The New England Journal of Medicine

Drug Therapy domly distributed along the heparin chains. Approx-

imately one third of the chains of unfractionated
heparin, but only 15 to 25 percent of the chains of
A L A S T A I R J . J . W O O D , M . D. , Editor low-molecular-weight heparins, contain the penta-
saccharide sequence.6
L OW -M OLECULAR -W EIGHT H EPARINS Binding of the pentasaccharide to antithrombin
causes a conformational change in antithrombin that
JEFFREY I. WEITZ, M.D. accelerates its interaction with thrombin and activat-
ed factor X (factor Xa) by about 1000 times.5 The
chief difference between unfractionated heparin and

A
FTER almost two decades of intensive re- low-molecular-weight heparins is in their relative in-
search, low-molecular-weight heparins have hibitory activity against factor Xa and thrombin.6 Any
established their niche as an important class pentasaccharide-containing heparin chain can inhibit
of antithrombotic compounds. The demonstration the action of factor Xa simply by binding to an-
that these compounds are safe and effective for the tithrombin and causing a conformational change (Fig.
prevention and treatment of venous thromboembo- 1). In contrast, to inactivate thrombin, heparin must
lism has led to the licensing of several of them in Eu- bind to both antithrombin and thrombin, thereby
rope and North America. In addition, danaparoid forming a ternary complex.7 This complex can be
sodium, which is a mixture of dermatan sulfate, hep- formed only by pentasaccharide-containing heparin
aran sulfate, and chondroitin sulfate, is often used for chains composed of at least 18 saccharide units.
the treatment of heparin-induced thrombocytope- Whereas most of the chains of unfractionated heparin
nia.1 Low-molecular-weight heparins have replaced are at least 18 saccharide units long, fewer than half
unfractionated heparin in many parts of Europe but of those of low-molecular-weight heparins are of suf-
are only now finding their place in North America. ficient length to bind to both antithrombin and
Their use is likely to increase, however, because two thrombin.8 Consequently, unlike unfractionated hep-
recent studies show that about half of all patients arin, which has equivalent activity against factor Xa
with venous thrombosis can be safely treated with and thrombin, low-molecular-weight heparins have
low-molecular-weight heparins without hospital ad- greater activity against factor Xa.
mission,2,3 and heparin-induced thrombocytopenia, a Tissue-factor-pathway inhibitor may also contrib-
dangerous complication of unfractionated-heparin ute to the inhibitory activity of low-molecular-weight
therapy, occurs less frequently with low-molecular- heparins and unfractionated heparin against factor
weight heparins.4 Xa.9 First, tissue-factor-pathway inhibitor forms a
complex with and inactivates factor Xa, and then the
MECHANISMS OF ACTION OF LOW- complex inactivates factor VIIa.10 Both unfractionat-
MOLECULAR-WEIGHT HEPARINS ed heparin and low-molecular-weight heparins re-
Like unfractionated heparin, low-molecular-weight lease tissue-factor-pathway inhibitor from endothe-
heparins are glycosaminoglycans consisting of chains lium10,11 and enhance its inhibitory activity against
of alternating residues of D-glucosamine and uronic factor Xa.12
acid, either glucuronic acid or iduronic acid.5 Unfrac- The relative importance of inhibition of factor Xa
tionated heparin is a heterogeneous mixture of poly- and inhibition of thrombin in mediating the anti-
saccharide chains ranging in molecular weight from thrombotic effect of unfractionated heparin and
about 3000 to 30,000. Low-molecular-weight hep- low-molecular-weight heparins is unclear, but there
arins are fragments of unfractionated heparin pro- is evidence that both are necessary. In vitro, throm-
duced by controlled enzymatic or chemical depoly- bin is the most important target, because inhibition
merization processes that yield chains with a mean of thrombin prevents feedback activation of factors
molecular weight of about 5000 (Table 1). Both un- V and VIII,13,14 but inhibition of factor Xa also con-
fractionated heparin and low-molecular-weight hep- fers antithrombotic activity.15
arins exert their anticoagulant activity by activating
antithrombin (previously known as antithrombin PHARMACOKINETICS OF LOW-
III). Their interaction with antithrombin is mediat- MOLECULAR-WEIGHT HEPARINS
ed by a unique pentasaccharide sequence that is ran- Low-molecular-weight heparins produce a more
predictable anticoagulant response than unfraction-
ated heparin,16 reflecting their better bioavailabili-
ty, longer half-life, and dose-independent clearance.
From the Hamilton Civic Hospitals Research Centre and McMaster Uni- Thus, when low-molecular-weight heparins are given
versity, Hamilton, Ont., Canada. Address reprint requests to Dr. Weitz at subcutaneously in low doses, the recovery of anti–
the Hamilton Civic Hospitals Research Centre, 711 Concession St., Hamil-
ton, ON L8V 1C3, Canada. factor Xa activity approaches 100 percent, as com-
©1997, Massachusetts Medical Society. pared with about 30 percent with unfractionated

688  Se p te m b e r 4 , 1 9 9 7

The New England Journal of Medicine

Downloaded from nejm.org at LINKOPING UNIVERSITY on September 8, 2013. For personal use only. No other uses without permission.
Copyright © 1997 Massachusetts Medical Society. All rights reserved.
 D R U G T H E R A PY

TABLE 1. COMPARISON OF LOW-MOLECULAR-WEIGHT HEPARIN PREPARATIONS.

MEAN ANTI-XA:
MOLECULAR ANTI-IIA
PREPARATION METHOD OF PREPARATION WEIGHT RATIO*

Ardeparin (Normiflo) Peroxidative depolymerization 6000 1.9

Dalteparin (Fragmin) Nitrous acid depolymerization 6000 2.7
Enoxaparin (Lovenox) Benzylation and alkaline depolymerization 4200 3.8
Nadroparin (Fraxiparine) Nitrous acid depolymerization 4500 3.6
Reviparin (Clivarine) Nitrous acid depolymerization, chromatographic 4000 3.5
purification
Tinzaparin (Innohep) Heparinase digestion 4500 1.9

*The ratios were calculated by dividing the anti–factor Xa (anti-Xa) activity by the antithrombin
(anti-IIa) activity. The ratios are based on information provided by the manufacturers.

Pentasaccharide
sequence Factor Xa
Unfractionated
heparin

Antithrombin

Thrombin

Low- Pentasaccharide
molecular- sequence
weight
heparin Factor Xa

Antithrombin

Figure 1. Catalysis of Antithrombin-Mediated Inactivation of Thrombin or Factor Xa by Unfractionated Heparin or Low-

Molecular-Weight Heparins.
The interaction of unfractionated heparin and low-molecular-weight heparins with antithrombin is mediated by the pen-
tasaccharide sequence of the drugs. Binding of either to antithrombin causes a conformational change at its reactive
center that accelerates its interaction with factor Xa. Consequently, both unfractionated heparin and low-molecular-
weight heparins catalyze the inactivation of factor Xa by antithrombin. In contrast to factor Xa inhibition, catalysis of
antithrombin-mediated inactivation of thrombin requires the formation of a ternary heparin–antithrombin–thrombin
complex. This complex can be formed only by chains at least 18 saccharide units long. This explains why low-molec-
ular-weight heparins have less inhibitory activity against thrombin than unfractionated heparin.

