Self-Managed Long-Term Low-Molecular-Weight Heparin Therapy: The Balance of Benefits and Harms

doi:10.1016/j.amjmed.2006.03.030M
M 1
0 d
LINICAL RESEARCH STUDY
elf-Managed Long-Term Low-Molecular-Weight Heparin herapy: The Balance of Benefits and Harms
ussell D. Hull, MBBS, MSc,a Graham F. Pineo, MD,a Rollin F. Brant, PhD,b Andrew F. Mah, BSc,a
atasha Burke, BSc,a Richard Dear, MD,a Turnly Wong, MD,c Roy Cook, MD,a Susan Solymoss, MD,d
an-Chiu Poon, MD, MSc,a Gary Raskob, PhD,e for the LITE Trial Investigators University of Calgary, Calgary, AB, Canada; bUniversity of British Columbia, Vancouver, BC, Canada; cUniversity of Manitoba, Winnipeg,
AJM Theme Issue: Cardiology
N, Canada; dM e kla.
P v a e s M m R 3 i o a e f C a c t s
E-mail address
cGill University, Montreal, PQ, Canada; University of Oklahoma Health Sciences Center, Oklahoma City, O
ABSTRACT
URPOSE: A substantial clinical need exists for an alternate to vitamin K antagonists for treating deep ein thrombosis in many patients. Long-term low-molecular-weight heparin (LMWH), body-weight djusted, avoids anticoagulant monitoring and may be associated with less bleeding. We evaluated the ffectiveness and safety of long-term LMWH compared with vitamin K antagonist therapy in a broad pectrum of patients with proximal vein thrombosis. ETHODS: We performed a multicenter, randomized, open-label clinical trial using objective outcome easures comparing therapy for 3 months. Outcomes were assessed at 3 and 12 months. ESULTS: Of 737 patients, 18 of 369 receiving tinzaparin (4.9%) had recurrent venous thromboembolism at months compared with 21 of 368 (5.7%) receiving usual care (absolute difference, 0.8%, 95% confidence
nterval 4.1-2.4). Hemorrhagic complications occurred less frequently in the LMWH group largely because f less minor bleeding: 48 of 369 patients (13.0%) versus 73 of 368 patients (19.8%) receiving usual-care nticoagulation (absolute difference 6.8%; P .011; risk ratio 0.66). New major bleeding events ceased arly (by day 23, P .034) for patients receiving LMWH but persisted throughout the study treatment interval or patients receiving vitamin K antagonist therapy. No mortality advantage was shown for LMWH. ONCLUSION: Our study shows that LMWH is similar in effectiveness to the usual-care vitamin K ntagonist treatment for preventing recurrent venous thromboembolism in a broad spectrum of patients. It auses less harm and enhances the clinicians’ therapeutic options for patients with proximal deep vein hrombosis. Our findings reported here suggest the possibility of a broader role for long-term LMWH in elected patients. © 2007 Elsevier Inc. All rights reserved.
KEYWORDS: Long-term low-molecular-weight heparin; Broad spectrum of patients; Usual-care anticoagulation; Vitamin K antagonist therapy
T v o u b t w s i r a
The study was supported by a Medical Research Council (now Canadian nstitutes for Health Research) and Industry grant (Leo Pharmaceutical Prod- cts Ltd A/S of Denmark). Additional funding was provided by Pharmion and upont Pharmaceuticals. Leo provided study drug and drug safety monitoring. he funding organization(s) and sponsor(s) did not have a role in the design nd conduct of the study; collection, management, analysis, and interpretation f the data; and preparation or approval of the article.
Study design table and participating sites appendix available online. Requests for reprints should be addressed to Russell D. Hull, MBBS,
Sc, Thrombosis Research Unit, 601 South Tower, Foothills Hospital, 403 29th Street NW, Calgary, Alberta, Canada T2N 2T9.
t: [email protected].
