|
Fondaparinux Compared with Enoxaparin for the
Prevention of Venous Thromboembolism after Hip-Fracture Surgery
Bengt I. Eriksson, M.D., Kenneth A. Bauer, M.D., Michael R.
Lassen, M.D., Alexander G.G. Turpie, F.R.C.P., for the Steering Committee of
the Pentasaccharide in Hip-Fracture Surgery Study
ABSTRACT
Background Surgery for
hip fracture carries a high risk of venous thromboembolism, despite
the use of current thromboprophylactic treatments. Fondaparinux, a
synthetic pentasaccharide, is a new antithrombotic agent that may
reduce this risk.
Methods In a
double-blind study, we randomly assigned 1711 consecutive patients
undergoing surgery for fracture of the upper third of the femur to
receive subcutaneous doses of either 2.5 mg of fondaparinux once
daily, initiated postoperatively, or 40 mg of enoxaparin once daily,
initiated preoperatively, for at least five days. The primary
efficacy outcome was venous thromboembolism up to postoperative day
11. Venous thromboembolism was defined as deep-vein thrombosis
detected by mandatory bilateral venography, documented symptomatic
deep-vein thrombosis, or documented symptomatic pulmonary embolism.
The main safety outcomes were major bleeding and mortality from all
causes. The duration of follow-up was six weeks.
Results The
incidence of venous thromboembolism by day 11 was 8.3 percent (52 of
626 patients) in the fondaparinux group and 19.1 percent (119 of 624
patients) in the enoxaparin group (P<0.001). The reduction in
risk with fondaparinux was 56.4 percent (95 percent confidence
interval, 39.0 to 70.3 percent). There were no significant
differences between the two groups in the incidence of death or
clinically relevant bleeding.
Conclusions In patients
undergoing surgery for hip fracture, fondaparinux was more effective
than enoxaparin in preventing venous thromboembolism and was equally
safe.
Patients undergoing surgery for hip
fracture are in the highest category of risk for postoperative
venous thromboembolism.1,2
Fatal pulmonary embolism occurs in 3.6 to 12.9 percent of patients who
have not received prophylaxis against thromboembolism.1
There are few data on thromboprophylaxis after surgery for hip fracture,
and recommendations are based mainly on expert opinion.1
Even with current methods of thromboprophylaxis, the incidence of
venographically confirmed deep-vein thrombosis is 24 to 34 percent.1,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21
Fondaparinux is a new synthetic pentasaccharide that causes selective
inhibition of activated factor X (factor Xa).22,23,24,25
A recent study of patients undergoing major orthopedic procedures suggested
that a once-daily subcutaneous injection of fondaparinux reduces the
risk of venous thromboembolism more than does low-molecular-weight heparin.26
We conducted a multicenter, randomized, double-blind trial to
compare two types of thromboprophylaxis after hip-fracture surgery: a
once-daily subcutaneous injection of fondaparinux, initiated postoperatively,
and a once-daily subcutaneous injection of enoxaparin, initiated
preoperatively.
Methods
Patients
Patients were considered for inclusion if they were at least 18
years of age and were scheduled to undergo standard surgery for
fracture of the upper third of the femur, including the femoral head
and neck, within 48 hours after admission.
The main reasons for exclusion were multiple trauma affecting
more than one organ system; an interval of more than 24 hours between
the injury and hospital admission; pregnancy; active bleeding; a
documented congenital or acquired bleeding disorder; current
ulcerative or angiodysplastic gastrointestinal disease; a history of
hemorrhagic stroke or brain, spinal, or ophthalmologic surgery
within the previous three months; planned use of an indwelling
intrathecal or epidural catheter for more than six hours after
surgery; hypersensitivity to heparin, low-molecular-weight heparins,
porcine products, or iodinated contrast medium; a contraindication
to anticoagulant therapy; a current addictive disorder; a serum
creatinine concentration above 2 mg per deciliter (177 µmol per
liter) in a well-hydrated patient; and a platelet count below
100,000 per cubic millimeter. Patients who required anticoagulant
therapy or received dextran or any type of anticoagulant or
fibrinolytic therapy from admission to the time of first
administration of the study drug or surgery were also excluded.
Study Design
Within 24 hours after admission and before surgery, patients were
randomly assigned to treatment groups in blocks of four, with
stratification according to center, with the use of a computer-generated randomization
list. Patients were assigned to receive once-daily subcutaneous
injections of either 2.5 mg of fondaparinux (Arixtra, Sanofi–Synthelabo,
Paris, and NV Organon, Oss, the Netherlands) and a placebo or 40 mg
of enoxaparin (Clexane/Lovenox, Aventis Pharmaceuticals,
Bridgewater, N.J.) and a placebo. In the enoxaparin group, the first
active dose was given 12±2 hours preoperatively and the second 12 to
24 hours postoperatively, according to the recommendation of the
manufacturer. Since fondaparinux is a new compound, which differs
from enoxaparin in its mechanism of action and pharmacokinetic
properties, the starting time after surgery and the dose were
determined during the early development of the drug26;
the first dose of fondaparinux was administered 6±2 hours
postoperatively and the second 12 hours or more after the first.
However, if surgery was delayed until 24 to 48 hours after
admission, administration of fondaparinux was initiated 12±2 hours
before surgery. In both groups, omission of preoperative injections
was recommended if spinal or epidural anesthesia or catheterization
was planned, and any indwelling intrathecal or epidural catheter was
to be removed at least two hours before the first postoperative
injection.
