximately 30% of patients with DVT. If one counts asymptomatic events, some 50-60% of patients with DVT develop PE (1). Hospital-based epidemiologic data suggest that VTE affects one in 1,000 persons yearly in North America and Europe (2, 3), but this incidence is likely an underestimate because an unknown number of patients with this condition are undiagnosed or misdiagnosed. Pulmonary embolism is estimated to cause 50,000 deaths in the United States every year (4). Most deaths that are directly attributable to acute PE occur rapidly before the diagnosis can be confirmed and effective treatment implemented, which makes prevention in the high-risk patient imperative.
In the condition known as thrombophilia, VTE is recurrent and can result in significant disability or death, which in many cases happens late in the course of the disease from thromboembolic pulmonary hypertension and cor pulmonale (5). Venous thromboembolism occurs more often in Caucasians than in other races, and its risk increases with age, although the disease is seen in all ages, even children (6, 7). Aside from an increased risk during and shortly after pregnancy, there is probably no sex-linked predilection for VTE. This review gives current recommendations for clinical risk assessment and offers a practical approach to diagnosis, prevention, and treatment. Special emphasis will be placed on recent knowledge about pathogenesis and management. Whenever possible, recommendations given in this article are based on the findings of two or more large randomized clinical trials that show a concordant and significant result or have sufficient power to demonstrate equivalence. Recommendations based on less firm evidence will be identified as such. In this context, the reader should recognize that any recommendation, no matter how strongly supported by randomized clinical trials, must accommodate individual patients and circumstances.
Updates on this topic have appeared in this journal for over 20 years, with the last update being published in 1990 (8). During the last few decades, great strides have been made in all areas of VTE management. The prevalence of VTE, however, appears to be steady and may actually be increasing. Possible explanations for this increase include an aging population and longer survival of many patients with cancer (9).
Venous thromboembolism has been studied with well- designed, randomized clinical trials to the extent that recommendations about diagnosis, prevention, and treatment can be made with a higher degree of confidence than applies to many other conditions managed by pulmonologists and critical care medicine specialists.
Because of these studies, relatively well-defined standards of care have emerged. As a result of these findings and standards and also because failure to diagnose this disease can have catastrophic consequences, VTE has become one of the most common diseases involved in medical malpractice litigation in the United States.
| |
DEEP VENOUS THROMBOSIS |
|---|
|
TOP
CONTENTS
DEEP VENOUS THROMBOSIS
PULMONARY EMBOLISM
TREATMENT
COMPLICATIONS AND SPECIAL...
FUTURE DIRECTIONS
REFERENCES
|
|---|
Risk Factors and Prevalence
Deep venous thrombosis most often originates in the deep veins of the major calf muscles (10). The plexus of multiple branching veins at this anatomic site seem to favor thrombosis because of a predilection to venous stasis with patient immobility (11). Whether these small veins have some additional intrinsic defect such as decreased fibrinolytic activity has never been established. Major predisposing conditions to DVT are venous stasis from any cause, any type of trauma including surgery and childbirth, and increasing age (4). All cancers increase risk, although adenocarcinoma of the visceral organs is the most recognized malignancy associated with DVT. At this time, lung cancer is the most common malignancy associated with VTE. Other acquired conditions that predispose patients to DVT include elevated levels of antiphospholipid antibodies, hyperhomocysteinemia, and certain chronic diseases such as primary polycythemia.
Thrombophilia is a relatively new word coined to describe recurrent VTE from inherited causes, but the term is now widely applied to anyone with recurrent venous or arterial thrombosis from inherited or acquired causes. Our understanding of venous thrombophilia has recently advanced with the description of activated protein C resistance (12). This inherited abnormality, known as factor V Leiden, involves a point mutation (adenine for guanine) that results in the substitution of glutamine for arginine at position 506 on factor V. The substituted glutamine of factor V Leiden occurs at one of the three cleavage targets for activated protein C and renders the coagulation factor relatively resistant to degradation. Approximately 5% of Caucasians in Europe and North America are heterozygous for this genetic defect.
This mutation has been found in nearly all cases of activated protein C resistance associated with venous thrombosis. The heterozygous state carries a three- to fivefold increased lifetime risk for VTE, and the homozygous state confers a significantly higher risk. Table 1 shows prevalence and risk comparisons for the most common inherited conditions that lead to VTE. Deficiencies of antithrombin or protein C or S, although associated with a higher relative risk for VTE, are much less prevalent than factor V Leiden (11). Consequently, screening for thrombophilia after recurrent or idiopathic VTE should begin with the polymerase chain reaction-based assay for factor V Leiden, which can be obtained during an acute episode of VTE or when the patient is receiving long-term warfarin therapy. The abnormality will be identified in some 20% of patients who have had one episode and about 50% of those with recurrent disease. When the condition is identified, decisions about duration of treatment, intensity of prophylaxis, and family counseling must be made. Recommendations regarding these issues will be given in this review.
|
Despite its high prevalence, routine screening for factor V Leiden is not advised because most individuals with this abnormality will never suffer VTE. It is probably also not cost-effective to screen young women for factor V Leiden before they begin using oral contraceptives. Whether a personal or family history of VTE should prompt screening for factor V Leiden in individuals about to undergo high-risk surgery remains debatable (13, 14). In patients who undergo hip or knee replacement, the presence of factor V Leiden does not significantly increase the risk of venous thrombosis as long as effective prophylaxis is implemented (14). A prudent recommendation would be to prolong prophylaxis when a patient with factor V Leiden is placed in a high-risk situation.
Recently, a genetic variant that increases plasma prothrombin concentration has been associated with VTE (15). This abnormality appears to be present in 1-2% of the population and increases relative risk two- or three-fold for VTE.
Risk factors for VTE act cumulatively. An individual with a phospholipid antibody is placed at especially high risk during and after a surgical procedure, with its accompanying immobility and bed rest. The older patient with factor V Leiden who takes a long automobile or airplane trip is another example of cumulative risk. In many cases, risk status can be prospectively quantified and appropriate preventive measures prescribed, but in other cases risk cannot be accurately assessed because an underlying thrombophilic condition is undetected.
Prevention
Anticoagulants and other antithrombotics form the basis for prophylaxis (4). These drugs are usually begun when the higher risk (e.g., surgery) begins and continued for at least 5- 7 d. In high-risk patients, such as those undergoing a major orthopedic procedure on a lower extremity, prophylaxis is recommended for at least 7-10 d (16). Some high-risk patients, especially the elderly and those with a history of VTE, probably benefit from even longer periods of prophylaxis (17, 18). Such patients include the very elderly, those with prolonged postoperative immobilization, and those with known thrombophilia.
Mechanical prophylactic methods range from graded elastic compression hose to full-leg pneumatic compression devices. Calf-high pneumatic compressive devices appear to be sufficient in most patients and are especially useful in patients who cannot receive anticoagulant prophylaxis. Foot-pump devices appear promising but cannot yet be widely recommended for high-risk patients until additional studies confirm their benefits (19). Table 2 gives general recommendations for prophylaxis by risk group.
|
It should be noted that in North America, 7-10 d of prophylaxis following orthopedic surgery or other high-risk situations usually implies a component of outpatient drug administration. The question remains unanswered of whether 4-5 d of prophylaxis, currently the average length of hospitalization after hip or knee replacement, will be as effective as the longer period.
Combinations of anticoagulant and mechanical methods to prevent DVT have been shown to give additive effects in patients undergoing elective abdominal surgery, coronary artery bypass grafting, and neurosurgery (20). Similar combinations, such as a warfarin and pneumatic compression devices, are often applied to patients undergoing hip or knee replacement, but there is no conclusive data yet of their added efficacy.
