The pattern and prognostic impact of “early” acute kidney injury (AKI) after primary percutaneous coronary intervention (PCI) in patients with ST-segment elevation myocardial infarction have not been well established. From November 2005 to November 2011, 971 post–myocardial infarction patients who underwent primary PCI were analyzed. Early AKI was defined using absolute change in serum creatinine (SCr; SCr <24 hours after primary PCI minus admission SCr) as follows: no early AKI (SCr change <0.3 mg/dl), mild early AKI (SCr change 0.3 to <0.5 mg/dl), moderate early AKI (SCr change 0.5 to <1.0 mg/dl), and severe early AKI (SCr change ≥1.0 mg/dl). One-year major adverse cardiac events were defined as death, nonfatal myocardial infarction, and revascularizations. Overall, 9.6% had early AKI, including 5.7% with mild, 2.5% with moderate, and 1.4% with severe early AKI. Diabetes mellitus (odds ratio 1.84, p = 0.042), the left ventricular ejection fraction (odds ratio 0.97, p = 0.042), and hemoglobin levels (odds ratio 0.84, p = 0.039) were independently associated with early AKI. Early AKI (adjusted hazard ratio 2.80, p = 0.005) was an independent predictor of 1-year major adverse cardiac events. The adjusted hazard ratios of 1-year major adverse cardiac events from the lowest (reference) to the highest quartile of early AKI were as follows: 1, 2.87 (p = 0.012), 3.22 (p = 0.021), and 5.83 (p = 0.004), respectively. In conclusion, early dynamic change in renal function after primary PCI can sensitively predict worse outcomes.
Highlights
- •
Early dynamic change in serum creatinine level <24 hours after primary PCI is a frequent complication of patients with STEMI.
- •
Diabetes mellitus, LVEF, and hemoglobin level were independent predictors of early AKI.
- •
Quick increases in serum creatinine level can sensitively predict 1-year clinical outcomes.
- •
Early AKI was an independent predictor of 1-year clinical outcome.
- •
A graded relation exists between the severity of early AKI and 1-year clinical outcome.
Although long-term prognosis as well as short-term risk for acute kidney injury (AKI) after acute myocardial infarction (AMI) have been examined, most studies have excluded patients who died or were discharged in the first 48 hours. Therefore, the true incidence and severity of AKI were likely underestimated in these previous studies. Moreover, a comprehensive assessment of the pattern and prognostic impact of “early” AKI—renal function deterioration during the first 24 hours after primary percutaneous coronary intervention (PCI)—in patients with ST-segment elevation myocardial infarction (STEMI) has not been well established. Thus, the aims of this study were to evaluate the pattern of early AKI and to examine the 1-year clinical outcomes associated with early AKI in patients who underwent primary PCI after STEMI.
Methods
This observational study included 971 consecutive post–myocardial infarction (MI) patients who underwent primary PCI who were enrolled in the Korea AMI Registry (KAMIR) from our single center from November 2005 to November 2011. KAMIR is a Korean, prospective, open, observational, multicenter on-line registry of AMI with the support of the Korean Society of Cardiology, in operation since November 2005. Details of KAMIR have been published. Renal data early after primary PCI were retrospectively collected, because they had not been entered into the KAMIR database. AMI was diagnosed by characteristic clinical presentation, serial changes on electrocardiography suggesting infarction, and increases in cardiac enzyme levels. We analyzed baseline demographic characteristics, initial presentation, initial vital signs, electrocardiographic findings, results of laboratory tests, procedural data, and medications. Assessment of the left ventricular ejection fraction (LVEF) was determined by 2-dimensional echocardiography on days 2 and 3 of hospitalization.
Early AKI was defined as absolute change in serum creatinine (SCr) using the modified AKI Network (AKIN) definition as SCr <24 hours after primary PCI minus admission SCr. We defined no early AKI as a change in SCr of <0.3 mg/dl <24 hours after primary PCI, mild early AKI as a change in SCr of 0.3 to <0.5 mg/dl <24 hours after primary PCI, moderate AKI as a change in SCr of 0.5 to <1.0 mg/dl <24 hours after primary PCI, and severe AKI as a change in SCr of ≥1.0 mg/dl <24 hours after primary PCI.
Baseline laboratory data including SCr, except for lipid measurements, were collected before primary PCI on admission. Overnight fasting blood was also sampled for lipid levels. Venous blood samples were obtained at 24, 48, and 72 hours thereafter. Within the first 24 hours of primary PCI, SCr measurements were routinely checked.
One-year major adverse cardiac events (MACEs) were defined as death, nonfatal MI, and revascularizations, including repeat PCI and coronary artery bypass grafting. During the follow-up period, follow-up data were obtained by reviewing medical records and telephone interviews with patients. All data were recorded on an electronic Web-based case report form.
