Author + information
- Received December 2, 2010
- Revision received April 21, 2011
- Accepted May 24, 2011
- Published online September 1, 2011.
- Elias B. Hanna, MD⁎,⁎ (, )
- Anita Y. Chen, MS†,
- Matthew T. Roe, MD†,
- Stephen D. Wiviott, MD‡,
- Caroline S. Fox, MD§,‖ and
- Jorge F. Saucedo, MD¶
- ↵⁎Reprint requests and correspondence
: Dr. Elias B. Hanna, Department of Medicine, Cardiovascular Section, Louisiana State University, 1542 Tulane Avenue, Room 323, New Orleans, Louisiana 70112
Objectives This study sought to evaluate the characteristics, therapies, and outcomes of patients with chronic kidney disease (CKD) presenting with non–ST-segment elevation myocardial infarction (NSTEMI) and managed with percutaneous coronary intervention (PCI). This specific population has not been evaluated previously.
Background Among patients with acute coronary syndrome, the presence of renal dysfunction is associated with an increased risk of death and major bleeding.
Methods We examined data on 40,074 NSTEMI patients managed with PCI who were captured by the ACTION (Acute Coronary Treatment and Intervention Outcomes Network) registry. Patients were divided according to baseline renal function in 4 groups: no CKD and CKD stages 3, 4, and 5.
Results Overall, 31.1% (n = 12,045) of patients with NSTEMI undergoing PCI had CKD. Compared with patients with normal renal function, CKD patients managed with PCI had significantly more history of myocardial infarction, heart failure, and more 3-vessel coronary artery disease. They received fewer antithrombotic therapies but were treated more frequently with bivalirudin. In addition, they had significantly higher rates of in-hospital mortality and major bleeding. CKD stage 4 was associated with the highest risk of adverse events relative to no CKD. The multivariable adjusted odds ratios of in-hospital mortality for CKD stages 3, 4, and 5 relative to no CKD were 2.0, 2.8, and 2.6, respectively (global p value <0.0001), and the analogous adjusted odds ratios of major bleeding were 1.5, 2.8, and 1.8, respectively (global p value <0.0001).
Conclusions CKD patients presenting with NSTEMI and managed with PCI have more comorbidities and receive guideline-recommended therapies less frequently than do patients without CKD. CKD is strongly associated with in-hospital mortality and bleeding in NSTEMI patients undergoing PCI.
- acute coronary syndrome
- chronic kidney disease
- myocardial infarction
- percutaneous coronary intervention
Among patients with acute coronary syndromes (ACS), the presence of renal dysfunction is associated with an increased risk of death and major bleeding (1–9). Registry data suggest that 14.5% to 42.9% of patients with non–ST-segment elevation myocardial infarction (NSTEMI) have moderate to severe chronic kidney disease (CKD) (2,3). Despite the high prevalence of CKD, large randomized trials that have addressed the benefit of an invasive strategy in ACS have excluded patients with advanced CKD (10,11). Subgroup analyses of these trials suggested a benefit of an invasive strategy in patients with mild CKD (12), but data assessing the interaction between CKD, outcomes, and the use of an invasive strategy, particularly percutaneous coronary intervention (PCI), is lacking. Although several retrospective analyses have suggested an increased risk of death and cardiac events in patients with CKD undergoing PCI as compared to patients with normal renal function (13,14), or in patients with CKD presenting with ACS (1–9), they have not specifically studied ACS population with CKD undergoing PCI.
Therefore, the aim of our study is to evaluate the baseline characteristics, treatment patterns, and in-hospital outcomes of patients with various stages of CKD presenting with NSTEMI who are managed with PCI. We focused specifically on patients who underwent PCI, a more homogeneous and potentially healthier population than the general NSTEMI population, and to examine the impact of CKD stage on medication use and outcomes within this select population.
The NCDR (National Cardiovascular Data Registry)–ACTION-GWTG (Acute Coronary Treatment and Intervention Outcomes Network–Get With The Guidelines) registry is an ongoing national database of NSTEMI and ST-segment elevation myocardial infarction (STEMI) that began enrolling patients January 1, 2007. The population for this study was derived from the 162,361 patients enrolled in the ACTION-GWTG registry from January 1, 2007, to December 31, 2009, at 397 ACTION-GWTG hospitals. Patients were included in the registry if they had acute ischemic symptoms, positive cardiac biomarkers, and were admitted within 24 h of these symptoms. Patients within the ACTION-GWTG registry were included in this analysis if they had NSTEMI and underwent an invasive strategy with subsequent PCI. Patients were ineligible for this analysis if they presented with STEMI (n = 63,816), did not undergo cardiac catheterization or the catheterization status was missing (n = 24,899), were not found to have significant coronary artery disease (CAD) on catheterization (n = 6,387), or had significant CAD but underwent coronary artery bypass grafting (n = 10,288) or no revascularization (n = 16,067), were admitted to facilities with no cardiac catheterization capability or were transferred in from facilities with no cardiac catheterization capability (n = 247), or if data for serum creatinine was missing (n = 583). Thus, the overall sample for our analysis consisted of 40,074 patients.