No prophylactic technique is completely effective against VTE (3). We now have results from clinical follow-up in large cohorts who underwent hip or knee replacement and received 7-10 d of prophylaxis with either warfarin or low-molecular-weight heparin. The incidence of symptomatic VTE in these patients was approximately 1-2%, and fatal PE occurred in approximately 0.1-0.2% (24). Prior venographic studies in similar patients who received the same prophylaxis showed a follow-up venographic incidence of DVT of approximately 10-20% (3). From these findings, it appears that a given high-risk population will show a tenfold drop in incidence from venographic DVT to clinical VTE and a further tenfold decrease from clinical VTE to fatal PE. These findings strongly suggest that some patients suffer asymptomatic or subclinical DVT and PE and as a result do not receive treatment. Whether such patients are at higher risk for long-term post-thrombotic complications remains unanswered. Future studies of prophylactic measures will likely focus on reducing venographic DVT, because the number of patients required to prove a mortality effect below the present 0.1-0.2% will be prohibitively large. Questions regarding the adequacy of 5-7 d of prophylaxis versus the currently validated 7-10 d will probably also be addressed with venographic studies.
Vena caval filters are essentially prophylactic devices for pulmonary embolism. A filter should be used in the patient with proximal DVT who cannot receive anticoagulants or who has failed this therapy. A filter should also be inserted in the patient who undergoes pulmonary embolectomy for acute disease and probably in those who undergo pulmonary endarterectomy for chronic thromboembolic pulmonary hypertension (25). A number of devices exist and improvements to make insertion easier have increased their use. However, in the only controlled study of vena caval filters, insertion of a filter in patients after a first episode of DVT reduced the incidence of pulmonary embolism within the following 2 wk but did not improve either short-term or long-term mortality (26). In fact, in this study there was increased recurrent DVT over the ensuing 2 yr in the group that underwent filter placement. The practice of routinely inserting a filter in patients with free-floating caval thrombi is probably not warranted (27). Despite increased use of filters, further controlled studies are needed before indications can be broadened.
Diagnosis
In studies of patients with suspected DVT, only one in four will prove to have the condition (28). Common maladies that are often confused clinically with DVT include cellulitis, heart failure with edema, ruptured Baker's cyst, and chronic venous insufficiency. Duplex ultrasound with manual compression has become widely used to diagnose DVT (29). The test currently incorporates color Doppler flow with a B-mode image of the vein projected either longitudinally or transversely. Despite the procedure's frequent inability to image a thrombus directly, noncompressibility of a venous segment has been shown to be reliable indirect evidence of venous thrombosis. This technique is most useful with large veins that are easily compressed, and consequently, it has found its greatest application in the thigh. Ultrasound examination is less reliable in the small overlapping veins of the calf plexus and in noncompressible pelvic veins (30, 31). Ultrasound is also less reliable as a case-finding tool in asymptomatic patients, including those at high risk (31). The test appears to be better than impedance plethysmography and continues to be improved (32). Figure 1 gives a diagnostic approach to DVT using clinical inference and duplex ultrasound.
|
Contrast venography is useful when ultrasound results are equivocal or when clinical suspicion is high in the face of a negative ultrasound study. Repeated noninvasive studies also appear to be useful in the latter circumstance (29, 33, 34). Venography also remains the standard test for validating new diagnostic procedures. It is useful for diagnosing recurrent DVT, another setting where the noninvasive tests are less reliable.
Magnetic resonance imaging (MRI) appears promising for diagnosing DVT and may prove most useful in the calf and pelvis, where ultrasound is less reliable (35, 36). Magnetic resonance imaging may prove especially useful for diagnosing recurrent DVT, and it will improve further with the advent of better vascular imaging agents. Various other imaging techniques are under study. These usually consist of a radionuclide tagged to a small molecule that localizes to some component of a thrombus. At present, none of these newer tests appears poised to replace duplex ultrasound with compression.
The fibrin degradation product, D-dimer, has shown some utility as a screening agent when combined with a noninvasive test such as ultrasound or impedance plethysmography (37, 38). In this way a low D-dimer level (< 500 µg/ml), when combined with a negative noninvasive study, reduces further the likelihood of a false negative finding. False positives are quite common with elevated D-dimer levels, and the test cannot be used to conclusively diagnose DVT. The test does not appear to be useful unless combined with a noninvasive leg study. Furthermore, the various methodologies have not been standardized, and results from the most sensitive method, the ELISA, are not immediately available to aid in clinical decisions (39, 40).
Natural History
When acute proximal DVT is untreated, clinical PE occurs in one-third to one-half of patients (41, 42), and another one-third appear to have subclinical PE. Untreated PE in turn tends to be recurrent over days to weeks and can either abate spontaneously or result in death. Even when DVT is properly treated with anticoagulants, considerable disruption of the involved venous architecture with loss of deep venous valvular function can occur (43). The terms "post-phlebitic" or "post-thrombotic syndrome" describe the pain, edema, skin discoloration, and ulceration associated with chronic venous insufficiency that follows DVT. Depending on the severity of venous insufficiency, this syndrome occurs in 10-30% of patients with clinical DVT. Post-thrombotic syndrome develops slowly and can wax and wane. Recently, a prospective study has reemphasized the prevalence of the syndrome in properly anticoagulated patients (43). Ipsilateral recurrence of DVT seems to be a major predictor of risk for this chronic syndrome, which puts thrombophilic patients at especially high risk. Management approaches are discussed in the section on treatment.
| |
PULMONARY EMBOLISM |
|---|
|
TOP
CONTENTS
DEEP VENOUS THROMBOSIS
PULMONARY EMBOLISM
TREATMENT
COMPLICATIONS AND SPECIAL...
FUTURE DIRECTIONS
REFERENCES
|
|---|
Clinical presentations of PE are notoriously deceptive and nonspecific, which makes clinical diagnosis unreliable. As with DVT, when PE is clinically suspected approximately one in four patients will prove to have it (44). Rarely, pulmonary embolism presents in such a dramatic fashion that the diagnosis is intuitively obvious and empiric treatment will be started, but the usual presentation is sufficiently nonspecific that further testing is necessary to establish or exclude the diagnosis. Occasionally, the condition presents atypically so that even experienced clinicians overlook its presence. The clinical presentations can be classified into three large groups. The first and most common presentation is dyspnea with or without pleuritic chest pain and hemoptysis. The second presentation is hemodynamic instability and syncope, which is usually associated with massive embolism; the third and least common presentation mimics indolent pneumonia or heart failure, especially in the elderly.
The most common symptom of acute PE is dyspnea, which
occurs in 70-80% of patients with angiographically documented disease (45). Pleuritic chest pain and tachypnea are
also common. The PIOPED study found dyspnea, pleuritic
pain, and/or respiratory rate 
20 breaths/min in 97% of patients with angiographically proven pulmonary embolism. The
absence of this triad has become useful in minimizing the clinical probability of PE (46). Often these findings seem to wax
and wane, which supports the clinical perception that pulmonary emboli occur in clusters.
Because ventilation-perfusion (V/Q) lung scans are often indeterminate in patients with suspected pulmonary embolism, the construction of a clinical probability scale (clinical prior) has been recommended to refine the diagnostic utility of the V/Q scan (47). Various efforts have been made, and several schemes that utilize neural networks have been advocated (48, 49). In practice, clinical probability is usually determined after careful consideration of risk factors, presentation, and screening laboratory results. Clinicians at every level of training have been shown to quickly gain expertise with this method of inference. Table 3 gives the approach I favor.
|
Massive pulmonary embolism is present when two or more lobar vessels are occluded on angiogram or more than 50% of perfusion is lost on the lung scan. Patients with massive PE are more likely to present with syncope, increased pulmonic component of second heart sound (S2[P]), and hemodynamic instability (46). However, the majority of patients with massive disease indicated by imaging studies do not develop hemodynamic instability.
In the elderly patient, PE can be confused with unresolving pneumonia or refractory heart failure. Again, the most useful screening symptom is episodic unexplained dyspnea, and the presence of this complaint should always alert the clinician to the possibility of PE (50).
