OUP user menu

Non-ST-segment elevation acute coronary syndrome in patients with renal dysfunction: benefit of low-molecular-weight heparin alone or with glycoprotein IIb/IIIa inhibitors on outcomes. The Global Registry of Acute Coronary Events

Jean-Philippe Collet, Gilles Montalescot, Giancarlo Agnelli, Frans Van de Werf, Enrique P. Gurfinkel, Jose López-Sendón, Christopher V. Laufenberg, Martin Klutman, Neelam Gowda, Dietrich Gulba
DOI: http://dx.doi.org/10.1093/eurheartj/ehi337 2285-2293 First published online: 2 June 2005

Abstract

Aims To determine whether low-molecular-weight heparin (LMWH)+glycoprotein (GP) IIb/IIIa inhibitors provide greater benefit than unfractionated heparin (UFH)+GP IIb/IIIa inhibitors, irrespective of renal status.

Methods and results Patients in the Global Registry of Acute Coronary Events (GRACE) were divided into three groups according to creatinine clearance (CrCl): normal renal function (CrCl >60 mL/min), moderate renal dysfunction (30<CrCl≤60 mL/min), and severe (CrCl≤30 mL/min) renal dysfunction. Data were analysed from 11 881 patients with acute coronary syndrome (ACS). Patients with moderate (n=3705) or severe (n=982) renal dysfunction were at higher risk of adverse outcomes than those with normal renal function. Decreasing CrCl was an independent predictor of mortality at 30 days and in-hospital major bleeding. LMWH+GP IIb/IIIa inhibitors were used significantly less frequently in patients with severe (2.0%) or moderate (3.1%) renal dysfunction than in those with normal function (3.9%, P=0.0056). LMWH alone was more beneficial than UFH alone, irrespective of renal status. LMWH alone was an independent predictor of 30 day survival [odds ratio (OR) 0.56; 95% confidence interval (CI) 0.43–0.73] and lower risk of in-hospital bleeding (OR 0.66; 95% CI 0.48–0.92). Bleeding rates were significantly lower with LMWH+GP IIb/IIIa inhibitors than those with UFH+GP IIb/IIIa inhibitors. Use of UFH+GP IIb/IIIa inhibitors was an independent predictor of bleeding (OR 2.02; 95% CI 1.42–2.90) compared with UFH alone.

Conclusion In patients with renal dysfunction and non-ST-segment elevation ACS, bleeding complications are more frequent and outcomes appear worse in individuals treated with UFH compared with LMWH. Combination therapy with LMWH and GP IIb/IIIa inhibitors appears to be better tolerated than with UFH and GP IIb/IIIa inhibitors.

  • Renal dysfunction
  • Antithrombotic
  • Bleeding
  • Mortality

This paper was guest edited by Prof. Bernard J. Gersh, Mayo Clinic, Rochester, USA

Introduction

Strong evidence exists to show that subcutaneous enoxaparin is superior to unfractionated heparin (UFH) in patients with unstable angina or non-ST-segment elevation myocardial infarction (NSTEMI).1 However, the use of low-molecular-weight heparin (LMWH) remains disproportionately low, probably because these data were obtained from selected populations. At least 40% of patients with unstable angina or NSTEMI seen in routine clinical practice would have been excluded from randomized pivotal trials performed with LMWH.2 Severe renal dysfunction is one of the most critical exclusion criteria and is found in 40% of excluded patients.2 It is consistently used as an exclusion criterion in randomized trials evaluating antithrombotic drugs. Renal status is a major indicator of vascular risk but is also an independent predictor of bleeding.26 As a consequence, the safety of LMWH has been questioned in patients with a low creatinine clearance (CrCl) primarily because of the risk of accumulation over time. In contrast, whether more aggressive antithrombotic approaches, including the combination of LMWH and glycoprotein (GP) IIb/IIIa inhibitors, would be beneficial in these patients remain to be established given the fact that this group is at higher risk of ischaemic events than patients with normal renal function.2,5,6

The aim of this non-randomized, prospective study was to evaluate if the use of LMWH alone or with GP IIb/IIIa inhibitors was associated with a greater benefit than UFH alone or with GP IIb/IIIa inhibitors, irrespective of renal status, in unselected patients with a non-ST-segment elevation acute coronary syndrome (NSTE-ACS) enroled in the Global Registry of Acute Coronary Events (GRACE).

