ORIGINAL INVESTIGATION. Low-Molecular-Weight Heparin vs Heparin in the Treatment of Patients With Pulmonary Embolism

Transcription

1 ORIGINAL INVESTIGATION Low-Molecular-Weight Heparin vs Heparin in the Treatment of Patients With Pulmonary Embolism Russell D. Hull, MBBS, MSc; Gary E. Raskob, PhD; Rollin F. Brant, PhD; Graham F. Pineo, MD; Gregory Elliott, MD; Paul D. Stein, MD; Alexander Gottschalk, MD; Karen A. Valentine, MD, PhD; Andrew F. Mah; for the American-Canadian Thrombosis Study Group Background: Pulmonary embolism (PE) occurs in 50% or more of patients with proximal deep-vein thrombosis. Low-molecular-weight heparin treatment is effective and safe in patients with deep vein thrombosis and may also be so in patients with PE. Recent rigorous clinical trials have established objective criteria for determining a high probability of PE by perfusion lung scanning. Objective: To compare low-molecular-weight heparin with intravenous heparin for the treatment of patients with objectively documented PE and underlying proximal deep vein thrombosis. Methods: In a multicenter, double-blind, randomized trial, we compared fixed-dose subcutaneous lowmolecular-weight heparin (tinzaparin sodium) given once daily with dose-adjusted intravenous heparin given by continuous infusion using objective documentation of clinical outcomes. Pulmonary embolism at study entry was documented by the presence of high-probability lung scan findings. Results: Of 200 patients with high-probability lung scan findings at study entry, none of the 97 who received lowmolecular-weight heparin had new episodes of venous thromboembolism compared with 7 (6.8%) of 103 patients who received intravenous heparin (95% confidence interval for the difference, 1.9%-11.7%; P =.01). Major bleeding associated with initial therapy occurred in 1 patient (1.0%) who was given low-molecularweight heparin and in 2 patients (1.9%) given intravenous heparin (95% confidence interval for the difference, 2.4% to 4.3%). Conclusions: Low-molecular-weight heparin administered once daily subcutaneously was no less effective and probably more effective than use of dose-adjusted intravenous unfractionated heparin for preventing recurrent venous thromboembolism in patients with PE and associated proximal deep vein thrombosis. Our findings extend the use of low-molecular-weight heparin without anticoagulant monitoring to patients with submassive PE. Arch Intern Med. 2000;160: The affiliations of the authors and list of members of the American-Canadian Thrombosis Study Group appear in the acknowledgment section of the article. RESULTS OF recent populationbased studies 1,2 demonstrate an annual incidence of deep vein thrombosis (DVT) and pulmonary embolism (PE) of 48 and 69 per persons, respectively. Asymptomatic PE occurs in 50% or more of patients with symptomatic proximal DVT. 3,4 Results of natural history studies 3-8 show that proximal deepvein thrombi pose a major threat of pulmonary embolization. Pulmonary embolism commonly occurs in patients presenting to the hospital. 5,9,10 Use of accurate objective tests to detect venous thromboembolism (VTE) 6-8,11-18 has led to randomized trials of treatment for venous thrombosis. These trials have shown that the initial heparin treatment intensity must be sufficient to prevent recurrent VTE. 22,28-32 Patients with proximal DVT who receive inadequate anticoagulant drug treatment have a 20% to 50% risk of recurrent VTE. 22,28-32 The standard treatment for acute VTE has been initial therapy with dose-adjusted continuous intravenous heparin, followed by long-term oral anticoagulant drug treatment ,25,26,33 Low-molecular-weight heparin fractions have a mean molecular weight of 4000 to 5000 d (by comparison, conventional heparin has a mean molecular weight of to d) Pharmacokinetic studies of low-molecularweight heparin show high bioavailability after subcutaneous injection and a longer half-life than unfractionated heparin. Anticoagulant monitoring of certain lowmolecular-weight heparin fractions is 229

2 PATIENTS AND METHODS STUDY DESIGN The American-Canadian Thrombosis Study 45 was a multicenter, randomized, double-blind clinical trial comparing unfractionated continuous intravenous heparin therapy with once-daily subcutaneous low-molecular-weight heparin therapy in patients with acute proximal DVT. The protocol mandated objective testing for PE in all patients at study entry. Fifteen centers in the United States and Canada participated in the trial. The protocol was approved by the institutional review board at each center. PATIENTS Consecutive eligible patients aged 18 years and older with proximal DVT (thrombosis of the popliteal or more proximal veins of the legs) documented by venography were enrolled in the study. Patients were eligible if they had none of the following: active bleeding or disorders contraindicating anticoagulant drug therapy; allergy to heparin, bisulfites, or fish; pregnancy; 2 or more previously documented episodes of DVT or PE; history of protein C deficiency; history of heparin-associated thrombocytopenia; severe malignant hypertension (diastolic blood pressure 130 mm Hg); severe hepatic failure (hepatic encephalopathy); severe renal failure necessitating dialysis; or