ront matter © 2007 Elsevier Inc. All rights reserved. ed.2006.03.030
he classic long-term treatment for deep vein thrombosis is itamin K antagonist therapy overlapped with initial heparin r low-molecular-weight heparin (LMWH) therapy.1 The se of accurate objective tests to detect venous thromboem- olism has led to randomized trials evaluating short-term herapy or long-term anticoagulants for venous thrombosis, hich have advanced our therapeutic understanding. Initial
hort-term LMWH therapy is effective and preferred over ntravenous heparin because anticoagulant monitoring is not equired, facilitating outpatient therapy.2,3 For vitamin K ntagonist therapy,1,4-6 the importance of maintaining a
herapeutic international normalized ratio (INR) (2.0-3.0)
i m
t w L s i w n o g t t l b q t i a t f t t r L a l w o t d v e l
l t L s i t
M
c d a
t t a h o o B b T s
t w o c b r t a
S A i e v
73Hull et al Long-term LMWH Therapy
s well documented; this necessitates frequent INR onitoring. There is a need for an effective and safe alternate therapy
o vitamin K antagonists in a broad spectrum of patients ith venous thrombosis. It is possible that long-term MWH, although administered ubcutaneously, is a viable option n such patients. A key concern ith long-term vitamin K antago- ists is harm resulting from hem- rrhagic complications.7,8 The ag- regate data,7,9-15 as reported by he “Cochrane Database of Sys- ematic Reviews,” identify that ong-term LMWH may cause less leeding, but further study is re- uired in the individual trial set- ing to determine whether safety is ndeed improved. The Cochrane uthors also report uncertainty on he efficacy of long-term LMWH or preventing recurrent venous hromboembolism, but the authors note that the prophylac- ic LMWH doses used in some of these trials may have esulted in inadequate therapy. Subsequently, long-term MWH was shown to be more effective than vitamin K ntagonists for preventing recurrent venous thromboembo- ism in patients with cancer and venous thromboembolism16
ithout increased bleeding. Whether this benefit applies to ther patient groups with deep vein thrombosis is uncertain; here are insufficient data in patients without cancer with eep vein thrombosis.7 Indeed, the authors of a recent re- iew article17 conclude that further clinical trials are nec- ssary, because LMWH may prove to be an appropriate ong-term therapy.
We conducted a large multicenter, randomized, open- abel clinical trial in a broad spectrum of patients to evaluate he effectiveness (benefits) and harm (safety) of long-term MWH, using a therapeutic dose of tinzaparin once daily ubcutaneously, compared with usual care consisting of ntravenous heparin and long-term vitamin K antagonist herapy.
ETHODS
tudy Design he study design, patient eligibility and allocation, and
reatment regimens are shown in the table available nline.1-5,18-25
urveillance and Follow-up atients were instructed to seek care immediately if they ad symptoms or signs of venous thromboembolism or leeding. Patients presenting with clinically suspected re- urrent venous thromboembolism underwent objective test- ng. Patients routinely attended the clinic at 12 weeks. At 1
CLINICAL SIGNIF
During 3 month venous thrombos tion of low-mo was equivalent t onism in preven thrombosis.
The 2 therapies spect to mortali test period.
ear, all patients or their primary care physicians were (
ontacted, determining whether the patient had experienced ocumented recurrent venous thromboembolism and was live.
Primary outcome measures were assessed at 3 months nd included objectively documented recurrent venous
thromboembolism or death. Pa- tients were then followed by tele- phone at 1 year and assessed for objectively documented venous thromboembolism or death. Re- current venous thrombosis was di- agnosed when a previously com- pressible proximal vein segment was not compressible on ultra- sonography2,3,26,27 or by the pres- ence of a constant intraluminal filling defect in the deep veins that was not present on the baseline venogram.28,29 For patients with clinically suspected pulmonary embolism, the diagnosis was con- firmed by high-probability lung
can findings;30,31 a nondiagnostic lung scan with docu- ented new deep vein thrombosis;30 spiral computed to- ography32 showing thrombus in the central pulmonary
rteries; pulmonary angiography30,33 revealing a constant ntraluminal filling defect or cutoff of a vessel greater than .5 mm in diameter; or by pulmonary embolism found at utopsy.
