Author + information
- Received January 21, 2000
- Revision received April 19, 2000
- Accepted June 19, 2000
- Published online November 1, 2000.
- Luis Gruberg, MD∗,
- Gary S Mintz, MD∗,* (, )
- Roxana Mehran, MD†,
- George Dangas, MD, PhD†,
- Alexandra J Lansky, MD†,
- Kenneth M Kent, MD, PhD∗,
- Augusto D Pichard, MD∗,
- Lowell F Satler, MD∗ and
- Martin B Leon, MD†
- ↵*Reprint requests and correspondence:
Dr. Gary S. Mintz, Washington Hospital Center, 110 Irving Street NW, Suite 4B-1, Washington, DC 20010
Acute deterioration in renal function is a recognized complication after coronary angiography and intervention.
The goal of this study was to determine the impact on acute and long-term mortality and morbidity of contrast-induced deterioration in renal function after coronary intervention.
We studied 439 consecutive patients who had a baseline serum creatinine ≥1.8 mg/dL (159.1 μmol/L) who were not on dialysis who underwent percutaneous coronary intervention in a tertiary referral center. All patients were hydrated before the procedure, and almost all received ioxaglate meglumine; 161 (37%) patients had an increase in serum creatinine ≥25% within 48 h or required dialysis and 278 (63%) did not. In-hospital and out-of-hospital clinical events (death, myocardial infarction, repeat revascularization) were assessed by source documentation.
Independent predictors of renal function deterioration were left ventricular ejection fraction (p = 0.02) and contrast volume (p = 0.01). In-hospital mortality was 14.9% for patients with further renal function deterioration versus 4.9% for patients with no creatinine increase (p = 0.001); other complications were also more frequent. Thirty-one patients required hemodialysis; their in-hospital mortality was 22.6%. Four patients were discharged on chronic dialysis. The cumulative one-year mortality was 45.2% for those who required dialysis, 35.4% for those who did not require dialysis and 19.4% for patients with no creatinine increase (p = 0.001). Independent predictors of one-year mortality were creatinine elevation (p = 0.0001), age (p = 0.03) and vein graft lesion location (p = 0.08).
For patients with pre-existing renal insufficiency, renal function deterioration after coronary intervention is a marker for poor outcomes. This is especially true for patients who require dialysis.
Acute deterioration in renal function is a recognized complication after coronary angiography, particularly for patients with pre-existing chronic renal insufficiency (CRI) (1–4). Previous studies have shown that 12 to 14% of patients who develop acute renal insufficiency during hospitalization do so after procedures involving radiographic contrast (5,6). For patients with abnormal baseline renal function, the incidence of progressive deterioration can be as high as 42% (1,7). For hospitalized, critically ill patients, this carries a poor prognosis, especially if dialysis becomes necessary (5–10).
The expanding use of diagnostic and therapeutic percutaneous interventions makes it important to understand the potential risks involved with these procedures. The purpose of this study was to review the in-hospital and long-term clinical outcomes of 439 consecutive patients with impaired baseline renal function who underwent percutaneous coronary intervention and had significant deterioration in renal function after the procedure.
The cohort is a consecutive series of 439 patients who had a baseline serum creatinine ≥1.8 mg/dL (159.1 μmol/L), but were not on dialysis, who underwent percutaneous coronary intervention. Patients were divided into two groups: 161 (37%) patients had an increase in serum creatinine ≥25% within 48 h or required dialysis (group 1) and 278 (63%) did not (group 2).
All patients underwent cardiac catheterization and intervention by standard techniques. All patients received intravenous hydration before the procedure. Hydration included 0.5 normal saline solution at 75 ml/h to 100 ml/h for 6 to 12 h before the intervention and for 6 h after the intervention. There was no formal protocol for use of periprocedural diuretics or other potentially renal-protective agents. The choice of radiocontrast material was left to the discretion of the operating physician; however, in over 95% of cases an ionic low osmolar contrast agent was used. Interventional treatment strategy (stents, rotational atherectomy, directional coronary atherectomy or excimer laser angioplasty) was performed by the operator based on vessel size, coronary anatomy and preprocedural lesion morphology. During the procedure heparin was administered to achieve an activated clotting time of approximately 300 s; heparin was not continued routinely after the procedure unless there were other indications for its use. Aspirin 325 mg was started at least 24 h before the procedure and continued indefinitely. After stent implantation patients received either ticlopidine (250 mg orally twice a day for two to four weeks) or clopidogrel (75 mg daily for two to four weeks). Other medications (that is, nitrates, calcium channel antagonists, etc.) were administered at the discretion of the operator.