A widened (A
a)O2 difference, usually associated with a
low PO2, is the most common arterial blood gas abnormality in
patients with PE, but arterial blood gases can be normal with
PE in previously healthy patients, particularly the young. Although the PCO2 is most often low, it can be elevated in patients
with massive PE. The chest X-ray is often unremarkable, but
focal atelectasis, small pleural effusions, and ill-defined pleural-based infiltrates are often present on close inspection of
the film. Rarely, one sees an engorged central pulmonary artery with a paucity of peripheral vessels (Westermark sign).
The most frequent electrocardiographic finding is nonspecific
ST-T-wave changes. Right-axis deviation, S1-Q3-T3 pattern,
and P-pulmonale are seen much less frequently, with each occurring in less than 12% of patients (51).
All of these findings are nonspecific, and their presence or absence cannot be used with any certainty in diagnostic reasoning. However, otherwise unexplained findings associated with high clinical probability of PE combined with a risk factor, as shown in Table 3, predict the disease in some 60% of patients (44). Consequently, it is recommended that clinicians use clinical information and screening laboratory data to arrive at a clinical estimate of prior probability before learning the results of the V/Q scan or other imaging study.
Diagnosis of PE continues to be problematic under the best of circumstances. Ventilation-perfusion lung scanning remains the standard screening test for PE, although the results are more often than not indeterminate (44). The hallmark of acute PE on the lung scan is two or more moderate-to-large perfusion defects (> 25% of a lung segment) with intact ventilation and a clear chest X-ray in the involved area. The only clinically important cause of a false-positive study is prior PE that has not resolved, leaving a lung scan pattern that remains in the high-probability category. Unfortunately, the disease often accompanies prior cardiopulmonary disease, which can make the lung scan less reliable. Even in the normal lung, PE does not always cause the classic V/Q mismatch because the disease often causes pleural effusions or infiltrates in an area of involvement, either of which can make the ventilation scan abnormal. Furthermore, through vagal mediation of bronchoconstriction, PE can directly induce ventilation abnormalities in the involved area.
Despite over 40 years of effort to validate a less invasive test, pulmonary angiography continues to be the final arbiter for diagnosing PE. The procedure is expensive and requires special equipment and expertise, but with proper precautions and training it has proven quite safe (44). The definitive finding of PE on angiography is an intravascular filling defect seen on two or more projections. Frank vascular cut-offs are less reliable, and vascular pruning and hypovascularity are nonspecific. Despite the complexity of this procedure and our reluctance to perform it, pulmonary angiography remains the most useful test when lung scanning and leg studies prove inconclusive.
The Integrated Approach
Because lung scanning by itself is often inconclusive and high-quality pulmonary angiography is not widely available, interest has turned to combinations of leg and lung studies in an effort to narrow the indications for angiography. This approach is rational because DVT and PE are manifestations of a single disease. In addition, knowledge of the leg study in a patient with suspected PE yields useful prognostic information about risk for recurrent PE. In this way, the clinical prior, lower extremity ultrasound, and V/Q scans are used to identify patients who need immediate treatment and to risk-stratify the remainder for angiography or follow-up without anticoagulation therapy (52). Reviewers seem to agree on this diagnostic approach (55, 56). An example of such a pathway is given in Figure 2.
|
The ultrasound study on impedance plethysmography is also useful when performed sequentially to risk-stratify patients. In this way, patients with suspected pulmonary embolism but a nondiagnostic scan and who have repeatedly negative leg studies over 7-10 d can avoid long-term anticoagulation treatment (52, 53).
Newer imaging methods are under study, and spiral computed tomography (CT), electron beam CT, and MRI appear most promising now (57, 58). The number of patients studied is relatively small, but most studies of spiral CT report sensitivities greater than 80% and specificities greater than 90% in comparison to pulmonary angiography (59). These newer imaging techniques appear to be particularly useful for identifying large vessel (segmental or more proximal) PE. These techniques are being continually improved, but the time has arrived for a controlled comparison of either spiral CT or MRI to pulmonary angiography in patients suspected of having PE. The protocol should require a prior commitment to manage the patient based on the findings of the CT or MRI. In this way, careful clinical follow-up can determine if missing small vessel emboli on spiral CT or MRI results in an adverse clinical outcome.
The syndrome of chronic thromboembolic pulmonary hypertension (CTPH) develops in less than 1% of patients who suffer PE (60). From retrospective inquiries, approximately one-half of those with the syndrome appeared to have had undiagnosed PE and then survived to develop pulmonary hypertension. Others develop the syndrome despite seemingly adequate treatment, although they may have had additional undiagnosed episodes. The hallmark of the syndrome seems to be recurrent or unresolved PE. Patients with recurrent PE have thrombophilia by definition, although the underlying pathogenesis may not be defined. The presence of a lupus anticoagulant appears to confer the highest risk for this debilitating condition, but in one large series this abnormality was present in only 10% of patients (61). Most patients will not have an identified cause of thrombophilia. The role of factor V Leiden in the syndrome appears to be minor.
Patients with CTPH usually present with increasing dyspnea, which inexorably progresses from dyspnea only with exertion to constant dyspnea. The diagnosis of CTPH is usually entertained only after excluding other causes of dyspnea and hypoxemia. The syndrome should be considered in anyone with unexplained dyspnea on exercise. The chest X-ray will often be unremarkable, but it may show asymmetrically enlarged central pulmonary vasculature and areas of focal oligemia. Surface echocardiography is useful to assess right ventricular function and estimate right ventricular systolic pressure. The V/Q lung scan, the most important screening test, will show multiple large defects. In contrast, typical lung scans of patients with primary pulmonary hypertension are nearly normal or show only minimal perfusion defects. However, it should be noted that the perfusion scan often underestimates the extent of vascular obstruction in CTPH. Treatment for CTPH is described in the following section.
| |
TREATMENT |
|---|
|
TOP
CONTENTS
DEEP VENOUS THROMBOSIS
PULMONARY EMBOLISM
TREATMENT
COMPLICATIONS AND SPECIAL...
FUTURE DIRECTIONS
REFERENCES
|
|---|
Initial treatment of proximal DVT and PE is remarkably similar, whereas the treatment of deep calf vein thrombosis limited to the veins below the trifurcation remains controversial. In the only randomized trial of treatment for symptomatic calf DVT, patients who did not receive out-of-hospital therapy had a 29% recurrence rate (8 of 28) as contrasted to no recurrences in the 23 patients who received anticoagulation for 3 mo (62). The prospective study by Prandoni and colleagues also showed a high rate of recurrence of symptomatic calf DVT (43). Based on the available evidence, symptomatic patients with deep calf DVT should receive at least 3 mo of anticoagulant therapy. An alternate management possibility is to follow such patients with noninvasive tests and treat only those whose thrombi show progression over the next 7-10 d. In the United States, the latter approach has never been practical and has recently lost favor with increasing recognition that, over the long term, calf DVT can be as problematic as proximal DVT (63, 64).
