Abstract
Venous thromboembolism (VTE) accounts for a significant amount of morbidity and mortality in the United States. The diagnostic and therapeutic management has never been so varied with the numerous options now available.
My purpose is to briefly review epidemiology, pathogenesis, prophylaxis, diagnosis, and treatment of VTE, highlighting the important studies and consensus recommendations informing current clinical practice. Invasive, noninvasive, direct, and indirect diagnostic modalities are reviewed with an evidence-based theme.
Certainly, all patients should undergo a rigorous prophylaxis risk assessment for which various pharmacologic and nonpharmacologic options are available with varied clinical efficacy and safety profiles.
Unfractioned heparins, low molecular weight heparins (LMWH), and thrombolytic agents play different roles as we enter the next millennium. Optimal outcomes with high patient satisfaction can be achieved with appropriate use of LMWHs for both deep vein thrombosis and pulmonary emboli.
Venous thromboembolism (VTE) is an important cause of morbidity and mortality in the United States. Pulmonary embolism (PE) afflicts over 500,000 American lives annually, causes 10% of all in-hospital deaths, and remains the number one medical cause of maternal deaths associated with live births in the United States (1–4). PE is not bound by any age restriction, and anyone from infants to the most aged members of our population can be afflicted with increasing prevalence (5). Unfortunately, autopsy studies continue to show that most cases of fatal PE are unrecognized and undiagnosed (6).
The explosion of excellent scientific data on both diagnostic and therapeutic fronts makes PE a particularly exciting topic to review. Diagnostic and therapeutic options for developing a rational approach to patient management are many and varied. Diagnostic options for deep vein thrombosis (DVT) and PE include noninvasive venous Doppler and compression ultrasound, impedance plethysmography, 125 I fibrinogen leg scanning, and invasive venography. Therapeutic options include unfractionated (UH) and low-molecular-weight heparins (LMWHs), catheter-directed and systemic thrombolytics, and surgical thrombectomy.
Pathogenesis
Thrombi typically form along the venous valve cusps within the soleal sinuses of the calf as a result of platelet aggregation and can overwhelm the endogenous fibrinolytic system in minutes. The clot's propensity to embolize is greatest in the early or “loose” phase (first 7 days) when the thrombus is comprised of red cells, white blood cells, and platelets within a fibrin mesh. The clot is infiltrated by histiocytes and fibroblasts, becoming “adherent” and the organizational process continues through collateralization or retraction, ultimately leading to an irregular, often irreversible, intimal damage in the recanalization phase.
The risk of development of a VTE is directly related to three pathologic factors first identified by Virchow in the 19th century and now known as Virchow's triad: vessel damage, blood hypercoagulability, and stasis. It is now recognized that stasis primarily plays the role of a permissive factor and most research has concentrated on hypercoagulability, which can be acquired or congenital.
Prophylaxis
Risk factors can be roughly classified by the patient's history and condition and additional factors resulting from disease or surgical procedure as shown in Table 1. These risk factors were qualitative only and not well quantifiable until the fifth American College of Chest Physician's (ACCP) conference, published in the November 1998 supplement of Chest (7). This group examined hundreds of case and cohort studies and randomized placebo-controlled trials. The ACCP evaluated the data based on study design to assign a level of evidence and made graded prophylaxis and treatment recommendations. Four risk categories for VTE prophylaxis were established: low, moderate, high, and highest risk (Table 2) (7).
Risk factors for venous thromboembolism (VTE) in inpatients
Classification of level of risk (based on published data)
Even the most technically advantaged surgeon has an incidence of postoperative DVT of 10–30% for those undergoing general surgery, 29% for neurosurgical procedures, and 50% to 65% for hip and knee surgery, respectively, without appropriate prophylaxis (7). Thus, the recommendation is to assess all hospitalized medical and surgical patients for their risk of VTE and develop protocols with appropriate pharmacologic and non-pharmacologic prophylaxis, using good audit practices to ensure high-level care. The choice of optimal preventive therapy is determined by a patient's individual risk profile (Table 3) (7).
Prevention of venous thromboembolism
Diagnosis: DVT
The clinical diagnosis of DVT is challenging given the low sensitivity and specificity of bedside examination skills. The sensitivity is low given many large potentially dangerous venous thrombi are clinically silent. The specificity is low because many nonthrombotic disorders can cause clinical symptoms and signs similar to DVT. Disorders such as ruptured muscles, tendons, or popliteal cysts; superficial thrombophlebitis; cutaneous vasculitis; lymphedema; or cellulitis can all mimic DVT (8); therefore, objective testing is mandatory. Classical clinical features in combination with at least one risk factor have a high probability (78%) of DVT. Patients have a low risk probability (5%) if they have atypical or minimal symptoms and no risk factors (Table 4). Clinical probability can be combined with objective testing to assist in clinical decision making (9).