The primary safety end point for assessing harm was he occurrence of bleeding (all, major, or minor) during he 12-week treatment interval. Bleeding was classified s major if it was overt and associated with a decrease in emoglobin of 2 g/dL or more, if it led to the transfusion f 2 or more units of blood; and if it was retroperitoneal, ccurred into a major joint, or was intracranial.4,34-37
leeding was defined as minor if it was clinically overt ut did not meet other criteria for major bleeding.4,34-37
hese criteria were used successfully in previous tudies.4,34-37
All suspected events, including recurrent deep vein hrombosis, pulmonary embolism, bleeding, or death, ere interpreted independently without knowledge of the ther findings by a central, independent adjudication ommittee. Adjudication was made by 2 committee mem- ers not involved in the patient’s care, and disputes were esolved independently by a third member. Members of he committee were unaware of the patients’ treatment ssignments.
tatistical Analysis sample size of 455 patients in each treatment group was
nitially chosen to provide 80% power for a 2-sided test ( .05) to detect a 50% reduction in mortality from the 9.6%
xperienced in a previous trial suggesting a mortality ad- antage.35 The findings of a companion randomized trial
CE
l vitamin K antag- recurrent venous
ot differ with re- ring the 3-month
ICAN
did n ty du
o b t r 7 ( ( p
v g t t c i h a f d M p C
i t b s t d D C
R
S A i o c p t g r s p
R T s
u s (
74 The American Journal of Medicine, Vol 120, No 1, January 2007
utcome suggested rebound recurrent venous thromboem- olism immediately after cessation of long-term LMWH herapy.18,38,39 Because of both safety and regulatory affair equirements, the trial reported here was terminated when 37 patients were enrolled. Because our confidence interval CI) for relative mortality risk excludes a 50% reduction favoring either group), we conclude that the achieved sam- le size is sufficient for our primary aims.
The comparison of the frequency of events (recurrent enous thromboembolism, bleeding, death) between the 2 roups in the first 12 weeks was based on the chi-square est and associated CIs. Cumulative incidence estimates o 1 year were derived from Kaplan-Meier survival urves and compared using the log-rank test after assess- ng the proportional hazard assumption.40 Proportional azards test diagnostics were based on weighted residu- ls.40 CIs were derived from standard errors calculated rom Greenwood’s formula. The meta-analysis was con- ucted following a fixed-effects approach based on the antel-Haenszel method as implemented in the metan
rocedure41 of Stata software, release 6.0 (Stata Corp, ollege Station, Tex).
The protocol was designed by 3 investigators. The steer- ng committee, central adjudication committee, and statis- ical analysis were independent of the sponsor. The Throm- osis Research Unit, University of Calgary, coordinated the tudy and carried out the data management and administra- ive duties. Statistical analysis was carried out indepen- ently of the industry sponsor by Rollin F. Brant, PhD, epartment of Community Health Sciences, University of algary.
369 Assigned to receive low- molecular-weight heparin
1 Lost to Follow-up at 3 months (+ 2 at 12 months)
1 withdrew consent at 3 months
369 Included in Analysis
2212 assessed for elig
tudy Population total of 737 consecutive patients were recruited beginning
n 1994 at 30 centers across Canada (see Appendix available nline) and randomized to tinzaparin (369 patients) or usual- are anticoagulants (368 patients). A broad spectrum of atients were evaluated. Figure 1 shows the patient flow for he eligible patients and those randomized to the treatment roups. The tinzaparin and usual-care groups were compa- able at entry (Table 1). Because of difficulties with occa- ional patient follow-up, which were overcome, successful atient follow-up was completed in July 2003.
ecurrent Venous Thromboembolism he outcomes for recurrent venous thromboembolism are hown in Figure 2 and Table 2.
At 3 months, of 737 patients, 18 of 369 patients eceiving tinzaparin (4.9%) had recurrent venous throm- oembolism compared with 21 of 368 patients (5.7%) eceiving usual-care anticoagulation (absolute difference, 0.8%, 95% CI, 4.1-2.4). Thus, LMWH was unlikely
o be more effective against recurrent venous thrombo- mbolism than usual-care anticoagulation by more than .1% (absolute risk difference), and usual-care anticoag- lation was unlikely to be more effective than LMWH by ore than 2.4%. The activated partial thromboplastin time and INR val-
es for patients receiving usual-care anticoagulation are hown according to test sequence in the Appendix figure available online).
patients were excluded: 575 received heparin, low-molecular- weight heparin or oral anticoagulant therapy for more than 2 days 900 were unable or declined to give written consent
368 Assigned to receive usual- care with intravenous heparin and warfarin
1 Lost to Follow-up at 3 months (+2 at 12 months)
3 Withdrew consent at 3 months (+2 at 12 months)
368 Included in Analysis
75Hull et al Long-term LMWH Therapy
igure 2 Time to event analysis for patients who had recurrent venous thromboembolism At 3 months, of 737 patients, 18 of 369 patients eceiving tinzaparin (4.9%) had recurrent venous thromboembolism compared with 21 of 368 (5.7%) receiving usual care (absolute
Table 1 Clinical Characteristics of Patients with Proximal Vein Thrombosis Treated with Long-Term Low-Molecular-Weight Heparin or Oral Anticoagulant Therapy
Characteristic
No. of Patients (%)
Age 60 y, 60 y 187, 182 151, 217 Sex (M, F) 207, 162 188, 180 Status at entry
Symptomatic deep vein thrombosis 350 (94.9) 346 (94.0) Symptoms of pulmonary embolism 98 (26.6) 82 (22.3)
Previous venous thromboembolism 64 (17.3) 74 (20.1) Factor V Leiden gene mutation 56 (15.2) 51 (13.9) High risk for bleeding* 144 (39.0) 146 (39.7) Clinical measures at entry
Surgery and/or trauma in past 6 mo 162 (43.9) 167 (45.4) Cancer 100 (27.1) 100 (27.2) Coronary heart disease 65 (17.6) 91 (24.7)
Diabetes 41 (11.1) 45 (12.2) Chronic obstructive pulmonary disease 33 (8.9) 29 (7.9) Peripheral vascular disease 28 (7.6) 21 (5.7) Congestive heart disease 14 (3.8) 19 (5.2) Leg paralysis 21 (5.7) 16 (4.4) Liver disease 10 (2.7) 9 (2.4)
*With regard to the risk of bleeding, patients were stratified into groups according to the absence (low risk) or presence (high risk) of 1 or more risk factors for bleeding including surgery or trauma within the previous 14 d, history of peptic ulcer disease, bleeding into the gastrointestinal or genitourinary tract, thrombotic stroke within the previous 14 d, platelet count less than 150 109/L, or miscellaneous reasons (predisposing disorders) for a high risk of bleeding. In the LMWH group, 1 patient withdrew, 1 patient was lost to follow-up at 3 mo, and 1 patient was lost to follow-up at 1 y. In the usual-care group, 3 patients withdrew and 1 patient was lost to follow-up at 3 mo; 2 additional patients withdrew consent and 2 were lost to follow-up at 1 year. Thus, 726 of the 737 patients (98.5%) randomized completed the 3-mo study interval and subsequent follow-up to 12 mo.
ifference, 0.8%, 95% CI, 4.1-2.4). IV intravenous; LMWH low-molecular-weight heparin.
76 The American Journal of Medicine, Vol 120, No 1, January 2007
Table 2 Outcomes at 3 Months
Tinzaparin N 369 n (%) Usual Care* N 368 n (%) Absolute Difference (95% CI) P Value
New episodes of venous thromboembolism†
At 3 mo 18 (4.9) 21 (5.7) 0.8 (4.1-2.4) At 12 mo 33 (8.9) 36 (9.8) 0.8 (5.5-3.5) At 12 mo
Death At 3 mo 25 (6.8) 24 (6.5) 0.3 (3.4-3.9) At 12 mo 60 (16.3) 59 (16.0) 0.2 (5.4-5.4) At 12 mo
Bleeding complications Frequency of bleeding
All 48 (13.0) 73 (19.8) 6.8 (12.4-1.5) P .011 Major 12 (3.3) 17 (4.6) 1.4 (4.3-1.4) Minor 36 (9.8) 56 (15.2) 5.5 (10.4-0.6) P .022
Stratified by risk of bleeding All bleeding
High risk 31/144 (21.5%) 39/146 (26.7%) 5.2 (15%-4.6%) Low risk 17/225 (7.6%)‡ 34/222 (15.3%)§ 7.8 (13.6%-1.9%) P .01
Major bleeding High risk 10/144 (6.9%) 13/146 (8.9%) 2.0 (8.2%-4.3%) Low risk 2/225 (0.9%) 4/222 (1.8%)¶ 0.9 (3.1%-1.2%)
Minor bleeding High risk 21/144 (14.6%) 26/146 (17.8%) 3.2 (11.7%-5.3%) Low risk 15/225 (6.7%)** 30/222 (13.5%) 6.9 (12.4%-1.3%) P .018
Other findings Thrombocytopenia†† Platelet count
100 109/L 10 (2.7%) 4 (1.1%) 1.6 (3.6-0.3) 150 109/L 21 (5.7%) 9 (2.4%) 3.3 (6.1-0.4) P .039
Bone fractures out to 12 mo‡‡ 4 (1.1%) 7 (1.9%) 0.8 (0.9-2.6)
CI confidence interval. *Unfractionated heparin and vitamin K antagonist therapy. †Recurrent venous thromboembolism at 3 mo: 18 patients in the LMWH group 10 with pulmonary embolism (documented by autopsy, n 1; high-probability