Quantitative angiographic analysis was done at a dedicated Angiographic Core Laboratory by observers who were unaware of the clinical data and purpose of this study. Quantitative coronary angiographic analysis was performed using automated edge-detection algorithm (CMS, Medis Medical Imaging Systems, Ridgefield, Connecticut). Using the contrast filled catheter as the calibration standard, minimal lumen diameters, reference diameter and percent diameter stenosis before and after intervention were measured from multiple projections, and the results from the single “worst” and least foreshortened view were recorded. Standard qualitative morphologic criteria were tabulated (11,12). Left ventricular ejection fraction was determined by ventriculography or by echocardiography.
Baseline laboratory values were obtained for all patients upon admission. Creatinine clearance (CrCl) was calculated by applying the Cockcroft-Gault formula to the baseline serum creatinine: CrCl = ([140 − age] × weight/serum creatinine × 72) with female gender adjustment (CrClfemale = CrCl × 0.85) (13).
Procedural success was defined as <50% residual diameter stenosis and Thrombolysis In Myocardial Infarction (TIMI) flow >grade 2 in the absence of major in-hospital complications: death, MI or urgent bypass surgery (within 24 h after procedure). Q-wave MI was identified by the presence of new Q-waves. Non-Q-wave MI was defined as the elevation of creatine kinase-MB fraction ≥5 × normal without the appearance of new Q-waves.
Follow-up clinical events were assessed by serial telephone interviews at 1, 3, 6, 12 months and yearly thereafter. All events were source-documented including death, MI and repeat intervention or bypass surgery as related to the treated lesion.
Statistical analyses were performed using SAS 6.10 (SAS Institute). Continuous variables were presented as mean ± 1 standard deviation and were compared using Student t test. Categorical variables were presented as percentages and were compared using the chi-square statistics or the Fisher exact test. Kaplan-Meier survival curves were used to compare freedom from death and late events (death, MI or revascularization). Multivariate logistic regression analysis was used to determine predictors of renal function deterioration. Cox proportionate hazard model was used to determine predictors of late mortality. A p value <0.05 was considered statistically significant. Odds ratios (OR) and 95% confidence intervals (CI) are reported.
Patient characteristics and procedural variables
Patients with postprocedural renal function deterioration (group 1) were compared with patients with no increase in serum creatinine (group 2). The group with renal function deterioration included more diabetics and had worse left ventricular function. Other clinical characteristics were similar (Table 1).
Baseline serum creatinine was lower in group 1 (Table 2). However, CrCl (estimated by the Cockcroft-Gault formula) was similar for both groups, indicating no difference in baseline renal function when age and body weight were taken in account. The majority of the patients in both groups received ioxaglate meglumine, an ionic low osmolar contrast agent (Hexabrix, Mallinkrodt Medical Inc., St Louis, Missouri), 93% versus 90%, p = NS. There was a larger volume of contrast used in group 1 patients. Angiographic analysis is shown in Table 3. There was a higher frequency of saphenous vein graft lesions in group 1. This was reflected in larger reference vessel size and larger final minimal lumen diameter.
Predictors of renal function deterioration
The following variables were entered into a multivariate model for predicting renal function deterioration: age, left ventricular ejection fraction, diabetes mellitus, hypertension, creatine kinase-MB >5× normal, baseline creatinine, CrCl, angiographic contrast volume, hematocrit drop and need for blood transfusion. The only independent predictors of renal function deterioration were blood transfusion (OR = 1.55, 95% CI = 1.27 to 1.91, p < 0.0001), contrast volume (OR = 1.003, 95% CI = 1.001 to 1.006, p = 0.01) and left ventricular ejection fraction (OR = 0.97, 95% CI = 0.95 to 0.99, p = 0.02).