geographic inaccessibility preventing follow-up visit attendance. Eligible patients were excluded if they received treatment with warfarin sodium, low-molecular-weight heparin, or heparinoids within 7 days before study entry; if they received treatment with therapeutic subcutaneous heparin within the preceding 12 hours; were receiving intravenous heparin; or if they declined to give written informed consent. Before randomization, patients were stratified into groups according to the study center where they were treated, presence or absence of previous VTE, and presence or absence of 1 or more risk factors for bleeding (surgery within the previous 14 days, history of peptic ulcer disease, thrombotic stroke within the previous 14 days, and a platelet count of /L). A randomized, computerderived treatment schedule was used to assign patients to receive intravenous heparin or subcutaneous low-molecularweight heparin. Within each stratum, the randomization schedule was balanced in blocks of 4. Approximately 50% to 60% of patients with acute proximal DVT had asymptomatic PE at presentation. 3,4 Perfusion lung scanning was performed on study entry to document the presence of PE and to compare against lung scan abnormalities found at any subsequent presentation with symptoms or signs of PE. In each patient, anticoagulant drug therapy was started as soon as possible after proximal DVT had been documented objectively, by ascending contrast venography 8,11 or by impedance plethysmography 6,7,12 or B-mode imaging using venous compression. 16,17 In patients studied by these noninvasive methods, the diagnosis was also confirmed as soon as possible by venography. REGIMENS Patients in the intravenous heparin group received an initial bolus dose of 5000 US Pharmocopeia units of heparin followed by continuous intravenous infusion of heparin. The initial dose was U every 24 hours for patients without the designated risk factors for bleeding and U every 24 hours for those with 1 or more designated risk factors. The doses were chosen to minimize the risk of insufficient heparin treatment during the first 24 hours of therapy 22,28-32 and to avoid high initial doses of heparin in patients with designated risk factors for bleeding. 30 The dose of intravenous heparin was adjusted according to the results of laboratory monitoring using the activated partial thromboplastin time (APTT). This was obtained 4 hours after starting heparin administration and was repeated every 4 to 6 hours until the result was within the prescribed therapeutic range ( 1.5 to 2.5 times the mean control value of 30 seconds obtained with a thromboplastin reagent [Actin FS; Dade Behring, Deerfield, Ill]). Thereafter, the APTT was measured once daily; if the result was subtherapeutic, the test was repeated every 4 hours until the therapeutic range was regained. Patients receiving low-molecular-weight heparin were given a fixed dose of 175 International Factor Xa Inhibitory Units per kilogram of body weight subcutaneously once every 24 hours. This regimen was chosen because results of pharmacokinetic studies in normal subjects demonstrated unnecessary because of a predictable anticoagulant response when administered subcutaneously in weightbased doses ,44 Low-molecular-weight heparin treatment has been shown to be effective and safe in patients with DVT and may also be so in patients with PE. We conducted a double-blind, randomized trial comparing low-molecular-weight heparin (tinzaparin sodium) with intravenous heparin treatment in patients with objectively documented proximal DVT. 45 A priori, all patients underwent baseline lung scanning, and those with subsequent episodes of suspected recurrent VTE underwent repeated objective testing. Recent rigorous clinical trials 18,53-55 have established objective criteria for determining a high probability of PE by perfusion lung scanning. Almost half (47.7%) of the study population (200 of 419 patients) had high-probability lung scan findings at randomization. This finding and the randomized trial design allowed us to compare low-molecular-weight heparin vs unfractionated heparin treatment in patients with objectively documented PE and proximal DVT. Our objective was to determine whether low-molecular-weight heparin administered subcutaneously once daily without anticoagulant monitoring is effective and safe in such patients. RESULTS PATIENTS Of 432 consecutive patients with proximal DVT enrolled in the study, 419 (97.0%) underwent lung scans. Of these 419 patients, 200 (47.7%) had high-probability 230