Vo l u m e 3 3 7 Nu m b e r 1 0  689

The New England Journal of Medicine

Downloaded from nejm.org at LINKOPING UNIVERSITY on September 8, 2013. For personal use only. No other uses without permission.
Copyright © 1997 Massachusetts Medical Society. All rights reserved.
 The New England Journal of Medicine

fibronectin; to platelet factor 4,21 which is released

TABLE 2. MECHANISMS RESPONSIBLE FOR THE PHARMACOKINETIC from activated platelets; and to high-molecular-
ADVANTAGES OF LOW-MOLECULAR-WEIGHT HEPARINS
OVER UNFRACTIONATED HEPARIN.
weight multimers of von Willebrand factor,22,23 the
storage form of von Willebrand factor that is re-
leased from platelets and endothelial cells. Binding
ADVANTAGE MECHANISM of unfractionated heparin to plasma proteins reduces
More predictable anticoagulant Less binding to plasma proteins and to its anticoagulant activity, because less is available to
response proteins released from activated interact with antithrombin,24 and the unpredictable
platelets and endothelial cells
anticoagulant response reflects the wide variability in
Better bioavailability at low doses Less binding to endothelium
Dose-independent clearance Less binding to macrophages
plasma concentrations of heparin-binding proteins.24
mechanism Some of these heparin-binding proteins are acute-
Longer half-life Less binding to macrophages phase reactants, the concentrations of which increase
in ill patients, whereas others, like platelet factor 4
and von Willebrand factor, are released during the
clotting process. Because of the unpredictable anti-
coagulant response,25 careful laboratory monitoring
TABLE 3. COMPARISON OF MONITORING REQUIREMENTS is essential when unfractionated heparin is given in
FOR CURRENTLY AVAILABLE ANTICOAGULANTS. therapeutic doses (Table 3).
The reduced binding of low-molecular-weight
INDICATION MONITORING REQUIREMENTS
heparins to plasma proteins26,27 and endothelium28
UNFRACTIONATED LOW-MOLECULAR- ORAL ANTI-
accounts for their better bioavailability. Their re-
HEPARIN WEIGHT HEPARINS COAGULANTS* duced binding to macrophages explains why they are
Prophylaxis None None INR
not cleared by hepatic mechanisms to the same extent
Treatment APTT or plasma heparin None‡ INR
as unfractionated heparin and why renal clearance is
concentrations† slower than hepatic uptake, thereby accounting for
the longer plasma half-life of low-molecular-weight
*INR denotes international normalized ratio. For both prophylaxis and
treatment, the dose of oral anticoagulant is usually adjusted to achieve an
heparins. The better bioavailability, dose-independ-
INR of 2.0 to 3.0. ent clearance, and decreased affinity for heparin-
†APTT denotes activated partial-thromboplastin time. Plasma heparin binding proteins make the anticoagulant response to
concentrations can be measured by a chromogenic anti–factor Xa assay or low-molecular-weight heparins more predictable than
by protamine titration.
that to unfractionated heparin. Consequently, labo-
‡Plasma anti–factor Xa concentrations should be monitored in patients
with renal insufficiency and possibly in those weighing less than 50 kg or ratory monitoring is unnecessary except in patients
more than 80 kg. with renal insufficiency29 and possibly those with a
body weight of less than 50 kg or more than 80 kg
(Table 3).
Low-molecular-weight heparins cause less bleed-
ing than unfractionated heparin in laboratory ani-
heparin.17 The plasma half-life of low-molecular- mals,30 for several reasons. First, low-molecular-
weight heparins is two to four times as long as that weight heparins inhibit platelet function less than
of unfractionated heparin, ranging from two to four unfractionated heparin31 because they bind less to
hours after intravenous injection and from three to platelets.32 Second, unlike unfractionated heparin,
six hours after subcutaneous injection.6,18,19 The in- low-molecular-weight heparins do not increase mi-
hibitory activity of low-molecular-weight heparins crovascular permeability.33 Third, because of their
against factor Xa persists longer than their inhibitory lower affinity for endothelial cells, high-molecular-
activity against thrombin, reflecting the more rapid weight forms of von Willebrand factor, and plate-
clearance of longer heparin chains. In contrast, un- lets,22,23,30 low-molecular-weight heparins are less
fractionated heparin is eliminated in two phases in a likely to interfere with the interaction between plate-
dose-dependent fashion: a rapid, saturable phase re- lets and vessel walls.
flecting hepatic uptake, and a slower phase corre-
sponding to renal clearance.20 CLINICAL STUDIES
The pharmacokinetic differences between low- Low-molecular-weight heparins are safe and effec-
molecular-weight heparins and unfractionated hepa- tive for the prevention and treatment of venous
rin can be explained by the decreased propensity of thromboembolism (Table 4). They have also been
the former to bind to plasma proteins, endothelial used successfully in patients with unstable angina or
cells, and macrophages (Table 2). In contrast to low- acute thrombotic stroke. The key clinical studies of
molecular-weight heparins, unfractionated heparin low-molecular-weight heparins are summarized be-
binds to endogenous plasma proteins, such as histi- low. On the basis of these studies, low-molecular-
dine-rich glycoprotein, polymeric vitronectin, and weight heparins are at least as safe and effective as

690  Se p te m b e r 4 , 1 9 9 7

The New England Journal of Medicine

Downloaded from nejm.org at LINKOPING UNIVERSITY on September 8, 2013. For personal use only. No other uses without permission.
Copyright © 1997 Massachusetts Medical Society. All rights reserved.
 D R U G T H E R A PY

TABLE 4. ADVANTAGES OF LOW-MOLECULAR-WEIGHT HEPARINS AND RECOMMENDED DOSES FOR THE PREVENTION
AND TREATMENT OF THROMBOSIS.

ADVANTAGES OF LOW-MOLECULAR-
INDICATION WEIGHT HEPARINS RECOMMENDED DOSES*

Prevention
General surgery At least as effective as low-dose unfraction- Low risk†
ated heparin34 but can be given once Dalteparin, 2500 U 1–2 hr before surgery and once daily after surgery
daily and cause fewer hematomas at Enoxaparin, 2000 U 1–2 hr before surgery and once daily after surgery
injection sites35,36 Nadroparin, 3100 U 2 hr before surgery and once daily after surgery
Tinzaparin, 3500 U 2 hr before surgery and once daily after surgery
High risk‡
Dalteparin, 5000 U 10–12 hr before surgery and once daily after surgery
Enoxaparin, 4000 U 10–12 hr before surgery and once daily after surgery
Orthopedic More effective than low-dose unfraction- Ardeparin, 50 U/kg twice daily starting 12–24 hr after surgery
surgery ated heparin34,37,38; more effective than Dalteparin, 5000 U 8–12 hr before surgery and once daily starting 12 hr after
warfarin in patients undergoing total surgery
knee replacement39-44; no monitoring Enoxaparin, 3000 U twice daily starting 12–24 hr after surgery or 4000 U
required once daily starting 10–12 hr before surgery
Nadroparin, 40 U/kg starting 2 hr before surgery and once daily after surgery
for 3 days; the dose is then increased to 60 U/kg once daily
Tinzaparin, 50 U/kg 2 hr before surgery and once daily after surgery or
75 U/kg once daily starting 12–24 hr after surgery
Acute spinal Apparently effective,45,46 whereas low-dose Enoxaparin, 3000 U twice daily
injury unfractionated heparin47 is not, and
higher doses of unfractionated heparin
cause excessive bleeding48
Multiple trauma More effective than unfractionated Enoxaparin, 3000 U twice daily
heparin49
Medical As effective as low-dose unfractionated Dalteparin, 2500 U once daily
conditions heparin but can be given once daily 50,51 Enoxaparin, 2000 U once daily
Treatment
Venous throm- At least as safe and effective as unfraction- Dalteparin, 100 U/kg twice daily
boembolism ated heparin52-54 but can be given subcu- Enoxaparin, 100 U/kg twice daily
taneously without laboratory monitor- Nadroparin, 90 U/kg twice daily
ing, thereby allowing out-of-hospital Tinzaparin, 175 U/kg once daily
treatment2,3
Unstable At least as effective as unfractionated Dalteparin, 100 U/kg twice daily
angina heparin55-57 but can be given sub- Enoxaparin, 100 U/kg twice daily
cutaneously without monitoring

*Doses are shown in anti–factor Xa units. Low-molecular-weight heparins are given subcutaneously for both prophylaxis and treatment.
The prophylactic doses recommended for each low-molecular-weight heparin preparation are slightly different, but a common rationale un-
derlies these regimens. Lower doses are used for low-risk general surgical or medical patients, whereas higher doses are used for high-risk
general surgical or orthopedic surgical patients. When relatively large doses of low-molecular-weight heparins are started preoperatively, the
dose is given 10 to 12 hours before surgery, to avoid excessive intraoperative bleeding. Lower doses of low-molecular-weight heparins can be
given one to two hours before surgery. The doses used for the treatment of venous thromboembolism or for unstable angina are higher than
those used for prophylaxis, and similar regimens are used for each of the low-molecular-weight heparins.
†Low-risk general surgical patients are those undergoing uncomplicated abdominal or pelvic surgery lasting 30 minutes or more.
‡High-risk general surgical patients are those undergoing abdominal or pelvic surgery for cancer or those with previous venous throm-
boembolism.