Day 1 was defined as the day of surgery. Treatment was scheduled
to continue until day 5 to day 9, and the primary efficacy outcome was
assessed between day 5 and day 11. Patients were then followed up in
person, by mail, or by telephone between day 35 and day 49. During
follow-up, patients were instructed to report any symptoms or signs
of venous thromboembolism or bleeding and any other clinical event
occurring since the completion of treatment. Investigators could
extend prophylaxis during follow-up with any currently available
therapy, but only after venography had been performed. If venous
thromboembolism occurred during the study, treatment was left to the
discretion of the investigators.
The study was conducted according to the ethical principles stated
in the Declaration of Helsinki and local regulations. The protocol
was approved by independent ethics committees, and written informed
consent was obtained from all patients before randomization.
Medications
Study medications were packaged in boxes of identical appearance,
each containing 10 prefilled, single-dose syringes of active treatment
and 10 prefilled, single-dose syringes of matching placebos. Each
syringe contained either 2.5 mg of fondaparinux sodium in 0.25 ml of
water for injectable preparations (a concentration of 10 mg per
milliliter), 40 mg of enoxaparin sodium in 0.4 ml of water for
injectable preparations (a concentration of 100 mg per milliliter),
or placebo (0.25 or 0.4 ml of isotonic saline).
Throughout the treatment period, the use of intermittent pneumatic
compression, dextran, and thrombolytic, anticoagulant, or antiplatelet
agents was prohibited. Centers were advised to avoid giving patients
aspirin or nonsteroidal antiinflammatory drugs whenever possible.
The use of graduated compression stockings and physiotherapy was
recommended.
Outcome Measures
The primary efficacy outcome was assessed by the rate of venous
thromboembolism (defined as deep-vein thrombosis, pulmonary embolism,
or both) up to day 11. Secondary efficacy outcomes were total,
proximal, or distal deep-vein thrombosis or symptomatic venous
thromboembolism up to day 11 and symptomatic venous thromboembolism up
to day 49. Patients were examined for deep-vein thrombosis by
systematic bilateral ascending venography of the legs27
between day 5 and day 11, but no more than two days after the last
dose of study drug, or earlier if thrombosis was clinically
suspected. Symptomatic pulmonary embolism was confirmed by a lung
scan indicating a high probability of pulmonary embolism, by
pulmonary angiography,28
by helical computed tomography, or at autopsy.
The primary safety outcome was the incidence of major bleeding,
which included fatal bleeding; bleeding that was retroperitoneal, intracranial,
or intraspinal or that involved any other critical organ; bleeding
leading to reoperation; and overt bleeding with a bleeding index of
2 or more. The bleeding index was calculated as the number of units
of packed red cells or whole blood transfused plus the hemoglobin
values before the bleeding episode minus the hemoglobin values after
the episode (in grams per deciliter). Secondary safety outcomes were
death, minor bleeding, a need for transfusion, thrombocytopenia, and
any other adverse event. Minor bleeding was defined as overt
bleeding that did not meet the criteria for major bleeding.
Efficacy and safety outcomes were adjudicated by a central
independent committee whose members were unaware of the treatment
assignments and included review of all venograms and reports of
bleeding and death.
Statistical Analysis
Assuming an incidence of venous thromboembolism by day 11 of 22
percent in the enoxaparin group7 and
a risk reduction of about 30 percent (i.e., an incidence of 15
percent in the fondaparinux group), 600 patients were needed per
group to provide the study with a power of 85 percent. The target
number of recruited patients was 1700, a number that allowed for
failure to obtain primary efficacy data in approximately 30 percent
of patients.
The analysis of the primary efficacy outcome included data on
all patients who had received at least one dose of study medication,
had undergone the appropriate surgery, and had had an adequate assessment
for venous thromboembolism by day 11. The analysis of safety
included data on patients who had received at least one dose of
study medication.
A two-tailed P value of less than 0.05 was considered to indicate
statistical significance. The analysis of the primary efficacy outcome
was performed with the use of a two-sided Fisher's exact test. Exact
95 percent confidence intervals for the absolute difference between
fondaparinux and enoxaparin and the risk ratio were calculated. The
treatment effect was also analyzed according to predefined
categorical covariates with use of a logistic-regression model.
The study was supervised by a steering committee of 10 people,
which included 6 representatives of the sponsor (Sanofi–Synthelabo and
NV Organon). The committee designed the study, interpreted the data,
and wrote the article. The final statistical analysis was performed
by the sponsor. The central adjudication committee and the
data-monitoring committee operated independently of the sponsor. One
planned interim analysis was conducted when half the projected patient
population had been enrolled, for reestimation of the sample size,
since the rate of venous thromboembolism in patients undergoing
hip-fracture surgery was uncertain. Simulations demonstrated that
the predefined procedure did not inflate the type I error. No change
in the sample size was found to be necessary, and the study
continued as planned.
Results
Study Population
Between November 1998 and October 1999, 1711 patients were
enrolled in 99 centers in 21 countries (listed in the Appendix).