3 Frequencies of recurrent VTE are shown in Table 2. All patients presented with overt signs and symptoms of VTE. Analysis using the log-rank test, which takes into account the length of time to an event, indicated a statistically significant difference (P =.009) between groups in the frequency of recurrent thromboembolic events (Table 2 and Figure 1). Of 7 patients in the intravenous heparin group with new episodes of VTE, 4 had new episodes of PE (all identified by new high-probability lung scan findings). Recurrent venous thrombosis was documented by venography in 1 patient and by impedance plethysmography in the remaining 2. The APTT during initial heparin treatment was in the therapeutic range in 6 of 7 patients. Durthat it produced a sustained anticoagulant response (antifactor Xa activity) throughout the 24-hour dosing period and did not produce a substantial accumulation of the anticoagulant effect when given for 5 to 6 days. All patients received long-term therapy with warfarin sodium for at least 3 months. The initial dose was 10 mg given on the second day of initial therapy, which was then adjusted to maintain the international normalized ratio between 2.0 and ,33 After the first 6 days, the dose was adjusted weekly by the patient s primary care physician. Treatment with intravenous heparin or subcutaneous low-molecular-weight heparin was discontinued on the sixth day provided that the international normalized ratio was 2.0 or more. The study used a double-blind design. Patients who were randomly assigned to receive intravenous heparin also received a subcutaneous placebo injection once every 24 hours. Patients assigned to receive subcutaneous lowmolecular-weight heparin also received an intravenous placebo bolus and a continuous intravenous infusion of placebo throughout initial therapy. To maintain doubleblinding, APTTs were reported only to a member of the health care team not involved in assessing the patient s outcome. The APTT was not recorded on the patient s medical chart during the study or reported to any other member of the health care team. Adjustments in the rate of intravenous infusion of heparin or placebo were made by an unblinded physician according to dosing schedules established before the trial began. Use of drugs containing aspirin was prohibited during the study. Use of sulfinpyrazone, dipyridamole, and indomethacin was strongly discouraged. SURVEILLANCE AND FOLLOW-UP All patients were examined daily during initial therapy; symptoms or signs of recurrent DVT, PE, or bleeding were sought. Perfusion lung scanning was performed in all patients within 48 hours of study entry. The diagnosis of PE on study entry was established according to published criteria for perfusion lung scanning. 18,53-55 When a subsequent episode of PE was suspected based on clinical signs or symptoms, the diagnosis was confirmed by lung scanning (indicating a new perfusion defect with a high probability of PE) 18,53-55 or by pulmonary angiography (revealing a constant intraluminal filling defect on multiple films), 56 performed when lung scanning did not indicate a high probability of PE. 18,53-55 All patients were followed up for 3 months to assess whether inadequate initial therapy could lead to recurrent thromboembolism during long-term therapy with warfarin. 31,32 Patients were asked to go to the hospital immediately if symptoms or signs of recurrent DVT or PE developed. Patients with suspected recurrent PE underwent objective testing as described immediately above. Those with suspected recurrent venous thrombosis underwent impedance plethysmography and venography; the diagnostic criteria are described elsewhere. 6,7,11,14 Bleeding was classified as major or minor according to criteria described elsewhere. 24,26 Data on the outcome measures of effectiveness (recurrent VTE), safety (bleeding complications), and patient deaths were interpreted by a central adjudicating committee. Adjudication was made by 2 committee members not involved in the patient s care; disputes were resolved independently by a third. Objective test results were interpreted independently and without the interpreter s knowledge of the patient s other results, clinical findings, or treatment group. STATISTICAL ANALYSIS We estimated that a sample of 200 patients per group would be large enough that a 95% confidence interval for the difference in frequencies of recurrent VTE would exclude a true difference of 5% or more, assuming observed frequencies of 5% in both treatment groups. Uncorrected 2 and Fisher exact tests were used to compare the frequencies of death, recurrent VTE, and bleeding in both treatment groups. Ninety-five percent confidence limits for the true incidences of recurrent VTE and bleeding complications were calculated from the binomial distribution. Confidence intervals for the difference between the 2 treatment groups in the incidence of recurrent VTE and bleeding complications were calculated using the normal approximation to the binomial distribution. The log-rank test was used to assess differences in the cumulative incidence of death and recurrent VTE. Values for the APTT and anti-factor Xa levels obtained in a given test are displayed as box plots. 57 perfusion lung scan findings, 183 (43.7%) had nondiagnostic perfusion lung scan patterns, and 36 (8.6%) had normal perfusion scan results. Lung scanning was not feasible in 13 (3.0%) of 432 patients. High-probability perfusion lung scan findings were present in 103 (47.0%) of 219 patients assigned to receive intravenous heparin and 97 (45.5%) of 213 patients assigned to receive low-molecular-weight heparin. The treatment groups were comparable at entry except for age; there were more elderly patients in the lowmolecular-weight heparin group (Table 1). To assess the possible effect of this age imbalance, multiple logistic regression was used; no significant effect of age was found. All patients were followed up during initial therapy and during 3 months of long-term therapy; none were lost to follow-up. RECURRENT VTE 231