unfractionated heparin and are more convenient to ing.58 Like unfractionated heparin, low-molecular-
use, because they can be given subcutaneously with- weight heparins also are given subcutaneously 2 to
out laboratory monitoring. 12 hours before surgery but are given only once dai-
ly postoperatively. They are marginally better than
PROPHYLAXIS AGAINST low-dose unfractionated heparin at preventing venous
THROMBOEMBOLISM thromboembolism59 and cause fewer wound hemato-
General Surgery mas.34,35
Low-dose unfractionated heparin (5000 U given
Orthopedic Surgery of the Lower Limb
subcutaneously 2 hours before surgery and every
8 to 12 hours postoperatively) provides safe and ef- Without prophylaxis, deep-vein thrombosis occurs
fective prophylaxis for patients undergoing general in 50 to 70 percent of patients undergoing total hip
surgery, reducing the risk of venous thromboembo- replacement, total knee replacement, or surgery for
lism and fatal pulmonary embolism by 70 percent hip fractures. Low-molecular-weight heparins are
and 50 percent, respectively, with minimal bleed- safe and effective in these high-risk patients.

Vo l u m e 3 3 7 Nu m b e r 1 0  691

The New England Journal of Medicine

Downloaded from nejm.org at LINKOPING UNIVERSITY on September 8, 2013. For personal use only. No other uses without permission.
Copyright © 1997 Massachusetts Medical Society. All rights reserved.
 The New England Journal of Medicine

Total Hip Replacement to a similar extent.64 No trial has yet compared low-
As compared with placebo in randomized clinical molecular-weight heparins with low-intensity war-
trials,36,60 low-molecular-weight heparins significantly farin. Low-molecular-weight heparins are a good
reduced the risk of deep-vein thrombosis (range of choice for prophylaxis in patients undergoing sur-
risk reduction, 31 percent to 79 percent) without gery for hip fracture. Treatment should be started
increasing bleeding. Low-molecular-weight heparins preoperatively if a delay in surgery is expected. Al-
were more effective than low-dose unfractionated though warfarin is also effective in these patients, it
heparin59 and equal61 or superior 62 to adjusted-dose is less convenient, because the timing of surgery is
unfractionated heparin (heparin started preoperative- often difficult to predict.
ly at a dose of 5000 U subcutaneously and continued Acute Spinal Cord Injury
three times daily postoperatively, with the dose ad-
justed to maintain the activated partial-thromboplas- Deep-vein thrombosis develops in about 40 per-
tin time near the upper range of normal). cent of patients with acute spinal cord injuries. The
In three studies that compared low-molecular- period of greatest risk is within two weeks after inju-
weight heparins with low-intensity warfarin (with ry,65 when the incidence of symptomatic venous
the dose adjusted to reach an international normal- thrombosis and pulmonary embolism may be as high
ized ratio of 2.0 to 3.0), there was no difference in as 14.5 percent and 4.6 percent, respectively. Two
the rates of thrombosis or bleeding.39,40,63 A meta- small trials suggest that low-molecular-weight hepa-
analysis41 comparing several prophylactic regimens rins are effective in patients with acute spinal cord in-
found that low-molecular-weight heparins were the juries.45,66 Adjusted-dose unfractionated heparin may
most effective, although their advantage over war- also be effective when given in doses sufficient to
farin and adjusted-dose unfractionated heparin was produce an activated partial-thromboplastin time in
small. Of these prophylactic options, however, low- the lower therapeutic range,46 but this regimen caus-
molecular-weight heparins are the easiest to admin- es an unacceptably high rate of bleeding. Although
ister, because no monitoring is required. neither intermittent pneumatic compression48 nor
low-dose unfractionated heparin67 is effective alone,
Total Knee Replacement intermittent pneumatic compression appears to be
Low-molecular-weight heparins given after total effective when combined with low-dose unfraction-
knee replacement are safe and effective, but the ab- ated heparin and the use of elastic stockings.47
solute incidence of deep-vein thrombosis remains
Multiple Trauma
high (25 to 30 percent, with one quarter of the
thromboses being proximal). In all six trials in which A prospective cohort study of patients with major
low-molecular-weight heparins were compared with trauma found a 50 percent incidence of deep-vein
low-intensity warfarin,37-40,42,63 low-molecular-weight thrombosis documented by venography.68 A recent
heparins were superior. Warfarin was relatively in- randomized study of 344 patients with major trau-
effective, because the incidence of venous thrombo- ma and without evidence of intracranial bleeding
sis was 45 to 50 percent (10 percent of thromboses compared low-dose unfractionated heparin with low-
were proximal). Two studies demonstrated a small molecular-weight heparin started within 36 hours
but significant increase in postoperative bleeding after injury.49 As compared with low-dose unfrac-
with low-molecular-weight heparins as compared tionated heparin, low-molecular-weight heparin re-
with warfarin,38,63 which is not surprising, because duced the overall rate of venous thrombosis from 44
the onset of anticoagulation with warfarin is de- percent to 31 percent (P0.014) and lowered the
layed. A recent audit examining the cause of post- incidence of proximal thrombosis from 15 percent
operative bleeding in patients treated with low- to 6 percent (P0.09). Major bleeding occurred in
molecular-weight heparins suggests that up to 80 six patients (1.7 percent), five of whom had received
percent of bleeding episodes are associated with ini- low-molecular-weight heparin.
tiation of treatment too soon after surgery (Cooley
Medical Conditions
M, Rhone–Poulenc Rorer: personal communication).
Low-molecular-weight heparins should not be given Patients with ischemic stroke have an overall inci-
for at least 12 hours after surgery. dence of deep-vein thrombosis of 42 percent in the
paretic or paralyzed leg.69 In a randomized trial,
Surgery for Hip Fracture low-molecular-weight heparin was better than place-
As compared with placebo, both low-dose unfrac- bo in reducing the incidence of venous thrombosis,
tionated heparin43 and low-molecular-weight hep- and it did not increase the incidence of bleeding.70
arins44 result in a 45 percent reduction in the In another trial, there was no difference in the rates
incidence of deep-vein thrombosis in patients under- of thrombosis between patients receiving once-daily
going surgery for hip fracture. Low-intensity war- low-molecular-weight heparin and those receiving
farin decreases the incidence of venous thrombosis placebo,71 but the dose was very low. Finally, danap-

692  Se p te m b e r 4 , 1 9 9 7

The New England Journal of Medicine

Downloaded from nejm.org at LINKOPING UNIVERSITY on September 8, 2013. For personal use only. No other uses without permission.
Copyright © 1997 Massachusetts Medical Society. All rights reserved.
 D R U G T H E R A PY