Thirty-eight patients did not receive either study drug (Table 1). Two
patients did not undergo the appropriate surgery, and primary
efficacy had not been assessed by day 11 in 421 patients. Thus, 1250
patients (73.1 percent) were included in the primary efficacy analysis,
a percentage in line with other large multicenter studies that used
venography after orthopedic surgery.29,30,31
The characteristics of patients excluded from the primary efficacy analysis
did not differ from those of patients included in the analysis (data
not shown).
|
Base-line characteristics did not differ significantly between the
two groups of patients included in the analysis of safety (Table 2) or
primary efficacy (data not shown). A total of 551 and 569 patients
underwent surgery within 24 hours after admission in the
fondaparinux and enoxaparin groups, respectively. Among the 626
patients in the fondaparinux group who were included in the primary
efficacy analysis, fondaparinux was given preoperatively to 68 (10.9
percent) because surgery was delayed until 24 to 48 hours after
admission; enoxaparin was given postoperatively, rather than
preoperatively, to 464 of 624 patients assigned to that drug (74.4
percent) because of very early surgery after admission or planned
regional anesthesia. The median time between surgery and the
assessment of primary efficacy was eight days in both groups; most
patients were assessed between day 5 and day 11 as planned. The two
groups did not differ significantly with regard to the last day of
active treatment or the use of concomitant treatments up to day 11 (Table 3).
|
|
Overall, 829 patients treated with fondaparinux and 840 patients treated
with enoxaparin returned for the follow-up visit on day 49. The
duration of follow-up was similar between the two groups. During
follow-up of patients who did not receive treatment for an acute
thromboembolic event, 58.5 percent of patients treated with
fondaparinux and 55.8 percent of patients treated with enoxaparin
received prolonged thromboprophylaxis, primarily with a preparation
of heparin or a vitamin K antagonist, after the study treatment.
Incidence of Venous Thromboembolism
The incidence of venous thromboembolism by day 11 was 8.3 percent
in the fondaparinux group (52 of 626 patients) and 19.1 percent in
the enoxaparin group (119 of 624 patients). This was a decrease of
10.8 percentage points, or a relative reduction in risk of 56.4
percent (95 percent confidence interval, 39.0 to 70.3 percent; P<0.001)
(Table 4). A
similar result was found in sensitivity analyses when patients who
had had no primary efficacy assessment by day 11 were included in
the primary efficacy analysis (data not shown). The incidence of
total, proximal, and distal-only deep-vein thrombosis was
significantly lower in the fondaparinux group (P<0.001 for all
three comparisons). The incidence of symptomatic venous
thromboembolism was low (6.5 percent), with no difference between
the two groups.
|
The superior efficacy of fondaparinux over enoxaparin was found when
patients were grouped according to age, sex, body-mass index (the
weight in kilograms divided by the square of the height in meters
[<30 vs. 30]), type of anesthesia
(general, regional, or both), type of hip fracture (cervical,
trochanteric, or subtrochanteric), type of surgery (implantation of
half prosthesis, implantation of total prosthesis, or
osteosynthesis), the use or nonuse of cement, or whether or not the
patient had had previous venous thromboembolism (data not shown).
The number of patients treated by participating physicians for a
venous thromboembolic event by day 11 was significantly lower in the
fondaparinux group (6.1 percent [43 of 702]) than in the enoxaparin
group (11.7 percent [84 of 716], P<0.001).
By day 49, the incidence of symptomatic venous thromboembolism
was similar in the fondaparinux group (2.0 percent [17 of 831 patients])
and the enoxaparin group (1.5 percent [13 of 840 patients]). Fatal
pulmonary embolism occurred in 8 of 831 patients in the fondaparinux
group and 7 of 840 patients in the enoxaparin group; nonfatal
pulmonary embolism occurred in 3 of 831 patients and 4 of 840
patients, respectively.
Safety Outcomes
Major bleeding occurred by day 11 in 18 of 831 patients treated
with fondaparinux and 19 of 842 patients treated with enoxaparin (P=1.00)
(Table 5).
Most of these episodes occurred at the surgical site (14 of 18 patients
in the fondaparinux group and 14 of 19 patients in the enoxaparin
group). Minor bleeding occurred more often in the fondaparinux group
(P=0.02). By day 49, three patients in the fondaparinux group and
six patients in the enoxaparin group underwent reoperation because
of bleeding. Transfusion requirements and the incidence of other
adverse events during treatment or follow-up did not differ
significantly between groups. The platelet count was lower than
100,000 per cubic millimeter in 40 of 822 patients in the
fondaparinux group (4.9 percent) and 44 of 831 patients in the
enoxaparin group (5.3 percent). No episode of decreased platelet
count was reported as a serious adverse event in either group. The
incidence of wound infection was low and was the same in both groups
(0.7 percent [6 of 831 in the fondaparinux group and 6 of 842 in
the enoxaparin group]). By day 49, 38 patients in the fondaparinux group
(4.6 percent) and 42 in the enoxaparin group (5.0 percent) had died.
|
Discussion
This large study demonstrates that fondaparinux is significantly
more effective than enoxaparin in preventing postoperative venous thromboembolism
after surgery for hip fracture. The 19.1 percent incidence of venous
thromboembolism in the enoxaparin group by day 11 is consistent with
the results of previous studies of enoxaparin after hip-fracture
surgery.7,8,9,10
By contrast, 8.3 percent of patients given fondaparinux had
postoperative venous thromboembolism. Moreover, proximal deep-vein
thrombosis, which is prone to embolize, occurred in 6 of 650
patients in the fondaparinux group and 28 of 646 patients in the
enoxaparin group (P<0.001).32,33,34
Three other large studies in patients undergoing elective knee35
or hip-replacement36,37
surgery also showed the superiority of fondaparinux over enoxaparin
in preventing venous thromboembolism. The efficacy of fondaparinux
may be attributed to its ability to inhibit factor Xa rapidly and
selectively, its predictable linear pharmacokinetics, and its
relatively long half-life, which permits the drug to achieve an
antithrombotic effect for 24 hours.