Methods

Population study

GRACE is a large, prospective, multinational, observational registry of the entire spectrum of ACS.7,8 Patients entered into the registry had to be at least 18 years old and alive at the time of hospital presentation, be admitted for ACS as a presumptive diagnosis (i.e. have symptoms consistent with acute ischaemia), and have at least one of the following: electrocardiographic changes consistent with ACS, serial increases in serum biochemical markers of cardiac necrosis, and/or documentation of coronary artery disease. The qualifying ACS must not have been precipitated by significant non-cardiovascular comorbidity (i.e. trauma or surgery). Standardized definitions of all patient-related variables and clinical diagnoses were used. NSTEMI was defined as the presence of at least one positive cardiac biochemical marker of necrosis without new ST-segment elevation seen on the index or qualifying electrocardiogram. Unstable angina was defined as the absence of both ST-segment elevation on the electrocardiogram and serum biochemical markers indicative of myocardial necrosis within each hospital laboratory's normal range but with a discharge diagnosis of ACS. Patients originally admitted for unstable angina but in whom myocardial infarction occurred during the hospital stay were classified as having a myocardial infarction. Where required, local approval from institutional review boards was obtained.

To ensure the enrolment of an unbiased population, the first 10–20 consecutive patients (depending on each site's patient throughput) were recruited from each site per month. Data were collected at each site by a trained coordinator using a standardized, case-report form. Demographic characteristics, medical history, presenting symptoms, duration of pre-hospital delay, biochemical and electrocardiographic findings, treatment practices, and a variety of hospital-outcome data were collected.

Patients were divided into three groups according to renal status: normal renal function (CrCl >60 mL/min), moderate renal dysfunction (30<CrCl≤60 mL/min), and severe (CrCl ≤30 mL/min) renal dysfunction. For practical purposes, estimated CrCl is used as a correlate of glomerular filtration rate and is commonly estimated using the Cockcroft–Gault formula without the need for 24 h urine collection.911 A standardized, core case record was used to collect information on patient demographics, symptoms, medical history, time of presentation, and clinical, electrocardiographic, and treatment characteristics.

Clinical endpoints

The endpoints of the study were mortality at 30 days and in-hospital major bleeding. Mortality was defined as all-cause mortality at 30 days from hospital admission. Major bleeding was defined as life-threatening bleeding, a bleed requiring a transfusion of two or more units of packed red cells, or a bleed resulting in an absolute decrease in haematocrit of ≥10% or resulting in death. Major bleeding also included intracranial bleeding.

Statistical analysis

Comparisons between groups were made using the two-tailed Kruskal–Wallis and Wilcoxon rank sum tests for continuous variables and χ2 test or Fisher's exact test for categorical variables, as appropriate. Multivariable logistic regression analysis was carried out to determine the relation between mortality at 30 days from hospital admission and in-hospital major bleeding against renal status and type of anticoagulants. The model for death was based on GRACE risk model for in-hospital death12 (c-statistic=0.83) and was adjusted for age, Killip class, systolic blood pressure, ST-segment deviation, cardiac arrest at hospital presentation, serum creatinine level, positive initial cardiac enzymes, and heart rate. The model for major bleeding was based on the GRACE risk model for in-hospital bleeding5 (c-statistic=0.75) and was adjusted for age, sex, medical history of bleeding and renal insufficiency, diastolic blood pressure, in-hospital use of diuretics, other vasodilators, and intravenous inotropics, right heart catheterization, and percutaneous coronary intervention.

All tests were double-sided and considered statistically significant at α≤0.05. The analyses were performed using the SAS software package (version 9.1, SAS Institute, Cary, NC, USA).