aroid sodium was superior to low-dose unfractionat- post-treatment thrombus size is uncertain, meta-
ed heparin in reducing the incidence of deep-vein analyses of trials comparing the effects of low-molec-
thrombosis in one trial.72 On the basis of these data, ular-weight heparins and unfractionated heparin on
low-molecular-weight heparins appear to be the best the incidence of recurrent venous thromboembolism
prophylaxis for patients with ischemic stroke. suggest that the former are more effective.52-54 For
In a study comparing low-molecular-weight hep- example, in one overview analysis,53 2.7 percent of
arin with placebo in medical patients older than 65 the patients treated with low-molecular-weight hep-
years,73 low-molecular-weight heparin reduced the arins had recurrences, as compared with 7.0 percent
rate of thrombosis detected by fibrinogen leg scan- of those given unfractionated heparin (P0.001).
ning from 9.1 percent to 3.0 percent (P0.03) with- Pooled analyses also suggest that low-molecular-
out any increase in bleeding. In two randomized weight heparins are safer.53,54 In the nine studies that
studies comparing low-dose unfractionated heparin could be evaluated, major bleeding occurred in 0.9
with low-molecular-weight heparins, the rates of percent of the patients treated with low-molecular-
venous thrombosis and bleeding were similar.50,51 weight heparins, as compared with 3.2 percent of
Thus, both therapies provide effective prophylaxis those given unfractionated heparin (P0.005). The
for medical patients. pooled long-term mortality rate was lower in pa-
tients treated with low-molecular-weight heparins
Patency of Femoropopliteal Bypass Grafts than in those given unfractionated heparin (4.3 per-
A recent study compared the effect of low-molec- cent vs. 8.1 percent, P0.03). This effect was almost
ular-weight heparin with that of aspirin and dipyrid- entirely attributable to differences in the subgroup
amole on the patency of femoropopliteal bypass of patients with cancer53 and was heavily influenced
grafts.74 Graft survival at one year was 78 percent in by the results of two studies.77,78 Nevertheless, the
the patients given low-molecular-weight heparin, as lower mortality in these trials may reflect superior
compared with 64 percent in those given aspirin and antithrombotic activity of low-molecular-weight hep-
dipyridamole (P0.03). arins in high-risk patients.
In most trials, low-molecular-weight heparins were
Restenosis after Angioplasty given in fixed or weight-adjusted doses by subcuta-
Restenosis occurs in up to 40 percent of patients neous injection either once or twice daily without
after successful coronary angioplasty. Two random- laboratory monitoring. Two recent trials conducted
ized trials evaluated the effect of short-term treat- among patients with deep-vein thrombosis took ad-
ment with low-molecular-weight heparins on the in- vantage of the predictable anticoagulant response to
cidence of restenosis.75,76 As compared with placebo, compare the effects of low-molecular-weight hepa-
neither low-molecular-weight heparin alone75,76 nor rins given subcutaneously twice daily to outpatients
the combination of low-molecular-weight heparin with the effects of unfractionated heparin given by
and fish oil76 reduced the incidence of restenosis af- continuous intravenous infusion to inpatients.2,3 The
ter coronary angioplasty. rates of recurrent thromboembolism and major
bleeding were similar with both treatments (Table
TREATMENT OF THROMBOSIS 5). On the basis of these results, unmonitored out-
patient therapy with low-molecular-weight heparin
Venous Thrombosis
appears to be as safe and effective as in-hospital in-
Low-molecular-weight heparins have been com- travenous unfractionated heparin in selected patients
pared with unfractionated heparin for the treatment with proximal-vein thrombosis. Two recent studies
of patients with established deep-vein thrombosis. indicate that unmonitored low-molecular-weight hep-
In eight trials, the effect of therapy on thrombus size arin is also as safe and effective as intravenous un-
was assessed by comparing pretreatment venograms fractionated heparin in patients with pulmonary em-
with those obtained after 5 to 10 days of treatment. bolism. The first study compared these agents in
On the basis of meta-analyses of these studies,52-54 1021 patients with venous thromboembolism, 26
low-molecular-weight heparins prevented thrombus percent of whom had pulmonary embolism, and re-
growth more than unfractionated heparin. A re- ported similar rates of recurrent thromboembolism
duction in thrombus size occurred in 64 percent of (5.3 percent and 4.9 percent, respectively) and ma-
the patients treated with low-molecular-weight hep- jor bleeding (3.1 percent and 2.3 percent).79 In the
arins, as compared with 50 percent of those given second study, 912 patients with pulmonary embo-
unfractionated heparin (P0.001). Furthermore, lism were randomly assigned to receive unfractionat-
only 6 percent of the patients treated with low- ed heparin or low-molecular-weight heparin. The
molecular-weight heparins had an increase in throm- incidence of death, recurrent venous thromboembo-
bus size, as compared with 12 percent of those given lism, or major bleeding was essentially the same in
unfractionated heparin (P0.001). both groups (2.9 percent and 3.0 percent, respec-
Although the clinical importance of changes in tively).80 These findings may shift the management

Vo l u m e 3 3 7 Nu m b e r 1 0  693

The New England Journal of Medicine

Downloaded from nejm.org at LINKOPING UNIVERSITY on September 8, 2013. For personal use only. No other uses without permission.
Copyright © 1997 Massachusetts Medical Society. All rights reserved.
 The New England Journal of Medicine

TABLE 5. RATES OF RECURRENT THROMBOEMBOLISM, MAJOR BLEEDING, AND DEATH

IN TWO RANDOMIZED TRIALS COMPARING LOW-MOLECULAR-WEIGHT HEPARIN
GIVEN TO OUTPATIENTS WITH UNFRACTIONATED HEPARIN GIVEN TO INPATIENTS
FOR PROXIMAL DEEP-VEIN THROMBOSIS.

EVENT LEVINE ET AL.2 KOOPMAN ET AL.3

LOW-MOLECULAR- UNFRACTIONATED LOW-MOLECULAR- UNFRACTIONATED

WEIGHT HEPARIN HEPARIN WEIGHT HEPARIN HEPARIN
(N 247) (N 253) (N 202) (N 198)

percent

Recurrent thrombo- 5.3 6.7 6.9 8.5

embolism
Major bleeding 0.5 2.0 2.0 1.2
Death 6.9 8.0 4.0 6.3

of venous thromboembolism from the inpatient to compared with unfractionated heparin given by
the outpatient setting. continuous intravenous infusion.56,57 The treatment
Once-daily subcutaneous low-molecular-weight with low-molecular-weight heparin was not moni-
heparin has also been compared with warfarin for tored, whereas the dose of unfractionated heparin
secondary prophylaxis of venous thrombosis after a was titrated to achieve a therapeutic activated par-
10-day course of continuous intravenous unfraction- tial-thromboplastin time. In one study,56 the inci-
ated heparin.81 The incidence of recurrent venous dence of death, myocardial infarction, or recurrent
thromboembolism was similar in patients given low- angina in the patients receiving low-molecular-weight
molecular-weight heparin for three months and those heparin was similar to that in the patients receiving
given warfarin (6 percent and 4 percent, respectively), unfractionated heparin (9.3 percent and 7.8 percent,
but bleeding was less frequent with low-molecular- respectively; P0.42), as was the need for urgent
weight heparin than with warfarin (4 percent and 13 revascularization procedures (5.2 percent and 5.8
percent, P0.04). Thus, low-molecular-weight hepa- percent, P0.48). In contrast, in the other study,57
rins may be a reasonable alternative to warfarin in the incidence of death, myocardial infarction, or re-
patients at high risk for bleeding or in whom mon- current angina was 17 percent lower in the patients
itoring is difficult. Low-molecular-weight heparins given low-molecular-weight heparin than in those
may be better than unfractionated heparin for sec- given unfractionated heparin (16.5 percent vs. 19.8
ondary prophylaxis, because monitoring is unneces- percent, P0.02), whereas the incidence of major
sary and the risk of osteoporosis appears to be lower bleeding was similar in both groups (6.5 percent
(see below). and 7.0 percent, respectively). On the basis of these
data, low-molecular-weight heparins appear to be at
Unstable Angina least as effective as unfractionated heparin in pa-
The combination of unfractionated heparin and tients with unstable angina, and they are more con-
aspirin is the current treatment of choice for pa- venient, because no monitoring is necessary. The
tients with unstable angina. In a small, open trial dose of low-molecular-weight heparin should not
comparing low-molecular-weight heparin plus aspi- exceed 100 anti–factor Xa units per kilogram of
rin, unfractionated heparin plus aspirin, and aspirin body weight, because higher doses appear to cause
alone, low-molecular-weight heparin reduced the risk excessive bleeding.83
of myocardial infarction.82 These promising results
Ischemic Stroke
prompted three large, randomized trials of low-
molecular-weight heparins in patients with unstable In a study of 312 patients with acute ischemic
angina who were treated with aspirin.55-57 In the first stroke, patients were randomly assigned to one of
study, low-molecular-weight heparin or placebo was two regimens of subcutaneous low-molecular-weight
given for 35 to 45 days.55 At day 6, the incidence of heparin (4100 anti–factor Xa units once or twice
death or myocardial infarction was lower in the pa- daily) or to placebo within 48 hours after the onset
tients given low-molecular-weight heparin than in of symptoms.84 The patients were treated for 10 days
those given placebo (1.8 percent vs. 4.7 percent, and followed for 6 months. Low-molecular-weight
P0.001). In the second and third studies, twice-dai- heparin was superior to placebo; 45 percent of the
ly subcutaneous low-molecular-weight heparin was patients given low-molecular-weight heparin twice