Physicians have been uncertain about effective and safe
thromboprophylaxis after hip-fracture surgery.1,2
Warfarin is moderately effective,1,15,16,17,18,19
and aspirin is not recommended in patients undergoing such surgery.1,20,38,39
Promising results have been reported in small studies of 40 mg
of enoxaparin administered once daily, with treatment initiated preoperatively.7,8,9,10
In our study, because of planned regional anesthesia, early surgery
after admission, or both, only 25.6 percent of patients received the
preoperative injection of enoxaparin. This indicates the difficulty
of administering low-molecular-weight heparin preoperatively in
emergency situations.
In our study, symptomatic events were rare during the treatment
period, with a 0.2 percent incidence of fatal pulmonary embolism —
similar to that reported in the large Pulmonary Embolism Prevention
trial.20
However, the incidence of symptomatic events in our study should be interpreted
with caution. Early detection by venographic screening and prolonged
prophylaxis in nearly 60 percent of our patients probably prevented
symptomatic venous thromboembolism. The incidence of fatal pulmonary
embolism by day 49 was nevertheless nearly 1.0 percent in both
groups. The duration of treatment may have been too short for some
patients who were still at risk for venous thromboembolism when
treatment was discontinued.
Our study demonstrates that prophylactic fondaparinux is more
effective than enoxaparin in preventing venous thromboembolism in
patients undergoing hip-fracture surgery and does not increase the
risk of clinically relevant bleeding.
Supported by NV Organon and Sanofi–Synthelabo. All authors have
served as consultants to NV Organon and Sanofi–Synthelabo.
Presented in abstract form at the 42nd Annual Meeting of the American
Society of Hematology, San Francisco, December 1–5, 2000 (Blood
2000;9:490A, A2110).
* Participants in the study are listed in
the Appendix.
Source Information
From the Department of Orthopedics, Sahlgrenska University Hospital–Östra,
Göteborg, Sweden (B.I.E.); the Department of Medicine, Veterans Affairs Boston
Healthcare System and Beth Israel Deaconess Medical Center, Boston (K.A.B.);
the Department of Orthopedics, Hillerųd University, Hillerųd, Denmark (M.R.L.);
and the Department of Medicine, Hamilton Health Sciences Corporation–General
Division, Hamilton, Ont., Canada (A.G.G.T.).
Address reprint requests to Dr. Eriksson at the Orthopedics
Department, Sahlgrenska University Hospital–Östra, S-41685 Göteborg, Sweden, or
at [log in to unmask].
References
Fondaparinux Compared with Enoxaparin for the
Prevention of Venous Thromboembolism after Elective Major Knee Surgery
Kenneth A. Bauer, M.D., Bengt I. Eriksson, M.D., Michael R.
Lassen, M.D., Alexander G.G. Turpie, F.R.C.P., for the Steering Committee of the
Pentasaccharide in Major Knee Surgery Study
ABSTRACT
Background Despite
thromboprophylaxis, major knee surgery carries a high risk of venous
thromboembolism. Fondaparinux, the first of a new class of synthetic
antithrombotic agents, may reduce this risk.
Methods In a
double-blind study, we randomly assigned 1049 consecutive patients
undergoing elective major knee surgery to receive subcutaneous doses
of either 2.5 mg of fondaparinux once daily or 30 mg of enoxaparin
twice daily, with both treatments initiated postoperatively. The
primary efficacy outcome was venous thromboembolism up to postoperative
day 11, defined as deep-vein thrombosis detected by mandatory
bilateral venography, documented symptomatic deep-vein thrombosis,
or documented symptomatic pulmonary embolism. The primary safety
outcome was major bleeding.
Results The primary
efficacy outcome was assessed in 724 patients. The fondaparinux
group had a significantly lower incidence of venous thromboembolism
by day 11 (12.5 percent [45 of 361 patients]) than the enoxaparin
group (27.8 percent [101 of 363 patients]; reduction in risk, 55.2
percent; 95 percent confidence interval, 36.2 to 70.2; P<0.001).
Major bleeding (including overt bleeding with a bleeding index of 2
or more) occurred more frequently in the fondaparinux group
(P=0.006), but there were no significant differences between the two
groups in the incidence of bleeding leading to death or reoperation
or occurring in a critical organ.
Conclusions In patients
undergoing elective major knee surgery, postoperative treatment with
2.5 mg of fondaparinux once daily was significantly more effective
in preventing deep-vein thrombosis than 30 mg of enoxaparin twice
daily.
Venous thromboembolism is a frequent,
life-threatening postoperative complication of
total-knee-replacement surgery.1,2
Without thromboprophylaxis, the prevalence rate is 40 to 84 percent
for venographically verified postoperative deep-vein thrombosis and
2 to 7 percent for pulmonary embolism.1
Thromboprophylaxis that is effective in hip-replacement surgery,
such as low-dose heparin, low-molecular-weight heparin, or warfarin,
is less successful in knee-replacement surgery and reduces the
prevalence of deep-vein thrombosis only to 31 to 47 percent.1 For
this reason, more effective antithrombotic prophylaxis is needed in
knee-replacement surgery.
Fondaparinux is one of a new class of antithrombotic agents, the
selective inhibitors of activated factor X (factor Xa).3,4,5,6
Fondaparinux is an entirely synthetic pentasaccharide that is structurally
related to the antithrombin-binding site of heparin. In contrast to
heparin, which interacts with many plasma components, the
pentasaccharide selectively binds to antithrombin, causing it to
rapidly inhibit factor Xa, a key enzyme in the coagulation pathway.
Recent dose-ranging studies suggested that a once-daily subcutaneous
injection of 2.5 mg of fondaparinux can prevent venous
thromboembolism after hip-replacement7 or
knee-replacement surgery (unpublished data).
