Author + information
- Received March 10, 1999
- Revision received December 16, 1999
- Accepted February 21, 2000
- Published online June 1, 2000.
- ↵*Reprint requests and correspondence: Dr. Sanjay Kaul, Division of Cardiology, 5314 North Professional Tower, Cedars-Sinai Medical Center, 8700 Beverly Boulevard, Los Angeles, California 90048
This article will review the results of recent clinical trials evaluating low molecular weight heparins (LMWHs) in the management of patients with acute coronary syndromes of unstable angina and non-ST segment elevation MI. Low molecular weight heparins are a new class of anticoagulants that have a number of advantages over unfractionated heparin (UFH) leading to their increasing use for thrombotic vascular disorders. There is convincing evidence that LMWH is more effective than placebo and at least as effective as UFH in reducing the hard end points of death and recurrent myocardial infarction. Convincing evidence for a superior efficacy is mostly limited to the least robust but most prevalent end point of recurrent angina, and benefits appear to be confined predominantly to high-risk patients. The benefits are sustained long-term, but there appears to be no incremental benefit with prolonged treatment. The risk for major bleeding is approximately equivalent to UFH, but minor hemorrhage is clearly increased, especially with vascular instrumentation. The increased bleeding risk together with its long half-life and absence of specific antidote warrants exercising caution when using LMWH with coronary intervention. Low molecular weight heparins have the potential of being cost-neutral or even cost-saving by reducing resource utilization, especially in the setting of aggressive interventional practice pattern. Last, the issue of whether one LMWH preparation is more effective and cost-effective than others remains an open question that can be answered only by direct head-to-head comparison of different LMWH preparations in randomized trials. In conclusion, subcutaneous weight-adjusted LMWH is as effective and safe as intravenous UFH in the management of patients with acute coronary syndromes. The logistic ease of administration without the need for monitoring anticoagulation appears to be the major advantage over UFH.
Low molecular weight heparins (LMWHs) are rapidly emerging as an alternative form of anticoagulant therapy to the standard unfractionated heparin (UFH). They are formed by controlled enzymatic or chemical depolymerization of UFH producing monosaccharide chains of varying lengths (3 to 7 kD) but with a mean molecular weight of ∼5 kD (1). Similar to UFH, LMWHs exert their anticoagulant activity by activating ATIII. The principal difference between LMWHs and UFH lies in their relative abilities to catalyze inactivation of factor-Xa and factor-IIa, which is dependent upon the relative composition of molecules with high affinity to ATIII called high-affinity molecules (1). They exist in two functionally different forms: below critical length molecules (BCLM) (5–17 monosaccharide units <5.4 kD) that catalyze factor-Xa inactivation but not that of factor-IIa and above critical length molecules (ACLM) (>18 monosaccharide units > 5.4 kD) that catalyze the inactivation of both factor-Xa and thrombin and inhibit thrombin generation. Unfractionated heparin is mostly composed of ACLM, whereas less than half of the chains of LMWH contains ACLM (1). Thus, unlike UFH, which has equivalent activity against factor-Xa and thrombin, LMWHs have greater activity against factor-Xa. In general, heparin preparations composed mostly of ACLM chains exhibit both anti-Xa and anti-IIa activity, bind nonspecifically to plasma proteins, endothelial cells and platelets, are sensitive to inactivation by platelet factor 4 (PF4) and are less efficient at inhibiting the generation of thrombin. In contrast, LMWHs containing primarily BCLM chains inhibit factor-Xa, are less bound to plasma proteins, endothelial cells and platelets, are resistant to inactivation by PF4 and are efficient inhibitors of thrombin generation.
Low molecular weight heparins offer a number of pharmacological advantages over the parent compound, which are listed in detail in Table 1. Low molecular weight heparins have a longer half-life (2 to 4 times that of UFH) and a superior bioavailability on subcutaneous administration (>90% compared with 30%) making once or twice daily administration feasible (2). In addition, LMWHs also have a reduced potential to cause bleeding compared with UFH because they are less likely to increase microvascular permeability or interfere with platelet-vessel wall interaction (1). Other features of LMWHs that are of particular clinical relevance are a decreased sensitivity to PF4 and lower rates of unwanted effects of heparin such as platelet activation (3), heparin-induced thrombocytopenia (4) and osteoporosis (1).
In addition to the anti-Xa and anti-IIa activity of heparin, there is also evidence to suggest that heparin may exert its antithrombotic effect by stimulating the release of an endogenous tissue factor pathway inhibitor (TFPI) (5). Tissue factor is abundantly present in the lipid-rich core of exposed atherosclerotic plaque (6), and it promotes activation of factor VII and stimulation of the extrinsic coagulation pathway, ultimately leading to formation of factor Xa. Tissue factor pathway inhibitor binds to factor Xa and inactivates the tissue factor:VIIa:Xa complex (5). Given the higher bioavailability of the heparin-like activity of LMWHs, one might speculate that they may also release TFPI more efficiently than UFH. Another potential mechanism by which LMWH (enoxaparin) exert their antithrombotic effect is through inhibition of interaction of von Willebrand factor (vWF) with platelets, resulting in reduced vWF-dependent platelet adhesion and aggregation (7,8).
Are all LMWHs the same?