4 Table 1. Clinical Characteristics of Patients With High-Probability Lung Scan Findings Treated With Continuous Intravenous Heparin or Low-Molecular-Weight Heparin Characteristic Intravenous Heparin (n = 103) Patients, No. ing long-term follow-up, subtherapeutic prothrombin times were noted before or at the time of the recurrent thromboembolic event in only 2 of 7 patients who were receiving intravenous heparin. There were no episodes of recurrent VTE during initial treatment or long-term follow-up of patients receiving low-molecular-weight heparin. BLEEDING COMPLICATIONS Low Molecular- Weight Heparin (n = 97) Age, 60 y/ 60 y 41/62 30/67 Sex, M/F 50/53 38/59 Status at study entry Symptomatic deep vein thrombosis Symptoms of pulmonary embolism Previous deep vein thrombosis Previous pulmonary embolism 6 4 Possessing 1 designated risk factors for bleeding Clinical measures at study entry Surgery or trauma in past 6 mo Cancer Coronary heart disease Chronic obstructive pulmonary 14 7 disease Diabetes mellitus Leg paralysis Peripheral vascular disease 7 10 Congestive heart failure 4 8 Liver disease 4 2 The frequency of bleeding complications during or immediately after initial treatment is shown in Table 2. Type of bleeding and predisposing disorders are shown in Table 3. Major bleeding occurred in 1 patient (1.0%) receiving low-molecular-weight heparin and in 2 (1.9%) receiving intravenous heparin (neither patient had an APTT in the supratherapeutic range) (Table 2). Minor bleeding occurred in 1 patient (1.0%) receiving low-molecular-weight heparin and in 3 (2.9%) receiving intravenous heparin (the APTT was in the supratherapeutic range in 2 of the 3 patients) (Table 2). Type of bleeding and predisposing disorders for bleeding complications that occurred during long-term warfarin therapy and remote from initial therapy are shown in Table 4. Major bleeding remote from the time of initial therapy occurred during long-term warfarin therapy in 3 patients (3.1%) receiving low-molecularweight heparin and in none receiving intravenous heparin. The bleeding was a muscle hematoma on day 56 and hematemesis on days 24 and 51. The international Proportion of Patients Table 2. Outcome Events and Combined Outcome Events in the 2 Treatment Groups Patients, No. (%) Intravenous Heparin (n = 103) normalized ratio was more than 3.0 at or before bleeding in 2 of the 3 patients. Minor bleeding remote from the time of initial therapy occurred during long-term warfarin therapy in 3 patients (3.1%) receiving low-molecular-weight heparin (hemoptysis, vaginal bleeding, and hematuria on days 23, 38, and 44, respectively) and in 3 (2.9%) receiving intravenous heparin (hematochezia, epistaxis, and hematochezia on days 41, 49, and 67, respectively). The international normalized ratio was more than 3.0 at or before bleeding in 1 of 3 patients receiving low-molecular-weight heparin and in 1 of 3 receiving intravenous heparin. DEATH Three patients (3.1%) taking low-molecular-weight heparin and 1 (1.0%) taking intravenous heparin had thrombocytopenia. Low- Molecular- Weight Heparin (n = 97) The proportions of patients who died are shown in Table 2. The causes and timing of death are shown in Table 5. THROMBOCYTOPENIA Difference, H L (95% CI), %* Recurrent venous 7 (6.8) (1.94 to 11.70) thromboembolism Bleeding Major 2 (1.9) 1 (1.0) 0.91 ( 2.43 to 4.25) Minor 3 (2.9) 1 (1.0) 1.88 ( 1.94 to 5.70) Death 9 (8.7) 6 (6.2) 2.55 ( 4.71 to 9.81) Death/recurrent venous thromboembolism 13 (12.6) 6 (6.2) 6.44 ( 1.57 to 14.40) *H L indicates intravenous heparin low molecular-weight heparin. P =.01 by Fisher exact test and P =.009 by log-rank analysis. Bleeding complications occurring during or immediately after initial therapy. Intravenous Heparin Low-Molecular-Weight Heparin Days Figure 1. Objectively documented recurrent venous thromboembolism occurred in none of 97 patients receiving low-molecular-weight heparin and 7 of 103 patients receiving intravenous heparin (6.8%) ( P =.009). 232

5 ANALYSIS OF APTTs AND Xa LEVELS The APTTs are shown in Figure 2, and the chromogenic Xa assay results are shown in Figure 3 for all patients during initial therapy. COMMENT Table 3. Bleeding Complications During or Immediately After Initial Treatment in the 2 Groups Site of Bleeding Major bleeding Hematuria, hematemesis Intracranial Minor bleeding, epistaxis Predisposing Disorder Intravenous Heparin Group Patients, No. (n = 7) Days After Start of Therapy Carcinoma of the bladder 1 2 Metastatic 1 7 carcinoma None 3 6, 3, 4 Low-Molecular-Weight Heparin Group Major bleeding, Gastric ulcer 1 3 hematemesis Minor bleeding, hematuria None 1 4 Our findings demonstrate that low-molecular-weight heparin administered once daily subcutaneously was no less effective and probably more effective than doseadjusted intravenous unfractionated heparin treatment for preventing recurrent VTE in patients with PE and associated proximal DVT. Low-molecular-weight heparin therapy has the advantage of using a fixed dose of weightbased antithrombotic therapy (thus avoiding the potential pitfalls of anticoagulant monitoring and dose adjustment inherent with unfractionated