Once the diagnosis of proximal DVT or PE is suspected, intravenous unfractionated heparin should be administered until the diagnosis is ruled out. When the diagnosis is confirmed, either low-molecular-weight heparin or unfractionated heparin should be administered. Unfractionated heparin is usually given by constant intravenous infusion and monitored by the activated partial thromboplastin time (APTT). The APTT correlates in a linear fashion with plasma heparin concentration up to one unit per milliliter, but different reagents and coagulation timers make the APTT quite variable relative to a given heparin concentration within this range. Consequently, the current recommendation is to give sufficient heparin to prolong the APTT into a range that corresponds to a plasma heparin level of 0.2-0.4 U/ml by protamine sulfate titration (0.3-0.60 U/ml by amidolytic assay [65]). This relationship can be established a priori by a simultaneous comparison of APTT and plasma heparin levels in 20-30 patients receiving heparin. Once the therapeutic range for the APTT is known, further monitoring of plasma heparin levels is probably unnecessary, except in the occasional patient with "heparin resistance" who requires more than 40,000 U/d. In these patients, plasma heparin levels determined either by protamine sulfate titration or with a chromogenic assay can be used to guide therapy (66). If the monitoring laboratory changes its coagulation timer or uses a different thromboplastin for the APTT, the correlation between APTT and plasma heparin levels should be reestablished. Table 4 gives a popular weight-based dosing and monitoring regimen for initial treatment of acute VTE with unfractionated heparin (67).
|
Intravenous unfractionated heparin has proven efficacy in the treatment of acute PE. The study by Barritt and Jordan (41) showed a nearly complete risk reduction for both recurrence and death from PE when heparin and oral anticoagulation were administered. In the untreated group, death from recurrent PE occurred in 25% of patients and nonfatal recurrence, in another 25%. Subsequent studies in the same era before sensitive diagnostic tests in patients with suspected VTE, who were intentionally not treated or unintentionally undertreated, have confirmed these high rates of recurrence and death (66, 70). More recent studies have shown a lower rate of recurrence in patients who were mistakenly not treated, but the thrombus load was quite low in these patients (44, 71). Two studies suggest that not all patients with proximal DVT require anticoagulation, but the patient numbers were low and this approach has achieved little acceptance (72, 73). Current trials have shown that VTE does recur even with adequate treatment (5-7%) but that the death rate directly attributable to recurrent PE is well under 5% (74). Based on these studies, it can be stated that prompt and adequate therapy with heparin or low-molecular-weight heparin, followed by oral anticoagulation for at least 3 mo, results in an 80-90% risk reduction for both recurrent VTE and death.
Intravenous heparin is widely recognized to be a difficult therapy to administer (78). The clearance of the drug is rapid and dose-dependent, and its kinetics are unpredictable because of variable interactions with plasma proteins and cells. This variability is seen even with weight-based dosing. Furthermore, a minimum or threshold effect of heparin is required to achieve an antithrombotic effect (79). Consequently, monitoring of drug effect and subsequent dose adjustment have proven necessary when treating acute VTE. Treatment doses of unfractionated heparin can be given subcutaneously, but monitoring and dose adjustment are still necessary. Prospective studies of heparin compliance often show great difficulty in achieving the therapeutic range (83), and ultrasound imaging studies of patients receiving heparin show a 10-20% incidence of thrombus growth despite careful attention to monitoring and dose adjustment (87, 88). Poor results in the latter studies are probably directly related to dosing and monitoring problems identified in the former studies but are also likely caused by the inability of heparin to inhibit thrombin that is bound to clot.
Because of the unfavorable properties of unfractionated heparin, interest has centered for several decades on the development of rapidly acting anticoagulants with more predictable dose-response characteristics. Low-molecular-weight heparins were the first class of new agents to become widely available (89, 90). These heparins are fractionated from the parent molecules in various ways to have mean molecular weights of approximately 4 kD with a range of 3-7 kD. Despite their small size, they have longer effective plasma half-lives than unfractionated heparin because low molecular weight heparins interact less with cells, platelets, and proteins and more with antithrombin. However, below a molecular weight of about 5.6 kD, low-molecular weight heparins lose anti-IIa activity, which means they cannot be reliably monitored with an APTT. The drugs can be monitored with an anti-Xa assay, but since they have favorable dose-response characteristics, monitoring is unnecessary in most patients. Because of these properties, low-molecular-weight heparins are typically dosed subcutaneously without monitoring or dose adjustment. As a class, they are much less likely to cause type II thrombocytopenia (HIT) than is unfractionated heparin (91). Long-term use of low-molecular-weight heparin also appears to cause less osteopenia than unfractionated heparin (92). Earlier studies also suggested that low-molecuar-weight heparins would result in less bleeding and thrombus growth than unfractionated heparin but more recent, larger studies have not confirmed this expectation (74)
At this time, low-molecular-weight heparins have mostly prophylactic indications in the U.S., but treatment results look promising and one drug (enoxaparin) has recently received the indication for treatment of unstable angina. It is expected that these drugs will soon enter wide use in treatment as has already occurred in Europe and elsewhere. In treatment regimens, low-molecular-weight heparins are most often given subcutaneously in a weight-based dose either once or twice daily. The weight-based dose is usually decreased somewhat in the massively obese and in those with significant renal insufficiency (creatinine > 2.0 mg/dl), since these drugs are cleared exclusively by the kidneys. Because they are usually dosed subcutaneously without monitoring or dose adjustment, low-molecular-weight heparins have ushered in the era of outpatient treatment of VTE. At the least, these agents should allow earlier hospital discharge with a period of home treatment with low-molecular-weight heparin. Table 5 gives prophylactic and treatment doses of the three low-molecular-weight heparins currently available in the U.S. In this regard, enoxaparin will likely be approved for treatment of DVT in either a twice-daily or once-daily dose. Dalteparin sodium has been evaluated with either twice-daily (93) or once-daily dosing (94). The treatment dose of ardeparin sodium has not yet been established, but early studies suggest it is approximately 130 anti-Xa U/kg twice a day.
|
|
Coumarin derivatives remain the drugs of choice for long-term outpatient treatment, and in North America sodium warfarin is used almost exclusively. Warfarin is usually begun with
or soon after heparin on the first day of therapy in a starting
dose of 5 mg (97). Subsequent dose adjustment is guided
by the international normalized ratio (INR), which should be
maintained in a therapeutic range of 2.0 to 3.0. Recent studies
have emphasized the increased risk of thrombosis when the
INR consistently falls below 2.0 (100). For this reason, authorities are now recommending a target INR of 2.5, which is
associated with no more bleeding than an INR of 2.0. Blood
for the prothrombin time should be carefully collected in a
3.2% citrate tube, not refrigerated, and the INR measured with a thromboplastin with an international sensitivity index (ISI) near 1.0 (103). Highly sensitive thromboplastins (ISI 
1) insure a more linear relationship between the INR and
the prothrombin time, whereas insensitive thromboplastins
(ISI > > 1) do not (104).
Diet, comorbidity, and numerous drugs interact with warfarin so that careful monitoring is necessary to keep the INR in the recommended therapeutic range of 2.0-3.0. These interactions have been extensively reviewed (105). Recently, generic warfarins have been reintroduced, which will likely make management of warfarin therapy even more problematic (106).
Because of possible teratogenic and fetopathic effects, warfarin is usually not administered during pregnancy or to women who can become pregnant (107, 108). Alternative treatment regimens are discussed in the section on pregnancy. Adjunctive recommendations for managing VTE include bed rest until heparin is therapeutic, and the use of fitted elastic stockings when the DVT patient becomes ambulatory. The latter recommendation is extremely important in patients with extensive DVT, since stockings appear to decrease the likelihood of subsequently developing the post-thrombotic syndrome (109).
The duration of anticoagulation therapy in patients with VTE remains controversial. In the past, 3-6 mo of therapy was recommended after a first episode of VTE and a year or more of therapy after recurrent disease. However, recent studies have shown that DVT is often a recurrent disease, especially when it occurs without an identifiable risk factor, a situation termed idiopathic or primary VTE (43). Longer therapy results in a lower recurrence rate at the cost of a slight increase in bleeding (43, 110). Therapeutic benefit appears to accrue mostly to those with idiopathic disease (111). The decision to implement prolonged anticoagulant therapy should take into account the patient's underlying risk status, age, family history, and individual preferences. Table 6 recommends therapy durations based on these considerations. For many of the conditions listed in the table, there are no randomized clinical trials to support the recommendations, so the clinician must exercise clinical judgment in partnership with the individual patient before determining length of therapy.
|
In this regard, it should be noted that the single best predictor of increased risk for VTE is a prior episode. Patients who have had one episode are at high risk to have another, whether or not they have a defined thrombophilic state (12). Similarly, patients with a family history of VTE also appear to be at higher risk for recurrence. Increasing age is a major risk factor for recurrence and should always be considered when determining duration of therapy.
Thrombolytic agents actively dissolve thrombi and are usually reserved for patients with massive PE who are hemodynamically unstable and for patients with extensive iliofemoral thrombosis. A recent report on the prevalence of the post-thrombotic syndrome after DVT suggests that we may need to increase use of lytic therapy in patients with large proximal DVT (43).