There are currently at least seven different forms of LMWHs available throughout the world. There are significant differences among LMWHs, which are thought to primarily emerge from differences in the manufacturing processes (either depolymerization or fractionation) resulting in different compositions, molecular weight distributions and molecular end-structures (1,2,9). These variations confer important differences among the LMWHs with respect to the ratio of anti-Xa to anti-IIa activity, bioavailability after subcutaneous administration, elimination half-life and interaction with plasma proteins, endothelial cells and platelets. Because these pharmacokinetic and pharmacodynamic properties of LMWH are critically dependent upon the distribution of ACLM and BCLM, an LMWH with the highest percentage of BCLM and the lowest percentage of ACLM is likely to exhibit superior pharmacologic efficacy than an LMWH with the lowest percentage of BCLM and the highest percentage of ACLM (1,2,9). Of the three LMWHs that have been investigated in acute coronary syndrome (ACS), enoxaparin has the highest percentage of BCLM (72%), and dalteparin has the lowest percentage (44%). Predictably, enoxaparin has the highest anti-Xa:IIa ratio of 3.9, and dalteparin has the lowest ratio of 2.2 (Table 2). In a direct head-to-head comparison, enoxaparin was shown to provide the fastest peak of anti-Xa activity (3 to 4 h), the highest bioavailability (as determined by the area under the curve) and the longest duration of anti-Xa activity (∼12 h) after subcutaneous injection compared with dalteparin and nadroparin (10). Thus, these data confirm the pharmacokinetic and pharmacodynamic advantages conferred by the presence of larger amounts of BCLM in LMWH preparations. The key question, whether these pharmacologic differences translate into differences in clinical outcomes, remains largely unresolved.
Clinical trials of LMWHs in ACS
Several clinical trials have evaluated LMWHs for the treatment of venous thrombosis and pulmonary embolism. The findings from these trials have been reviewed extensively (11,12) and indicate that subcutaneous weight-adjusted LMWH is as effective and safe as intravenous UFH. Low molecular weight heparin, however, is much more convenient.
Three LMWHs have been evaluated in seven clinical trials in ACS of unstable angina (UA) and non Q-wave myocardial infarction (NQMI), two each with nadroparin (13,14), three with dalteparin (15–18) and two with enoxaparin (19–21). Each of these trials is reviewed briefly below.
The first published LMWH trial compared subcutaneously administered nadroparin (fraxiparin, anti-Xa:IIa ratio of 3.6:1) plus aspirin against intravenous dose-adjusted UFH plus aspirin, or aspirin alone, in 219 patients with UA treated for five to seven days (13). The results indicated that nadroparin plus aspirin was the most effective regimen in reducing adverse clinical outcomes, including myocardial infarction (MI). The triple end point of death, MI or recurrent angina was significantly reduced in patients receiving LMWH and aspirin compared with either of the other two regimens. Although this study was small and open-labeled, it provided the first evidence for improved outcome in UA with an LMWH, but larger studies were required to confirm the results.
Fraxiparine in Ischemic Syndromes (FRAXIS)
The FRAXIS study compared the efficacy of nadroparin versus unfractionated heparin in 3,468 patients with UA/NQMI (14). Patients were randomized to three groups: intravenous heparin (5,000 IU bolus followed by maintenance infusion, n = 1,151), six-day treatment with nadroparin (86 IU/kg IV bolus followed by 86 IU/kg SC twice daily, n = 1,166) and 14-day treatment with nadroparin (86 IU/kg IV bolus followed by 86 IU/kg SC twice daily, n = 1151). The preliminary results indicate that there were no differences with regards to the primary combined end point of cardiovascular death, MI and recurrent/refractory angina at 14 days between the treatment groups (18.1%, 17.8% and 20%, respectively). In contrast, a significant increase in bleeding was observed with nadroparin after 14 days of treatment compared with UFH (3.5% vs. 1.8%). Interestingly, the incidence of cardiac events was significantly increased at three months of follow-up in patients who received nadroparin for 14 days (26.2% vs. 22.2% UFH). The incidence of bleeding was also increased for these patients (4.0% vs. 2.4%). The preliminary results of this study were presented at the European Society of Cardiology meeting in Vienna, Austria, August 1998 (14), and detailed data await publication.
The first large-scale randomized clinical trial to test LMWH in ACS was the FRISC (Fragmin during Instability in Coronary Artery Disease) study that evaluated dalteparin (fragmin, anti-Xa:IIa ratio of 2.2:1) (15). It was a double-blind, placebo-controlled trial in which 1,506 patients with UA/NQMI were randomized to dalteparin (120 IU/kg SC twice daily for the first six days and then once daily at 7,500 IU for approximately six weeks) plus aspirin and placebo plus aspirin. The results demonstrated that dalteparin was superior to placebo for the acute-phase management of UA/NQMI. The rate of death or MI was significantly reduced at six days (1.8% vs. 4.8%, risk ratio of 0.37, 95% confidence interval of 0.20 to 0.68). At 40 days, the difference in rates of death and MI and of the composite end point persisted. However, with longer-term follow-up and continued treatment with a once-daily injection of dalteparin, event rates for the dalteparin and placebo groups began to converge, and there was no significant difference in event rates in the two groups by 150 days. Reduction in the dose as well as increased dosing interval have been implicated in the clustering of events (reactivation of ischemia and occurrence of reinfarction) in the patients assigned dalteparin, especially among smokers, possibly related to their tendency to hypercoagulability.