heparin). Hemorrhagic complications were infrequent in both groups. Our study evaluated patients with PE documented by high-probability lung scan findings. These emboli were not minor, since by definition, they resulted in perfusion deficits of 75% or more of a lung segment. There was an important risk (6.8%) of recurrent VTE among patients receiving standard treatment with intravenous heparin. Thus, although most patients had symptoms of venous thrombosis rather than PE at study entry, the outcomes on follow-up indicate that our study population was composed of patients with clinically important venous thromboembolic disease consisting of PE and underlying proximal DVT. Diagnosis of PE at study entry was based on a high-probability interpretation of a perfusion lung scan and a regionally normal finding on a chest radiograph. 55 The report by the American College of Chest Physicians Consensus Committee on Pulmonary Embolism 58 identifies that if the perfusion scan is interpreted as high probability for PE and the chest radiograph findings are regionally normal, a ventilation scan is unnecessary. In the Prospective Investigation of Pulmonary Embolism Diagnosis, using pulmonary angiography as the reference standard, a high-probability perfusion lung scan pattern combined with a chest radiograph had a high positive predictive value for acute PE that was no less predictive than a high-probability, combined ventilation-perfusion lung scan finding. 53,55 As such, we used valid criteria for establishing the presence of PE at study entry. Table 4. Long-term Bleeding Complications After Initial Treatment in the 2 Groups Days After Site of Bleeding Predisposing Disorder Patients, No. Start of Therapy Intravenous Heparin Group Minor bleeding Rectal Colon cancer with 1 41 secondary metastasis Back of throat Hemorrhoids 1 67 and rectal Epistaxis None 1 49 Low-Molecular-Weight Heparin Group Major bleeding Hematemesis Gastric ulcer 1 24 Gastrointestinal Duodenal ulcer 1 51 tract Hamstring Quadriplegic 1 56 hematoma Minor bleeding Vaginal Postexcisional vaginal 1 38 polyp Hemoptysis Endobronchial 1 23 carcinomatosis Hematuria Suprapubic catheter 1 44 Table 5. Causes and Timing of Death in the 2 Treatment Groups* Intravenous Heparin Group Low Molecular-Weight Heparin Group Cause of Death Patients Who Died, No. Days After Start of Therapy Patients Who Died, No. Days After Start of Therapy Pulmonary embolism (abrupt) NA Metastatic carcinoma (abrupt) 5 10, 12, 16, 75, Metastatic carcinoma (insidious) 3 61, 71, , 32, 58, 77 Amyotrophic lateral sclerosis (insidious) 0 NA 1 2 *Patients were categorized according to whether they died abruptly or insidiously. Patients who died insidiously had progressive declines in their health, and their immediate death was anticipated, whereas most patients who died abruptly did so without being anticipated. One patient receiving low-molecular-weight heparin died abruptly (1.0%) compared with 6 patients receiving dose-adjusted unfractionated heparin (5.8%) (95% confidence limits for the difference, 0.156% to 9.74%; P =.06 by uncorrected 2 ). NA indicates not applicable. 233

6 APTT, s 100 Xa Units/mL Test Figure 2. Activated partial thromboplastin times (APTTs) for the study patients during initial treatment according to treatment group and test sequence. Black boxes indicate patients receiving low-molecular-weight heparin; gray boxes, patients receiving intravenous heparin; top and bottom of each box, upper and lower quantities, respectively, of the values for the sample; circles, medians; and bars above and below each box, maximal and minimal values, respectively, in the sample or, if there are extreme data points, to limits based on the interquartile range, defined as the distance from the lower quartile to the upper quartile. Outliers beyond these limits are plotted separately. The first 3 tests reflect the results of the 4 to 6 daily tests obtained on the first day; thereafter the tests were performed daily Test Figure 3. Chromogenic Xa assay findings for the study patients during initial treatment according to treatment group and test sequence. Black boxes indicate patients receiving low-molecular-weight heparin; gray boxes, patients receiving intravenous heparin; top and bottom of each box, upper and lower quantities, respectively, of the values for the sample; circles, medians; and bars above and below each box, maximal and minimal values, respectively, in the sample or, if there are extreme data points, to limits based on the interquartile range, defined as the distance from the lower quartile to the upper quartile. Outliers beyond these limits are plotted separately. The first 3 tests reflect results of the 4 to 6 daily tests obtained on the first day; thereafter the tests were performed daily. The scintigraphic diagnosis of acute PE is compromised in patients with a past history of PE or DVT; the positive predictive value of high-probability lung perfusion scan patterns for PE is considerably less in these patients. 53 Most of our patients (82.5%) did not have the confounding history of previous DVT or PE. None of the patients who had recurrent VTE also had a previous history of DVT or PE. Care was taken throughout the study to ensure that adequate doses of intravenous heparin were administered. The standardized protocol used has been shown to achieve therapeutic levels in 90% or more of patients during the first 24 hours and maintains them thereafter. 