Glomerular filtration rate (GFR) was estimated via the Modification of Diet in Renal Disease equation. Stage 3 CKD is defined as GFR between 30 and 59 ml/min/1.73 m2, stage 4 CKD is defined as GFR between 15 and 29 ml/min/1.73 m2, and stage 5 CKD as GFR <15 ml/min/1.73 m2 or dialysis therapy. Patients were divided based on baseline renal function in 4 groups: 1) no CKD or CKD stage 1 or 2 (referred to as “no CKD” for the purpose of this study); 2) CKD stage 3; 3) CKD stage 4; and 4) CKD stage 5.
The primary outcome was the short-term in-hospital death. Moreover, we examined rates of major bleeding. Major bleeding was defined as an absolute hemoglobin drop of ≥4 g/dl (baseline to nadir), intracranial hemorrhage, documented or suspected retroperitoneal bleed, any red cell blood transfusion with a baseline hemoglobin ≥9 g/dl, or any red cell transfusion with a baseline hemoglobin <9 g/dl and a suspected bleeding event. Bleeding in coronary artery bypass graft patients was included as a bleed only if it occurred before surgery. Bleeding during or after surgery was not included in the bleeding definition.
Baseline demographics, clinical characteristics, in-hospital treatments, and outcomes are shown across baseline CKD groups and p-trend values are reported. Continuous variables are reported as medians and 25th and 75th percentiles and categorical variables are reported as percentages. The p values are based on chi-square rank-based group means score statistics for all categorical variables and on chi-square 1° of freedom rank correlation statistics for all continuous or ordinal variables.
Logistic generalized estimating equations modeling was used to estimate the adjusted associations of CKD stages with outcomes. Variables used for in-hospital mortality adjustment were from the validated ACTION-GWTG in-hospital mortality model (15): age, prior peripheral arterial disease, systolic blood pressure (SBP) on presentation, heart rate on presentation, heart failure (HF) or shock on admission (HF only, shock only, or HF with shock, vs. none), electrocardiographic findings (STEMI, ST-segment depression, or transient ST-segment elevation vs. no ST-segment changes), initial serum creatinine, and initial troponin ratio. Variables used for major bleeding adjustment were from the validated ACTION-GWTG in-hospital major bleeding model (16): age, female sex, diabetes, prior peripheral arterial disease, home warfarin use, body weight, heart rate on presentation, SBP on presentation (≤130 mm Hg, 130 to 160 mm Hg, and ≥160 mm Hg), HF on presentation (HF only, shock only or HF with shock, vs. none), electrocardiographic findings (STEMI, ST-segment changes vs. no ST-segment changes), initial serum creatinine, and baseline hemoglobin <12 g/dl (vs. ≥12 g/dl). STEMI and initial serum creatinine were dropped from both models because STEMI patients were excluded from this analysis and CKD is a function of serum creatinine. Adjusted associations for outcomes were displayed as odds ratios (ORs) (95% confidence intervals [CIs]) and a global comparison p value was calculated and reported. All analyses were performed using SAS software (version 9.2, SAS Institute, Cary, North Carolina).
Study sample characteristics
Overall, 40,074 patients with NSTEMI undergoing PCI were included; the peak troponin median was 30.4 times the upper limit of normal (Online Table 1). Of those patients, 28,029 patients (69.9%) had no CKD on presentation, 10,168 patients (25.4%) had CKD stage 3, 930 patients (2.3%) had CKD stage 4, and 947 patients (2.4%) had CKD stage 5, of which 790 patients were on chronic dialysis. With progressively more advanced CKD, patients were more likely to have a history of diabetes mellitus, prior heart failure, prior myocardial infarction, prior coronary revascularization, and peripheral arterial disease. As compared to patients with no CKD, patients with CKD were more likely to have HF and shock on presentation (Table 1).
After accounting for patient-level contraindications to individual medication classes, the early use of antiplatelet and anticoagulant therapies was lower among patients with CKD. Only bivalirudin was used more frequently in patients with progressively more advanced CKD, including CKD stage 5. In addition, the use of approved ACS therapies (statin and beta-blockers) was lower among patients with worse CKD (Table 2). The use of angiotensin-converting enzyme inhibitors and/or angiotensin-receptor blockers was lowest among patients with CKD stage 4.