Fragmin in unstable coronary artery disease (FRIC)
The FRIC study randomized 1,482 patients in two phases (16). In the open, acute phase (days 1 to 6), patients were randomized to dalteparin (120 IU/kg SC twice daily) plus aspirin or UFH plus aspirin (Table 3). In the double-blind, prolonged treatment phase (days 6 to 45), patients received either dalteparin (7,500 IU once daily) plus aspirin (75 to 165 mg) or placebo plus aspirin. The principal outcomes of the study were the composite end points of death, MI and recurrent angina during the double-blind phase (days 6 to 45). The trial demonstrated equivalence between dalteparin and UFH during the acute-phase management, but there was no reduction in event rates over aspirin alone when dalteparin was continued for approximately six weeks at a once daily maintenance dose. The trial was underpowered to reveal significant differences between dalteparin and UFH, especially in view of the fact that most of the events (about 75%) occurred during the open phase of the trial. Moreover, the dose of dalteparin in the prolonged-treatment phase of the FRIC trial may have been either too low or the dosing interval too long to provide adequate 24-h anticoagulation coverage. It is also likely that the combination treatment with LMWH plus aspirin would have been more effective in a high-risk subset of patients that were filtered out during the acute phase of this study.
Fragmin and fast revascularization during instability in coronary artery disease (FRISC II)
The FRISC II study was a prospective, randomized, multicenter trial in which invasive and noninvasive treatments were compared by factorial design in patients with UA/NQMI (17,18). After open label treatment with dalteparin (120 IU/kg SC bid × five days) or intravenous unfractionated heparin in all patients, half of the patients in each group were randomly assigned long-term treatment with subcutaneous weight-adjusted dalteparin (7,500 IU SC bid) or placebo for three months. The comparison of the invasive and noninvasive strategies was open, and the comparison of long-term dalteparin treatment with placebo was double-blind.
The results are summarized in Table 4. In the double-blind phase of the study, treatment with dalteparin produced a nonsignificant reduction in the primary double end point of death or MI at 90 days. However, death or MI was significantly reduced early on at 30 and 45 days, as was the triple composite end point of death, MI or revascularization at three months. In the invasive versus noninvasive phase of the study, the primary double end point of death or MI at six months was significantly lowered in the invasive arm, mostly due to reduction in nonfatal MI alone. In addition, anginal symptoms and readmission rates were halved by the invasive strategy. Invasive treatment provided the greatest benefit in patients with high-risk features such as old age, chest pain at rest and ST segment depression. Dalteparin treatment provided no benefit at six months in the invasive group regardless of the presence or absence of high-risk features such as elevated troponin T. Treatment with dalteparin was associated with an increased risk of major (3.3 vs. 1.5%) and minor (23 vs. 8.4%) bleeding episodes. While the rate of total strokes (n = 24) did not differ between the treatment groups, a higher rate of hemorrhagic strokes were observed in the dalteparin group (8 vs. 0, 1 in the open-label phase, 2 in the invasive and 5 in the noninvasive group of the double-blind phase). It must be emphasized that the FRISC II trial was not a comparison of LMWH with UFH. Nevertheless, it does provide meaningful insight into the role of LMWH in the treatment of patients with ACS.
Based on the data derived from the FRISC, FRIC and FRISC II trials, we can conclude that weight-adjusted administration of dalteparin (administered in a dose of 120 anti-Xa IU/kg SC twice daily) is an effective and safe alternative to intravenous UFH in the acute treatment of patients with UA/NQMI. A lower maintenance dose of 7,500 IU of dalteparin once daily provides inadequate protection beyond the acute phase. There is a potential of benefit with prolonged treatment with a higher maintenance dose (7,500 IU SC twice daily) in high-risk patients, at least during the first six weeks of therapy. The initial benefits are not sustained during longer-term follow-up if patients are managed with a noninvasive treatment strategy. However, after an invasive procedure the risk of new cardiac events is low in patients previously treated with dalteparin.
Efficacy and safety of subcutaneous enoxaparin in non-Q wave coronary events (ESSENCE)
A large double-blind multicenter study, ESSENCE evaluated the efficacy and safety of subcutaneous enoxaparin (lovenox, anti-Xa:IIa of 3.9:1) 1 mg/kg every 12 h plus aspirin relative to intravenous UFH plus aspirin in 3,171 patients with UA/NQMI (19). The period of administration ranged from 2 to 8 days (median 2.6 days). Analysis of the triple composite end point of death, MI or recurrent angina after 14 days (the primary end point) revealed a statistically significant benefit favoring enoxaparin over UFH (16.6% vs. 19.8%; relative risk reduction of 16.2%; odds ratio, 0.80; 95% confidence interval, 0.67 to 0.96; p = 0.02) (Table 5). This was mainly due to a lower incidence of recurrent angina, the least robust but the most prevalent (up to 75% of all events) of the composite end points, in the enoxaparin group (12.9% vs. 15.5%). The treatment effect was not evident at 48 h, but once it became apparent, it persisted through day 30. Nonsignificant trends favoring enoxaparin over UFH were evident for the double composite end point of death or MI at 14 and 30 days and for the single end points of death (30 days) and MI (14 and 30 days). Enoxaparin recipients also had significantly reduced need for revascularization than UFH recipients after 30 days (27% vs. 32.2%; relative risk reduction of 16%; p = 0.001). This was mostly driven by reduction in rates of percutaneous transluminal coronary angioplasty (PTCA) (14.7% vs. 18.7%; relative risk reduction of 22%; p = 0.002) but not coronary artery bypass grafting (CABG) (12.3% vs. 13.7%; relative risk reduction of 10%; p = 0.25). There was no increase in the risk of serious bleeding, but there was about 4% excess in minor bleeding, including ecchymoses at the site of subcutaneous injection (Table 6).