30 Thus, our results cannot be attributed to inadequate initial therapy with intravenous heparin. The study was a multicenter, double-blind clinical trial. To avoid a selection bias, care was taken to ensure that participating physicians adhered to the protocol. Before the study, the criteria for eligibility were specified; 51% of eligible patients were randomized. Baseline perfusion lung scanning was mandated a priori and obtained in 97.0% of patients. Well-validated criteria were used to determine the presence of both PE on study entry and of recurrent VTE and hemorrhagic complications. The clinical characteristics for each patient group were similar; thus, our findings cannot be attributed to bias caused by underlying patient variables. Our analysis adhered to published methodological criteria 59 for a valid subgroup analysis. Our findings support and extend those of Simonneau et al (Tinzaparine ou Heparine Standard: Evaluations dans l Embolie Pulmonaire Study Report [THÉSÉE]), 60 who observed that tinzaparin treatment was as effective and safe as intravenous unfractionated heparin treatment in patients with acute PE. These authors reported an unexpectedly low event rate in the overall study population, which markedly reduced the statistical power of their study to detect a significant difference between treatment groups. 60 They concluded that a much larger study would be necessary to show a statistically significant difference. Our study differs with regard to important variables documented by the THÉSÉE Study Group 60 : all our patients had associated proximal DVT compared with approximately 50% in the THÉSÉE Study report 60 ; predisposing factors, including previous surgery, trauma, or cancer, were present in a higher proportion of our patients; the event rates in the control group (intravenous unfractionated heparin) for recurrent VTE and death were considerably higher in our study; and our study was doubleblind, protecting against diagnostic suspicion bias. Furthermore, the THÉSÉE investigators treated most patients with therapeutic doses of unfractionated heparin before initiating low-molecular-weight heparin treatment; in our study, patients who were randomly assigned to receive low-molecular-weight heparin did not receive unfractionated heparin. Accordingly, our study had greater power to detect a difference between treatment groups because of the greater burden of illness of the study population, resulting in a higher rate of recurrent thromboembolic events in the unfractionated heparin comparison group. Our findings are strengthened by the findings that patients in either study who received low-molecular-weight heparin (tinzaparin sodium) had a similar low frequency of recurrent VTE. The therapeutic role of low-molecular-weight heparin in patients with massive PE who are hemodynamically unstable remains to be determined. Most patients in our study presented with symptomatic proximal DVT, and were found on entry by objective testing to have PE; symptoms and signs of PE were identified in only 15% of these patients. For this reason, our findings although applicable to patients with PE who have clinical characteristics similar to those in our study should not be generalized to patients with massive embolism who are hemodynamically unstable. 234

7 Accepted for publication April 6, From the Divisions of General Internal Medicine and Hematology, Thrombosis Research Unit, Department of Community Health Sciences, University of Calgary, Calgary, Alberta (Drs Hull, Brant, Pineo, Valentine, and Mah); Departments of Biostatics and Epidemiology and Medicine, University of Oklahoma Health Sciences Center, Oklahoma City (Dr Raskob); LDS Hospital, University of Utah, Salt Lake City (Dr Elliott); Cardiac Wellness Center, Henry Ford Hospital, Detroit, Mich (Dr Stein); and Department of Radiology, Michigan State University, East Lansing (Dr Gottschalk). This study was supported in part by a grant from the Heart and Stroke Foundation of Alberta, Edmonton, and by Novo Nordisk, Bagsvaerd, Denmark. We thank the medical, surgical, emergency, nursing, pharmacy, and support staff of all the sites participating in the study, and Jennifer White, Jeanne Sheldon, Margot McDonald, and Victoria Stagg. The following persons also participated in the American- Canadian Thrombosis Study: University of Calgary, Peter Lougheed Centre site, Calgary, Alberta: W. Blahey, J. F. T. Thaell, R. Dear, T. Godinez, P. Hardin, D. Linden, D. McKeage, and A. Wilson; Foothills Hospital site, Calgary, Alberta: B. Baylis, N. Campbell, A. Ferland, C. B. Hatfield, C. D. Thompson, L. Styner, and R. Mulcair; Northwestern University, Rehabilitation Institute of Chicago, Chicago, Ill: H. Kohl, N. Reynolds, and J. Deutsche; Texas A&MUniversity, Scott and White Clinic, Temple: E. J. Schoolar and B. Woodruff; University of Utah, LDS Hospital, Salt Lake City: M. Suchyta, S. Yeates, M. Boyer, and N. Kitterman; New York Medical College, Westchester County Medical Center, Valhalla: J. Nelson, E. Scorcia, and P. Pugni; Methodist Hospital of