Clinical outcome (Fig. 1)
The combined (cardiac and noncardiac) in-hospital mortality in patients with further renal function deterioration was 14.9% compared with 4.9% in patients with no increase in creatinine (Table 4). The percentage of deaths from cardiac causes was also elevated in patients with renal function deterioration. Other complications (pulmonary edema, gastrointestinal bleeding and need for blood transfusions) were also more frequent for patients with renal function deterioration.
Thirty-one patients required hemodialysis during hospitalization. Their in-hospital mortality was 22.6%. Four patients were discharged on chronic dialysis. The cumulative one-year mortality was significantly higher in patients with renal function deterioration (37.7% vs. 19.4%, p = 0.001, Table 5and Fig. 2) —35.4% for those who did not require dialysis and 45.2% for those who required dialysis. Of the 17 patients who required in-hospital dialysis and who were alive at one year, 3 required chronic dialysis.
When the one-year mortality was compared with regard to the percent increase in creatinine, there was a noticeable difference in patients with a 25% increase or greater (Fig. 3). Myocardial infarction and revascularization events were similar in both groups.
Predictors of mortality
We separated the two groups into those patients with and those patients without diabetes (Fig. 4). There was no difference in long-term survival between diabetic patients versus patients who were not diabetic with no increase in creatinine or between diabetic patients versus patients who were not diabetic with a further deterioration in renal function.
The following variables were entered into a multivariate model predicting death during the follow-up period: age, male gender, diabetes mellitus, baseline CrCl, left ventricular ejection fraction, left main and vein graft lesion location, stent deployment, final minimum lumen diameter, creatine kinase-MB release, in-hospital dialysis, contrast volume, creatinine elevation, hematocrit drop and blood transfusion. Independent predictors of death were creatinine elevation (OR = 3.86, 95% CI = 1.96 to 7.58, p = 0.0001), age (OR = 1.05, 95% CI = 1.05 to 1.09, p = 0.03) and vein graft lesion location (OR = 1.55, 95% CI = 0.95 to 2.52, p = 0.08).
Patients with baseline azotemia and postangioplasty deterioration in renal function had worse short-term and long-term outcomes compared with patients who did not. These patients had significant comorbidities (Table 1) and a significant increase in most in-hospital adverse events: death, non-Q-wave MI, pulmonary edema and bleeding. At one-year follow-up, mortality rates were high for both groups, but especially for patients with deterioration in renal function (47.7%). Using multivariate analysis, creatinine elevation, advanced patient age and vein graft lesion location were independent predictors of death in these patients.
Renal function deterioration postcontrast
Renal function deterioration after exposure to radiographic contrast agents is common in patients with impaired renal function (1,2). Diabetes mellitus is one of the strongest predictors of acute renal failure after coronary intervention (6). In a study of patients with diabetic nephropathy and azotemia who underwent prerenal transplant coronary angiography, 50% had a serum creatinine increase >25%, and 12% required dialysis within a week (4). This was not substantiated in this study. Other reported risk factors include chronic renal insufficiency; dehydration; ionic, high osmolar contrast agents and congestive heart failure (14–17).
The renal function deterioration is an important predictor of in-hospital mortality. Rich et al. (2) reported 183 patients ≥ 70 years of age undergoing cardiac catheterization—including patients with prior renal insufficiency (mean serum creatinine 1.3 ± 0.7 mg/dL); 21/183 patients (11.5%) developed progressive renal dysfunction (serum creatinine elevation >0.5 mg/dL), which persisted in five of them (24%). The overall in-hospital mortality was 7.1% to 14.3% for those who developed progressive renal dysfunction and 6.2% for those who did not. Levy et al. (10) reported 183 patients who developed contrast-mediated renal dysfunction and compared these index patients against 174 “controls” matched for age and baseline serum creatinine. The in-hospital mortality was 34% in the index cases versus 7% in the control population.
A 25% increase in creatinine was the typical definition of renal function deterioration used in the preceding studies. The use of this “cutoff” as a predictor of cumulative mortality was substantiated in this study (Fig. 3).
Patients with azotemia were at a higher risk for bleeding complications and required blood transfusions more often. Bleeding, hypovolemia and hypotension may lead to further deterioration in renal function. In this study blood transfusion, but not bleeding, was an independent predictor of creatinine elevation, nor was it a predictor of mortality.