694  Se p te m b e r 4 , 1 9 9 7

The New England Journal of Medicine

Downloaded from nejm.org at LINKOPING UNIVERSITY on September 8, 2013. For personal use only. No other uses without permission.
Copyright © 1997 Massachusetts Medical Society. All rights reserved.
 D R U G T H E R A PY

daily and 52 percent of those given low-molecular- their lower affinity for platelet factor 4 results in the
weight heparin once daily died or became depend- formation of fewer complexes.
ent with regard to activities of daily living, as com- Low-molecular-weight heparins should not be
pared with 65 percent of the patients given placebo given to patients with established heparin-induced
(P0.005). The rate of hemorrhagic transforma- thrombocytopenia, because they have a high degree
tion of the infarct during the 10-day treatment pe- of in vitro cross-reactivity with the antibody that
riod was similar in the three groups. causes this disorder1 and they can cause heparin-
induced thrombocytopenia in patients with a history
REMAINING QUESTIONS of it.88-90 Danaparoid sodium, which cross-reacts
Preoperative or Postoperative Dosing with Low-Molecular- minimally with heparin antibodies in vitro, has been
Weight Heparins used successfully in this setting.1 Alternatively, direct
Although anticoagulant prophylaxis started post- thrombin inhibitors such as hirudin, bivalirudin, or
operatively reduces the incidence of venous throm- argatroban can be used.
boembolism after hip or knee arthroplasty, break-
Neutralization of Low-Molecular-Weight Heparins with
through venous thromboembolism develops in about Protamine Sulfate
one fifth of patients. This probably occurs because
thrombi form during surgery. Whether treatment When given in equimolar concentrations, prota-
with low-molecular-weight heparins started preoper- mine sulfate neutralizes the anti-thrombin activity of
atively is more effective is not known. Preoperative low-molecular-weight heparins but only partially re-
dosing may be a problem in patients undergoing spi- verses their anti–factor Xa activity,91 probably be-
nal anesthesia, although spinal cord bleeding is rare.85 cause it fails to bind to the very-low-molecular-
Furthermore, preoperative low-molecular-weight hep- weight heparin chains. Although protamine sulfate
arins might increase intraoperative bleeding. blocks bleeding induced by low-molecular-weight
heparins in laboratory animals,92 there have been no
Once-Daily versus Twice-Daily Dosing studies in humans.
With its apparent half-life of three to four hours,
Safety of Low-Molecular-Weight Heparins in Pregnancy
administering low-molecular-weight heparins twice
daily should be optimal. However, low-molecular- Unfractionated heparin is the anticoagulant of
weight heparins are effective when treatment is start- choice in pregnant women, because unlike warfarin,
ed preoperatively and continued once daily postop- it does not cross the placenta. Low-molecular-weight
eratively in patients undergoing general surgery or heparins also do not cross the placenta,93 and descrip-
total hip replacement.59 When started postoperative- tive studies suggest that they are both safe and effec-
ly, twice-daily treatment with low-molecular-weight tive in pregnancy.94
heparin was more effective than once-daily therapy in
Risk of Heparin-Induced Osteoporosis
patients undergoing orthopedic procedures.39,86 On
the basis of these results, twice-daily low-molecular- When given for more than one month, un-
weight heparin is preferred for postoperative therapy fractionated heparin can cause osteoporosis. This
in patients who have had orthopedic surgery. complication may be less frequent with low-molec-
For the treatment of venous thrombosis, low- ular-weight heparins. In a case series, low-molecular-
molecular-weight heparins were given twice daily in weight heparin was used successfully in patients with
most studies. However, once-daily low-molecular- established heparin-induced osteoporosis,94 and in a
weight heparin was at least as safe and effective as recent three-month trial of 80 patients, the inci-
unfractionated heparin in one study,77 and once-dai- dence of osteoporosis was lower in patients given
ly dosing is more convenient. low-molecular-weight heparin than in those given
unfractionated heparin (2.6 percent vs. 17.6 percent,
Risk of Heparin-Induced Thrombocytopenia
P0.05).95
Heparin-induced thrombocytopenia, which can
cause devastating thrombotic complications, is trig- Comparability of Various Preparations of Low-Molecular-
Weight Heparin
gered by antibodies directed against complexes of
heparin and platelet factor 4 that form on the sur- Although low-molecular-weight heparins have sim-
face of platelets and activate their Fc receptors.87 In ilar mechanisms of action, their molecular-weight dis-
a recent randomized trial, the incidence of heparin- tributions vary (Table 1), causing differences in their
induced thrombocytopenia was significantly lower inhibitory activities against factor Xa and thrombin,96
in patients given prophylaxis with low-molecular- the extent to which they bind to plasma proteins,26,27
weight heparin than in those receiving unfractionat- and their plasma half-lives.6,17,18 There are few studies
ed heparin.4 These findings may reflect the fact that comparing different low-molecular-weight heparins.
low-molecular-weight heparins cause less activation In a study of enoxaparin and reviparin for the pre-
of platelets and release of platelet factor 4 and that vention of venous thromboembolism in 416 pa-

Vo l u m e 3 3 7 Nu m b e r 1 0  695

The New England Journal of Medicine

Downloaded from nejm.org at LINKOPING UNIVERSITY on September 8, 2013. For personal use only. No other uses without permission.
Copyright © 1997 Massachusetts Medical Society. All rights reserved.
 The New England Journal of Medicine

tients undergoing total hip replacement, the inci- replaced unfractionated heparin in many parts of
dence of venous thrombosis was similar in both Europe, they are only beginning to find their niche
groups (9 percent and 10 percent, respectively), as in North America. Because they cause less heparin-
was the rate of major bleeding (1 percent in both induced thrombocytopenia and possibly less os-
groups).35 Thus, despite their different methods of teoporosis, the use of low-molecular-weight heparins
preparation and small differences in their specific ac- is likely to increase over the coming years.
tivities, the two agents were equally safe and effective There still is room for anticoagulants that are
when equivalent doses in terms of anti–factor Xa units more potent. Like unfractionated heparin, low-
were given. These results suggest that low-molecular- molecular-weight heparins are unable to inactivate
weight heparins with similar molecular-weight pro- thrombin bound to fibrin,100 which may be an im-
files, and hence similar specific activities, are equally portant trigger for clot extension at sites of vascular
effective. injury. This may explain why it has been difficult to
show an advantage of low-molecular-weight hepa-
Cost Effectiveness rins over unfractionated heparin in patients with
In North America, low-molecular-weight heparins unstable angina.53-57 The limitations of both low-
are 10 to 20 times as expensive as unfractionated molecular-weight heparins and unfractionated hep-
heparin. Accordingly, from an economic viewpoint, arin have stimulated the development of new
it is difficult to justify their routine use in general antithrombotic drugs, including hirudin, factor Xa
surgical or medical patients, in whom their advan- inhibitors, tissue-factor-pathway inhibitor, and an-
tages over unfractionated heparin are minimal. Low- tagonists of glycoprotein IIb/IIIa, the platelet fi-
molecular-weight heparins are more effective than brinogen receptor. With these and newer drugs on
unfractionated heparin for patients undergoing or- the horizon, our ability to prevent and treat throm-
thopedic surgery41,59 and more effective than war- botic diseases is likely to improve substantially.
farin for patients undergoing total knee replace-
ment.37-40,42,63
Consequently, fewer patients require treatment Supported by grants from the Medical Research Council of Canada and
for postoperative venous thrombosis, thereby ex- the Heart and Stroke Foundation of Ontario. Dr. Weitz is the recipient of
a Career Investigator Award from the Heart and Stroke Foundation of On-
plaining why analyses of cost effectiveness favor tario.
low-molecular-weight heparins over unfractionated
heparin97 or warfarin.98,99 The validity of these stud- I am indebted to Drs. J. Ginsberg and J. Hirsh for their critical
review of the manuscript, Dr. M. Crowther for assistance with the
ies is open to question, however, because venogra- bibliography, and Mrs. S. Crnic for her help preparing the paper.
phy was used to detect venous thrombosis. Since
only 10 percent of patients with venographically de- REFERENCES
tected thrombosis after orthopedic surgery have
clinically evident thrombosis, and since venography 1. Chong BH, Ismail F, Cade J, Gallus AS, Gordon S, Chesterman CN.
Heparin-induced thrombocytopenia: studies with a new low molecular
is not routinely performed in this setting, these weight heparinoid, Org 10172. Blood 1989;73:1592-6.
studies overestimate the cost differences between 2. Levine M, Gent M, Hirsh J, et al. A comparison of low-molecular-
weight heparin administered primarily at home with unfractionated hepa-
the treatments. rin administered in the hospital for proximal deep-vein thrombosis. N Engl
The use of low-molecular-weight heparins to treat J Med 1996;334:677-81.
selected patients with venous thromboembolism out- 3. Koopman MMW, Prandoni P, Piovella F, et al. Treatment of venous
thrombosis with intravenous unfractionated heparin administered in the
side the hospital has the potential to dramatically re- hospital as compared with subcutaneous low-molecular-weight heparin ad-
duce the cost of health care. If other studies confirm ministered at home. N Engl J Med 1996;334:682-7.
the effectiveness of low-molecular-weight heparins 4. Warkentin TE, Levine MN, Hirsh J, et al. Heparin-induced thrombo-
cytopenia in patients treated with low-molecular-weight heparin or unfrac-
for the outpatient management of venous throm- tionated heparin. N Engl J Med 1995;332:1330-5.
boembolism,2,3 and if these findings can be extrapo- 5. Rosenberg RD, Bauer KA. The heparin–antithrombin system: a natural
anticoagulant mechanism. In: Colman RW, Hirsch J, Marder VJ, Salzman
lated to patients with arterial thrombosis, enormous EW, eds. Hemostasis and thrombosis: basic principles and clinical practice.
cost savings may be realized. 3rd ed. Philadelphia: J.B. Lippincott, 1994:837-60.
6. Harenberg J. Pharmacology of low molecular weight heparins. Semin
SUMMARY Thromb Hemost 1990;16:Suppl:12-8.
7. Danielsson A, Raub E, Lindahl U, Bjork I. Role of ternary complexes,
Low-molecular-weight heparins have proved to in which heparin binds both antithrombin and proteinase, in the accelera-
be both safe and effective for the prophylaxis and tion of the reactions between antithrombin and thrombin or factor Xa.
J Biol Chem 1986;261:15467-73.
treatment of venous thromboembolism and show 8. Jordan RE, Oosta GM, Gardner WT, Rosenberg RD. The kinetics of
promise for prophylaxis and treatment of arterial hemostatic enzyme-antithrombin interactions in the presence of low mo-
lecular weight heparin. J Biol Chem 1980;255:10081-90.
thrombosis. Their pharmacokinetic advantages over 9. Abildgaard U, Lindahl AK, Sandset PM. Heparin requires both anti-
unfractionated heparin have obviated the need for thrombin and extrinsic pathway inhibitor for its anticoagulant effect in hu-
laboratory monitoring and have set the stage for the man blood. Haemostasis 1991;21:254-7.
10. Sandset PM, Abildgaard U, Larsen ML. Heparin induces release of
outpatient management of venous thrombosis. Al- extrinsic coagulation pathway inhibitor (EPI). Thromb Res 1988;50:803-
though low-molecular-weight heparins have largely 13.

696  Se p te m b e r 4 , 1 9 9 7

The New England Journal of Medicine

Downloaded from nejm.org at LINKOPING UNIVERSITY on September 8, 2013. For personal use only. No other uses without permission.
Copyright © 1997 Massachusetts Medical Society. All rights reserved.
 D R U G T H E R A PY