Patients with progressively more advanced CKD had more 3-vessel CAD and more left main disease than did patients with no CKD (Table 3). Patients with CKD stage 4 had the highest prevalence of 3-vessel CAD.
Patients with various stages of CKD had higher rates of in-hospital mortality and major bleeding than did patients with no CKD (Table 4). Notably, patients with CKD stage 4 had the highest rates of these adverse in-hospital clinical events, with strikingly worse outcomes than patients with CKD stage 5. The association of CKD stages 3, 4, and 5 with increased risks of death and major bleeding relative to no CKD remained significant after adjustment for baseline characteristics using the multivariable models (Table 4).
Overall, 30.1% of patients with NSTEMI undergoing PCI in this real-world registry had CKD. This is lower than the prevalence of CKD among all NSTEMI patients enrolled in the ACTION registry (overall prevalence of CKD: 42.9%) (3) and is probably explained by the fact that an invasive strategy, particularly PCI, is less frequently used in patients with CKD. Patients with CKD undergoing PCI had more extensive CAD than patients without CKD, with a 3-vessel CAD prevalence of 35% to 41% and left main disease prevalence of ∼10%. In addition, CKD patients undergoing PCI had significantly more comorbidities than did patients without CKD, and the present study documented lower use of antithrombotic therapies and cardioprotective medications among CKD patients. Thus, there seems to be a paradigm on the management of CKD patients, where the highest-risk patients receive less guideline-recommended measures, even when those patients are determined to be “healthy” enough to undergo PCI procedures. This study is unique in that all patients underwent PCI, yet there were significant disparities in the concomitant use of evidence-based medications by CKD stage.
The rates of adverse events were markedly higher in patients with CKD as compared to patients without CKD. This has been suggested by other registry data (1,17). Notably, patients with CKD stage 4 had a higher rate of mortality and major bleeding as compared to CKD stages 3 or 5. This finding has not been described previously in this patient population, and although noteworthy, it should be interpreted carefully. CKD stages 4 and 5 subgroups consisted of fewer patients than other subgroups in this analysis, which precludes us from drawing definitive conclusions about this result. Also, patients with CKD stage 4 in this study were older than patients with CKD stage 5. In fact, after risk adjustment for multiple covariates that included older age of patients with CKD stage 4, the mortality OR of patients with CKD stage 4, when compared with patients with no CKD, was similar to that of CKD stage 5. By contrast, several factors may contribute to better outcomes in patients with CKD stage 5. It has been previously shown that patients with CKD stages 3 and 4 are 5× to 10× more likely to die than they are to reach CKD stage 5 or hemodialysis (18,19). As survivors of earlier stages of CKD, patients with CKD stage 5 may thus represent a selection bias. Furthermore, as compared to CKD stage 4 patients, contrast-induced nephropathy may be less consequential in CKD stage 5 patients who are already receiving dialysis, which is the case of most patients with CKD stage 5 in this study. In addition, dialysis improves the hemostatic defect associated with advanced CKD, which may explain the lower bleeding risk associated with CKD stage 5 (20).
Although CKD was associated with an increased mortality risk in patients with NSTEMI undergoing PCI in our study, this result neither opposes nor favors the performance of PCI in this population. In fact, our reported mortality rates are lower than those reported for the overall CKD population with NSTEMI in the ACTION registry (the mortality rates for no CKD, CKD stages 3a, 3b, 4, and 5 were 1.8%, 4.8%, 8.6%, 13.4%, and 12.4%, respectively) (3). This may be related to the selection bias, wherein only the healthier-perceived patients, those with less extensive comorbidities, or those with more favorable anatomy undergo PCI as has been described previously (21), but it may also be related to the more aggressive revascularization therapy in these patients.
There are several potential explanations for the increased risk of adverse outcomes across all CKD categories undergoing PCI. Platelet dysfunction related to CKD and dosing errors of antithrombotic agents are factors that increase bleeding risk among CKD patients and contribute to increased mortality (22). In fact, excess dosing was shown to be common among patients presenting with ACS, particularly if they also had CKD, and was a major contributor to major bleeding events and subsequent mortality (22). By contrast, CKD patients have more extensive atherosclerosis and more comorbidities including history of CAD, HF, peripheral arterial disease, and anemia, which partly explain their prognosis. In addition, the lower usage of proven medical therapies may have played a role in worsening outcomes. Furthermore, contrast-induced nephropathy occurs more frequently in patients with CKD and is strongly associated with mortality (23–26). However, because mortality of the overall CKD population is high regardless of invasive management, the isolated impact of contrast-induced nephropathy on outcomes needs to be further investigated.