Probability of pulmonary embolism (PE). (Data from 15, 25)
Objective tests include the invasive venogram, which is the gold standard but is expensive, technically difficult (10% reported failure rate), and associated with a low incidence of post-venography thrombosis. A positive test result is identified by an intraluminal filling defect in the deep venous tree.
The noninvasive modalities include impedance plethysmography (IPG) and ultrasound. IPG is performed by placing a blood pressure cuff around the thigh and a set of electrodes on the calf. As the cuff is inflated, the augmented blood volume is recorded by the electrodes. When the thigh cuff is deflated, the blood volume should reduce precipitously; however, if it is impeded, there is evidence an occlusion is present and a DVT is diagnosed. This modality is less sensitive and specific than venous ultrasound and when negative should be repeated on days 1, 3, 5 and 7 (10–12).
Ultrasound is the most cost-effective first step for diagnosing a DVT. Color duplex is state-of-the-art and highly accurate for both occlusive and nonocclusive thrombi in the proximal veins, but calf veins are difficult to visualize.
While nuclear venography remains available, it has largely fallen out of favor and for good reason. This test often takes 24–48 hours for accurate interpretation to incorporate the radiolabeled fibrinogen into a vein and carries an infectious disease risk as it comes from whole blood products.
A common challenging scenario is the symptomatic patient with a high clinical probability of DVT and a negative ultrasound. The recommendation is to repeat the ultrasound on days 3, 5, and 7 and, if serial testing is impossible due to compliance or other factors, an invasive venogram should be pursued (Figure 1).
Algorithm for differential diagnosis of recurrent venous thromboembolism
Recurrent VTE and postphlebitic syndrome can manifest similarly. The current criteria for diagnosing recurrent disease includes enlargement of a thrombus by more than 2 mm (12,13), a change in the echogenicity of the existing thrombus, or a thrombus in a segment of vein previously deemed clear (Figure 2) (14).
Algorithm for differential diagnosis of deep venous thromboembolism
Dr. Deitelzweig is the Head of Ochsner's Section on Hospital Based Internal Medicine
Diagnosis: PE
PE can present in a rather fulminant and abrupt fashion or somewhat chronically and insidiously. Similar to DVT, there are a host of disorders that can mimic a PE including pneumonia, chronic obstructive lung disease, acute MI, pleurisy, congestive heart failure, angina, and lung cancer (15).
As the incidence and mortality rate of PE increases with age, so must our suspicion in the elderly whose only manifestation may be arrhythmia, fever, delirium, resistant heart failure, or bronchospasm. Dyspnea, tachypnea, or chest pain will be present in 97% of patients with proven PE (16).
The initial approach to diagnosing a PE entails a complete history and physical, chest x-ray, EKG, and perfusion lung scan. The EKG may show the classic S1, Q3, T3 pattern (right heart stain, right bundle branch pattern)or simply a sinus tachycardia. Findings on the radiograph might include Westermarks sign, Hampton's hump, or peripheral oligemia.
As the clinical presentation of PE is nonspecific, objective evaluations must be performed to clarify the clinical picture. Various invasive and noninvasive diagnostic methods have been developed, incorporating both direct and indirect approaches to diagnosis.
The direct approaches would involve either the invasive pulmonary angiogram or a perfusion lung scan. The value of the ventilation/perfusion lung scan has been extensively studied. Certainly, a normal result excludes PE.
More recently, there have been increasing reports of utilizing plasma D-dimer assays, spiral CT scans, and lower-limb ultrasonography to render a PE diagnosis. The D-dimer is a sensitive marker for fibrinolysis and the assay can be by enzyme-linked immunoadsorbent assay (ELISA), latex, or whole blood agglutination (17). The D-dimer by ELISA is useful in the outpatient setting if the D-dimer is less than 500 mg/L. This has been associated with a 90% negative predictive value in excluding a pulmonary embolus (18). The inpatient use of D-dimer is useless as it is often elevated whether a surgical procedure has been performed or not and should not be ordered for PE diagnosis (19).
Since a DVT is present in 70% of patients by classic venography with confirmed PE, the finding of a DVT in a patient with pulmonary symptomatology of PE confirms a PE diagnosis (20).
Spiral computerized tomography and magnetic resonance imaging are being suggested as modalities for diagnosing PE (21,22), though neither technique has yet demonstrated sufficient evidence to recommend their use, nor been validated by pulmonary angiography for subsegmental emboli (23). Spiral CT is criticized for missing central clots in the right middle and left lingular pulmonary arteries due to their nearly horizontal take-off from the hila (24). Hence, both false positives and false negatives have occurred at unacceptably high percentages.
The Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) studies combined clinical suspicion with the lung scan to determine the likelihood of PE and obviate the need for pulmonary angiography (25).
As shown in Table 4, a decision of whether or not to treat most these patients for PE can be made definitively. Fifty percent of patients, however, will have a PE in the setting of a low suspicion with a high probability scan, or in a high suspicion setting with an intermediate or low probability scan. It is in these individuals that either a pulmonary angiogram or another indirect approach may be used to confirm a diagnosis (26). As of January 2000, no predictive rule for assessing the clinical probability of PE in a standardized fashion has been validated.
Treatment
Various pharmacologic options have been studied with different efficacy and safety results including the anticoagulants UH, LMWH, and warfarin, and thrombolytic therapy. Surgical venous thrombectomy is rarely used because the risks outweigh the benefits. The de-endothelialized venous surface is highly thrombogenic and recurrence is common. While this procedure can be a life-saving resource, personnel and facilities are difficult to mobilize in time to treat those patients who might be candidates. In addition, the long-term efficacy of this procedure has not been proven.
One early question that was answered by Barritt and Jordan in 1960 (27) was, “should patients with an acute PE be anticoagulated?” In a thankfully abbreviated trial of only 35 patients, a greater than 25% mortality from PE was demonstrated in the untreated group versus no deaths from PE in the anticoagulated group. The current practice remains to anticoagulate these patients.
Unfractionated Heparin (UH)
The importance of achieving an adequate intensity of initial and maintenance anticoagulation with heparin was emphasized by noting a recurrent VTE rate of at least 29% without therapeutic anticoagulation (28,29). This can be achieved by initiating heparin with an intravenous bolus of 80 U/kg followed by an intravenous infusion of 18 U/kg/hr or subcutaneous injection of approximately 17,500 units twice daily. Using the partial tissue thromboplastin (APTT), the dosage of heparin should be adjusted to maintain an anticoagulant intensity above the lower limit of a defined therapeutic range (30). Warfarin can be started within the first 24–48 hours at a dosage of 5.0 mg to 7.5 mg a day, given the potential shortcomings of the 10 mg dose, notably a 36% overshoot phenomenon at 60 hours, requiring correction (31). Both heparin and warfarin should be overlapped for at least 4 days until the international normalized ratio (INR) is within therapeutic range (INR at 2.0 to 3.0), preferably for 2 consecutive days, at which time the heparin dosage can be discontinued (32). This is important because heparin's mechanism of action depends on activating antithrombin (antithrombin III previously) to inactivate both factors IIa and Xa. The mechanism by which this occurs differs. Heparin can bind both antithrombin and IIa (thrombin), create a physical bond inactivating IIA, or the heparin chain can bind with antithrombin inducing a conformational change. Factors II and X have respective plasma half-lives of 60 hours and 40 hours. In addition, when Coumadin is initiated, factor VII is the first factor reduced with a half-life of 6 hours, leading to an early prolongation of the INR. Therefore, the INR could be increased early when the patient still has a risk of clotting. Thus, the 4–5 days of overlapping heparin and warfarin (33).
Despite physicians being most comfortable with an APTT of 1.5–2.5 times the control as a therapeutic range for heparin, there are certainly alternatives, such as heparin levels via either thrombin/protamine titration with a target of 0.2 to 0.4 U/mL, or an anti-Xa level of 0.3 to 0.7 U/mL.
One of the problems most often confronted clinically with UH is heparin resistance. Heparin binds to various proteins in the plasma (platelets, factor VIII) and the cell wall leading to a variable dose response. Another is phlebotomy, which should be performed every 6 hours after the initial bolus. In addition, a notable adverse effect with all forms of heparin is heparin-induced thrombocytopenia (HIT). HIT occurs at an incidence of 3.5% with UH and 0.6% with LMWH (34). It appears as if UH establishes an intimate relationship with platelet factor 4 and that bond induces an antibody resulting in HIT (33). If this complication occurs, a direct thrombin inhibitor such as hirudin needs to be administered because of the potential for cross reactivity with another heparin. Furthermore, the risk of osteoporosis appears less with prolonged administration of LMWH than with UH (35).
Currently, the indications for lifelong warfarin include: (1) more than two episodes of recurrent VTE, (2) proven idiopathic VTE and thrombolytic disorders, and (3) continuing underlying risk factor(cancer).