11. Broze GJ Jr. Tissue factor pathway inhibitor. Thromb Haemost 1995; boembolism after knee arthroplasty: a randomized, double-blind trial com-
74:90-3. paring enoxaparin with warfarin. Ann Intern Med 1996;124:619-26.
12. Huang ZF, Wun T-C, Broze GJ Jr. Kinetics of factor Xa inhibition by 38. Spiro TE, Fitzgerald RH, Trowbridge AA, et al. Enoxaparin a low mo-
tissue factor pathway inhibitor. J Biol Chem 1993;268:26950-5. lecular weight heparin and warfarin for the prevention of venous thrombo-
13. Ofosu FA, Hirsh J, Esmon CT, et al. Unfractionated heparin inhibits embolic disease after elective knee replacement surgery. Blood 1994;84:
thrombin-catalysed amplification reactions of coagulation more efficiently Suppl 1:246A. abstract.
than those catalysed by factor Xa. Biochem J 1989;257:143-50. 39. RD Heparin Arthroplasty Group. RD heparin compared with warfarin
14. Beguin S, Mardiguian J, Lindhout T, Hemker HC. The mode of ac- for prevention of venous thromboembolic disease following total hip or
tion of low molecular weight heparin preparation (PK10169) and two of knee arthroplasty. J Bone Joint Surg Am 1994;76:1174-85.
its major components on thrombin generation in plasma. Thromb Hae- 40. Hamulyak K, Lensing AWA, van der Meer J, Smid WM, van Ooy A,
most 1989;61:30-4. Hoek JA. Subcutaneous low-molecular weight heparin or oral anticoagu-
15. Barrowcliffe TW, Merton RE, Havercroft SJ, Thunberg L, Lindahl U, lants for the prevention of deep-vein thrombosis in elective hip and knee
Thomas DP. Low-affinity heparin potentiates the action of high-affinity replacement? Thromb Haemost 1995;74:1428-31.
heparin oligosaccharides. Thromb Res 1984;34:125-33. 41. Imperiale TF, Speroff T. A meta-analysis of methods to prevent venous
16. Handeland GF, Abildgaard U, Holm HA, Arnesen K-E. Dose adjusted thromboembolism following total hip replacement. JAMA 1994;271:
heparin treatment of deep venous thrombosis: a comparison of unfraction- 1780-5. [Erratum, JAMA 1995;273:288.]
ated and low molecular weight heparin. Eur J Clin Pharmacol 1990;39: 42. Heit JA, Berkowitz SD, Bona R, et al. Efficacy and safety of low mo-
107-12. lecular weight heparin (ardeparin sodium) compared to warfarin for the
17. Bara L, Billaud E, Gramond G, Kher A, Samama M. Comparative prevention of venous thromboembolism after total knee replacement sur-
pharmacokinetics of a low molecular weight heparin (PK 10 169) and un- gery: a double-blind, dose-ranging study. Thromb Haemost 1997;77:32-8.
fractionated heparin after intravenous and subcutaneous administration. 43. Moskovitz PA, Ellenberg SS, Feffer HL, et al. Low-dose heparin for
Thromb Res 1985;39:631-6. prevention of venous thromboembolism in total hip arthroplasty and sur-
18. Bara L, Samama MM. Pharmacokinetics of low molecular weight hep- gical repair of hip fractures. J Bone Joint Surg Am 1978;60:1065-70.
arins. Acta Chir Scand Suppl 1988;543:65-72. 44. Barsotti J, Gruel Y, Rosset P, et al. Comparative double-blind study of
19. Matzsch T, Bergqvist D, Hedner U, Ostergaard P. Effects of an enzy- two dosage regimens of low-molecular weight heparin in elderly patients
matically depolymerized heparin as compared with conventional heparin in with a fracture of the neck of the femur. J Orthop Trauma 1990;4:371-5.
healthy volunteers. Thromb Haemost 1987;57:97-101. 45. Green D. Prophylaxis of thromboembolism in spinal cord-injured pa-
20. Bjornsson TO, Wolfram KM, Kitchell BB. Heparin kinetics deter- tients. Chest 1994;102:Suppl A:649S-651S.
mined by three assay methods. Clin Pharmacol Ther 1982;31:104-13. 46. Green D, Lee Y, Ito VY, et al. Fixed- vs adjusted-dose heparin in the
21. Lane DA. Heparin binding and neutralizing proteins. In: Lane DA, prophylaxis of thromboembolism in spinal cord injury. JAMA 1988;260:
Lindahl U, eds. Heparin: chemical and biological properties, clinical appli- 1255-8.
cations. Boca Raton, Fla.: CRC Press, 1989:363-91. 47. Merli GJ, Crabbe S, Doyle L, Ditunno JF, Herbision GJ. Mechanical
22. Sobel M, McNeill PM, Carlson PL, et al. Heparin inhibition of von plus pharmacological prophylaxis for deep vein thrombosis in acute spinal
Willebrand factor-dependent platelet function in vitro and in vivo. J Clin cord injury. Paraplegia 1992;30:558-62.
Invest 1991;87:1787-93. 48. Green D, Rossi EC, Yao JST, Flinn WR, Spies SM. Deep vein throm-
23. de Romeuf C, Mazurier C. Heparin binding assay of von Willebrand bosis in spinal cord injury: effect of prophylaxis with calf compression, as-
factor (vWF) in plasma milieu — evidence of the importance of the mul- pirin, and dipyridamole. Paraplegia 1982;20:227-34.
timerization degree of vWF. Thromb Haemost 1993;69:436-40. 49. Geerts WH, Jay RM, Code KI, et al. A comparison of low-dose hep-
24. Young E, Prins M, Levine MN, Hirsh J. Heparin binding to plasma arin with low-molecular-weight heparin as prophylaxis against venous
proteins, an important mechanism for heparin resistance. Thromb Hae- thromboembolism after major trauma. N Engl J Med 1996;335:701-7.
most 1992;67:639-43. 50. Bergmann JF, Neuhart E. A multicenter randomized double-blind
25. Hirsh J, van Aken WG, Gallus AS, Dollery CT, Cade JF, Yung WL. study of enoxaparin compared with unfractionated heparin in the preven-
Heparin kinetics in venous thrombosis and pulmonary embolism. Circula- tion of venous thromboembolic disease in elderly in-patients bedridden for
tion 1976;53:691-5. an acute medical illness. Thromb Haemost 1996;76:529-34.
26. Young E, Wells P, Holloway S, Weitz J, Hirsh J. Ex-vivo and in- 51. Harenberg J, Kallenbach B, Martin U, et al. Randomized controlled
vitro evidence that low molecular weight heparins exhibit less binding to study of heparin and low molecular weight heparin for prevention of deep-
plasma proteins than unfractionated heparin. Thromb Haemost 1994;71: vein thrombosis in medical patients. Thromb Res 1990;59:639-50.
300-4. 52. Leizorovicz A, Simonneau G, Decousus H, Boissel JP. Comparison of
27. Young E, Cosmi B, Weitz J, Hirsh J. Comparison of the non-specific efficacy and safety of low molecular weight heparins and unfractionated
binding of unfractionated heparin and low molecular weight heparin heparin in initial treatment of deep venous thrombosis: a meta-analysis.
(enoxaparin) to plasma proteins. Thromb Haemost 1993;70:625-30. BMJ 1994;309:299-304.
28. Barzu T, Molho P, Tobelem G, Petitou M, Caen J. Binding and en- 53. Siragusa S, Cosmi B, Piovella F, Hirsh J, Ginsberg JS. Low-molecular-
docytosis of heparin by human endothelial cells in culture. Biochem Bio- weight heparins and unfractionated heparin in the treatment of patients
phys Acta 1985;845:196-203. with acute venous thromboembolism: results of a meta-anlysis. Am J Med
29. Cadroy Y, Pourrat J, Baladre MF, et al. Delayed elimination of enox- 1996;100:269-77.
aparin in patients with chronic renal insufficiency. Thromb Res 1991;63: 54. Lensing AWA, Prins MH, Davidson BL, Hirsh J. Treatment of deep
385-90. venous thrombosis with low-molecular-weight heparins: a meta-analysis.
30. Carter CJ, Kelton JG, Hirsch J, Cerskus AL, Santos AV, Gent M. The Arch Intern Med 1995;155:601-7.
relationship between the hemorrhagic and antithrombotic properties of 55. Fragmin during Instability in Coronary Artery Disease (FRISC) Study
low molecular weight heparin in rabbits. Blood 1982;59:1239-45. Group. Low-molecular-weight heparin during instability in coronary artery
31. Salzman EW, Rosenberg RD, Smith MH, Lindon JN, Favreau L. Ef- disease. Lancet 1996;347:561-8.
fect of heparin and heparin fractions on platelet aggregation. J Clin Invest 56. Klein W. Low molecular weight heparin in the initial and prolonged
1980;65:64-73. treatment of unstable coronary artery disease — the fragmin in unstable
32. Horne MK III, Chao ES. The effect of molecular weight on heparin coronary heart disease study (FRIC). Eur Heart J 1996;17:Suppl:306. ab-
binding to platelets. Br J Haematol 1990;74:306-12. stract.
33. Blajchman MA, Young E, Ofosu FA. Effects of unfractionated heparin, 57. Cohen M, Demers C, Gurfinkel EP, et al. A comparison of low-molec-
dermatan sulfate and low molecular weight heparin on vessel wall perme- ular-weight heparin with unfractionated heparin for unstable coronary ar-
ability in rabbits. Ann N Y Acad Sci 1989;556:245-54. tery disese. N Engl J Med 1997;337:447-52.
34. Kakkar W, Cohen AT, Edmonson RA, et al. Low molecular weight 58. Collins R, Scrimgeour A, Yusuf S, Peto R. Reduction in fatal pulmo-
versus standard heparin for prevention of venous thromboembolism after nary embolism and venous thrombosis by perioperative administration of
major abdominal surgery. Lancet 1993;341:259-65. subcutaneous heparin: overview of results of randomized trials in general,
35. Boneu B. An international multicentre study: Clivarin in the preven- orthopedic, and urologic surgery. N Engl J Med 1988;318:1162-73.
tion of venous thromboembolism in patients undergoing general surgery: 59. Nurmohamed MT, Rosendaal FR, Buller HR, et al. Low-molecular-
report of the International Clivarin Assessment Group. Blood Coagul Fi- weight heparin versus standard heparin in general and orthopaedic surgery:
brinolysis 1993;4:Suppl 1:S21-S22. a meta-analysis. Lancet 1992;340:152-6.
36. Turpie AGG, Levine MN, Hirsh J, et al. A randomized controlled trial 60. Torholm C, Broeng L, Jorgensen PS, et al. Thromboprophylaxis by
of a low-molecular-weight heparin (enoxaparin) to prevent deep-vein low-molecular-weight heparin in elective hip surgery: a placebo controlled
thrombosis in patients undergoing elective hip surgery. N Engl J Med study. J Bone Joint Surg Br 1991;73:434-8.
1986;315:925-9. 61. The German Hip Arthroplasty Trial (GHAT) Group. Prevention of
37. Leclerc JR, Geerts WH, Desjardins L, et al. Prevention of venous throm- deep vein thrombosis with low molecular-weight heparin in patients un-

Vo l u m e 3 3 7 Nu m b e r 1 0  697

The New England Journal of Medicine

Downloaded from nejm.org at LINKOPING UNIVERSITY on September 8, 2013. For personal use only. No other uses without permission.
Copyright © 1997 Massachusetts Medical Society. All rights reserved.
 The New England Journal of Medicine