This multicenter, randomized, double-blind trial was part of a
program that also evaluated fondaparinux as prophylaxis against venous
thromboembolism in patients undergoing surgery for hip fracture8 and
elective hip replacement.9,10
The aim of the study was to compare the efficacy and safety of a
once-daily subcutaneous injection of 2.5 mg of pentasaccharide with
twice-daily subcutaneous injections of 30 mg of enoxaparin for the
prevention of venous thromboembolism after elective major knee
surgery.
Methods
Patients
Patients were considered for inclusion if they were at least 18
years of age and were undergoing elective major knee surgery — that
is, surgery requiring resection of the distal end of the femur or
proximal end of the tibia or revision of at least one component of a
previously implanted total-knee prosthesis.
Patients were excluded if surgery in the contralateral knee was
performed at the same time or within two weeks after enrollment. Women
were excluded if they were pregnant or not using effective contraception.
Other main reasons for exclusion were active bleeding; a documented
congenital or acquired bleeding disorder; current ulcerative or
angiodysplastic gastrointestinal disease; hemorrhagic stroke or
brain, spinal, or ophthalmologic surgery within the previous three
months; insertion of an indwelling intrathecal or epidural catheter
during the treatment period; unusual difficulty in administering
epidural or spinal anesthesia (e.g., more than two attempts);
hypersensitivity to heparin, low-molecular-weight heparins, porcine
products, or iodinated contrast medium; a contraindication to
anticoagulant therapy; a current addictive disorder; a serum
creatinine concentration above 2 mg per deciliter (177 µmol per
liter) in a well-hydrated patient; and a platelet count below
100,000 per cubic millimeter. Finally, patients who required
anticoagulant therapy were also excluded. The use within one week
before randomization of dextran or any type of anticoagulant,
fibrinolytic, or antiplatelet agent was discouraged.
Study Design
Immediately after surgery, patients were randomly assigned (in
a ratio of 1:1 in blocks of four, stratified according to center), through
a central computer-derived randomization scheme to receive subcutaneous
doses of either 2.5 mg of fondaparinux (Arixtra, NV Organon, Oss,
the Netherlands, and Sanofi–Synthelabo, Paris) once daily and a
placebo once daily or 30 mg of enoxaparin (Clexane/Lovenox, Aventis
Pharmaceuticals, Bridgewater, N.J.) twice daily. In the enoxaparin
group, the first dose was given between 12 and 24 hours after
surgery, according to the recommendation of the manufacturer. Since
fondaparinux is a new compound, which differs from enoxaparin in its
mechanism of action and pharmacokinetic properties, the starting
time after surgery and the dose were determined during the early
development of the drug7;
the first postoperative injection was administered 6±2 hours after
surgery and the second injection 12 hours or more after the first.
The day of surgery was defined as day 1. Treatment was scheduled
to continue until day 5 to day 9, and the primary efficacy outcome was
assessed between day 5 and day 11. Patients were then followed up in
person, by mail, or by telephone between day 35 and day 49. During
follow-up, patients were instructed to report any symptoms or signs
of venous thromboembolism or bleeding and any other clinical event
occurring since the completion of treatment. Investigators could
extend prophylaxis during follow-up with any currently available
therapy, but only after venography had been performed. If venous
thromboembolism occurred during the study, treatment was left to the
discretion of the investigator.
The study was conducted according to the ethical principles stated
in the Declaration of Helsinki and local regulations. The protocol
was approved by independent local institutional review boards, and
written informed consent was obtained from all patients before
randomization.
Medications
Study medications were packaged in boxes of identical appearance,
each containing 19 prefilled, single-dose syringes: 10 syringes of
fondaparinux and 9 syringes of placebo for each patient assigned to
fondaparinux, or 18 syringes of enoxaparin and 1 syringe of placebo
for each patient assigned to enoxaparin. Each syringe contained 2.5
mg of fondaparinux sodium in 0.25 ml of water for injectable
preparations (a concentration of 10 mg per milliliter), 30 mg of
enoxaparin sodium in 0.3 ml of water for injectable preparations (a
concentration of 100 mg per milliliter), or placebo (0.25 or 0.3 ml
of isotonic saline). Each syringe was loaded inside an opaque
autoinjector (Autoject, Owen Mumford, Woodstock, United Kingdom) to
maintain blinding.
Throughout the treatment period, the use of intermittent pneumatic
compression, dextran, and any other anticoagulant, thrombolytic, or
antiplatelet agent was prohibited. Centers were advised to avoid
giving patients aspirin or nonsteroidal antiinflammatory drugs
whenever possible. The use of graduated-compression stockings and
physiotherapy was recommended.
Outcome Measures
The primary efficacy outcome was assessed by the rate of venous
thromboembolism (defined as deep-vein thrombosis, pulmonary embolism,
or both) up to day 11. Secondary efficacy outcomes were total,
proximal, or distal deep-vein thrombosis or symptomatic venous
thromboembolism up to day 11 and symptomatic venous thromboembolism up
to day 49. Patients were examined for deep-vein thrombosis by
systematic bilateral ascending venography of the legs11
between day 5 and day 11, but no more than two days after the last
injection of study drug, or earlier if thrombosis was clinically
suspected. Symptomatic pulmonary embolism was confirmed by a lung
scan indicating a high probability of pulmonary embolism, pulmonary
angiography,12
or helical computed tomography or at autopsy.