Urokinase, streptokinase, and tissue plasminogen activator (tPA) are currently approved for use in PE and streptokinase for treatment of DVT. All three thrombolytic agents have roughly equal thrombolytic effects, although tPA is infused for a shorter period and can be expected to cause more rapid thrombolysis (114). Many authorities agree that thrombolytic therapy should be reserved for the most severely ill patients. Those at low risk for bleeding with hemodynamically unstable PE or massive ileofemoral DVT appear to be the best candidates (117, 118). Other authorities widen the indication to patients with PE who have echocardiographic evidence of right ventricular dysfunction (119, 120).
The unstable patient who is suspected of having massive PE and is in or near frank shock presents a special problem. Often such patients are deemed to be too ill for transport to lung scanning or pulmonary angiography. Recently, transesophageal and surface echocardiography have been reported to be useful to diagnose both massive pulmonary embolism and free-floating right heart thrombi. In the selected patient, echocardiography may expedite diagnosis and allow more prompt institution of therapy, especially when thrombolytic therapy or surgical embolectomy is being considered (121).
In the few patients who fail thrombolytic therapy, mechanical and surgical means of thrombus extraction have been used. Various devices for transvenous extraction or fragmentation of thromboemboli have been described (122). Pulmonary embolectomy using cardiopulmonary bypass has been employed when other means either fail or are unavailable (125). Neither catheter manipulation nor surgical embolectomy has been tested in a controlled fashion against thrombolytic therapy or conventional anticoagulant therapy, and it is likely that no large controlled studies will ever be performed because of the relative scarcity of these patients. Depending on patient selection, a 20-75% mortality rate can be expected in such patients no matter what management approach is used.
A few patients develop the syndrome of CTPH. The frequency of this condition is disputed, but it appears to occur in less than 1% of patients with PE. It is sufficiently rare that only a few centers have evaluated large numbers of these patients. In selected patients with chronic pulmonary thromboembolism in proximal vessels, pulmonary endarterectomy has dramatically improved pulmonary artery pressure and functional status (126, 127). This surgery is delicate and fraught with hazard, both during the procedure and in the immediate recovery period, so that only specialized centers should perform pulmonary endarterectomy. A vena cava filter is often inserted during surgery, and the patient should subsequently receive life-long anticoagulant therapy.
| |
COMPLICATIONS AND SPECIAL CIRCUMSTANCES |
|---|
|
TOP
CONTENTS
DEEP VENOUS THROMBOSIS
PULMONARY EMBOLISM
TREATMENT
COMPLICATIONS AND SPECIAL...
FUTURE DIRECTIONS
REFERENCES
|
|---|
Hemorrhage
Bleeding is the predominant risk associated with all antithrombotic therapy (128). Major risk factors for bleeding are intensity and duration of therapy and patient factors, predominantly increasing age and significant renal or hepatic dysfunction (129). Despite the common fear of inducing bleeding with heparin, it appears that undertreatment with a subtherapeutic APTT in the first few days is more hazardous than overtreatment with a prolonged APTT (130). However, long periods with an overly prolonged APTT undoubtedly predispose patients to bleeding (131). The same relationships among dose, anticoagulant effect, and bleeding appear to hold for low-molecular-weight heparin (132).
In the case of life-threatening bleeding, heparin can be rapidly reversed with protamine sulfate given in a slow (10-20 min) intravenous infusion. One milligram of protamine neutralizes approximately 100 units of unfractionated heparin, and no more than 50 mg should be given with a single infusion unless a large overdose of heparin is known to have occurred. Protamine sulfate only partially reverses low-molecular-weight heparin, suppressing the drug's anti-IIa activity but with less effect on anti-Xa activity. Protamine sulfate should always be infused slowly, and resuscitative equipment should be immediately available. In most cases, bleeding associated with unfractionated heparin or low-molecular-weight heparin can be managed by stopping the drug and supporting the patient with blood transfusions and other factors as necessary without resorting to the use of protamine sulfate.
Long-term treatment with warfarin predisposes patients to bleeding when the INR is consistently above 4.0 (133). Careful monitoring is thus necessary. When significant bleeding occurs in the face of an excessively prolonged INR, either fresh-frozen plasma or vitamin K can be administered. If vitamin K is given intravenously, it should be diluted and infused over a period of at least 10 min. Again, resuscitative equipment should be immediately available. Many times an elevated INR can be corrected by holding doses or giving a small amount (1-2 mg) of oral vitamin K.
Thrombolytic therapy clearly increases bleeding risk, and patients should be carefully selected for this therapy (134). The elderly, patients with uncontrolled hypertension, and those with a recent stroke or craniotomy appear to be at especially high risk for intracerebral bleeding. When serious bleeding occurs, the drug should be stopped and the patient supported with blood transfusions and blood products (cryoprecipitate, fresh-frozen plasma) as needed. Intracranial bleeding may require surgical intervention. Thrombolytic agents can be reversed with inhibitors such as epsilon aminocaproic acid, but the rapid decrease of the thrombolytic effect usually makes administration of a direct inhibitor unnecessary.
Venous Thromboembolism and Pregnancy
Pregnancy increases the risk of VTE, with the highest risk period coming in the last trimester and the first few weeks postpartum (135). In regard to the diagnosis of suspected VTE in a pregnant patient, the diagnostic pathways outlined in Figures 1 and 2 are applicable with several caveats (136). In the third trimester, increased blood volume and obstruction to venous return by the gravid uterus can make noninvasive leg studies less reliable. Lung scanning is safe in the pregnant patient, although the dose of the radionuclide is usually reduced. With proper precautions, pulmonary angiography and venography expose the fetus to minimal radiation risk, and these tests should be performed when indicated.
When VTE is diagnosed during pregnancy, warfarin is usually not administered because of its fetopathic and teratogenic effects. Either heparin or low-molecular-weight heparin can be administered in place of warfarin for long-term treatment of VTE during pregnancy, since neither drug crosses the placenta (137, 138). One or the other of these drugs should be given subcutaneously in a treatment dose. At term, the heparin product should be discontinued for 24 h and labor then induced so that delivery can occur without any lingering anticoagulant effect, especially if epidural anesthesia is to be given (139). After delivery, heparin and warfarin are given for the standard 4-5 d overlap period and warfarin subsequently continued for at least 4 wk postpartum (140).
Heparin, Thrombocytopenia, and Thrombosis
Heparin-induced thrombocytopenia (HIT) develops in 3-4% of patients who receive unfractionated heparin for at least 7- 10 d (141). This process typically appears between 5 and 10 d of therapy, is immune-mediated, and can result in either paradoxical thrombosis or bleeding. The thrombotic complications are fearful and should be suspected whenever the platelet count falls to less than 50% of its baseline value or below 100,00 per µl. This severe thrombocytopenia should not be confused with the transient mild thrombocytopenia that often occurs in the first few days of heparin therapy. The diagnosis should always be suggested when the platelet count falls precipitously, because laboratory testing for the antibody is unreliable and time-consuming. When HIT is suspected, heparin should be immediately discontinued. A recent report has suggested that administering warfarin in the acute phase of HIT may actually aggravate the thrombotic tendency, possibly by suppressing protein C synthesis, and in this way lead to venous thrombosis (142).
Low-molecular-weight heparins crossreact with the antibody in vitro in some 90% of cases and should not be substituted in the acute setting (143). A newly released heparinoid, danaparoid, has a low crossreactivity with the antibody (< 10%) and is a logical substitute for unfractionated heparin when HIT develops in the presence of thrombosis (144). Danaparoid should be administered in a treatment dose and warfarin held for several days until the thrombotic episode appears to have resolved. Recombinant hirudin (145) has recently been approved for treatment of HIT with thrombosis. Table 7 gives some recommended dosing regimens with danaparoid and recombinant hirudin. Plasmapheresis and immune IgG infusion may also be effective in the treatment of thrombosis associated with HIT (146).