Results

Patient population

Data from 24 309 patients with ACS enroled in the GRACE registry between April 1999 and September 2002 were analysed (Figure 1). A total of 4168 patients were excluded from the analysis because of missing CrCl data. However, this population had a similar profile to patients with CrCl data (data not shown). Patients with moderate or severe renal dysfunction were at higher risk of ischaemic events than those with normal renal function; these groups included a significantly higher proportion of older patients, women, and patients with diabetes, prior myocardial infarction or stroke, heart failure on admission, and left bundle branch block at the qualifying electrocardiogram (Table 1). In addition, patients with moderate or severe renal dysfunction had a higher TIMI risk score,13 and displayed a higher risk of major bleeding than those with normal renal function, as reflected by their higher incidence of a prior history of bleeding and the higher proportion of women and older patients (Table 1).

Figure 1 Flow chart of patients who entered the study.

View this table:
Table 1

Patient baseline characteristics and outcomes

VariablesCrCl (mL/min)P value (three-way)
≤30 (n=982)>30 to ≤60 (n=3705)>60 (n=7194)
Characteristics
 Age (mean±SD), years78±1175±961±11<0.0001
 >80 years (%)49272<0.0001
 Women (%)574823<0.0001
Risk factors
 Current smoker (%)8.491228<0.0001
 Diabetes (%)332925<0.0001
 Body mass index (mean±SD)24±526±429±6<0.0001
Prior cardiac history
 Myocardial infarction (%)424134<0.0001
 CABG (%)171916<0.0001
 PCI (%)151821<0.0001
 Aspirin (%)525449<0.0001
 Past history of bleeding (%)3.371.970.99<0.0001
Admission characteristics
 ST-segment depression (%)443627<0.0001
 LBBB (%)118.073.35<0.0001
 Killip class ≥3 (%)167.382.47<0.0001
 Cardiogenic shock (%)1.880.780.33<0.0001
 Positive troponin test (%)796765<0.0001
 Left ventricular ejection fraction <40% (%)393118<0.0001
 TIMI-risk score for NSTE-ACS (mean±SD)3.26±13.13±12.51±1<0.0001
Comorbidities
 Recent surgery (%)0.310.410.250.38
 Stroke (%)17136.72<0.0001
 Peripheral artery disease (%)23169.06<0.0001
 Cancer (%)9.937.524.67<0.0001
Patient outcomes
 Mortality at 30 days (%)165.681.91<0.0001
 Major bleeding (%)9.474.162.37<0.0001

CAGB, coronary artery bypass grafting; LBBB, left bundle branch block; PCI, percutaneous coronary intervention; TIMI, Thrombolysis in myocardial infarction.

Antithrombotic regimen

Patients with moderate or severe renal dysfunction were treated with a less aggressive antithrombotic regimen and less frequently underwent invasive interventions and revascularization than those with normal renal function (Table 2). The use of LMWH alone was similar irrespective of renal status, whereas the use of UFH alone was higher in patients with severe renal impairment relative to those with moderate impairment or normal renal function. A greater proportion of patients received LMWH alone compared with UFH alone (Table 2). Less than 1% of patients with severe renal dysfunction were not given any anticoagulants during their hospital stay. LMWHs+GP IIb/IIIa inhibitors were used significantly less frequently in patients with severe (2.0%) or moderate (3.1%) renal dysfunction than in those with normal renal function (3.8%, P=0.0056). UFH was combined with GP IIb/IIIa inhibitors in 6.9% of patients with severe renal dysfunction, 7.2% with moderate renal dysfunction, and 10.8% of patients with normal renal function (P<0.0001).

View this table:
Table 2

Use of medications and interventions during hospital stay

CrCl (mL/min)P value (three-way)
≤30 (n=982)>30 to ≤60 (n=3705)>60 (n=7194)
Pharmacologic intervention (%)
 LMWH455354<0.0001
 UFH475054<0.0001
 GP IIb/IIIa inhibitors131522<0.0001
 Thienopyridine283239<0.0001
Invasive intervention (%)
 Coronary angiography334660<0.0001
 Multiple-vessel disease736858<0.0001
 PCI172231<0.0001
 CABG679<0.0001
Treatment combinations (%)
 UFH alone282422<0.0001
 LMWH alone3031300.31
 UFH+GP IIb/IIIa (no LMWH)6.887.2311<0.0001
 LMWH+GP IIb/IIIa (no UFH)1.953.103.780.0056
 LMWH+GP IIb/IIIa+UFH2.873.515.53<0.0001
 LMWH+UFH (no GP IIb/IIIa)8.9313130.0005
 GP IIb/IIIa (no LMWH or UFH)211614<0.0001
 No LMWH or UFH or GP IIb/IIIa0.821.201.360.34

CAGB, coronary artery bypass grafting; PCI, percutaneous coronary intervention.