The increased use of bivalirudin in patients with CKD stages 4 and 5 is noteworthy. Even though bivalirudin was associated with a lower bleeding risk than unfractionated heparin and low-molecular-weight heparin in patients without CKD as well as in patients with CKD stage 3 in the ACUITY (Acute Catheterization and Urgent Intervention Triage Strategy) trial (27), this trial excluded patients with CKD stages 4 and 5. Similarly, other large trials that addressed bivalirudin therapy excluded patients with creatinine >4 mg/dl (28,29). Thus, the safety of bivalirudin in patients with advanced CKD is not well established and the dosage adjustment of the infusion is unclear, particularly as bivalirudin is renally cleared and its half-life is significantly increased in advanced CKD.
Study strengths and limitations
The major strength of this study is the use of a large real-world multicenter registry. However, certain limitations are notable. Being a retrospective registry analysis, there is an inherent selection bias, under-reporting bias, and confounding related to unmeasured variables. We selected NSTEMI patients who underwent PCI, an inherently healthier population than the overall NSTEMI population; thus, our outcome data may not apply to all patients with NSTEMI and CKD and neither oppose nor favor the performance of PCI in these patients. Moreover, we only evaluated short-term in-hospital outcomes, but previous evidence suggests that these poor outcomes are likely seen over the long-term follow-up as well (1,2,6–8). The relatively small number of patients with CKD stages 4 and 5 precludes definite conclusions about outcomes in these subgroups.
CKD patients presenting with NSTEMI and managed with PCI have more comorbidities than patients without CKD have and are less frequently treated with the recommended ACS therapies. CKD is strongly associated with in-hospital mortality and bleeding events in patients undergoing PCI; CKD stage 4, in particular, is associated with the highest mortality and most adverse bleeding outcomes. Further research is needed to elucidate the role of aggressive revascularization strategies in improving outcomes of CKD patients and their potential risk.
For the troponin ratio data, please see the online version of this paper.
Dr. Roe has received research funding and has been a consultant and member of the Speakers' Bureaus for the companies that fund the ACTION registry via the American College of Cardiology—Bristol-Myers Squibb/sanofi aventis and Schering-Plough; has received research funding from Eli Lilly, Roche, Bristol-Myers Squibb, American College of Cardiology, and the American Heart Association; and consulting fees or honoraria from KAI Pharmaceuticals, Bristol-Myers Squibb, Sanofi Aventis, Merck, Orexigen, Holsinn Pharmaceuticals, AstraZeneca, and Regeneron. Dr. Wiviott has been a consultant to sanofi-aventis, Bristol-Myers Squibb, and AstraZeneca; has received research support and/or honoraria from Eli Lilly, Daiichi Sankyo, AstraZeneca, Schering-Plough, Merck, and Pfizer Inc.; and does CME speaking for Eli Lilly, AstraZeneca, and Daiichi Sankyo. Dr. Saucedo has been a consultant to Eli Lilly, Bristol-Myers Squibb/sanofi-aventis, and The Medicines Company; and has received research support and/or honoraria from Schering-Plough, Pfizer Inc., Merck, Eli Lilly, Abbott, Boston Scientific, and Medtronic. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- acute coronary syndromes
- coronary artery disease
- confidence interval
- chronic kidney disease
- glomerular filtration rate
- heart failure
- non–ST-segment elevation myocardial infarction
- odds ratios
- percutaneous coronary intervention
- systolic blood pressure
- ST-segment elevation myocardial infarction
- Received December 2, 2010.
- Revision received April 21, 2011.
- Accepted May 24, 2011.
- American College of Cardiology Foundation
- Fox C.S.,
- Muntner P.,
- Chen A.Y.,
- et al.,
- ACTION Registry
- Gibson C.M.,
- Pinto D.S.,
- Murphy S.A.,
- et al.,
- TIMI Study Group
- Gibson C.M.,
- Dumaine R.L.,
- Gelfand E.V.,
- et al.,
- TIMI Study Group
- Santopinto J.J.,
- Fox K.A.,
- Goldberg R.J.,
- et al.,
- GRACE Investigators
- Best P.J.,
- Lennon R.,
- Ting H.H.,
- et al.
- Chin C.T.,
- Chen A.Y.,
- Wang T.Y.,
- et al.
- Hirsch A.,
- Windhausen F.,
- Tijssen J.G.,
- et al.,
- ICTUS Investigators
- Rihal C.S.,
- Textor S.C.,
- Grill D.E.,
- et al.
- Lincoff A.M.,
- Bittl J.A.,
- Harrington R.A.,
- et al.,
- REPLACE-2 Investigators