Bone Loss with Heparin
Patients are at risk for heparin-induced osteopenia after unfractionated heparin has been administered for longer than 1 mo, and incidence and severity of osteopenia are directly related to both dose and duration of heparin therapy (147, 148). As osteopenia progresses, bone loss becomes less reversible. The patient receiving long-term heparin should be monitored with tests of axial bone density, and heparin should be discontinued when bone loss is shown to be progressive (149). No preventive therapy has been proven effective for heparin- induced osteopenia, although supplements of vitamin D and calcium are often given, especially to pregnant women. The incidence of osteopenia appears to be less with low-molecular-weight heparin (92).
Coumarin Skin Necrosis
Coumarin derivatives as a class occasionally initiate a prothrombotic tendency that has been termed coumarin purpura, but actually represents widespread subcutaneous microthrombosis (150, 151). Favored sites are breasts, the abdominal wall, and the lower extremities. The syndrome occurs most often in patients who are deficient in protein C, either on a congenital basis or from large loading doses of a coumarin derivative (152). The syndrome has also been described in a few patients with malignancy and may occur in patients with HIT who receive warfarin during an acute thrombotic episode (142). Avoiding large loading doses of warfarin and initiating therapy with the estimated daily maintenance dose should minimize this complication in all of these situations. In the setting of HIT, holding warfarin for several days is advisable. Prompt administration of a treatment dose of unfractionated heparin or low-molecular-weight heparin is indicated when this syndrome is suspected.
Lupus Anticoagulant and Antiphospholipid Syndromes
Managing warfarin anticoagulation in patients with elevated levels of an antiphospholipid antibody is difficult because these patients seem to remain at increased risk for recurrent disease when anticoagulation is maintained in the therapeutic range. The current recommendation is to prolong the INR to 3.0 or more (153). In some patients with a lupus anticoagulant, the baseline prothrombin time will be prolonged, which makes the INR less reliable for long-term monitoring. In this circumstance, a chromogenic Xa assay can be substituted, and Xa activity maintained at a target of 20% of mean normal for the laboratory (154).
Elective Surgery in Patients Receiving Warfarin
Most patients who are receiving long-term warfarin for a remote episode of VTE can undergo elective surgical procedures after stopping warfarin and instituting the standard prophylaxis recommended for the particular procedure (155).
Warfarin should be held for 2-3 d before surgery and the procedure performed when the INR has nearly normalized. If the
INR has not approached normal before surgery, small doses
of vitamin K (
2 mg) can be given either orally or subcutaneously to rectify any lingering hemostatic abnormality. Anticoagulation with warfarin can then be restarted after the surgical
procedure. Large preoperative doses of vitamin K (5-10 mg)
will make the patient difficult to anticoagulate postoperatively.
In a few instances, the patient may be at especially high risk, e.g., those with certain mechanical heart valves, proven inherited thrombophilia with multiple recurrences, or a recent acute episode of VTE. In these situations, warfarin should be withheld and a treatment dose of low-molecular-weight heparin or unfractionated heparin substituted, beginning 3-4 d before surgery and continuing until 12 h before the procedure. At that point, the heparin product is held during and after the procedure until the bleeding risk has subsided. Heparin and warfarin are then restarted together, and heparin is continued for at least 4-5 d or until warfarin has achieved a therapeutic INR for two consecutive days. Depending on risk assessment, warfarin is usually administered for at least 4-6 wk after surgery.
Venous Thromboembolism and Cancer
All malignancies increase the risk of VTE. Factors intrinsic to cancer, chemotherapy, and inanition and immobility all contribute to the pathogenesis in this setting. Venous thromboembolism can also precede cancer by several years. In one study of patients with recurrent idiopathic VTE, 18% developed clinical cancer over the following 3 yr (156). Patients with idiopathic VTE should always be evaluated for the possibility of occult malignancy. Initial screening should include a careful history and physical examination, including routine laboratory tests such as chest X-ray, fecal occult blood test, urinalysis, and liver function tests. Further evaluation should be tailored to any positive findings from these screening procedures. Patients with VTE and cancer generally have higher recurrence and bleeding rates and higher mortality than similar patients without cancer (76). Such patients are also more likely to fail warfarin therapy and often require long-term treatment with subcutaneous heparin or low-molecular-weight heparin. Controlled studies of therapy for VTE in cancer patients are currently needed.
| |
FUTURE DIRECTIONS |
|---|
|
TOP
CONTENTS
DEEP VENOUS THROMBOSIS
PULMONARY EMBOLISM
TREATMENT
COMPLICATIONS AND SPECIAL...
FUTURE DIRECTIONS
REFERENCES
|
|---|
The near future will likely see widespread use of low-molecular-weight heparin in treatment of VTE with transition from inpatient to outpatient management of many patients. As the heparinoids and direct thrombin inhibitors enter use, further refinements should allow for more protocol-driven therapy with reduced input from health care providers. Cost savings should prove substantial and will be directly proportional to number of hospital days avoided (157, 158). Decreased hemorrhage and less thromboembolic recurrence may also result as we gain experience with these new agents. The association between cancer and VTE looms ever larger as we have become more successful in preventing VTE in high-risk surgical patients. Studies of newer antithrombotic agents in treatment of cancer patients with VTE are needed.
Numerous questions about duration of therapy remain to be answered. Of these, perhaps the foremost is how aggressively to search for individuals who are heterozygous for factor V Leiden and what to do when we find them. Should screening for the defect be performed in individuals about to undergo high-risk surgery or begin use of oral contraceptives? Should patients with the abnormality receive lifetime anticoagulation after the first event? What other predisposing factors interact with the genetic defect to place particular patients at high risk for recurrence? Extremely high plasma levels of homocysteine may combine with factor V Leiden to increase risk synergistically (159). Other genetic and acquired conditions that predispose patients to VTE are likely to be identified. Recently, venous thrombosis has been associated with an inherited abnormality of prothrombin that is present in 1-2% of the population (15). Further prospective studies are needed to address these issues.
The time has come to assess ultrafast spiral CT, electron beam CT, and MRI against pulmonary angiography in patients with suspected PE. One or more of these newer diagnostic tests should be compared to angiography in a randomized controlled study where treatment decisions are based on results from the imaging study to which the patient is assigned. In this way we will learn if CT or MRI gives prognostic information that results in similar clinical outcomes to those for patients managed with pulmonary angiography or lung scans and noninvasive leg studies. To use CT or MRI in any other way, e.g., in risk stratification and clinical management of patients with indeterminant lung scans, runs the risk of introducing additional expensive technology without definitive evidence of its efficacy.
Major unmet needs remain for preventing VTE. Although there is clear evidence of the superiority of some prophylactic methods over others in high-risk patients (4), clinicians continue to prescribe inferior methods or practice no prevention at all. The common use of duplex ultrasound at discharge in high-risk patients to determine the need for further prophylaxis appears to be unwarranted (160), but this approach continues to be widely employed and prophylaxis halted at discharge in those with negative studies. Whether 5 d of prophylaxis is sufficient in high-risk orthopedic patients remains unproven, and validation studies of shortened periods of prophylaxis are urgently needed.
As new preventive strategies emerge, we must agree on methods of evaluating efficacy in a setting where primary event rates are very low. Proof of equivalent event rates to the established prophylaxis has become insufficient. Cost-benefit analysis must also be performed. In today's environment, cost-benefit analyses and quality-of-life measures have become necessary for every new assessment of prevention, diagnosis, and treatment.