Clinical outcomes

The rate of mortality at 30 days was significantly (P<0.0001) higher in patients with severe renal dysfunction when compared with patients with moderate or normal renal function (Table 1). The use of LMWH alone was associated with a significantly lower risk of 30 day mortality compared with UFH alone (4.2 vs. 6.2%, P<0.0001). Figure 2 illustrates the rates of 30 day mortality according to the type of treatment and renal status. The magnitude of this benefit varied according to renal function (Figure 2). Although the mortality rate in patients with severe renal dysfunction who received LMWH alone was lower than that in patients who received UFH only (15 vs. 19%), the difference was not statistically significant. The use of LMWH+GP IIb/IIIa inhibitors appeared to be associated with a higher mortality rate in patients with moderate or severe renal dysfunction compared with the use of UFH+GP IIb/IIIa inhibitors; however, these results are based on a very small number of patients and should be interpreted with caution.

Figure 2 Mortality at 30 days according to renal status and antithrombotic treatment. Table shows the absolute differences between treatment groups and the 95% CI. *P<0.0001 across the three groups for each type of antithrombotic treatment.

In-hospital major bleeding occurred more frequently in patients with severe or moderate renal dysfunction than in patients with normal renal function (Table 1). The use of LMWH alone was associated with significantly fewer in-hospital major bleeds than UFH alone (2.1 vs. 3.3%, P=0.0006). Figure 3 illustrates the rates of major in-hospital bleeding according to treatment and renal status. This benefit was found irrespective of renal status but remained non-significant in patients with severe renal dysfunction. There was also a significant reduction of in-hospital major bleeding with the combination of LMWH+GP IIb/IIIa inhibitors compared with UFH+GP IIb/IIIa inhibitors (4.3 vs. 6.9%, P=0.0420). This favourable trend was consistent across the renal function subgroups but the differences were not statistically significant (Figure 3).

Figure 3 In-hospital major bleeding according to renal status and antithrombotic treatment. Table shows the absolute differences between treatment groups and the 95% CI. *P<0.0001 across the three groups.

Multivariable analysis

Renal status was an independent predictor of mortality at 30 days (Figure 4A) and in-hospital major bleeding (Figure 4B). Patients with severe renal dysfunction had an almost four-fold increase (OR 3.45, 95% CI 2.52–472) and those with moderate renal dysfunction had an almost two-fold increase (OR 1.52; 95% CI 1.1–2.01) in the 30 day mortality rate when compared with patients with normal function (Figure 4A). The use of LMWH alone vs. UFH alone was an independent predictor of survival and lower risk of in-hospital major bleeding (Figure 4B). The use of UFH+GP IIb/IIIa inhibitors was an independent predictor of in-hospital major bleeding when considering UFH as the reference anticoagulation regimen (Figure 4B). There was no statistically significant interaction between the renal function and the anticoagulation regimen in the multivariable models for bleeding (P=0.17) or mortality at 30 days (P=0.62) (adjusted using the GRACE risk scores5,12).

Figure 4 Adjusted OR and CI (A) for mortality at 30 days and (B) for in-hospital major bleeding obtained with the multivariable analyses. In (A), *CrCl>60 mL/min is the reference value; †UFH is the reference value. In (B), *female sex is the reference value; †UFH is the reference value; ‡CrCl>60 mL/min is the reference value.

Renal function and clinical endpoints

CrCl was also considered as a continuous variable given the potential impact of the definition of renal failure on its power to predict clinical outcome. There was no cut-off point of CrCl for both endpoints but rather a gradient of increasing risk between 120 and 15 mL/min of CrCl (Figure 5). The estimated probability of mortality at 30 days according to CrCl was similar for UFH and LMWH, although there was a clear shift towards a higher probability of in-hospital bleeding for UFH compared with LMWH. Below the cut-off value of 60 mL/min, the estimated probability of 30 day mortality was higher than that of in-hospital bleeding. The risk–benefit ratio for bleeding or mortality appeared to be better with LMWH than those with UFH in patients with CrCl<60 mL/min.