Preliminary reports of one-year follow-up data for 2,915 patients in the ESSENCE study (92% of original subjects) suggest that the benefits of enoxaparin are maintained long-term (22). The triple end point of death, MI or recurrent angina was lower in enoxaparin-treated than in UFH-treated patients (32% vs. 35.7% relative risk reduction of 10.4%, p = 0.022) as was death or MI (11.5% vs. 13.5%, relative risk reduction of 14.8%, p = 0.082).
Prospective pharmacoeconomic substudies in the ESSENCE trial suggest that treatment with enoxaparin would provide cost savings compared with UFH (approximately $763 and $1,100 for initial hospitalization and 30-day cumulative costs, respectively) mainly as a result of reduced requirement for revascularization procedures (23).
TIMI-11A (thrombolysis in myocardial infarction)
The second clinical trial in which enoxaparin was evaluated in patients with UA/NQMI was the TIMI-11 study (20). It was conducted in two phases: TIMI-11A, an open-label, dose-ranging study that examined the safety of two doses of subcutaneous enoxaparin administered every 12 h–1.25 mg/kg and 1.0 mg/kg in 630 patients. During the acute hospital period (two to eight days), all patients received enoxaparin as an initial intravenous bolus of 30 mg followed by a subcutaneous injection of 1.25 mg/kg twice daily (tier one). Bleeding rates were higher than expected, and the dose was subsequently lowered to 1.0 mg/kg sc twice daily (tier two). The outpatient treatment (up to 14 days) dose was fixed, either 40 mg subcutaneously every 12 h for patients less than 65 kg or 60 mg every 12 h for those weighing ≥65 kg. The 6.5% rate of major hemorrhage (primary end point) in the 1.25 mg/kg group decreased to 1.9% when the dose was reduced to 1.0 mg/kg. The patients who had major hemorrhage tended to be older, weighed less and had higher plasma anti-Xa levels than those without major hemorrhage. Cardiovascular event rates (death, MI or recurrent angina requiring revascularization) were comparable in both dose tiers. A dose-related increase in anti-Xa levels was observed both at peak (1.0 and 1.5 U/ml at 1.0 and 1.25 mg/kg dose, respectively) and trough (0.5 and 0.6 U/ml at 1.0 and 1.25 mg/kg dose, respectively) with steady-state levels established by third dose of enoxaparin. Because efficacy is generally related to trough anti-Xa levels and bleeding complications related to peak anti-Xa levels, the optimum dose for enoxaparin is one that results in a trough anti-Xa levels of ∼0.5 units/ml and a peak level not exceeding 1.0 U/ml, i.e., 1.0 mg/kg. This was the dose that was subsequently chosen for the larger TIMI-11B trial.
The TIMI-11B trial is a randomized, double-blind, placebo-controlled study that compares prolonged enoxaparin therapy with standard dose-adjusted unfractionated heparin in 3,910 patients with UA/NQMI (21). The trial included two treatment phases, an acute phase (three to eight days) and a chronic outpatient phase up to 43 days (Table 5). In the acute hospitalization phase, eligible patients were randomized to either enoxaparin 30 mg intravenous bolus followed by 1 mg/kg subcutaneously every 12 h or to UFH administered in a bolus dose of 70 U/kg IV and an infusion rate of 15 U/kg/h adjusted to an activated partial thromboplastin time target of 1.5 to 2.5 times control (coordinated through a blinded third party as in the ESSENCE trial). In the chronic outpatient phase, patients were randomized to aspirin plus subcutaneous enoxaparin either 40 mg (<65 kg) or 60 mg (≥65 kg) every 12 h for a total treatment period of 43 days or to aspirin plus subcutaneous placebo. The median duration of treatment in the acute phase was 4.6 days with enoxaparin and three days with UFH. Analysis of the primary end point of death, MI and recurrent angina requiring urgent revascularization at day 14 revealed a significant reduction in the enoxaparin-treated patients (14.2% vs. 16.6%; relative risk reduction of 15%, p < 0.03) with the predominant effect evident in urgent revascularization (9.6% vs. 11.1%). Kaplan-Meier curves depicting event rates for the primary composite end point through day 14 revealed early separation with the largest risk reduction of about 26% at 48 to 72 h. At day 14, the differences were still significant although attenuated by almost 50%. There was a 50% increase in major bleeding episodes, which was statistically not significant (Table 6). However, it may be that the study was simply not powered to detect a clinically important increased risk of serious bleeding. As in the ESSENCE study, enoxaparin was particularly beneficial in patients presenting with ECG changes, especially ST depression and in patients with prior aspirin use.
In the chronic phase (60% of patients in each group progressed to this phase), the benefits of enoxaparin were still evident at day 43 (17.3% vs. 19.6%, a relative risk reduction of 12%, p = 0.049). Kaplan-Meier curves revealed that the two curves progressed in parallel suggesting no additional benefit with chronic therapy. Unlike the acute phase, outpatient enoxaparin treatment was associated with excess major hemorrhage, both spontaneous as well as instrumented (2.9% vs. 1.5%, p < 0.02) (Table 6).