The primary safety outcome was the incidence of major bleeding,
which included fatal bleeding; bleeding that was retroperitoneal, intracranial,
or intraspinal or that involved any other critical organ; bleeding
leading to reoperation; and overt bleeding with a bleeding index of
2 or more. The bleeding index was calculated as the number of units
of packed red cells or whole blood transfused plus the hemoglobin
values before the bleeding episode minus the hemoglobin values after
the episode (in grams per deciliter). Secondary safety outcomes were
death, other bleeding, a need for transfusion, thrombocytopenia, and
any other adverse event.
Efficacy and safety outcomes were adjudicated by a central
independent committee whose members were unaware of the treatment
assignments and included reviews of all venograms and reports of
bleeding and death.
Statistical Analysis
Assuming an incidence of venous thromboembolism by day 11 of 34
percent in the enoxaparin group and a risk reduction of about 30
percent (i.e., an incidence of 24 percent in the fondaparinux group),
319 patients were needed in each group (for a total of 638 patients)
to provide the study with a power of 85 percent. The target number
of recruited patients was 912, a number that allowed for failure to
obtain primary efficacy data in up to 30 percent of patients.
The analysis of the primary efficacy outcome included data on
all patients who had received at least one dose of study medication,
had undergone the appropriate surgery, and had had an adequate assessment
for venous thromboembolism by day 11. The analysis of safety
included data on patients who had received at least one dose of
study medication.
A two-tailed P value of less than 0.05 was considered to indicate
statistical significance. The analysis of the primary efficacy outcome
was performed with the use of a two-sided Fisher's exact test. Exact
95 percent confidence intervals for the absolute difference between
fondaparinux and enoxaparin and the risk ratio were calculated. The
treatment effect was also analyzed according to predefined
categorical covariates with the use of a logistic-regression model.
The study was supervised by a steering committee of 11 people,
which included 7 representatives of the sponsors (NV Organon and
Sanofi–Synthelabo). The committee designed the study, interpreted
the data, and wrote the article. The final statistical analysis was
performed by the sponsor. The central adjudication committee and the
data monitoring committee operated independently of the sponsor. One
planned interim analysis was conducted by an independent statistical
center when half the projected patient population had been enrolled,
for reestimation of the sample size, since the rate of venous
thromboembolism in patients undergoing knee surgery was uncertain. Simulations
demonstrated that the predefined procedure did not inflate the type
I error. No change in the sample size was found to be necessary, and
the study continued as planned.
Results
Study Populations
Between December 1998 and January 2000, 1049 patients were
enrolled in 64 centers in North America. Fifteen patients did not
receive any study drug, leaving 1034 available for the safety
analysis (Table 1);
primary efficacy had not been assessed by day 11 in 310 patients.
Thus, 724 patients (69.0 percent) were included in the primary
efficacy analysis, a percentage in line with other large multicenter
trials that used venography after orthopedic surgery.13,14,15,16
The characteristics of patients excluded from the primary efficacy
analysis did not differ from those of patients included in the
analysis (data not shown).
|
Base-line characteristics did not differ significantly between the
two groups of patients included in the analysis of safety (Table 2) or
primary efficacy (data not shown). Among patients analyzed for
primary efficacy, the median time between surgery and the qualifying
examination for venous thromboembolism was seven days in both
groups; most patients underwent this examination between day 5 and
day 11, as planned. The two groups did not differ with regard to the
last day of active treatment or the use of concomitant treatments up
to day 11 (Table
3).
|
|
Overall, 514 patients treated with fondaparinux and 511 patients treated
with enoxaparin returned for the follow-up visit on day 49. The
duration of follow-up was similar in the two groups. During
follow-up of patients who were not treated for an acute thromboembolic
event, 19.1 percent of patients assigned to fondaparinux (86 of 450)
and 20.2 percent of patients assigned to enoxaparin (82 of 406) received
prolonged thromboprophylaxis, primarily with a preparation of
heparin or a vitamin K antagonist.
Incidence of Venous Thromboembolism
The incidence of venous thromboembolism by day 11 was 27.8 percent
(101 of 363 patients) in the enoxaparin group and 12.5 percent (45
of 361 patients) in the fondaparinux group (reduction in risk, 55.2
percent; 95 percent confidence interval, 36.2 to 70.2 percent;
P<0.001) (Table
4). A similar result was found in sensitivity analyses when
patients who had had no primary efficacy assessment by day 11 were
included in the primary efficacy analysis (data not shown). As
compared with enoxaparin, fondaparinux reduced the incidence of
proximal deep-vein thrombosis by 54.5 percent (P=0.06) and distal
deep-vein thrombosis by 55.9 percent (P<0.001). The incidence of
symptomatic venous thromboembolism was low and did not differ
significantly between the two groups (Table 4).
Overall, the superiority of fondaparinux over enoxaparin with
respect to primary efficacy was consistent according to age, sex,
body-mass index (the weight in kilograms divided by the square of
the height in meters [<30 vs. 30]),
type of anesthesia (general, regional, or both), type of surgery
(primary or revision), use or nonuse of cement, and whether or not
patients had had previous venous thromboembolism (data not shown).
The number of patients treated by participating physicians for a
venous thromboembolic event by day 11 was significantly lower in
the fondaparinux group (15.1 percent [67 of 443]) than in the
enoxaparin group (25.1 percent [111 of 442], P<0.001).
|
By day 49, the incidence of symptomatic venous thromboembolism did
not differ significantly between the fondaparinux group (1.0 percent
[5 of 517 patients]) and the enoxaparin group (1.9 percent [10 of
517 patients]). Fatal pulmonary embolism occurred in one patient in
each group; nonfatal pulmonary embolism occurred in two patients in
the fondaparinux group and four in the enoxaparin group.
