Relationship between renal function and outcomes in high-risk patients with non-ST-segment elevation acute coronary syndromes: Results from SYNERGY

Results Of 9838 patients with a CrCl value, 70.6% ( N = 6950) had CrCl ≥ 60 mL/min, 27.8% ( N = 2732) had CrCl 30–59 mL/min, and 1.6% ( N = 156) had CrCl < 30 mL/min. No randomized treatment by CrCl interaction test was found to be statistically significant, suggesting renal insufficiency affected enoxaparin and unfractionated heparin outcomes similarly. After adjustment, CrCl was an independent predictor of 30-day death or MI (OR 1.06, 95% CI 1.03–1.09), TIMI major bleeding (OR 1.06, 95% CI 1.02–1.10), GUSTO severe bleeding (OR 1.10, 95% CI 1.03–1.17), and transfusion (OR 1.07, 95% CI 1.04–1.11). Conclusions Patients with CKD had higher rates of 30-day death or MI and bleeding than those without CKD, regardless of randomized antithrombin therapy. While this analysis suggests that there is a rise in bleeding events as CrCl falls for patients in either treatment group, it is unknown whether a reduction in dose would decrease bleeding risk. Keywords Acute coronary syndromes Bleeding Chronic kidney disease Enoxaparin Heparin Renal insufficiency 1 Introduction Chronic kidney disease (CKD), defined as a glomerular filtration rate < 60 mL/min/1.73 m 2 , is a well-recognized risk factor for the development of cardiovascular disease [1] . Prior studies have demonstrated that patients with renal insufficiency and either ST-segment elevation (STE) acute coronary syndrome (ACS) [2] , non-ST-segment elevation (NSTE) ACS [3,4] , or those undergoing percutaneous coronary intervention (PCI) [5–7] have less favorable outcomes than those with normal renal function. In the TACTICS TIMI 18 trial of 2220 high-risk patients with NSTE managed with unfractionated heparin (UFH) and tirofiban followed by either early angiography or a conservative strategy of medical management followed by symptom-driven or positive stress test-driven angiography, decreasing creatinine clearance (CrCl) was associated with an increased risk of death, myocardial infarction (MI), or rehospitalization for ACS and major bleeding but not with the end point of death or MI [4] . Previously, we described an increased risk of death, MI, or need for urgent revascularization and bleeding in 6826 lower-risk patients with NSTE ACS treated primarily with a medical management approach that included either enoxaparin or UFH [3] . In this analysis, we report the effect of renal function on efficacy and safety outcomes in 9838 high-risk patients with NSTE ACS randomly assigned to either enoxaparin or UFH enrolled in the Superior Yield of the New Strategy of Enoxaparin, Revascularization, and Glycoprotein IIb/IIIa inhibitors (SYNERGY) trial, which included administration of glycoprotein (GP) IIb/IIIa inhibitors in more than 50% of patients, surgical revascularization in 19%, and PCI in 47% of patients. 2 Methods The design and results of the SYNERGY trial have been previously described [8,9] . Briefly, 9978 patients with NSTE ACS were randomly assigned and treated with either enoxaparin (median duration 36 h [12, 81.2]) or UFH (median duration 35 h [16.9, 70.4]) and an intended early angiography treatment strategy in a multicenter, open-label study. Aspirin was administered to all patients, and the use of other medical therapies such as GP IIb/IIIa inhibitors and clopidogrel was encouraged. The dose of enoxaparin was 1 mg/kg subcutaneous every 12 h, and patients with CrCl < 30 mL/min were to have been excluded from the trial. Therefore, in accordance with enoxaparin product labeling, no dosing adjustments were recommended for enrolled patients. Eligible patients were those meeting at least 2 of the following criteria for enrollment: age 60 years or older, elevated troponin or creatine kinase elevation above the upper limit of normal, or electrocardiographic changes (either new ST-segment depression ≥ 1 mm or transient ST-segment elevation ≥ 1 mm in at least 2 contiguous leads). Patients with severe renal insufficiency (CrCl < 30 mL/min) were to have been excluded [8] . In SYNERGY, the median time to angiography was 22 h, more than 50% of patients received a GP IIb/IIIa inhibitor, 63% received clopidogrel, 47% underwent PCI, and 19% underwent surgical revascularization during the initial hospitalization. Overall, enoxaparin was noninferior to UFH for the prevention of death or MI at 30 days but increased the risk of major bleeding [9] . Because there are no published prospective trials and few published retrospective analyses of large clinical trials that include a comparison of both bleeding and efficacy outcomes between enoxaparin and UFH in patients with renal dysfunction, we conducted a post-hoc analysis of the SYNERGY database. Glomerular filtration rate at the time of randomization was estimated using the Cockcroft–Gault equation (CrCl = 140 – age [yrs] × actual body weight [kg]/72 × serum creatinine [mg/dL] for men or × 0.85 for women) [10] , as this is the recommended method to determine enoxaparin dosing. One-hundred and forty patients with missing values for CrCl were excluded from this analysis. End points were clinical events committee-adjudicated death or non-fatal MI at 30 days, TIMI major bleeding, GUSTO severe bleeding, and non-coronary artery bypass graft surgery (CABG) packed red blood cell transfusions during the initial hospitalization. Non-fatal MI was defined as an elevation in creatine kinase-MB > 2 times the upper limit of the normal range [8] . Previously developed multivariable logistic models for 30-day death or MI [9] and in-hospital post-randomization TIMI major bleeding [11] were used in this analysis. Randomized treatment, age, sex, height, weight, time from symptom onset to randomization, race, region, smoking status, baseline serum creatinine, Killip class, systolic blood pressure, diastolic blood pressure, presence of rales, T-wave inversion/ST segment elevation/depression on baseline 12-lead electrocardiogram, and history of diabetes, coronary artery disease, angina, peripheral arterial disease, MI, heart failure, CABG, PCI, as well as inclusion criteria at time of enrollment were evaluated using both stepwise and backward selection to determine the best model for the end point of 30-day death or MI. These covariates were evaluated for linearity to determine if a cubic spline or some other type of transformation was necessary prior to inclusion into the model. Other multivariable logistic models were developed in a similar manner for GUSTO severe bleeding and transfusion [11] . In order to evaluate the significance of renal insufficiency on death or MI and bleeding outcomes, CrCl was added to the model as a continuous variable for each of the outcomes after adjusting for a variety of baseline factors. The linearity assumption of CrCl with each of the outcomes was evaluated using restricted cubic splines. The model chi-square with the transformation included was compared to the model chi-square assuming linearity. If significant nonlinearity had been detected, the plot of the logit of the outcome versus CrCl would have been evaluated to help in finding the best transformation needed. Linearity was not violated in any of these 4 models, so no transformations were included. Adjusted odds ratios, as well as 95% Wald confidence intervals, were obtained from the models. Adjusted plots of predicted probability of 30-day death or MI and bleeding events by CrCl were produced by obtaining average predicted values across all patients. In order to obtain the 95% confidence intervals for the adjusted plots, average predicted values of 100 bootstrapped samples were obtained from the original data set. Analyses were performed using SAS version 8.2 (SAS, Cary, North Carolina). A p value ≤ 0.05 was considered statistically significant. 3 Results The demographic data, medications, and in-hospital procedures of patients with CKD enrolled in SYNERGY are described in Tables 1 and 2 . Of 9838 patients enrolled in SYNERGY with a CrCl value, 70.6% ( N = 6950) had CrCl ≥ 60 mL/min (normal or stage 2 CKD), 27.8% ( N = 2732) had CrCl 30–59 mL/min (stage 3 CKD, moderate renal insufficiency), and 1.6% ( N = 156) had CrCl < 30 mL/min (stages 4 or 5 CKD, severe renal insufficiency). The median CrCl was 74.9 mL/min (56.4, 97.2). The median initial dose of enoxaparin administered in all 3 groups was 1 mg/kg. Eight patients with CrCl < 30 mL/min received only 1 dose of enoxaparin, and the median number of doses received among the remainder of patients with CrCl < 30 mL/min was 5 doses (range 2 to 26 doses). The unadjusted rates of 30-day death or MI, as well as TIMI major bleeding, GUSTO severe bleeding, and transfusion, are described in Table 3 . Rates of 30-day death and transfusion were higher in patients with moderate or severe renal insufficiency. After adjusting for demographic and clinical variables, no treatment by CrCl interaction term was found to be significant, indicating similar relationships for enoxaparin and UFH with respect to all treatment outcomes and CrCl. This interaction term was removed from the models. Results of multivariate logistic regression suggest that for every 10-unit decrease in CrCl, there was a 6% greater risk of 30-day death or MI for the entire study population ( Table 4 ) (OR 1.06, 95% CI 1.03–1.09, p < 0.001), as well as a 7% increased risk for those patients who underwent diagnostic cardiac catheterization (95% CI 1.04–1.11, p < 0.001). Creatinine clearance was not a significant predictor of death or MI in the population of patients who underwent PCI (OR 1.06, 95% CI 0.97–1.16). The risk of bleeding as measured by the rates of either TIMI major bleeding ( Table 5 , OR 1.06, 95% CI 1.02–1.10), GUSTO severe bleeding ( Table 6 , OR 1.10, 95% CI 1.03–1.17), or risk of transfusion ( Table 7 , OR 1.07, 95% CI 