lung scan, n 4; pulmonary angiography, n 1; spiral computed tomography [CT], n 3; pulmonary angiography and lung scan, n 1); 8 with recurrent deep vein thrombosis (documented by duplex ultrasonography); 21 patients in the heparin/warfarin group 6 with pulmonary embolism (documented by autopsy, n 1; high-probability lung scan, n 5); 15 with recurrent deep vein thrombosis (documented by venography, n 3; duplex ultrasonography, n 12). At 12 mo: 33 patients in the LMWH group 12 with pulmonary embolism (documented by autopsy, n 2; lung scan, n 5; spiral CT scan, n 3, pulmonary angiography, n 1; pulmonary angiography and high-probability lung scan, n 1); 21 with deep vein thrombosis (documented by venography, n 1; duplex ultrasonography, n 20); 36 patients in the heparin/warfarin group 11 with pulmonary embolism (documented by autopsy, n 4; high-probability lung scan, n 5; spiral CT scan, n 2); 25 with deep vein thrombosis (documented by venography, n 7; duplex ultrasonography, n 18).
Long-term vitamin K antagonist therapy was administered by the primary care physician, if indicated, in patients such as those with recurrent or idiopathic deep vein thrombosis at entry or who had a continuing risk factor. The physician instituted or continued vitamin K antagonist therapy long term in 146 patients assigned to LMWH (mean duration 202 d, median duration 258 d) and in 250 patients assigned to intravenous heparin/warfarin (mean duration 156 d, median duration 147 d). In patients with cancer, 37 were assigned to LMWH (mean duration 215 d, median duration 272 d), and 57 were assigned to intravenous heparin/warfarin (mean duration 165 d, median duration 152 d).
The presence of a predisposing factor for bleeding (high risk for bleeding) on entry overwhelmed any potentially significant bleeding outcome differences between the groups. In patients without a predisposing factor for bleeding on entry (low risk for bleeding), long-term LMWH therapy use was associated with significantly less overall bleeding.
High risk for bleeding stratum compared with low risk: ‡P .001; §P .007; P .001; ¶P .002; **P .012. A preponderance of bleeding from mucosal sites was observed in the usual-care group compared with the LMWH group: nasopharyngeal (epistaxis), 20
patients (5.4%) versus 9 patients (2.4%), respectively; hemoptysis, 8 patients (2.2%) versus 4 patients (1.1%), respectively; and upper and lower digestive tract bleeding, 23 patients (6.3%) versus 13 patients (3.5%), respectively. Thus, hemorrhagic complications involving mucosal sites (nasopharyngeal, upper and lower digestive tract, and bronchial) occurred in 51 of 368 patients (13.9%) receiving vitamin K antagonist therapy compared with 26 of 369 patients (7%) receiving LMWH (absolute difference 6.8%, P .003).
Long-term LMWH use was associated with significantly less nasopharyngeal and upper and lower digestive tract bleeding (absolute difference 5.7%, P .01). On diagnostic pursuit of the causes of bleeding (in the absence of known predisposing causes), a substantive cause for bleeding was frequently not established. Bleeding complications from other diverse sites were rare or similar between the LMWH and usual-care groups: intracranial bleeding, 2 patients and 1 patient,
respectively; genitourinary, 12 and 10 patients, respectively; hematomas, 5 and 8 patients, respectively; and miscellaneous, 3 and 3 patients, respectively. ††Thrombocytopenia: LMWH group: Thrombocytopenia was associated with cancer or cancer therapy in 10 patients, 1 patient had systemic lupus
erythematosus, and 2 patients had heparin-associated thrombocytopenia. Usual-care group: Thrombocytopenia was associated with cancer or cancer therapy in 5 patients, and 1 patient had autoimmune pancytopenia.
‡‡Twelve-month data provided, because osteoporosis, if present, may have resulted in delayed fractures.
i a
B A p “
F H T r i f c p s i i a r
v 3 t p a
T i i c
P a c a c h t i p
S l
F c p r b L w 1…