Safety Outcomes
There were no instances of fatal bleeding or bleeding in a
critical organ in either treatment group; bleeding requiring
reoperation occurred in two patients in the fondaparinux group and
one in the enoxaparin group. In all three patients, drainage of a
knee effusion was performed. In the fondaparinux group, nine
episodes of overt bleeding were associated with a bleeding index of
2 or more; seven occurred at the surgical site, and only three led
to discontinuation of study treatment. The total for the primary
safety outcome was therefore 11 major bleeding episodes in the
fondaparinux group and 1 in the enoxaparin group (P=0.006) (Table 5). The
incidence of minor bleeding, a need for transfusion, and other
adverse events during treatment or follow-up did not differ
significantly between the two groups. Platelet counts lower than
100,000 per cubic millimeter were measured in 14 patients (2.7
percent) in the fondaparinux group as compared with 19 (3.7 percent)
in the enoxaparin group (P=0.27). No episode of a decreased platelet
count was reported as a serious adverse event in either group. By
day 49, two patients in the fondaparinux group (0.4 percent) and
three in the enoxaparin group (0.6 percent) had died from causes
unrelated to the treatment.
|
Discussion
This study demonstrates that fondaparinux is significantly more
effective than enoxaparin in preventing venous thromboembolism after
elective major knee surgery. Deep-vein thrombosis has been more
difficult to prevent with anticoagulation therapy after knee surgery
than after total hip replacement, even with low-molecular-weight
heparin, the most effective thromboprophylactic therapy to date.1 The
27.8 percent incidence of venous thromboembolic events in the
enoxaparin group by day 11 is consistent with the incidence of 19.0
to 25.0 percent in other trials of enoxaparin after knee surgery.16,17,18,19
In our study, the reduction to 12.5 percent in the fondaparinux
group is consistent with the results of three other large studies in
patients undergoing surgery for hip fracture8 or
elective hip replacement.9,10
The superior efficacy of fondaparinux may be related to its ability
to initiate selective inhibition of factor Xa, its predictable linear
pharmacokinetics, the choice of dose, and the starting time after
surgery.
The low incidence of symptomatic events in our study should be
interpreted with caution, as it is likely to be lower than would be
observed in typical clinical situations. As in other trials of
thromboprophylaxis, most of our asymptomatic patients with positive
venograms had been receiving an anticoagulant at a therapeutic dose.
Moreover, about 20 percent of our patients were receiving
prophylaxis after the study treatment period ended. Both of these
factors may have prevented symptomatic venous thrombosis.
Major bleeding was significantly more frequent in the fondaparinux
group (11 patients, including 9 with a bleeding index of 2 or more)
than in the enoxaparin group (1 patient). Administration of
fondaparinux was continued in six of the nine patients with a
bleeding index of 2 or more. Nevertheless, the two groups did not
differ significantly with respect to fatal bleeding, bleeding in
critical organs, or bleeding leading to reoperation.
Supported by NV Organon and Sanofi–Synthelabo. All authors have
served as consultants to NV Organon and Sanofi–Synthelabo.
Presented in abstract form at the 42nd Annual Meeting of the American
Society of Hematology, San Francisco, December 1–5, 2000 (Blood
2000;9:491A, A2111).
* Participants in the study are listed in
the Appendix.
Source Information
From the Department of Medicine, Veterans Affairs Boston
Healthcare System and Beth Israel Deaconess Medical Center, Boston (K.A.B.);
the Department of Orthopedics, Sahlgrenska University Hospital–Östra, Göteborg,
Sweden (B.I.E.); the Department of Orthopedics, Hillerųd University, Hillerųd,
Denmark (M.R.L.); and the Department of Medicine, Hamilton Health Sciences
Corporation–General Division, Hamilton, Ont., Canada (A.G.G.T).
Address reprint requests to Dr. Bauer at the Beth Israel Deaconess
Medical Center, 330 Brookline Ave., Boston, MA 02215, or at [log in to unmask].
References
Choosing a Parenteral Anticoagulant Agent
For decades, if a patient required a parenteral anticoagulant
agent, the choice was simple: unfractionated heparin was the only
such agent available. Unfractionated heparin has a long track record
of effectiveness in both the treatment of and prophylaxis against
arterial and venous thromboembolic disease. However, because of
differences among batches of heparin and problems related to the
bioavailability of the drug, monitoring of the anticoagulant effect
of heparin has been not only necessary but also problematic. In
addition, heparin-induced bleeding and thrombocytopenia can threaten
life and limb. Moreover, there have always been problems with the
use of unfractionated heparin to prevent thrombosis in high-risk
patients. In that group, it is clear that fixed-dose prophylaxis is
inadequate, yet treatment with adjusted-dose heparin or
adjusted-dose warfarin poses the additional problem of monitoring
therapy.
Since 1987, when the first low-molecular-weight heparin was approved
for use in the United States, there has been an explosion in the
number of available parenteral anticoagulant drugs. There are
currently four low-molecular-weight heparins, one heparinoid, two
hirudin derivatives, and one direct thrombin inhibitor approved for
use, all of which have defined roles in patients requiring anticoagulation.