1.04–1.11) increased with increasing renal dysfunction. The risk of TIMI major bleeding was greater for those randomized to enoxaparin, but this was not the case for GUSTO severe bleeding or transfusion ( Tables 5–7 ). The adjusted probability of 30-day death or MI, TIMI major bleeding, GUSTO severe bleeding, and transfusion by CrCl for patients randomized to UFH and enoxaparin are depicted in Figs. 1–4 . These adjusted analyses suggest similar probabilities of 30-day death or MI as well as transfusion across the spectrum of renal function with enoxaparin compared with UFH ( Figs. 1 and 4 ), while suggesting an increasingly greater probability of both TIMI major and GUSTO severe bleeding with enoxaparin as renal function worsens ( Figs. 2 and 3 ); albeit, no significant CrCl by treatment interaction could be detected in these 2 models. 4 Discussion In SYNERGY, the risk of TIMI major bleeding—but not GUSTO severe bleeding or transfusions—increased with enoxaparin as compared with UFH [9] . Patients with CKD—a known risk factor for bleeding with antithrombotic therapy—made up a large proportion (almost 30%) of patients enrolled in SYNERGY. While CKD increased the risk of 30-day death or MI and bleeding events, the adjusted risk was similar in patients with CKD randomized to either enoxaparin or UFH, as no interaction between CrCl and treatment was detected in the models. The absence of an interaction between anticoagulant strategy and CrCl was observed despite no enoxaparin dosing adjustment and with a longer median duration of treatment in the group of patients with renal dysfunction ( Table 2 ). Although statistically significantly associated with outcomes, the model chi-square for several baseline characteristics was greater than it was for CrCl. While CrCl was predictive of bleeding events in the population of patients undergoing angiography, it was not found to be predictive in the subgroup of patients undergoing PCI, most likely secondary to power. The risk of bleeding in patients with CrCl < 60 mL/min also increased when compared with patients without CKD in ACUITY [12] . However, rates of ACUITY major bleeding were reported to be similar in patients randomized to bivalirudin, UFH, or enoxaparin in this subgroup, perhaps because of the brief duration of study drug administration [12] . In OASIS-5, the rate of TIMI major bleeding with enoxaparin was 9.2% in 282 patients with CrCl < 30 mL/min [13] despite dosing reduction of enoxaparin to 1 mg/kg every 24 h (implemented during the course of the trial). This difference in major bleeding between enoxaparin and fondaparinux increased with increasing renal dysfunction [14] , perhaps related to the longer duration of study drug administration in OASIS-5 (median 5.2 days) compared with either SYNERGY or ACUITY. Patients with CKD have higher anti-Xa levels than patients without CKD. In TIMI 11A, 7 patients with NSTE ACS and CrCl < 40 mL/min who received a dose of enoxaparin of 1.25 mg/kg every 12 h had higher peak and trough anti-factor-Xa concentrations, as well as increased hemorrhagic events, compared with patients whose CrCl was ≥ 40 mL/min [15] . A dosing reduction of 20% of enoxaparin in patients with moderate renal dysfunction (CrCl 30–50 mL/min) to 0.8 mg/kg every 12 h, as well as a reduction of 34% to 0.66 mg/kg every 12 h for patients with CrCl < 30 mL/min, have been suggested based upon the results of a pharmacokinetic study of anti-factor Xa levels in 158 patients with NSTE ACS and moderate and severe renal dysfunction in a study by Hulot and colleagues [16] . This dose reduction was designed to keep anti-factor Xa levels between 0.5 IU/mL and 1.2 IU/mL. 4.1 Clinical implications The observations in this study that patients with renal dysfunction had worse outcomes is consistent with prior studies but now extended to a contemporary clinical trial with invasive management and aggressive use of evidence-based therapies. No statistical interaction was observed for the risk for cardiac outcomes (death or MI) and bleeding events, suggesting that the outcome by randomized treatment assignment and renal dysfunction were similar as that observed in the overall trial. These results are similar to those reported from ESSENCE and TIMI 11b [3] . These data support the use of enoxaparin in patients with moderate renal dysfunction (CrCl ≥ 30 mL/min). These analyses of CrCl as a continuous variable show the potential perils of dichotomizing such factors when making patient treatment decisions. Figs. 2 and 3 show that risk of bleeding increases as CrCl decreases across the entire spectrum. Although the current enoxaparin product label recommends a dosing adjustment to 1 mg/kg every 24 h in patients with CrCl < 30 mL/min, it is unknown whether a dosing reduction in patients with more moderate renal impairment would decrease the bleeding risk in these patients. No trials of dosing adjustments of UFH in patients with CKD have been reported. Additional trials comparing the efficacy, safety, and dosing of anticoagulants and antiplatelet agents in patients with CKD are warranted. The ongoing VALIDE trial, sponsored by the French Cardiology Society, is prospectively studying bleeding events and anti-factor Xa levels in patients with NSTE ACS and CrCl 30–50 mL/min receiving adjusted doses of enoxaparin compared with patients with normal renal function receiving enoxaparin 1 mg/kg twice daily (ClinicalTrials.gov Identifier: NCT00412802). 