There is convincing evidence that LMWH is more effective than placebo and at least as effective as UFH in reducing the hard end points of death and recurrent MI in patients with ACSs of UA and NQMI. Of the four large randomized clinical trials in patients with UA/NQMI where an LMWH was compared with UFH (the FRIC, ESSENCE, TIMI-11B and FRAXIS trials), only trials utilizing enoxaparin (the ESSENCE and TIMI-11B trials) have been successful in demonstrating superior efficacy of LMWH over UFH using the composite triple end point of death, MI or recurrent angina. In contrast, essentially negative data have been obtained with nadroparin (FRAXIS), despite a promising preliminary report (Gurfinkel et al. ) and with dalteparin (FRIC).
Because of the difference in design of individual trials, particularly the differences in entry criteria, management strategies, duration of the study drug treatment and the components and time of primary end point, direct comparisons of results are of limited utility and often misleading. Such issues are best resolved by head-to-head comparisons, which seldom attract the interest of pharmaceutical industry. Nonetheless, the differing results obtained from the studies of LMWHs in ACS, especially those with dalteparin (FRIC) and enoxaparin (ESSENCE, TIMI-11B) could possibly be explained by differences in pharmacological properties of LMWH preparation or may be related to the differences in trial population and design or to both.
It has been argued that the clinical superiority of enoxaparin compared with UFH is derived, in part, from its higher anti-Xa:IIa ratio of 3.9:1 compared with 1:1. Low molecular weight heparins with a lower anti-Xa:IIa ratio, such as dalteparin (2.2:1), appear to be equivalent to UFH in the acute phase of UA/NQMI management. However, lack of benefit with nadroparin (an anti-Xa:IIa ratio of 3.6:1 similar to enoxaparin) in the FRAXIS study is not consistent with this notion. Comparison of LMWHs based on the anti-Xa:IIa ratio could potentially be misleading. An elevated ratio could be either due to increased anti-Xa activity or due to reduced anti-IIa activity or due to both. Conversely, a depressed ratio could be either due to reduced anti-Xa activity or due to increased anti-IIa activity or due to both. Of the three LMWHs that have been studied in ACS, the anti-Xa activity is the highest for dalteparin > nadroparin = enoxaparin (Table 2). Yet the ratio is the highest for enoxaparin > nadroparin > dalteparin, which indicates that anti-IIa activity is the lowest for enoxaparin (Table 2). While an elevated anti-Xa activity is potentially an advantage, a concurrent high anti-IIa activity does not necessarily imply a disadvantage. It is difficult to compare the different LMWHs because the currently available compounds are standardized only in terms of anti-Xa activity. It is still unknown whether anti-Xa activity is the sole reflector of efficacy or if the efficacy is correlated to some other activity/activities, e.g., anti-IIa activity, inhibition of thrombin generation, TFPI activity or vWF-neutralizing activity (9). Some important components, such as TFPI activity, may vary by as much as 30% after administration of different LMWHs at the same anti-Xa dosage (9).
Another explanation for the variable clinical responses to different LMWHs could be the rapidity and the duration of anti-Xa activity reached in these trials. Whereas a trough anti-Xa activity of 0.6 U/ml was reached within 24 h in the ESSENCE trial, a median trough anti-Xa activity of 0.35 U/ml was reached during the FRIC trial. In a direct head-to-head comparison, enoxaparin was shown to provide the fastest peak of anti-Xa activity (3 to 4 h), the highest bioavailability (as determined by the AUC) and the longest duration of anti-Xa activity (∼12 h) after subcutaneous injection compared with dalteparin and nadroparin (10). Thus, differences in pharmacokinetic profiles, rather than anti-Xa:IIa ratios or anti-Xa activity alone, may critically modulate the clinical efficacy of LMWHs.
Trial population and design
Most discrepancies in the outcomes of drug trials are usually explained by differences in the patient populations, trial designs or both. As shown in Table 7, the patient populations in the FRIC and ESSENCE/TIMI-11B studies differed with respect to: 1) high-risk markers such as NQMI, history of previous MI and history of previous revascularization; 2) defining criterion for MI (relatively more robust in FRIC—any two of chest pain, ECG or CK elevation, than in ESSENCE and TIMI-11B—any one of ECG or CK elevation); and 3) short-term risk of death or MI in the UFH group—4.7% at 45 days in the FRIC versus 7.7% at 30 days in the ESSENCE and 8.9% at 43 days in the TIMI-11B trial. Thus, the frequency of death or MI, revascularization and patients enrolled with NQMI in the FRIC trial was roughly 50% of that reported in the enoxaparin trials (ESSENCE, TIMI-11B), indicating a low-risk subset of patients randomized in the FRIC study. The higher event rates in patients in the ESSENCE and TIMI-11B studies could possibly account for the more favorable results with the LMWH used in these studies. The magnitude of benefit in randomized clinical trials is generally greater in patients at high risk compared with those at low risk. Again, data from FRISC II trial provide evidence in support of this. In this trial the protective effects of long-term dalteparin were mostly confined to patients with elevated troponin T (a marker of high-risk) who constituted about 60% of the patient population. Thus, it appears that LMWH works better in high-risk patients with ACS.
Degree of anticoagulation
The FRIC and the ESSENCE/TIMI-11B studies differed significantly in terms of degree and duration of adequate anticoagulation achieved in the acute in-hospital as well as the chronic outpatient phase (Table 7). Only with enoxaparin (median trough levels of 0.6 and 0.5 anti-Xa U/ml in the ESSENCE and TIMI-11B trials, respectively) but not with dalteparin (median trough levels of 0.35 anti-Xa U/ml in the FRIC trial) was anticoagulation achieved within the optimal range in the acute in-hospital phase. This occurred despite a more aggressive anticoagulation regimen in patients assigned dalteparin in the FRIC study (120 anti-Xa IU/kg SC twice daily for six days) than those assigned enoxaparin in the ESSENCE study (100 anti-Xa IU/kg SC twice daily for two to eight days). These data reflect potential differences in bioavailability and pharmacokinetic properties between the LMWH preparations.
It has also been argued that the dose of dalteparin used in the prolonged-treatment phase of the FRIC trial was either too low or the dosing interval too long to provide adequate 24-h anticoagulation coverage. In contrast, the enoxaparin dose used in the chronic treatment phase of TIMI-11B trial was adjusted to provide adequate round-the-clock anticoagulation, which possibly could contribute to preservation of benefits with enoxaparin at day 45. Recent data from FRISC II study provide useful insight on this issue. A weight-adjusted dose of dalteparin in FRISC II (7,500 IU SC twice daily × 3 months) was associated with a significant reduction in triple end point of death, MI or revascularization at three months. The more robust double end point of death or MI was also significantly reduced at 30 and 45 days even though there was a nonsignificant difference in these outcomes at three months. Thus, the more intensive the anticoagulation regimen the greater is the likelihood of benefit.
One important issue that has been raised in the LMWH clinical trials is the difference in anticoagulation protocols both in the UFH and LMWH treatment arms. A weight-adjusted standard heparin protocol, which is likely to produce therapeutic activated partial thromboplastin times within the first 24 h (24) (a critical time for a beneficial therapeutic effect) was used in the FRIC and TIMI-11B trials but not in the ESSENCE trial. However, the aPTT levels were reasonably adequate in all the three trials with 40% to 50% patients achieving aPTT levels within the therapeutic range within the first 48 h. It is, therefore, unlikely that differences in the degree of therapeutic anticoagulation in the UFH-treated patients could account for the differences among the treatment arms in the FRIC study as compared with the ESSENCE and TIMI-11B studies. An interesting feature of TIMI-11B trial is that approximately one third of patients were receiving intravenous UFH for up to 24 h before randomization. It is conceivable that some of these individuals may represent treatment failures of UFH, thereby selecting out patients for potential benefit with LMWH. An additional distinctive feature of the TIMI-11B protocol was the bolus intravenous administration of enoxaparin that might have contributed to its early benefit at 48 h. Of note, patients were treated longer with enoxaparin than with UFH in TIMI-11B (median duration of 4.6 vs. three days), which could also potentially bias the results in favor of enoxaparin. In fairness, the point has been raised that this difference in the duration of treatment is unlikely to have played an important role in the observed benefit of enoxaparin because statistically significant treatment effect was already evident at 48 h (21).
Relative to the hard end points of death or MI, none of the LMWHs have been shown to be statistically superior to UFH in patients with ACS. Evidence for superiority of LMWH over UFH appears to emerge only when the least robust, somewhat subjective end point of recurrent angina is considered. In both the enoxaparin trials, recurrent angina constituted the bulk (>70% of all events) of the events. Consistent with this, treatment with dalteparin in FRISC II was associated with a significant reduction in triple end point of death, MI or revascularization at three months but not in the more robust double end point of death or MI. In contrast to the ESSENCE study, the criterion for recurrent angina was relatively more robust in the TIMI-11B trial where only new anginal pain requiring urgent revascularization was considered. This may help explain the lower incidence of recurrent angina in TIMI-11B compared with the ESSENCE trial (11.1% vs. 15.5% in the UFH arm at 14 days and 12.6% vs. 18% at 30 to 43 days) despite a higher proportion of high-risk patients, based on the inclusion criterion of NQMI enrolled in the former (35% vs. 21%).
It has been argued that neither the ESSENCE nor the TIMI-11B trial was designed with sufficient power to detect statistically significant reduction in the hard end points of death or MI. A prospectively planned combined analysis of the ESSENCE and TIMI-11B data (almost 7,000 patients) appears to favor enoxaparin over UFH in reducing the double end point of death or MI at four time points—days 2, 8, 14 and 43 (Table 8)(25). This meta-analysis, although suggestive of consistent and important treatment effects of enoxaparin, should be regarded as exploratory and not considered to be conclusive or definitive evidence of its superiority over UFH in reducing the hard end points. In addition, the scope of this meta-analysis would be significantly enhanced by including data from other LMWH trials such as FRIC (dalteparin) and FRAXIS (nadroparin, an agent with similar molecular weight and anti Xa:IIa ratio as enoxaparin). One such preliminary meta-analysis was recently presented at the 1999 American Heart Association scientific sessions (26). The meta-analysis demonstrated no difference in the end point of death or MI (OR 0.89; 95% CI: 0.70–1.14) or major bleeding (OR 0.97; 95% CI: 0.62–1.51) between LMWH and UFH during short-term therapy (day 2 to 8). However, LMWH was associated with a significantly increased risk of major bleeding during long-term therapy (OR 2.38; 95% CI: 1.59–3.54). Thus, based on this meta-analysis, there is no difference in efficacy and safety between LMWH and UFH in patients with UA/NQMI, a finding consistent with the venous thrombosis trial data.
Another trial component that was different between the FRIC study and the enoxaparin studies was the inclusion of periprocedural MI defined by arbitrary values for CPK/MB elevations after PTCA (>3 times the upper limit of normal) or CABG (>5 times the upper limit of normal) in the enoxaparin studies (19,21). The relevance of these enzyme elevations to coronary morbidity and mortality has recently been the subject of intense debate (27–29). Although a meta-analysis of several small retrospective studies seems to suggest a tight association between periprocedural creatine phosphokinase elevations and late excess mortality (27), a causal relationship has not been demonstrated, especially in prospectively conducted rigorously controlled trials (28,29). While the association might be true, CPK elevations may be a marker of high risk, such as large atherosclerotic burden, rather than the cause of excess mortality. Thus, the clinical significance of these periprocedural enzyme elevations “remains somewhat controversial,” and their effect on prognosis is unclear. In addition, almost a third (32.2%) of patients underwent revascularization in the ESSENCE study compared with 14.2% in the FRIC trial. This is reflective of the aggressive interventional practice pattern in North America from where 70% of patients were recruited in the ESSENCE compared with a predominantly European recruitment in the FRIC trial. The increased rates of revascularization in patients randomized in the ESSENCE trial from North America is also likely to have an impact on the positive pharmacoeconomic analysis, which was mostly a result of reduced requirement for revascularization procedures with the use of enoxaparin. Similarly, the lower bleeding complications in FRIC reflect reduced rates of revascularization because most bleeding episodes with LMWH occur at vascular access sites. It is well known that revascularization procedures in the setting of ACS appear to increase event rates (mostly defined by periprocedural enzyme elevations). Although the relative contribution of these periprocedural enzyme elevations to the reinfarction component of the composite end point has not been published previously, data on file at Rhone-Poulenc Rorer Pharmaceuticals Inc. shed an important light on this issue. As shown in Table 9, the differences in MIs in each arm of the ESSENCE trial were mostly due to the differences in postrevascularization MIs with post-CABG MIs, defined mostly by CK elevation criterion, constituting the bulk of these events (22/30 and 15/22 in the UFH and enoxaparin arm, respectively). Thus, it is likely that the increased rate of periprocedural MIs associated with increased frequency of revascularization procedures may have significantly biased the results in the direction of enoxaparin in the ESSENCE and TIMI-11B trials. Thus, with respect to the high-risk patients and the higher risk associated with the use of angiography and revascularization procedures, treatment with a LMWH appears to be superior to UFH.
LMWHs in invasive intervention
For LMWHs the subcutaneous route of administration, without the need for dose adjustment, has significant logistic advantages. However, with invasive intervention, the lack of a readily available rapid assay for anticoagulant activity results in uncertainty about the amount of adjunctive antithrombin treatment needed. Different methods for measuring anti-Xa or anti-IIa activity are currently available to assess anticoagulant efficacy. However, there are no clear indications of the usefulness of measuring these parameters to decrease hemorrhagic risk and increase antithrombotic efficacy of LMWHs except in patients with renal dysfunction or underweight or overweight patients. Specific guidelines and recommendations for managing patients with LMWH during diagnostic catheterization, percutaneous coronary interventions and CABG have been described (30), and preliminary results have been encouraging (31).
The reversal of anticoagulation, especially during bleeding complications, presents a dilemma. While protamine is effective in neutralizing antithrombin activity of heparin, it is not the optimal agent to antagonize LMWHs. The LMWH compounds with the highest anti-Xa activity exhibit the weakest neutralization with protamine. Recent evidence has demonstrated heparinase to be much more effective in neutralizing the LMWHs, but these data need to be confirmed.
Direct antithrombins versus LMWH
Are the more potent and direct antithrombins (hirudin, hirulog) more effective than LMWH in ACS, especially when optimal doses of these agents are used? This issue remains largely unresolved. However, experience with direct antithrombins has been disappointing to date with treatment benefit evident only at early time points (day 3) and not at longer follow-up (day 35) (32).
Synergism with antiplatelet agents such as ADP antagonists, ticlopidine and/or clopidogrel, glycoprotein IIb/IIIa antagonists and fibrinolytic agents
Exposure of thrombin is a very potent stimulus for platelet activation and, in turn, activated platelets trigger thrombin generation and activation (1). The antithrombin and antiplatelet approaches in ACS and in percutaneous coronary interventions are, therefore, complimentary and potentially synergistic. Several trials are evaluating the potential for this synergism, e.g., the ATLAST (Aspirin/Ticlopidine vs. LMWH (enoxaparin)/Aspirin/Ticlopidine High Risk Stent) (33), the ENTICES trial (the Enoxaparin and Ticlopidine after Elective Stenting) (34) and the NICE-4 (National Investigators Collaborating on Enoxaparin) trial with GP IIb/IIIa antagonist, abciximab (33). The feasibility of adjunctive use of LMWH (nadroparin) with fibrinolysis (tPA) were recently published (35) and a trial where adjunctive use of LMWH with fibrinolysis is being compared with UFH (the HIPA trial) (30) is currently ongoing.
Extrapolation of results to low-risk patients
A “positive” result of a randomized clinical trial occurs as a consequence of benefit for those individuals within the cohort who are at a high intrinsic risk. Those at low intrinsic risk are far less likely to benefit (and may even suffer). Current results suggest that only a few patients with ACS progress to recurrent MI and death. Most of these patients are identifiable by their high-risk features such as age >65 years, ST segment depression, elevated enzymes or prior evidence of coronary artery disease. Hence, only a small proportion of the ACS cohort who are at high risk can benefit from the proposed intervention. Thus far, the LMWH trials (especially with enoxaparin) have deliberately selected patient populations who are at relatively higher risk. Thus, the results are applicable to ACS patients who present with a marker of high risk such as age >65 years, ST segment depression (ESSENCE, TIMI-11B), elevated enzymes (CPK in ESSENCE but not in TIMI-11B and troponin T in FRISC), prior aspirin use (ESSENCE, TIMI-11B) or other evidence of CAD. Whether benefits exceed the risk among low-risk patients remains to be determined. In medical practice the ethical, scientific and practical challenge is to offer treatment only to those who are reasonably likely to benefit and to withhold it from those unlikely to benefit or likely to be harmed.
In conclusion, LMWHs represent an important advance in the management of patients with ACS. The logistic ease of administration without the need for monitoring anticoagulation appears to be the major advantage over UFH. Convincing evidence for a superior efficacy is mostly limited to the least robust but most prevalent end point of recurrent angina, and benefits appear to be confined predominantly to high-risk patients. The benefits are sustained long-term, but there appears to be no incremental benefit with prolonged treatment. The risk for major bleeding is approximately equivalent to UFH, but minor hemorrhage is increased, especially with vascular instrumentation. The increased bleeding risk together with its long half-life and absence of specific antidote warrants exercising caution when using LMWH with coronary intervention. Last, but not least, they have the potential of being cost-neutral or even cost-saving by reducing resource utilization, especially in aggressive interventional strategy.
However, a critical analysis of the results of randomized clinical trials in the comparison of LMWH with UFH raises important issues. The robustness of the data is open to question. The benefit of LMWH may be inflated due to reliance on soft end points such as recurrent angina and urgent revascularization, which are somewhat subjective and driven by individual, institutional or geographic practice patterns. The CPK elevation criteria for periprocedural myocardial infarction are arbitrary and rely heavily on the assumption that these enzyme elevations are the cause of adverse clinical coronary events. The impact of increased rate of revascularization procedures and the associated enzyme elevations on clinical as well as on pharmacoeconomic outcomes is bound to be greater with aggressive interventional strategy (as in the ESSENCE and TIMI-11b trials) than conservative strategy (as in the FRIC trial). Last, the issue of whether one LMWH preparation is more effective and cost-effective than others remains an open question that can be answered only by direct head-to-head comparison of different LMWH preparations in randomized trials.
☆ This study was supported by an unrestricted educational grant from Pharmacia and Upjohn.
- above critical length molecules
- acute coronary syndrome
- area under the curve
- below critical length molecules
- coronary artery bypass grafting
- Efficacy and Safety of Subcutaneous Enoxaparin in Non-Q Wave Coronary Events
- Fraxiparine in Ischemic Syndromes
- Fragmin in Unstable Coronary artery disease
- Fragmin during InStability in Coronary artery disease
- FRISC II
- Fragmin and Fast Revascularization during InStability in Coronary artery disease
- low molecular weight heparin
- myocardial infarction
- non-Q-wave myocardial infarction
- platelet factor 4
- percutaneous transluminal coronary angioplasty
- tissue factor pathway inhibitor
- Thrombolysis in Myocardial Infarction 11B
- unstable angina
- unfractionated heparin
- von Willebrand factor
- Received March 10, 1999.
- Revision received December 16, 1999.
- Accepted February 21, 2000.
- American College of Cardiology
- Xiao Z,
- Theroux P
- Toschi V,
- Gallo R,
- Lettino M,
- et al.
- French Investigators of the ESSENCE Trial,
- Montalescot G,
- Philippe F,
- Ankri A,
- et al.
- Antman E.M,
- Handin R
- Kuijer P.M.M,
- Prins M.H,
- Buller H.R
- Hirsh J
- Gurfinkel E.P,
- Manos E.J,
- Mejail R.I,
- et al.
- Verheugt F.W.A
- FRIC Investigators,
- Klein W,
- Buchwald A,
- Hillis S.E,
- et al.
- Thrombolysis in Myocardial Infarction (TIMI) 11A Trial Investigators
- Antman E.M,
- McCabe C.H,
- Gurfinkel E.P,
- et al.
- Goodman S,
- Bigonzi F,
- Radley D.R,
- Le Louer V,
- Gosset F,
- Cohen M
- Mark D.B,
- Cowper P.A,
- Berkowitz S.D,
- et al.
- TIMI-11B and ESSENCE Investigators,
- Antman E.M,
- Cohen M,
- Radley D.R,
- et al.
- ↵Eikelboom JW, Anand SS, Malmberg KA, Weitz J, Ginsberg JS, Yusuf S. Unfractionated heparin (UFH) and low-molecular weight heparin (LMWH) in unstable angina and non-Q wave myocardial infarction: a meta-analysis (abstr). Circulation 1999;100:1–621.
- Califf R.M,
- Abdelmeguid A.E,
- Kuntz R.E,
- et al.
- Baim D.S,
- Cutlip D.E,
- Sharma S.K,
- et al.
- Kini A,
- Marmur J.D,
- Kini S
- ↵Lovenox [package insert]. Rhone-Poulenc Rorer Pharmaceuticals Inc., Collegeville, Pennsylvania.
- Deutsch E,
- Cohen M,
- Radley D.R,
- Tarazona N,
- Fromell G.J
- ↵Data on file, Rhone-Poulenc Rorer Pharmaceuticals Inc., Collegeville, Pennsylvania, 1999.
- Chamuleau S.A,
- de Winter R.J,
- Levi M,
- et al.