The absolute contraindications to anticoagulant therapy include intracranial hemorrhage, active internal bleeding, bleeding peptic ulcer, and malignant hypertension.
Low-Molecular-Weight Heparin (LMWH)
A new option has recently become available in LMWHs, which are chemically or enzymatically depolymerizations of UH possessing real advantages over UH. LMWH has favorable pharmacokinetics with 90% bioavailability at both low (prophylaxis) and high (treatment) doses (Table 5). A prolonged half-life, independent of dose between 2 hours and 4 hours, leads to subcutaneous injection once or twice daily with a predictable dose response, most often without monitoring (anti-Xa level). The response is so predictable that, with the exception of certain high-risk situations, no monitoring is necessary. LMWHs have fewer pentasaccharide units, the high affinity binding sites for antithrombin III, and the anti-factor Xa to IIa ratio is 1:1 for UH and from 2:1 to 4:1 for LMWHs (33).
Pharmacokinetic comparison of unfactionaeted heparin and low molecular weight heparin (LMWH)
A host of these products are available in the United States and Europe (Table 6). Currently, there is not felt to be any class effect as these products differ in their anti-Xa to anti-IIa ratio molecular weights, plasma half-lives, and US and European clinical trial results.
Available for low-molecular weight heparin product profiles
After LMWH was shown to be efficacious and safe in comparison with UH, the challenge was to compare the two on an outpatient basis. Two large randomized trials using enoxaparin and nadroparin demonstrated safety and efficacy in the outpatient setting with certain exclusionary criteria (36,37). When enoxaparin was administered subcutaneously at a dosage of 1 mg/kg twice daily, the investigators found 5.3% of the 247 LMWH patients developed recurrent thromboembolism compared with 6.7% of the 253 standard heparin patients (pNS). In addition, there was no significant difference between the two groups in the incidence of major bleeding. Patients in the LMWH group were discharged from the hospital after a mean of 1.1 days while the UH patients spent a mean of 6.5 days.
Currently, enoxaparin (Lovenox) is FDA-labeled at a dosage of 1 mg/kg, given subcutaneously twice daily or 1.5 mg/kg once daily to inpatients with or without DVT or PE, and to outpatients at a dosage of 1 mg/kg twice daily for DVT without PE.
Fragmin is not yet approved for the treatment of VTE but has been used in dosages of 100 anti-factor Xa units/kg given subcutaneously twice daily and 200 anti-Xa units/kg given once daily.
LMWH has a significant cost and patient convenience advantage over UH, but not all patients with VTE should be treated with LMWH in the outpatient setting. Certainly, with increasing experience, there appear to be fewer absolute contraindications, and Ochsner has taken the step of implementing an outpatient protocol. On average, at least 3 months of oral anticoagulant therapy with warfarin is appropriate if a reversible risk factor has been identified. Otherwise, at least 6 months of therapy is recommended.
Thrombolysis
Catheter-directed and systemic thrombolytic agents with activated plasminogen to form plasmin, which then lyses fibrin in a thrombus, have been investigated. There are three FDA-approved thrombolytic regimens for PE utilizing weight-based monograms (Table 7) and, for DVT, a catheter-directed dissolution of the thrombus can be performed by an invasive cardiologist or radiologist. (Catheter-directed thrombolysts have not yet been approved by the FDA.) Thrombolysis should be considered in acute thrombi (less than 28 days old) as Bjarnason (38) reported a 66-100% lysis compared with only 33% if the clot was present for more than 4 weeks. Theoretically, this will result in a reduced incidence of post-phlebitic syndrome and might be considered for iliofemoral thrombosis (39).
FDA-approved thrombolytic regimens for pulmonary embolism
Conclusion
The field of venous thromboembolism continues to evolve at a rapid pace. Our focus needs to be one of avoidance and a high degree of suspicion as VTE most often remains unrecognized and undiagnosed.
The initial or recurrent diagnosis needs to be made in the most expedient, cost-effective fashion, minimizing the likelihood of developing significant pathophysiological sequelae, namely pulmonary hypertension or post-phlebitic syndrome with a constellation of symptoms and signs including pain, edema, ulceration, and hyperpigmentation. Therapy with an anticoagulant (UH or LMWH), thrombolysis, or surgery must be initiated with an evidence-based approach. The LMWHs have been revolutionary, especially with DVT patients who can often be treated in the outpatient setting, increasing the convenience to the patient and substantially reducing the cost to the health care system.
- Ochsner Clinic and Alton Ochsner Medical Foundation
References
In this issue
Jump to section
Cited By...
- No citing articles found.