To this plethora of agents is now added fondaparinux, a synthetic
sulfated pentasaccharide that was derived from the activated factor
X (factor Xa)–binding moiety of unfractionated heparin (Table 1).1 In
this issue of the Journal, two
groups report that once-daily treatment with 2.5 mg of fondaparinux
initiated in the early postoperative period is more effective than
enoxaparin, a low-molecular-weight heparin, in preventing venous
thromboembolism after hip2 or
knee3
surgery and does not increase the risk of bleeding. If we assume
that there will be no problems with the Food and Drug Administration
approval process, fondaparinux will join the many other agents now
available to prevent and treat thromboembolic disease. The question
is which agent should be used for a particular indication.
|
The low-molecular-weight heparins are chemical or physical fractions of
unfractionated heparin. All are excreted by the kidney, and all have
greater bioavailability than unfractionated heparin. For this
reason, their anticoagulant activity is much more predictable than
that of unfractionated heparin, making administration without monitoring
a reality in the vast majority of patients. They are at least as
effective as adjusted-dose warfarin in preventing perioperative
venous thromboembolism in high-risk patients, especially when
administered preoperatively. They are also as effective as
fixed-dose unfractionated heparin for preventing venous
thromboembolism in patients at moderate risk for deep venous
thrombosis. Low-molecular-weight heparins are less likely to cause
thrombocytopenia than unfractionated heparin, since they bind poorly
to platelet surfaces. For these reasons, I recommend that the
low-molecular-weight heparins replace unfractionated heparin
altogether for prophylaxis in patients at moderate risk for venous
thromboembolic disease (for example, after a myocardial infarction
or abdominal surgery).
The value of low-molecular-weight heparins for perioperative prophylaxis
in high-risk patients is limited by the need to administer the drug
preoperatively to obtain maximal benefit. The move from general to
regional anesthesia for surgical procedures also limits the use of
the low-molecular-weight heparins. All these drugs (and danaparoid,
a heparinoid) carry a warning against their use in patients
undergoing regional anesthesia, because of the risk of epidural
hematomas. This potential complication has severely curtailed the
perioperative use of low-molecular-weight heparins. Fondaparinux may
find its initial niche in patients undergoing regional anesthesia, since
it seems to be at least as effective as the best current treatment,
even when it is given only postoperatively.
At higher doses, the low-molecular-weight heparins enoxaparin,
dalteparin, and tinzaparin are all effective for the treatment of
venous thromboembolism. The relatively long half-life of these drugs
and the fact that in most cases monitoring is unnecessary simplify
inpatient care and make outpatient treatment of uncomplicated venous
thromboembolic disease a reality. The true potential of low-molecular-weight
heparins will be seen when we develop better systems for the
outpatient management of thromboembolic disease.
Enoxaparin and dalteparin have now been approved for the treatment
of unstable angina. Here, the differences in efficacy and safety between
unfractionated heparin and low-molecular-weight heparins are modest,
and the long half-life of the newer drugs can be problematic. Most
procedure-oriented cardiology departments have shied away from these
agents, since unfractionated heparin is nearly as effective and its
anticoagulant effects can be neutralized should an urgent procedure
become necessary. Bivalirudin, a direct inhibitor of thrombin, is
also approved for patients with unstable angina who are undergoing
coronary angioplasty, but other uses of the drug are still being
determined.
The low-molecular-weight heparins are contraindicated in patients
with type II heparin-induced thrombocytopenia because the drugs cross-react
with antiheparin antibodies. Since we now recognize heparin-induced
thrombocytopenia as a prothrombotic rather than an antithrombotic
condition, the need for effective anticoagulant drugs to treat the
disorder has become much clearer. The three agents currently
approved for heparin-induced thrombocytopenia are danaparoid, a
heparinoid compound with little cross-reactivity with antiheparin
antibodies; lepirudin, a hirudin derivative; and argatroban, a
small-molecule direct thrombin inhibitor.
Danaparoid is the only agent approved for prophylaxis against
venous thromboembolism in high-risk patients with heparin-induced thrombocytopenia.
It is also often used for treating thrombosis in this situation.
Danaparoid has a prolonged half-life, and its effects are only
partially reversible — characteristics that make its use problematic
in patients in need of procedures. Since danaparoid is derived from
heparin, there is also the theoretical risk of cross-reactivity with
antiheparin antibodies (though this has not proved to be a
significant clinical problem).
Lepirudin and argatroban are both direct thrombin inhibitors that
exert their anticoagulant effect independently of antithrombin III.
They are not derived from heparin, so there is no chance of
cross-reactivity with antiheparin antibodies. Both are administered as
intravenous infusions, and both have a very short half-life (30
minutes to 1 hour). Both are monitored with the use of the activated
partial-thromboplastin time, in much the same way as heparin is
traditionally monitored. Lepirudin is excreted by the kidney in patients
with renal dysfunction; therefore, substantial dose reduction is
necessary. Argatroban is metabolized in the liver, so substantial
dose reduction is required in patients with liver disease. Neither
agent has reversible effects, and both interact with warfarin.
Lepirudin and argatroban seem to be equally effective in the
management of thrombosis due to heparin-induced thrombocytopenia.
Since many patients with heparin-induced thrombocytopenia have
evidence of multiorgan dysfunction, the choice of an agent often
depends on other conditions that the patient might have.
Is there ever a situation in which the use of unfractionated heparin
is appropriate? In the operating room, in the intensive care unit,
and in patients with renal failure (and in the absence of
heparin-induced thrombocytopenia), unfractionated heparin is still
the agent of choice, given its short half-life, easy reversibility,
and extrarenal metabolism. However, there are clear advantages to
the newer agents, and we should use them appropriately, especially
in the area of prophylaxis against venous thromboembolism.
David L. Diuguid, M.D.
College of Physicians and Surgeons of
Columbia University
New York, NY 10032
References
Edward E.
Rylander, M.D.
Diplomat American
Board of Family Practice.
Diplomat American
Board of Palliative Medicine.