dergoing total hip replacement: a randomized trial. Arch Orthop Trauma 81. Pini M, Aiello S, Manotti C, et al. Low molecular weight heparin ver-
Surg 1992;111:110-120. sus warfarin in the prevention of recurrences after deep vein thrombosis.
62. Dechavanne M, Ville D, Berruyer M, et al. Randomized trial of a low- Thromb Haemost 1994;72:191-7.
molecular-weight heparin (Kabi 2165) versus adjusted-dose subcutaneous 82. Gurfinkel EP, Manos EJ, Mejail RI, et al. Low molecular weight hep-
standard heparin in the prophylaxis of deep-vein thrombosis after elective arin versus regular heparin or aspirin in the treatment of unstable angina
hip surgery. Haemostasis 1989;19:5-12. and silent ischemia. J Am Coll Cardiol 1995;26:313-8.
63. Hull R, Raskob G, Pineo G, et al. A comparison of subcutaneous low- 83. The Thrombolysis in Myocardial Infarction (TIMI) 11A Trial Investi-
molecular-weight heparin with warfarin sodium for prophylaxis against gators. Dose-ranging trial of enoxaparin for unstable angina: results of
deep-vein thrombosis after hip or knee implantation. N Engl J Med 1993; TIMI 11A. J Am Coll Cardiol 1997;29:1474-82.
329:1370-6. 84. Kay R, Wong KS, Yu YL, et al. Low-molecular-weight heparin for the
64. Powers PJ, Gent M, Jay RM, et al. A randomized trial of less intense treatment of acute ischemic stroke. N Engl J Med 1995;333:1588-93.
postoperative warfarin or aspirin therapy in the prevention of venous 85. Bergqvist D, Lindblad B, Matzsch T. Risk of combining low molecular
thromboembolism after surgery for fractured hip. Arch Intern Med 1989; weight heparin for thromboprophylaxis and epidural or spinal anesthesia.
149:771-4. Semin Thromb Hemost 1993;19:Suppl 1:147-51.
65. Waring WP, Karunas RS. Acute spinal cord injuries and the incidence 86. Spiro TE, Johnson GJ, Christie MJ, et al. Efficacy and safety of enox-
of clinically occurring thromboembolic disease. Paraplegia 1991;29:8-16. aparin to prevent deep venous thrombosis after hip replacement surgery.
66. Green D, Lee MY, Lim AC, et al. Prevention of thromboembolism af- Ann Intern Med 1994;121:81-9.
ter spinal cord injury using low-molecular-weight heparin. Ann Intern Med 87. Amiral J, Bridey F, Wolf M, et al. Antibodies to macromolecular plate-
1990;113:571-4. let factor 4-heparin complexes in heparin-induced thrombocytopenia: a
67. Frisbie JH, Sasahara AA. Low dose heparin prophylaxis for deep study of 44 cases. Thromb Haemost 1995;73:21-8.
venous thrombosis in acute spinal cord injury patients: a controlled study. 88. Leroy J, Leclerc MH, Delahousse B, et al. Treatment of heparin-asso-
Paraplegia 1981;19:343-6. ciated thrombocytopenia and thrombosis with low molecular weight hep-
68. Geerts WH, Code KI, Jay RM, Chen E, Szalai JP. A prospective study arin (CY 216). Semin Thromb Hemost 1985;11:326-9.
of venous thromboembolism after major trauma. N Engl J Med 1994;331: 89. Vitoux JF, Mathieu JF, Roncato M, Fiessinger JN, Aiach M. Heparin-
1601-6. associated thrombocytopenia treatment with low molecular weight hepa-
69. Clagett GP, Anderson FA Jr, Heit J, Levine MN, Wheeler HB. Preven- rin. Thromb Haemost 1986;55:37-9.
tion of venous thromboembolism. Chest 1995;108:Suppl:312S-334S. 90. Horellou MH, Conard J, Lecrubier C, et al. Persistent heparin in-
70. Prins MH, Gelsema R, Sing AK, van Heerde LR, den Ottolander GJ. duced thrombocytopenia despite therapy with low molecular weight hep-
Prophylaxis of deep venous thrombosis with a low-molecular-weight hep- arin. Thromb Haemost 1984;51:134.
arin (Kabi 2165/Fragmin) in stroke patients. Haemostasis 1989;19:245- 91. Holst J, Lindblad B, Bergqvist D, et al. Protamine neutralization of
50. intravenous and subcutaneous low-molecular-weight heparin (tinzaparin,
71. Sandset PM, Dahl T, Stiris M, Rostad B, Scheel B, Abildgaard U. A Logiparin): an experimental investigation in healthy volunteers. Blood Co-
double-blind and randomized placebo-controlled trial of low molecular agul Fibrinolysis 1994;5:795-803.
weight heparin once daily to prevent deep-vein thrombosis in acute ische- 92. Van Ryn-McKenna J, Cai L, Ofosu FA, Hirsh J, Buchanan MR. Neu-
mic stroke. Semin Thromb Hemost 1990;16:Suppl:25-33. tralization of enoxaparine-induced bleeding by protamine sulfate. Thromb
72. Turpie AGG, Levine MN, Powers P, et al. A double blind randomized Haemost 1990;63:271-4.
trial of ORG 10172 low molecular weight heparinoid versus unfractionated 93. Forestier F, Daffos F, Capella-Pavlovsky M. Low molecular weight hep-
heparin in the prevention of deep vein thrombosis in patients with throm- arin (PK 10169) does not cross the placenta during the second trimester
botic stroke. Thromb Haemost 1991;65:753. abstract. of pregnancy: study by direct fetal blood sampling under ultrasound.
73. Dahan R, Houlbert D, Caulin C, et al. Prevention of deep vein throm- Thromb Res 1984;34:557-60.
bosis in elderly medical in-patients by a low molecular weight heparin: a 94. Melissari E, Parker CJ, Wilson NV, et al. Use of low molecular weight
randomized double-blind trial. Hemostasis 1986;16:159-64. heparin in pregnancy. Thromb Haemost 1992;68:652-6.
74. Edmondson RA, Cohen AT, Das SK, Wagner MB, Kakkar VV. Low- 95. Monreal M, Lafoz E, Olive A, del Rio L, Vedia C. Comparison of sub-
molecular weight heparin versus aspirin and dipyridamole after femoropop- cutaneous unfractionated heparin with a low molecular weight heparin
liteal bypass grafting. Lancet 1994;344:914-8. [Erratum, Lancet 1994; (Fragmin) in patients with venous thromboembolism and contraindica-
344:1307.] tions to coumarin. Thromb Haemost 1994;71:7-11.
75. Faxon DP, Spiro TE, Minor S, et al. Low molecular weight heparin in 96. Fareed J, Walenga JM, Hoppensteadt D, Huan X, Racanelli A. Com-
prevention of restenosis after angioplasty: results of Enoxaparin Restenosis parative study on the in vitro and in vivo activities of seven low-molecular-
(ERA) Trial. Circulation 1994;90:908-14. weight heparins. Haemostasis 1988;18:Suppl 3:3-15. [Erratum, Haemosta-
76. Cairns JA, Gill J, Morton B, et al. Fish oils and low-molecular-weight sis 1988;18:389a.]
heparin for the reduction of restenosis after percutaneous transluminal 97. Anderson DR, O’Brien BJ, Levine MN, Roberts R, Wells PS, Hirsh J.
coronary angioplasty: the EMPAR Study. Circulation 1996;94:1553- Efficacy and cost of low-molecular-weight heparin compared with standard
60. heparin for the prevention of deep vein thrombosis after total hip arthro-
77. Hull RD, Raskob GE, Pineo GF, et al. Subcutaneous low-molecular- plasty. Ann Intern Med 1993;119:1105-12.
weight heparin compared with continuous intravenous heparin in the treat- 98. O’Brien BJ, Anderson DR, Goeree R. Cost-effectiveness of enoxaparin
ment of proximal-vein thrombosis. N Engl J Med 1992;326:975-82. versus warfarin prophylaxis against deep-vein thrombosis after total hip re-
78. Prandoni P, Lensing AWA, Buller HR, et al. Comparison of subcuta- placement. Can Med Assoc J 1994;150:1083-90.
neous low-molecular-weight heparin with intravenous standard heparin in 99. Menzin J, Colditz GA, Regan MM, Richner RE, Oster G. Cost-effec-
proximal deep-vein thrombosis. Lancet 1992;339:441-5. tiveness of enoxaparin vs low-dose warfarin in the prevention of deep-vein
79. The Columbus Investigators. Low-molecular-weight heparin in the thrombosis after total hip replacement surgery. Arch Intern Med 1995;
treatment of patients with venous thromboembolism. N Engl J Med 1997; 155:757-64.
337:657-62. 100. Weitz JI, Hudoba M, Massel D, Maraganore J, Hirsh J. Clot-bound
80. Simonneau G, Sors H, Charbonnier B, et al. A comparison of low- thrombin is protected from inhibition by heparin-antithrombin III but is
molecular-weight heparin with unfractionated heparin for acute pulmonary susceptible to inactivation by antithrombin III-independent inhibitors.
embolism. N Engl J Med 1997;337:663-9. J Clin Invest 1990;86:385-91.

698  Se p te m b e r 4 , 1 9 9 7

The New England Journal of Medicine

Downloaded from nejm.org at LINKOPING UNIVERSITY on September 8, 2013. For personal use only. No other uses without permission.
Copyright © 1997 Massachusetts Medical Society. All rights reserved.