Figure 5 Kernel curves of in-hospital mortality (black) or bleeding (grey) according to the level of CrCl in patients treated with UFH (plain curves) and LMWH (dashed curves).

Discussion

This prospective but non-randomized study indicates that, despite the fact that renal dysfunction is associated with more serious presentation and a poorer prognosis than normal renal function, NSTE-ACS patients with renal dysfunction receive less aggressive medical therapy and are less likely to undergo catheterization or revascularization procedures than those with normal renal function. Importantly, this prospective observational registry also suggests that LMWH is better tolerated and more effective than UFH in patients hospitalized with NSTE-ACS in clinical practice. This benefit was found irrespective of renal function, although it remained non-significant in patients with severe renal dysfunction, probably because of the low number of events in this small subgroup of patients. In addition, combination therapy with LMWH+GP IIb/IIIa inhibitors appears to be better tolerated than with UFH+GP IIb/IIIa inhibitors; however, there was a non-significant increase in the mortality rate at 30 days in patients with moderate or severe renal impairment who were treated with LMWH+GP IIb/IIIa inhibitors.

Strong evidence exists to suggest that antithrombotic therapy is beneficial in NSTE-ACS patients and that LMWH is as effective as, or superior to, UFH in preventing ischaemic events in this setting.1,14 However, very few data are available on the safety and efficacy of LMWH in NSTE-ACS patients with renal dysfunction.2,15,16 This is mainly due to the fact that renal dysfunction is a risk factor for bleeding5 and that the safety of LMWH has been questioned in patients with a low CrCl because of the risk of accumulation over time. However, these patients have been safely treated with a reduced subcutaneous dose of enoxaparin according to the patient's CrCl.2,15,17,18 In the present study, >40% of patients with renal dysfunction (CrCl <60 mL/min) were given LMWHs (84% of which was enoxaparin), lending further evidence to suggest that LMWHs may be safer and more effective than UFHs in unselected populations.

The benefit of LMWH over UFH in randomized trials of NSTE-ACS patients is driven by the reduction of risk of myocardial infarction and the need for urgent revascularization.1 We report here that LMWH use is associated with a lower mortality rate when compared with UFH in NSTE-ACS, and that LMWH is an independent predictor of survival at 30 days in unselected patients. This finding is consistent with the findings of a systematic overview of patients in randomized trials, which reported a non-significant 12% relative risk reduction in mortality.1 The large number of patients enroled in the GRACE registry, combined with their higher risk compared with populations studied in randomized trials, may account for these unexpected findings.

Renal dysfunction (CrCl <60 mL/min), which was found in 40% of the entire GRACE population, is a major indicator of the risk profile of all comers with NSTE-ACS.3,4,6 This subgroup of patients tended to have other high-risk features, explaining the subsequent extraordinary rise in the rate of cardiovascular mortality at 30 days. Interestingly, multivariable analysis indicated that CrCl is the major predictor of cardiovascular death, along with ST-segment depression. The other independent predictors of cardiovascular death, including heart failure and cardiac arrest, were present in <10% of the whole population.

Not surprisingly, age and renal dysfunction were risk factors for bleeding.5 Both patients with severe and moderate renal dysfunction displayed a significantly higher rate of major bleeding when compared with those with normal renal function. This finding was irrespective of the antithrombotic regimen used. Renal dysfunction was also an independent predictor of major bleeding. The significantly better safety of LMWH compared with UFH was unexpected. Indeed, none of the randomized trials comparing the safety of LMWH and UFH without GP IIb/IIIa inhibitors in patients with NSTE-ACS demonstrated a significant difference.1,19 In addition, the recent SYNERGY trial reported a modest increase of major bleeding in patients treated with enoxaparin compared with UFH.20 The absence of dose adjustment (1 mg/kg subcutaneously every 12 h in all patients), the additional intravenous bolus of 0.3 mg/kg in patients referred to the catheterization laboratory >8 h after the subcutaneous injection, crossover therapy, and the inclusion of older patients with severe renal dysfunction (CrCl <30 mL/min) may account for these differences. Our results suggest that dose adjustment was performed in patients with impaired renal function, although we cannot provide the effective dose used in our registry.

The higher rate of bleeding in patients who received GP IIb/IIIa inhibitors was expected and further confirms the findings of most clinical trials. The combination of LMWH and GP IIb/IIIa inhibitors has been shown to be well tolerated in NSTE-ACS patients,1,21 with bleeding rates ranging from 0.3 to 2.8%,14 which are close to that of GP IIb/IIIa inhibitors and UFH (overall 2.4% rate of major bleeding).22 This favourable trend is confirmed in the present study in patients with normal renal function (Figure 3). In-hospital bleeding rates tended to be lower with LMWH+GP IIb/IIIa inhibitors than those with UFH+GP IIb/IIIa inhibitors. However, no definite conclusion can be drawn given the small number of patients and the fact that the study was not designed or powered to assess this specific issue. Importantly, the use of UFH and GP IIb/IIIa inhibitors was an independent predictor of bleeding events.

Several studies have shown a clear relation between excessive anticoagulation and increased risk of bleeding.2325 This relationship was not assessed in the present study, which is an important limitation given the fact that enoxaparin with dose adjustment in patients with renal dysfunction provides a similar anticoagulation profile and tolerance as in patients with normal renal function.2 Moreover, the results of one study suggest that low anti-Xa activity in enoxaparin-treated patients with ACS is strongly and independently associated with early mortality, highlighting the need for complete anticoagulation as often as possible in these high-risk patients.26 The higher probability of early death with UFH compared with that of major bleeding in patients with CrCl <30 mL/min, who carry the highest risk of both bleeding and ischaemic events, further emphasizes the need for an optimal anticoagulation in this subset of patients. Finally, the trend for a greater difference between the estimated probability of in-hospital bleeding and the early death with LMWH treatment compared with UFH in patients with CrCl <30 mL/min favours the use of LMWH in this subgroup of patients.

Limitations of the study

GRACE is the largest multinational registry of patients with the complete spectrum of ACS. Standardized criteria for defining ACS and hospital outcomes are employed, along with quality control and audit measures. Individuals enroled in GRACE are more representative of unselected patients with ACS in the community than those enroled in randomized clinical trials, which are subject to stringent exclusion criteria. Although registry studies, by nature, are subjected to potential biases, they are valuable for studying real-world practice patterns and for detecting rare side-effects of drugs and drug interactions.27 As mentioned earlier, we are unable to measure the relation between anticoagulant dose and risk of bleeding. Finally, because our study is observational, the results should be considered as hypothesis generating rather than conclusive.

Conclusion

Renal dysfunction in patients with NSTE-ACS is associated with an excess of both cardiovascular death at 30 days and major in-hospital bleeding. In addition, bleeding complications are more frequent and outcomes apparently worse in patients treated with UFH than in those treated with LMWH, especially in patients with normal renal function and moderate renal dysfunction. Combination therapy with LMWH+GP IIb/IIIa inhibitors appears to be better tolerated than therapy with UFH+GP IIb/IIIa inhibitors. These findings should encourage further evaluation of these agents in randomized clinical trials in this high-risk population.

Acknowledgements

The authors would like to express their gratitude to the physicians and nurses participating in GRACE. The authors thank Sophie Rushton-Smith for providing editorial assistance. Further information about the project, along with the complete list of participants, can be found at www.outcomes.org/grace. GRACE is funded by an unrestricted educational grant from sanofi aventis to the Center for Outcomes Research, University of Massachusetts, Worcester, MA, USA.

Appendix: GRACE Advisory Board

Keith A.A. Fox, Joel M. Gore (GRACE Co-Chairs); Kim A. Eagle, Gabriel Steg (Publication Committee Co-Chairs); Giancarlo Agnelli, Frederick A. Anderson Jr, Álvaro Avezum, David Brieger, Andrzej Budaj, Marcus D. Flather, Robert J. Goldberg, Shaun G. Goodman, Christopher B. Granger, Dietrich C. Gulba, Enrique P. Gurfinkel, Brian M. Kennelly, Werner Klein, José López-Sendón, Gilles Montalescot, and Frans Van de Werf.

References

View Abstract