Indiana, Indianapolis: S. Gamance, R. Hahn, S. Rolfe, K. Berron, K. Colvin, and S. Kerner-Slemons; University of British Columbia, Vancouver: S. Krinkler and S. Fowler; University of Colorado, University Hospital, Denver: B. Whitcomb, L. A. Barbour, D. Tanaka, R. G. Badget, and M. LoVerde; Penrose Hospital, Colorado Springs, Colo: J. Hillman and M. Dieterich; Mercy Hospital and Medical Center, San Diego, Calif: M. Sullivan and K. Engstrand; Jefferson Medical College, Philadelphia, Pa: L. Doyle and C. Pilgrim; Lions Gate Hospital, Vancouver: R. Cohen, M. Dart, and S. Regehr; Clinical Trials Group, University of Calgary: B. Doucette and L. Styner; Safety monitor: D. Bergqvist; Liaison with Novo Nordisk: A. Brusby, N. Griffin, K. Birch, S. Glazer, and U. Hedner. Reprints: Russell D. Hull, MBBS, MSc, Thrombosis Research Unit, 601 S Tower, Foothills Hospital, St NW, Calgary, AB Canada T2N 2T9 ( jeanne.sheldon@crha-health.ab.ca). REFERENCES 1. Anderson FA Jr, Wheeler B, Goldberg RJ, et al. A population-based perspective of the hospital incidence and case-fatality rates of deep-vein thrombosis and pulmonary embolism: the Worcester Deep-Vein Thrombosis Study. Arch Intern Med. 1991;151: Silverstein MD, Heit JA, Mohr DN, Petterson TM, O Fallon WM, Melton LJ III. Trends in the incidence of deep vein thrombosis and pulmonary embolism: a 25-year population-based study. Arch Intern Med. 1998;158: Moser KM, LeMoine JR. Is embolic risk conditioned by location of deep venous thrombosis? Ann Intern Med. 1981;94: Huisman MV, Buller HR, ten Cate JW, et al. Unexpected high prevalence of silent pulmonary embolism in patients with deep venous thrombosis. Chest. 1989;95: Dalen JE, Alpert TS. Natural history of pulmonary embolism. Prog Cardiovasc Dis. 1975;17: Hull RD, Hirsh J, Carter CJ, et al. Diagnostic efficacy of impedance plethysmography for clinically suspected deep-vein thrombosis: a randomized trial. Ann Intern Med. 1985;102: Huisman MV, Buller HR, ten Gate JW, Vreeken J. Serial impedance plethysmography for suspected deep venous thrombosis in outpatients: the Amsterdam General Practitioner Study. N Engl J Med. 1986;314: Hull RD, Hirsh J, Sackett DL, et al. Clinical validity of a negative venogram in patients with clinically suspected venous thrombosis. Circulation. 1981;64: Dismuke SE, Wagner EH. Pulmonary embolism as a cause of death: the changing mortality in hospitalized patients. JAMA. 1986;255: Stein PD, Henry JW. Prevalence of acute pulmonary embolism among patients in a general hospital and at autopsy. Chest. 1995;108: Rabinov K, Paulin S. Roentgen diagnosis of venous thrombosis in the leg. Arch Surg. 1972;104: Wheeler HB, O Donnell JA, Anderson FA Jr, Benedict K Jr. Occlusive impedance phlebography: a diagnostic procedure for venous thrombosis and pulmonary embolism. Prog Cardiovasc Dis. 1974;17: Hull RD, Hirsh J, Sackett DL, et al. Combined use of leg scanning and impedance plethysmography in suspected venous thrombosis: an alternative to venography. N Engl J Med. 1977;296: Hull RD, Carter CJ, Jay RM, et al. The diagnosis of acute, recurrent, deep-vein thrombosis: a diagnostic challenge. Circulation. 1983;67: Dalen JE, Brooks HL, Johnson LW, et al. Pulmonary angiography in acute pulmonary embolism: indications, techniques and results in 367 patients. Am Heart J. 1971;81: White RH, McGahan JP, Daschback MM, Hartling RP. Diagnosis of deep-vein thrombosis using duplex ultrasound. Ann Intern Med. 1989;111: Lensing AWA, Prandoni P, Brandjes D, et al. Detection of deep-vein thrombosis by real-time B-mode ultrasonography. N Engl J Med. 1989;320: Hull RD, Hirsh J, Carter C, et al. Diagnostic value of ventilation-perfusion lung scanning in patients with suspected pulmonary embolism. Chest. 1985;88: Salzman EW, Deykin D, Shapiro RM, Rosenberg R. Management of heparin therapy: controlled prospective trial. N Engl J Med. 1975;292: Glazier RL, Crowell EB. Randomized prospective trial of continuous vs intermittent heparin therapy. JAMA. 1976;236: Mant MJ, O Brien ED, Thong KL, et al. Haemorrhagic complications of heparin therapy. Lancet. 1977;1: Hull RD, Raskob GE, Hirsh J, et al. Continuous intravenous heparin compared with intermittent subcutaneous heparin in the initial treatment of proximal-vein thrombosis. N Engl J Med. 1986;315: Hull RD, Delmore T, Carter C, et al. Adjusted subcutaneous heparin versus warfarin sodium in the long-term treatment of venous thrombosis. N Engl J Med. 1982;306: Hull RD, Hirsh J, Jay R, et al. Different intensities of oral anticoagulant therapy in the treatment of proximal-vein thrombosis. N Engl J Med. 1982;307: Gallus AS, Jackaman J, Tillett J, et al. Safety and efficacy of warfarin started early after submassive venous thrombosis or pulmonary embolism. Lancet. 1986;2: Hull RD, Raskob GE, Rosenbloom D, et al. Heparin for 5 days as compared with 10 days in the initial treatment of proximal venous thrombosis. N Engl J Med. 1990;322: Hull RD, Delmore T, Genton E, et al. Warfarin sodium versus low-dose heparin in the long-term treatment of venous thrombosis. N Engl J Med. 1979;301: Brandjes DPM, Heijboer H, Buller HR, et al. Acenocoumarol and heparin compared with acenocoumarol alone in the initial treatment of proximal-vein thrombosis. N Engl J Med. 1992;327: Raschke RA, Reilly BM, Guidry JR, et al. The weight-based heparin dosing nomogram compared with a standard care nomogram. Ann Intern Med. 1993; 119: Hull RD, Raskob GE, Rosenbloom D, et al. Optimal therapeutic level of heparin therapy in patients with venous thrombosis. Arch Intern Med. 1992;152: Hull RD, Raskob GE, Brant RF, et al. The importance of initial heparin treatment on long-term clinical outcomes of antithrombotic therapy: the emerging theme of delayed recurrence. Arch Intern Med. 1997;157: Hull RD, Raskob GE, Brant RF, et al. Relation between the time to achieve the 235