5 Conclusions Renal dysfunction is common in high-risk ACS patients despite aggressive use of evidence-based therapies and invasive management strategies. The current analyses show that, after adjusting for demographic and clinical variables, no treatment by CrCl interaction term was found to be significant, indicating similar relationships for enoxaparin and UFH with respect to all treatment outcomes and CrCl, particularly in patients with modest renal dysfunction. More work is needed to understand if dose adjustment in patients with CrCl 30–60 mL/min may be prudent. Acknowledgement The authors of this manuscript have certified that they comply with the Principles of Ethical Publishing in the International Journal of Cardiology [17] . References [1] M.J. Sarnak A.S. Levey A.C. Schoolwerth Kidney disease as a risk factor for development of cardiovascular disease: a statement from the American Heart Association Councils on Kidney in Cardiovascular Disease, High Blood Pressure Research, Clinical Cardiology, and Epidemiology and Prevention Circulation 108 2003 2154 2189 [2] C.M. Gibson D.S. Pinto S.A. Murphy Association of creatinine and creatinine clearance on presentation in acute myocardial infarction with subsequent mortality J Am Coll Cardiol 42 2003 1535 1543 [3] S.A. Spinler S.M. Inverso M. Cohen Safety and efficacy of unfractionated heparin versus enoxaparin in patients who are obese and patients with severe renal impairment: analysis from the ESSENCE and TIMI 11B studies Am Heart J 146 2003 33 41 [4] J.L. Januzzi C.P. Cannon P.M. DiBattiste Effects of renal insufficiency on early invasive management in patients with acute coronary syndromes (The TACTICS-TIMI 18 Trial) Am J Cardiol 90 2002 1246 1249 [5] P.J. Best R. Lennon H.H. Ting The impact of renal insufficiency on clinical outcomes in patients undergoing percutaneous coronary interventions J Am Coll Cardiol 39 2002 1113 1119 [6] S.S. Naidu F. Selzer A. Jacobs Renal insufficiency is an independent predictor of mortality after percutaneous coronary intervention Am J Cardiol 92 2003 1160 1164 [7] D.P. Chew A.M. Lincoff H. Gurm Bivalirudin versus heparin and glycoprotein IIb/IIIa inhibition among patients with renal impairment undergoing percutaneous coronary intervention (a subanalysis of the REPLACE-2 trial) Am J Cardiol 95 2005 581 585 [8] SYNERGY Executive Committee The SYNERGY trial: study design and rationale Am Heart J 143 2002 952 960 [9] J.J. Ferguson R.M. Califf E.M. Antman Enoxaparin vs unfractionated heparin in high-risk patients with non-ST-segment elevation acute coronary syndromes managed with an intended early invasive strategy: primary results of the SYNERGY randomized trial JAMA 292 2004 45 54 [10] D.W. Cockcroft M.H. Gault Prediction of creatinine clearance from serum creatinine Nephron 16 1976 31 41 [11] S.V. Rao K.W. Mahaffey K. Chiswell Baseline predictors of bleeding in patients treated for acute coronary syndromes with either enoxaparin or unfractionated heparin: results from the superior yield of the new strategy of enoxaparin, revascularization and glycoprotein IIb/IIIa inhibitors (SYNERGY) Trial J Am Coll Cardiol 45 Suppl A 2005 247A Abstract 850-5 [12] G.W. Stone B.T. McLaurin D.A. Cox Bivalirudin for patients with acute coronary syndrome N Engl J Med 355 2006 2203 2216 [13] K.A. Fox S.R. Mehta Impact of renal function on the benefit-risk ratio of fondaparinux in patients with non-ST elevation acute coronary syndromes in the OASIS-5 trial Paper presented at: XXI International Society on Thrombosis and Haemostasis Congress; July 6–12 2007; Geneva, Switzerland 2007 Abstract P-M-682 [14] K.A.A. Fox J.P. Bassand S.R. Mehta Influence of renal function on the efficacy and safety of fondaparinux relative to enoxaparin in non-ST-segment elevation acute coronary syndromes Ann Intern Med 147 1997 304 310 [15] R.C. Becker F.A. Spencer M. Gibson Influence of patients characteristics and renal function on factor Xa inhibition pharmacokinetics and pharmacodynamics after enoxaparin administration in non-ST-segment elevation acute coronary syndromes Am Heart J 143 2002 753 759 [16] J.S. Hulot G. Montalescot P. Lechat P. Collet A. Ankri S. Urien Dosing strategy in patients with renal failure receiving enoxaparin for the treatment of non-ST-segment elevation acute coronary syndrome Clin Pharmacol Ther 77 2005 542 552 [17] A.J. Coats Ethical authorship and publishing Int J Cardiol 131 2009 149 150