Data are expressed as mean ± SD for continuous variables and percentages for categorical variables. All comparisons between baseline variables were assessed with the Student’s t test for continuous variables and Pearson’s chi-square test for categorical variables. Univariate analyses were performed to determine the clinical predictors of early AKI. Multivariate logistic regression modeling was used to determine independent predictors of early AKI. Variables with p values <0.05 on univariate analysis, such as age, gender, heart rate, Killip class, the LVEF, history of diabetes mellitus, blood levels of hemoglobin, uric acid, and log-transformed N-terminal pro–B-type natriuretic peptide, were entered into the logistic regression model. A Cox proportional-hazards model was used to determine independent predictors of 1-year MACEs. Variables with p values <0.05 on univariate analysis—including age, gender, systolic blood pressure, current smoking status, history of hyperlipidemia, Killip class, multivessel disease, early AKI, and blood levels of glucose, hemoglobin, uric acid, and log-transformed N-terminal pro–B-type natriuretic peptide—were selected and entered into Cox proportional-hazards models to estimate the hazard ratios (HRs) and 95% confidence intervals (CIs) of 1-year MACEs. The Hosmer-Lemeshow chi-square statistic, a measure of deviation between observed and predicted outcomes in deciles of predicted risk, was used to evaluate the calibration of the model.
Moreover, patients were categorized into 4 groups according to quartiles of early AKI: no early AKI (n = 878), mild early AKI (n = 55), moderate AKI (n = 24), and severe AKI (n = 14). The cumulative incidence rates of 1-year MACEs according to the early AKI quartiles were estimated by using the Kaplan-Meier method and compared by using the log-rank test. Cox proportional-hazards analyses were performed to estimate the HRs and 95% CIs of 1-year MACEs for increasing early AKI quartiles, where the lowest quartile was used as the reference. The multivariate model included early AKI quartiles and the aforementioned risk factors. For all analyses, 2-sided p values <0.05 were considered statistically significant. Statistical analysis was performed using SPSS version 18.0 for Windows (SPSS, Inc., Chicago, Illinois).
Results
The baseline characteristics of the study subjects are listed in Table 1 . Overall, 9.6% had early AKI by SCr level, including 5.7% with mild early AKI, 2.5% with moderate early AKI, and 1.4% with severe early AKI ( Figure 1 ). Among patients with early AKI, male gender was significantly less frequent, and LVEFs were significantly lower, whereas heart rate and Killip class were significantly higher as compared with patients without early AKI. A history of diabetes mellitus was more common among patients with early AKI than among those without early AKI. Laboratory data indicated that the SCr level <24 hours after primary PCI, serum uric acid level, and N-terminal pro–B-type natriuretic peptide level were significantly higher, whereas the hemoglobin level was significantly lower, in patients with early AKI than in those without early AKI. In multivariate logistic regression model, history of diabetes mellitus (odds ratio 1.84, 95% CI 1.02 to 3.32, p = 0.042), a lower LVEF (odds ratio 0.97, 95% CI 0.95 to 0.99, p = 0.042), and a lower hemoglobin level (odds ratio 0.84, 95% CI 0.71 to 0.99, p = 0.039) were independently associated with early AKI ( Table 2 ).
| Variable | Overall (n = 971) | Early Acute Kidney Injury | p Value | |
|---|---|---|---|---|
| No (n = 878) | Yes (n = 93) | |||
| Age (year) | 62.8 ± 12.2 | 62.6 ± 11.9 | 64.6 ± 14.3 | 0.199 |
| Men | 712 (73.3%) | 652 (74.3%) | 60 (64.5%) | 0.043 |
| Body mass index (kg/m 2 ) | 23.7 ± 3.1 | 23.7 ± 3.1 | 23.8 ± 3.7 | 0.777 |
| Initial presentation | ||||
| Typical chest pain | 939 (97.2%) | 851 (97.3%) | 88 (96.7%) | 0.760 |
| Dyspnea | 98 (10.6%) | 84 (10.0%) | 14 (16.3%) | 0.071 |
| Preinfarct angina pectoris | 560 (57.7%) | 502 (57.2%) | 58 (62.4%) | 0.335 |
| Systolic blood pressure (mmHg) | 131.0 ± 31.0 | 130.7 ± 30.3 | 134.0 ± 36.9 | 0.441 |
| Heart rate (beats/min) | 78.3 ± 20.7 | 77.7 ± 20.2 | 84.2 ± 23.7 | 0.021 |
| Killip class >1 | 255 (26.3%) | 211 (24.0%) | 44 (47.3%) | <0.001 |
| Left ventricular ejection fraction (%) | 49.5 ± 10.6 | 50.4 ± 10.4 | 44.0 ± 12.0 | <0.001 |
| Prior coronary heart disease | 96 (9.9%) | 86 (9.8%) | 10 (10.8%) | 0.769 |
| Hypertension | 664 (68.4%) | 595 (67.8%) | 69 (74.2%) | 0.205 |
| Diabetes mellitus | 233 (24.1%) | 200 (22.9%) | 33 (35.5%) | 0.007 |
| Hyperlipidemia | 199 (21.0%) | 186 (21.7%) | 13 (14.4%) | 0.108 |
| Current smoker | 470 (48.6%) | 427 (48.8%) | 43 (46.2%) | 0.638 |
| Multivessel disease | 341 (35.5%) | 302 (34.7%) | 39 (43.3%) | 0.102 |
| Hemoglobin (g/dL) | 13.8 ± 1.8 | 13.9 ± 1.8 | 13.3 ± 2.1 | 0.007 |
| Glucose (mg/dL) | 175.4 ± 79.9 | 173.8 ± 79.3 | 190.1 ± 84.2 | 0.060 |
| Uric acid (mg/dL) | 5.5 ± 1.8 | 5.4 ± 1.8 | 5.9 ± 2.1 | 0.036 |
| Serum creatinine at baseline (mg/dL) | 1.0 ± 0.7 | 1.0 ± 0.7 | 1.2 ± 1.1 | 0.112 |
| Serum creatinine within 24-hour after percutaneous coronary intervention (mg/dL) | 1.0 ± 0.7 | 1.0 ± 0.7 | 1.8 ± 1.3 | <0.001 |
| Estimated glomerular filtration rate (mL/min/1.73 m 2 ) | 85.7 ± 26.2 | 86.2 ± 25.3 | 80.8 ± 33.5 | 0.139 |
| Peak cardiac troponin I (ng/mL) | 97.3 ± 181.6 | 93.3 ± 176.1 | 135.7 ± 224.7 | 0.081 |
| Total cholesterol (mg/dL) | 181.4 ± 44.0 | 181.4 ± 43.1 | 181.0 ± 52.6 | 0.940 |
| Triglyceride (mg/dL) | 138.6 ± 136.1 | 137.3 ± 133.9 | 151.9 ± 156.7 | 0.355 |
| High-density lipoprotein cholesterol (mg/dL) | 43.9 ± 13.1 | 43.9 ± 12.8 | 43.7 ± 15.4 | 0.882 |
| Low-density lipoprotein cholesterol (mg/dL) | 117.9 ± 38.7 | 118.6 ± 37.4 | 111.4 ± 49.5 | 0.199 |
| Log N-terminal B-type natriuretic peptide (pg/mL) | 5.4 ± 2.1 | 5.4 ± 2.1 | 6.5 ± 2.7 | <0.001 |
| Previous medication on admission | ||||
| Aspirin | 46 (4.7%) | 41 (4.7%) | 5 (5.4%) | 0.760 |
| Clopidogrel | 22 (2.5%) | 21 (2.7%) | 1 (1.2%) | 0.416 |
| Angiotensin-converting enzyme inhibitors | 15 (1.7%) | 15 (1.9%) | 0 (0.0%) | 0.203 |
| Angiotensin II type 1 receptor blockers | 41 (4.7%) | 37 (4.7%) | 4 (4.8%) | 0.971 |
| Beta-blockers | 38 (4.4%) | 34 (4.3%) | 4 (4.8%) | 0.842 |
| Statins | 38 (4.4%) | 36 (4.6%) | 2 (2.4%) | 0.352 |
| Nitrates | 18 (2.4%) | 15 (2.2%) | 3 (3.8%) | 0.377 |
| Diuretics | 19 (2.2%) | 15 (1.9%) | 4 (4.8%) | 0.086 |
| Variables | Odds Ratio | 95% Confidence Interval | p Value |
|---|---|---|---|
| Age | 0.99 | 0.96–1.01 | 0.305 |
| Men | 0.91 | 0.46–1.81 | 0.784 |
| Heart rate | 1.01 | 0.99–1.02 | 0.253 |
| Killip class >1 | 1.58 | 0.85–2.94 | 0.153 |
| Diabetes mellitus | 1.84 | 1.02–3.32 | 0.042 |
| Left ventricular ejection fraction | 0.97 | 0.95–0.99 | 0.042 |
| Hemoglobin | 0.84 | 0.71–0.99 | 0.039 |
| Uric acid | 1.11 | 0.96–1.28 | 0.170 |
| Log N-terminal pro-B type natriuretic peptide | 1.09 | 0.92–1.28 | 0.314 |
During the 1-year follow-up period, 95 (11.3%) MACEs occurred. Patients with early AKI had a significantly higher rate of 1-year MACEs compared with those without early AKI (30.1% vs 7.0%, log-rank p <0.001; Figure 2 ). In the Cox proportional-hazards model, early AKI (HR 3.38, 95% CI 1.86 to 6.15, p <0.001) in addition to age (HR 1.03, 95% CI 1.00 to 1.06, p = 0.036), Killip class >I (HR 2.48, 95% CI 1.40 to 4.39, p = 0.002), and serum glucose level (HR 1.003, 95% CI 1.00 to 1.01, p = 0.028) was an independent predictor of 1-year MACEs after adjusting for confounding variables ( Table 3 ).