The risk of emergency dialysis after radiographic contrast ranges up to 12% in patients with baseline serum creatinine levels above 1.4 mg/dL (18). In-hospital mortality rates for patients who require emergency dialysis have been reported to be as high as 62%; conversely, in patients who developed renal failure but did not require dialysis, the mortality rate was lower (31%) (18). In one report of 132 critically ill patients in the intensive care unit who developed acute renal failure and required dialysis, the in-hospital mortality rate was 70%; of those who were discharged, 33% were dialysis dependent (19).
Using a definition similar to the current study (a 25% increase in creatinine), McCullough et al. (6) reported an incidence of acute renal failure after coronary interventions of 14.5%. The in-hospital mortality rate for these patients was 7.1 but increased to 35.7% for patients who required dialysis.
Whether these patients, especially patients who require dialysis, die from renal failure itself or from the underlying conditions is an unresolved issue (18). Analysis of our data and data from other studies suggests that comorbidities alone do not account for the increased mortality. When the effect of comorbidity was controlled in a multivariate model in this study, renal deterioration still predicted increased mortality.
At least four mechanisms have been implicated in contrast-induced nephropathy. After radiographic contrast exposure, there is a brief period of vasodilation followed by renal vasoconstriction leading to intense, but transient, reduction in renal blood flow, direct toxicity to renal tubular epithelium, tubular obstruction by protein precipitates and complement activation (14).
Recent works have suggested that endothelin, a potent endogenous vasoconstrictor, may be an important contributor to ischemic damage to the kidneys in certain circumstances and has been proposed as one of the potential mediators of renal vascular vasoconstriction. The exposure to large volumes of contrast material is associated with increased levels of circulating endothelin both in animal models and in humans. This is even more exaggerated in diabetic patients or patients with CRI (20,21). The simultaneous release of various endogenous renal vasodilators—such as atrial natriuretic peptide, nitric oxide and prostaglandins E2 and I2—may play an important role in counterbalancing the vasoconstrictive effects of endothelin by enhancing the glomerular filtration rate and urinary output (22).
Several measures have been tried in an attempt to prevent or reduce further renal function deterioration after contrast exposure: hydration, furosemide, mannitol, calcium channel-blockers, dopamine, atrial natriuretic peptide, endothelin inhibitors and theophylline (23–25). Hydration with 0.45% saline and the use of low osmolality contrast media have been shown to provide some protection (24,26). Recent reports have also suggested that forced diuresis with furosemide, mannitol and intravenous 0.45% saline may have a beneficial effect (27).
The current report is a retrospective analysis, and, therefore, the results and conclusions are subject to the limitations inherent in all such reports. location and poor left ventricular ejection fraction. In addition, this was a high-risk patient population with a majority of the patients having had previous bypass surgery. The use of glycoprotein IIb/IIIa inhibitors in this patient population was <2%. Serum creatinine was only measured for 48 h; we may have missed a later increase in creatinine in some of the 63% of the patients who did not have renal function deterioration within 48 h of their procedure.
These data show that in patients with pre-existing renal insufficiency, renal function deterioration (elevation of serum creatinine above 25% baseline levels or need for dialysis) after coronary intervention is a marker for poor in-hospital and long-term outcomes. Patients who require dialysis after the procedure are especially prone to a poor clinical outcome, and every measure should be taken to avoid such an end point. Even in the era of modern interventional cardiology and new contrast agents, the risk of renal injury remains high with important long-term consequences.
- confidence interval
- creatinine clearance
- chronic renal insufficiency
- myocardial infarction
- odds ratio
- Received January 21, 2000.
- Revision received April 19, 2000.
- Accepted June 19, 2000.
- American College of Cardiology
- Kleinknecht D,
- Jungers P,
- Chanard J,
- Barbane L,
- Ganeval D,
- Rondon-Nucete M
- Lansky A.J,
- Popma J.J
- Gross P,
- Bussemaker E
- Heyman S.N,
- Clark B.A,
- Kaiser N,
- et al.
- Stevens M.A,
- McCullough P.A,
- Tobin K.J,
- et al.