In closing this review, it is useful to recall where we started when, 40 years ago, we began to perform clinical trials in VTE. At that time we were attempting to manage a disease that was diagnosed clinically and was thought to require 2-3 wk of hospitalization for initial anticoagulant therapy. We now often diagnose and treat VTE patients with very short periods of hospitalization, and it appears that selected stable patients can be treated with home therapy from the onset. We have learned much about risk factors and the pathogenesis of VTE, which has led directly to significant improvement in prevention in high-risk patients. Progress has obviously been made, and randomized clinical trials have led the way. We can take pride in this progress as long as we remember that much remains to be done.
| |
Footnotes |
|---|
Correspondence and requests for reprints should be addressed to Thomas M. Hyers, M.D., C.A.R.E. Clinical Research, 533 Couch Avenue, Suite 140, St. Louis, MO 63122.
(Received in original form March 25, 1998 and in revised form August 12, 1998).
This review is dedicated to the memory of Kenneth M. Moser (1929-1997), the author of previous State of the Art articles on this subject.Dr. Hyers has received educational grants from DuPont Pharma, Pharmacia Upjohn, and Rhône-Poulenc Rorer.
| |
References |
|---|
|
TOP
CONTENTS
DEEP VENOUS THROMBOSIS
PULMONARY EMBOLISM
TREATMENT
COMPLICATIONS AND SPECIAL...
FUTURE DIRECTIONS
REFERENCES
|
|---|
1. Moser, K. M., P. F. Fedullo, J. K. LitteJohn, and R. Crawford. 1994. Frequent asymptomatic pulmonary embolism in patients with deep venous thrombosis. J.A.M.A. 271: 223-225 .
2. Nordstrom, M., B. Lindblad, D. Bergqvist, and T. Kjellstrom. 1992. A prospective study of the incidence of deep-vein thrombosis within a defined urban population. J. Intern. Med. 232: 155-160 .
3. Anderson, F. A., H. B. Wheeler, R. J. Goldberg, D. W. Hosmer, N. A. Patwardhan, B. Jovanovic, A. Forcier, and J. E. Dalen. 1991. A population-based perspective of the hospital incidence and case-fatality rates of deep vein thrombosis and pulmonary embolism. The Worcester DVT Study. Arch. Intern. Med. 151: 933-938 .
4. Clagett, G. P., F. A. Anderson, J. Heit, M. N. Levine, and H. B. Wheeler. 1995. Prevention of venous thromboembolism. Chest 108: 312S-334S .
5. Lindner, D. J., J. M. Edwards, E. S. Phinney, L. M. Taylor, and J. M. Porter. 1986. Long term hemodynamic and clinical sequelae of lower extremity deep vein thrombosis. J. Vasc. Surg. 4: 436-442 .
6. Rees, D. S., M. Cox, and J. B. Clegg. 1995. World distribution of factor V Leiden. Lancet 346: 1133-1134 .
7. Michelson, A. D., E. Bovill, and M. Andrew. 1995. Antithrombotic therapy in children. Chest 108: 506S-522S .
8. Moser, K. M.. 1990. State of the art: venous thromboembolism. Am. Rev. Respir. Dis. 141: 235-249 .
9. Alpert, J. S., and J. E. Dalen. 1994. Epidemiology and natural history of venous thromboembolism. Prog. Cardiovasc. Dis. 36: 417-422 .
10. Cogo, A., A. W. A. Lensing, P. Prandoni, and J. Hirsh. 1993. Distribution of thrombosis in patients with symptomatic deep vein thrombosis. Arch. Intern. Med. 153: 2777-2780 .
11. Nicholaides, A. N., V. V. Kakkar, E. S. Field, and J. T. G. Renney. 1971. The origin of deep vein thrombosis: a venographic study. Br. J. Radiol. 44: 653-663 .
12. Price, D. T., and P. M. Ridker. 1997. Factor V Leiden mutation and the risks for thromboembolic disease: a clinical perspective. Ann. Intern. Med. 127: 895-903 .
13. Middeldorp, S., C. M. A. Henkens, M. M. W. Koopman, E. C. M. van Pampus, K. Hamulýak, J. van der Meer, M. H. Prins, and H. R. Buller. 1998. The incidence of venous thromboembolism in family members of patients with factor V Leiden mutation and venous thrombosis. Ann. Intern. Med. 128:15-20.
14. Ryan, D. H., M. A. Crowther, J. S. Ginsberg, and C. W. Francis. 1998. Relation of factor V Leiden genotype to risk for acute deep venous thrombosis after joint replacement surgery. Ann. Intern. Med. 128: 270-276 .
15. Poort, R., F. R. Rosendaal, P. H. Reitsma, and R. M. Bertina. 1996. A common genetic variation in the 3'-untranslated region of the prothrombin gene is associated with elevated plasma prothrombin levels and an increase in venous thrombosis. Blood 88: 3698-3703 .
16.
Ricotta, S.,
A. Iorio,
P. Parise,
G. G. Nenci, and
G. Agnelli.
1996.
Post
discharge clinically overt thromboembolism in orthopaedic surgery
patients with negative venography
an overview analysis.
Thromb.
Haemost.
76:
887-892
.
17. Bergqvist, D., G. Benoni, O. Björgell, H. Fredin, U. Hedlundh, S. Nicolas, P. Nilsson, and G. Nylander. 1996. Low-molecular-weight heparin (enoxaparin) as prophylaxis against venous thromboembolism after total hip replacement. N. Engl. J. Med. 335: 696-700 .
18.
Dahl, O. E.,
G. Andreassen,
T. Aspelin,
C. Müller,
P. Mathiesen,
S. Nyhus,
M. Abdelnoor,
J. H. Solhaug, and
H. Arnesen.
1997.
Prolonged
thromboprophylaxis following hip replacement surgeryresults of a
double-blind prospective, randomized, placebo-controlled study with
dalteparin.
Thromb. Haemost.
77:
26-31
.
19. Westrich, G. H., and T. P. Sculco. 1996. Prophylaxis against deep venous thrombosis after total knee arthroplasty: pneumatic plantar compression and aspirin compared with aspirin alone. J. Bone Joint Surg. 78: 826-834 .
20. Wille-Jørgensen, P., J. Thorup, A. Fischer, J. Holst-Christensen, and R. Flamsholt. 1985. Heparin with and without graded compression stockings in the prevention of thromboembolic complications of major abdominal surgery: a randomized trial. Br. J. Surg. 72: 579-581 .
21. Wille-Jørgensen, P., O. Hauch, B. Dimo, S. W. Christensen, R. Jensen, and B. Hansen. 1991. Prophylaxis of deep venous thrombosis after acute abdominal operation. Surg. Gynecol. Obstet. 172: 44-48 .
22. Ramos, R., B. I. Salem, M. P. De Pawlikowski, C. Coordes, S. Eisenberg, and R. Leidenfrost. 1996. The efficacy of pneumatic compression stockings in the prevention of pulmonary embolism after cardiac surgery. Chest 109: 82-85 .
23. Agnelli, G., F. Piovella, P. Buoncristiani, P. Severi, M. Pini, A. D'Angelo, C. Beltrametti, M. Damiani, G. C. Andrioli, R. Pugliese, A. Iorio, and G. Brambilla. 1998. Enoxaparin plus compression stockings compared with compression stockings alone in the prevention of venous thromboembolism after elective neurosurgery. N. Engl. J. Med. 339: 80-85 .
24. LeClerc, J., M. Gent, J. Hirsh, W. H. Geerts, J. S. Ginsberg, and for the Canadian Collaborative Group. 1998. The incidence of symptomatic venous thromboembolism during and after prophylaxis with enoxaparin: a multi-institutional cohort study in patients who underwent hip or knee arthroplasty. Arch. Intern. Med. 158: 873-878 .
25. Hirsh, J., and J. Hoak. 1996. Management of deep vein thrombosis and pulmonary embolism: a statement for healthcare professionals. Circulation 93: 2212-2245 .
26. Decousus, H., A. Leizorovicz, F. Parent, Y. Page, B. Tardy, P. Girard, S. Laporte, R. Faivre, B. Charbonnier, F.-G. Barral, Y. Huet, and G. Simonneau. 1998. A clinical trial of vena cava filters in the prevention of pulmonary embolism in patients with proximal deep-vein thrombosis. N. Engl. J. Med. 338: 409-415 .
27. Pacouret, G., D. Alison, J.-M. Pottier, P. Bertrand, and B. Charbonnier. 1997. Free-floating thrombus and embolic risk in patients with angiographically confirmed proximal deep venous thrombosis. Arch. Intern. Med. 157: 305-308 .
28. Wells, P. S., J. Hirsh, D. R. Anderson, A. W. A. Lensing, G. Foster, C. Kearon, J. Weitz, R. D'Ovidio, A. Cogo, P. Prandoni, A. Girolami, and J. S. Ginsberg. 1995. Accuracy of clinical assessment of deep-vein thrombosis. Lancet 345:1326-1330. [Erratum, Lancet 1995;346: 516.]
29. Heijboer, H., H. R. Büller, A. W. A. Lensing, A. G. G. Turpie, L. P. Colly, and J. W. ten Cate. 1993. A comparison of real-time compression ultrasonography with impedance plethysmography for the diagnosis of deep-vein thrombosis in symptomatic outpatients. N. Engl. J. Med 329: 1365-1369 .
30. Pedersen, O. M., A. Aslaksen, H. Vik-Mo, and A. M. Bassoe. 1991. Compression ultrasonography in hospitalized patients with suspected deep venous thrombosis. Arch. Intern. Med. 151: 2217-2220 .
31. Davidson, B. L., C. G. Elliott, A. W. A. Lensing, and for the RD Heparin Arthroplasty Group. 1992. Low accuracy of color Doppler ultrasound in the detection of proximal leg thrombosis in asymptomatic high-risk patients. Ann. Intern. Med. 117: 735-738 .
32. Wells, P. S., J. Hirsh, D. R. Anderson, A. W. A. Lensing, G. Foster, C. Kearon, J. Weitz, A. Cogo, P. Prandoni, T. Minuk, G. Thomson, L. Benedetti, and A. Girolami. 1995. Comparison of the accuracy of impedance plethysmography and compression ultrasonography in out-patients with clinically suspected deep vein thrombosis: a two centre paired-design clinical trial. Thromb. Haemost. 74: 1423-427 .
33. Huisman, M. V., H. R. Büller, J. W. ten Cate, and J. Vreeken. 1986. Serial impedance plethysmography for suspected deep venous thrombosis in outpatients: the Amsterdam General Practitioner Study. N. Engl. J. Med. 314: 823-828 .
34. Hull, R. D., G. E. Raskob, and C. J. Carter. 1990. Serial impedance plethysmography in pregnant patients with clinically suspected deep-vein thrombosis. Ann. Intern. Med. 112: 663-667 .
35. Evans, A. J., H. D. Sostman, M. H. Knelson, C. E. Spritzer, G. E. Newman, S. S. Paine, and C. A. Beam. 1993. Detection of deep venous thrombosis: prospective comparison of MR imaging with contrast venography. Am. J. Roentgenol. 161: 131-139 .
36. Carpenter, J. P., G. A. Holland, R. A. Baum, R. S. Owen, J. T. Carpenter, and C. Cope. 1993. Magnetic resonance venography for the detection of deep venous thrombosis: comparison with contrast venography and duplex Doppler ultrasonography. J. Vasc. Surg. 18: 734-741 .
37. Wells, P. S., P. Brill-Edwards, P. Stevens, A. Panju, A. Patel, J. Douketis, P. Massicotte, J. Hirsh, J. I. Weitz, C. Kearon, and J. S. Ginsberg. 1995. A novel and rapid whole-blood assay for D-dimer in patients with clinically suspected deep vein thrombosis. Circulation 91: 2184-2187 .
38. Ginsberg, J. S., C. Kearon, J. Douketis, A. G. G. Turpie, P. Brill-Edwards, P. Stevens, A. Panju, A. Patel, M. Crowther, M. Andrew, P. Massicotte, J. Hirsh, and J. I. Weitz. 1997. The use of D-dimer testing and impedance plethysmographic examination in patients with clinical indications of deep venous thrombosis. Arch. Intern. Med. 157: 1077-1081 .
39. Becker, D. M., J. T. Philbrick, T. L. Bachhuber, and J. E. Humphries. 1996. D-dimer testing and acute venous thromboembolism: a shortcut to accurate diagnosis? Arch. Intern. Med. 156: 939-946 .
40. Moser, K. M.. 1994. Diagnosing pulmonary embolism: D-dimer needs rigorous evaluation. B.M.J. 309: 1525-1526 .
41. Barritt, D. W., and S. C. Jordan. 1960. Anticoagulant drugs in the treatment of pulmonary embolism: a controlled trial. Lancet 1: 1309-1312 .
42. Huisman, M. V., H. R. Büller, J. W. ten Cate, E. A. van Royen, J. Vreeken, M.-J. Kersten, and R. Bakx. 1989. Unexpected high prevalence of silent pulmonary embolism in patients with deep venous thrombosis. Chest 95: 498-502 .
43. Prandoni, P., A. W. A. Lensing, A. Cogo, S. Cuppini, S. Villalta, M. Carta, A. M. Cattelan, P. Polistena, E. Bernardi, and M. H. Prins. 1996. The long-term clinical course of acute deep venous thrombosis. Ann. Intern. Med. 125: 1-7 .
44.
A Collaborative Study by the PIOPED Investigators.
1990.
Value of
the ventilation/perfusion scan in acute pulmonary embolismresults
of the prospective investigation of pulmonary embolism diagnosis
(PIOPED).
J.A.M.A.
263:
2753-2759
.
45. Stein, P. D., M. L. Terrin, C. A. Hales, H. I. Palevsky, H. A. Saltzman, T. Thompson, and J. G. Weg. 1991. Clinical, laboratory, roentgenographic and electrocardiographic findings in patients with acute pulmonary embolism and no pre-existing cardiac or pulmonary disease. Chest 100: 598-603 .
46. Stein, P. D., H. A. Saltzman, and J. G. Weg. 1991. Clinical characteristics of patients with acute pulmonary embolism. Am. J. Cardiol. 68: 1723-1724 .
47. Stein, P. D., J. W. Henry, and A. Gottschalk. 1993. The addition of clinical assessment to stratification according to prior cardiopulmonary disease further optimizes the interpretation of ventilation/perfusion lung scans in pulmonary embolism. Chest 104: 1472-1476 .
48. Patil, S., J. W. Henry, M. Rubenfire, and P. D. Stein. 1993. Neural network in the clinical diagnosis of acute pulmonary embolism. Chest 104: 1685-1689 .
49. Tourassi, G. D., C. E. Floyd, H. D. Sostman, and R. E. Coleman. 1995. Artificial neural network for diagnosis of acute pulmonary embolism: effect of case and observer selection. Radiology 194: 889-893 .
50. Stein, P. D.. 1994. Acute pulmonary embolism. Disease a Month 40: 467-523 .
51. Stein, P. D., J. E. Dalen, K. M. McIntyre, A. A. Sasahara, N. K. Wenger, and P. W. Willis III.. 1975. The electrocardiogram in acute pulmonary embolism. Prog. Cardiovasc. Dis. 17: 247-257 .
52. Hull, R. D., G. E. Raskob, G. Coates, A. A. Panju, and G. J. Gill. 1989. A new noninvasive management strategy for patients with suspected pulmonary embolism. Arch. Intern. Med. 149: 2549-2555 .
53. Stein, P. D., R. D. Hull, and H. A. Saltzman. 1993. Strategy for diagnosis of patients with suspected acute pulmonary embolism. Chest 103: 1553-1559 .
54. Hull, R. D., W. Feldstein, and P. D. Stein. 1996. Cost-effectiveness of pulmonary embolism diagnosis. Arch. Intern. Med. 156: 68-72 .
55. Ginsberg, J. S.. 1996. Management of venous thromboembolism. N. Engl. J. Med. 335: 1816-1828 .
56. Hyers, T. M.. 1995. Diagnosis of pulmonary embolism. Thorax 50: 930-932