8 lower limit of the APTT therapeutic range and recurrent venous thromboembolism during heparin treatment for deep vein thrombosis. Arch Intern Med. 1997; 157: Hyers TM, Hull RD, Weg JG. Antithrombotic therapy for venous thromboembolic disease. Chest. 1998;114(suppl):561S-578S. 34. Salzman EW. Low-molecular-weight heparin: is small beautiful? N Engl J Med. 1986;315: Verstraete M. Pharmacotherapeutic aspects of unfractionated and low-molecularweight heparin. Drugs. 1990;40: Weitz JI. Low-molecular-weight heparins. N Engl J Med. 1997;337: Bergqvist D, Hedner U, Sjorin E, Holmer E. Anticoagulant effects of two types of low-molecular-weight heparin administered subcutaneously. Thromb Res. 1983; 32: Bara L, Billaud E, Gramond G, et al. Comparative pharmacokinetics of a lowmolecular-weight heparin (PK ) and unfractionated heparin after intravenous and subcutaneous administration. Thromb Res. 1985;39: Bratt G, Tornebohm E, Widlund L, Lockner D. Low-molecular-weight heparin (KABI 2165, Fragmin): pharmacokinetics after intravenous and subcutaneous administration in human volunteers. Thromb Res. 1986;42: Harenberg J, Wurzner B, Zimmermann R, Schettler G. Bioavailability and antagonization of the low-molecular-weight heparin CY 216 in man. Thromb Res. 1986;44: Frydman AM, Bara L, Le Roux Y, et al. The antithrombotic activity and pharmacokinetics of enoxaparin, a low-molecular-weight heparin, in humans given single subcutaneous doses of 20 to 80 mg. J Clin Pharmacol. 1988;28: Matzsch T, Bergqvist D, Hedner U, Ostergaard P. Effects of an enzymatically depolymerized heparin as compared with conventional heparin in healthy volunteers. Thromb Haemost. 1987;57: Aiach M, Michaud A, Ballan J-L, et al. A new low-molecular-weight heparin derivative: in vitro and in vivo studies. Thromb Res. 1983;31: Siegbahn A, Y-Hassan S, Boberg J, et al. Subcutaneous treatment of deep venous thrombosis with low molecular weight heparin: a dose finding study with LMWH-Novo. Thromb Res. 1989;55: Hull RD, Raskob GE, Pineo GF, et al. Subcutaneous low-molecular-weight heparin compared with continuous intravenous heparin in the treatment of proximalvein thrombosis. N Engl J Med. 1992;326: Prandoni P, Lensing AWA, Buller HR, et al. Comparison of subcutaneous lowmolecular-weight heparin with intravenous standard heparin in proximal deepvein thrombosis. Lancet. 1992;339: Lopaciuk S, Meissner AJ, Filipecki S, et al. Subcutaneous low-molecularweight heparin versus subcutaneous unfractionated heparin in the treatment of deep-vein thrombosis: a Polish multicentre trial. Thromb Haemost. 1992; 68: Simonneau G, Charbonnier B, Decousus H, et al. Subcutaneous LMWH compared with continuous intravenous unfractionated heparin in the treatment of proximal deep-vein thrombosis. Arch Intern Med. 1993;153: Lindmarker P, Holmstrom M, Granqvist S, et al. Comparison of once-daily subcutaneous Fragmin with continuous intravenous unfractionated heparin in the treatment of deep-vein thrombosis. Thromb Haemost. 1994;72: Levine M, Gent M, Hirsh J, et al. A comparison of LMWH administered primarily at home with unfractionated heparin administered in the hospital for proximal deep-vein thrombosis. N Engl J Med. 1996;334: Koopman MMW, Prandoni P, Piovella F, et al. Treatment of venous thrombosis with intravenous unfractionated heparin administered in the hospital as compared with subcutaneous LMWH administered at home. N Engl J Med. 1996; 334: Columbus Investigators. Low-molecular-weight heparin is an effective and safe treatment of deep-vein thrombosis and pulmonary embolism. Blood. 1996;88: PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA. 1990;263: PISA-PED Investigators. Value of perfusion lung scan in the diagnosis of pulmonary embolism: results of the Prospective Investigative Study of Acute Pulmonary Embolism Diagnosis. Am J Respir Crit Care Med. 1996;154: Stein PD, Terrin ML, Gottschalk A, Alavi A, Henry JW. Value of ventilation/ perfusion scans versus perfusion scans alone in acute pulmonary embolism. Am J Cardiol. 1992;69: Bookstein JJ, Silver TM. The angiographic differential diagnosis of acute pulmonary embolism. Radiology. 1974;110: Williamson DF, Parker RA, Kendrick JS. The box plot: a simple visual method to interpret data. Ann Intern Med. 1989;110: ACCP Consensus Committee on Pulmonary Embolism. Second report: opinions regarding the diagnosis and management of venous thromboembolic disease. Chest. 1998;113: Oxman AD, Guyatt GH. A consumer s guide to subgroup analyses. Ann Intern Med. 1992;116: Simonneau G, Sors H, Charbonnier B, et al. A comparison of low-molecularweight heparin with unfractionated heparin for acute pulmonary embolism. N Engl J Med. 1997;337: