Short- and Long-Term Clinical Outcomes of Coronary Drug- Eluting Stent Recipients Presenting with Chronic Renal Disease

Demographic, clinical and procedural characteristics of the study subgroups.
Lesion characteristics of each study subgroup.
Two-year actuarial survivals free from target vessel revascularization of the study subgroups.
Two-year actuarial survivals free from death + myocardial infarction of the study subgroups.
Two-year actuarial survivals free from major adverse cardiac events of the study subgroups.
In-hospital adverse clinical events.
One- and two-year Kaplan-Meier survival estimates in patients with a GFR > 60 ml/min versus a GFR < 60 ml/min.
Numbers (%) of definite and probable stent thromboses* in
patients with a GFR ≥ 60 ml/min versus a GFR < 60 ml/min.
Actuarial survivals free from target vessel revascularization at 18 months in recipients of sirolimus- versus paclitaxel-eluting stents,
grouped by a GFR ≥ 60 ml/min versus a GFR < 60 ml/min.

*†‡Gregory J. Mishkel, MD, ‡Joji J. Varghese, MD, †Anna L. Moore, MPH, *†‡Frank Aguirre, MD,
‡Stephen J. Markwell, MS, *†‡Marc Shelton, MD

Chronic renal disease (CRD) is a common health concern in wealthy countries, with approximately 20 million American adults currently affected.1 While the association between renal insufficiency and increased risk of adverse non-fatal and fatal cardiovascular events has been firmly established,2–5 the relationship between renal function and risk of restenosis after coronary artery stenting is less certain.6,7

Furthermore, a recent U.S. Food and Drug Administration panel examining the safety of drug-eluting stents (DES) focused on the proportion of current recipients who would have been excluded from the original randomized trials, and showed the superiority of DES over bare-metal stents (BMS).8 The majority of patients enrolled in the DES landmark pivotal trials had preserved renal function.9–12 The few studies of the effects of CRD on clinical outcomes after implantation of DES have included small numbers of patients, and have been limited by: (a) ≤ 1-year follow-up periods; (b) the exclusive use of a single stent type; or (c) the exclusion of patients with severe renal insufficiency.13–15

These limitations prompted us to analyze the relationship between baseline renal function and in-hospital and longterm clinical outcomes in a large series of consecutive patients who underwent implantations of DES. We therefore sought to review our database with the following goals: (1) to report 2-year clinical outcomes of a large cohort of DES patients; (2) to report outcomes based on the degree of renal insufficiency, including dialysis; (3) to compare sirolimus versus paclitaxel outcomes.

Patient Population and Methods

Our retrospective review of our hospital databases identified 2,758 consecutive patients who underwent DES placement in our tertiary care medical center between May 1, 2003 and December 31, 2004, and who met the following criteria: (a) underwent a first DES procedure; (b) did not have a BMS implanted in that procedure; and (c) a preprocedural measurement of serum creatine was available. The study protocol was approved by the local Investigational Review Board.

Renal function assessment. Glomerular filtration rate (GFR) was calculated using the Cockroft-Gault formula [(140 – age) (weight in kg) (0.85 for women) / (72 x serum creatine)],16 from measurements of height, weight and serum creatine (mg/dl) made before the procedure, and retrospectively obtained by chart review. Patients were stratified into 5 groups according to their renal function: (1) GFR ≥ 90 ml/min (normal function); (2) < 90 and ≥ 60 ml/min (mild insufficiency); (3) < 60 and ≥ 30 ml/min (moderate insufficiency); (4) < 30 ml/min and not on dialysis (severe insufficiency); and (5) renal failure treated with dialysis.17 The size of the severe renal insufficiency group was relatively small (n = 65), as was the dialysis group (n = 23). Therefore, for selected analyses, all patients with a GFR ≥ 60 ml/min (normal renal function) were compared with all patients with a GFR < 60 ml/min (chronic renal disease) who were not treated with long-term dialysis.

Interventional procedures. PCIs were performed according to the practices and preferences of each operator, including the administration of intraprocedural heparin or bivalirudin and the use of a glycoprotein (GP) IIb/IIIa inhibitor. After having received a first 325 mg dose prior to the procedure, all patients were advised to continue treatment with aspirin indefinitely. Clopidogrel, 300–600 mg, was administered at the time of PCI, and continued for ≥ 6 months after DES implantation. Quantitative coronary angiographic measurements were performed using a computer-based edge-detection system (HeartLab, Inc., Westerly, Rhode Island).

Clinical follow up and study endpoints. The procedural and in-hospital outcome data were obtained from our American College of Cardiology National Cardiovascular Data Registry and Society of Thoracic Surgeons databases. Patients undergoing implantation of DES are routinely followed at our institution at 6 months, 1 year and annually thereafter by means of telephone contact with the patient or referring physician, mailed survey, electronic chart review and/or review of death certificates. The 1- and 2-year study endpoints were all-cause mortality, target vessel revascularization (TVR), and major adverse cardiac events (MACE). TVR was defined as reintervention on the stented segment for clinical manifestations consistent with recurrent myocardial ischemia or > 70% stenosis on follow-up angiogram. Myocardial infarction (MI) was defined as the occurrence of chest pain accompanied by new, typical electrocardiographic changes and a rise in creatine kinase-MB above 3-fold the upper normal limit. Cardiac enzymes are not drawn systematically at our institution, but rather at the onset of chest pain and/or electrocardiographic (ECG) changes. MACE was a composite endpoint of all-cause mortality, non-fatal MI and TVR. Procedure- induced or aggravated renal dysfunction was defined as: (a) a postprocedural increase in serum creatine to > 2.0 mg/dl; (b) a ≥ 50% increase over an abnormal baseline value; or (c) new requirement for dialysis. Vascular complications included: (a) blood loss at the point of vascular access requiring a blood transfusion, or prolongation of hospitalization, orcausing a > 3.0 gm/dl decrease in blood hemoglobin content; and (b) vascular dissection or pseudoaneurysm documented by arteriography or ultrasound examination. The definitions of stent thrombosis developed by the Academic Research Consortium18 were applied.

Statistical analyses. Discrete variables are reported as frequencies (%), and continuous variables as mean ± standard deviation (SD). Chi-square tests of independence or exact tests, when appropriate, were used to compare discrete variables. Analysis of variance was used to compare continuous variables. The Kaplan-Meier method was used to produce actuarial survival estimates for death and MI, TVR and MACE. Overall and pair-wise log-rank testing was used to compare actuarial survivals among the renal function subgroups. For follow-up pair-wise comparisons of significant in-hospital outcomes, Bonferroni correction was used to adjust for multiple testing that yielded an alpha ≤ 0.005. Cox proportional hazards regression analyses were performed to examine the relationships between: (a) SES- versus PES-eluting stents; (b) multiple clinical variables; and (c) diabetes or anemia combined with level of renal function on the one hand, and (i) TVR, (ii) death, and (iii) MACE on the other hand. Hazard ratio (HR) and 95% confidence interval (CI) are presented. Results were considered statistically significant when p was < 0.05.


Baseline and procedural characteristics. The baseline and procedural characteristics of 2,758 consecutive DES recipients, subgrouped into 4 ranges of calculated GFR + 1 group of patients on long-term dialysis, are shown in Table 1. Patients with preserved renal function (> 60 ml/min) were more likely to be men, young, current smokers and treated with a GP IIb/IIIa inhibitor. Patients with impaired renal function (< 60 ml/min) were more likely to be women, hypertensive (data not shown), suffer from congestive heart failure and from peripheral or cerebral arterial disease, or operated coronary artery disease. Hemodialysis was used in 91.3% and peritoneal dialysis in 8.7% of dialyzed patients. Information regarding the administration of angiotensin-converting enzyme inhibitor and angiotensin receptor-blocker was available in 2,527 (91.6%) of the overall patient population; one or both were given in 70.9% of the normal renal function patients, and in 64.8%, 72.1%, 59.3% and 76.2% of patients in each respective decremental group. The mean left ventricular ejection fraction was 51.7 ± 10.9% in the normal group, and was 52.4 ± 12.1, 51.1 ± 13.1, 48.6 ± 14.6 and 50.7 ± 10.3% in each respective decremental group.

Lesion characteristics. The baseline characteristics of the 4,068 lesions treated are shown in Table 2. The majority were de novo lesions in native coronary arteries. Overall, the mean reference vessel diameter was 2.7 ± 0.6 mm, and the mean lesion length was 18.8 ± 9.9 mm.

In-hospital outcomes. The rates of in-hospital adverse clinical events are shown in Table 3. While no patient undergoing long-term dialysis suffered an early adverse clinical event, the rates of death among patients with moderate or severe renal insufficiency (0.7% and 4.6%, respectively) were significantly higher than in patients with a GFR > 60 ml/min. The incidence of vascular bleeding complications, dissection or pseudoaneurysm was significantly higher among patients with severe renal insufficiency than in patients with normal renal function, despite the more prevalent use of GP IIb/IIIa inhibitors among patients with preserved renal function.

Long-term survival and adverse clinical events. At a mean follow up of 706 ± 273 days, information was available for 98.8% of patients.

Target vessel revascularization, death and myocardial infarction. The actuarial survivals free from: (a) TVR, (b) death and MI, and (c) MACE in each study subgroup are shown in Figures 1 A, B and C, respectively. While the difference in the event-free survival distributions of TVR over 2 years did not reach statistical significance, the p-value indicates a trend toward increased clinical need for revascularizationamong patients with preserved renal function (p = 0.069; Figure 1A). Patients with worse renal function were observed to have higher rates of death and MI (p < 0.001; Figure 1B) and MACE (p < 0.001; Figure 1C).

Normal renal function versus chronic renal disease. Patients with preserved renal function had significantly lower 1- and 2-year rates of death, death + MI and MACE than the patients suffering from chronic renal disease. A nonsignificant trend was observed toward a lower incidence of 1- and 2-year rates of TVR in the GFR < 60 ml/min group.

Stent thrombosis. The incidence of early, late and very late, definite and probable stent thromboses in patients with GFR ≥ 60 ml/min versus GFR < 60 ml/min was similar (Table 5).

Sirolimus- versus paclitaxeleluting stents. The 18-month survival free from TVR after implantation of SES- versus PESeluting stents was similar, both in patients with a GFR ≥ 60 (p = 0.254) and patients with a GFR < 60 ml/min (p = 0.380) (Figure 2).

Post new comment

  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.

More information about formatting options

Enter the characters shown in the image.