Predictive value of two different definitions of contrast-associated acute kidney injury for long-term major adverse kidney events in patients with ST-segment elevation myocardial infarction undergoing primary percutaneous coronary intervention

Abstract

Background

It remains controversial whether contrast-associated acute kidney injury (CA-AKI) is associated with long-term major adverse kidney events (MAKE) in patients with ST-segment elevation myocardial infarction (STEMI) undergoing primary percutaneous coronary intervention (PCI).

Methods

By the Acute Kidney Injury Network (AKIN) criteria, CA-AKI was defined as an increase in serum creatinine ≥ 0.3 mg/dL or 50% from baseline within 48 h after PCI; or an increase in serum creatinine ≥ 0.5 mg/dL or 25% within 72 h by the contrast-induced nephropathy (CIN) criteria. The primary endpoint was 1-year MAKE, defined as a composite of all-cause mortality and persistent renal dysfunction.

Results

A total of 402 patients were finally included in this study. The primary endpoint occurred in 29 (7.2%) patients. There was a significant association between CA-AKI and 1-year MAKE assessed by both the AKIN (hazard ratios [HR]: 11.58, 95% confidence interval [CI]: 4.29–31.24, p = 0.000) and CIN (HR: 6.45, 95% CI: 2.56–16.25, p = 0.000) definitions. However, the AKIN definition (HR: 4.95, 95% CI: 1.17–21.02, p = 0.030) was more reliable in the prediction of persistent renal dysfunction than CIN definition (HR: 4.08, 95% CI: 0.99–16.87, p = 0.052). Additionally, the area under receiver operating characteristic curve was larger for predicting 1-year MAKE with the AKIN definition than CIN definition (0.742 vs. 0.727).

Conclusions

In patients with STEMI undergoing primary PCI, CA-AKI was significantly associated with 1-year MAKE. Moreover, the AKIN definition might be more reliable in the prediction of long-term prognosis.

Introduction

Contrast-associated acute kidney injury (CA-AKI), previously described as contrast-induced acute kidney injury (CI-AKI), is characterized by a decline in renal function that occurs within days after the intravascular administration of iodinated contrast media [1]. It is a common adverse complication in patients with ST-segment elevation myocardial infarction (STEMI) undergoing primary percutaneous coronary intervention (PCI), associated with a higher incidence of in-hospital and short-and long-term clinical outcomes [2–5]. Traditionally known as contrast-induced nephropathy (CIN), it is defined as an increase in serum creatinine ≥ 0.5 mg/dL or 25% from baseline within 48–72 h after contrast media exposure [6]. More recently, the Acute Kidney Injury Network (AKIN) recommends a novel standardized definition of CA-AKI, that is, an increase in serum creatinine ≥ 0.3 mg/dL or 50% from baseline within 48 h [7]. Irrespective of the definitions adopted, CA-AKI could be a useful predictor of long-term mortality in patients with STEMI treated by primary PCI [8, 9].

Recent observational studies reveal that CA-AKI also has a persistent impact on the impairment of renal function, such as progression to chronic kidney disease (CKD) or end-stage renal disease (ESRD) [10, 11]. However, depending on the definitions of CA-AKI used in other survival analyses, the results remain controversial [12]. Thus, this study aimed to assess the association between CA-AKI and long-term major adverse kidney events (MAKE) in patients with STEMI undergoing primary PCI, and compare the predictive value between the two different definitions of CA-AKI.

The following article is presented in accordance with the STARD reporting checklist.

Methods

Study population

A total of 441 consecutive patients diagnosed with STEMI undergoing primary PCI in Yuyao People’s Hospital of Zhejiang Province between January 2015 and January 2020 were prospectively enrolled in this study. Inclusion criteria were (i) chest pain persisting for over 30 min with ST-segment elevation on the electrocardiogram of ≥ 0.1 mV in at least two contiguous leads or left bundle branch block; (ii) onset of chest pain within 12 h, or 24 h in the presence of ongoing symptoms suggestive of ischemia. Exclusion criteria included death within 24 h after admission, cardiogenic shock requiring intra-aortic balloon pump (IABP), ESRD on maintenance dialysis, recent exposure to contrast media (within 72 h before/after the procedure), concomitant use of nephrotoxic agents (e.g., non-steroidal anti-inflammatory drugs or aminoglycoside antibiotics, which are thought to be associated with increased risk of AKI [13, 14]), or inability to obtain informed consent.

Study protocol

For all patients, a 12-lead electrocardiogram was conducted in the emergency room. After cardiac intervention, blood samples for serum creatinine measurement were collected at baseline and daily. The estimated glomerular filtration rate (GFR) was calculated using the CKD-EPI equation [15], which had less bias than the Modification of Diet in Renal Disease (MDRD) equation especially at higher GFR. Primary PCI was performed based on the current guidelines [16, 17]. Before the procedure, acetylsalicylic acid (300 mg) and clopidogrel (300–600 mg) or ticagrelor (180 mg) were routinely administered. The anticoagulant agents and the glycoprotein IIb/IIIa inhibitor use were determined at the discretion of the physicians. Nonionic, low-osmolality contrast media and standard intravenous hydration were used in all patients. Within 24 h after admission, the left ventricular ejection fraction (LVEF) was measured by echocardiography using the biplane Simpson’s method [18]. All patients were scheduled to return for follow-up by in-or outpatient clinic visits until 1 year after the procedure. The study protocol was approved by the Institutional Research Ethics Board. All patients provided informed consent prior to participation. All procedures were conducted in accordance with the principles outlined in the Declaration of Helsinki.

Definitions and endpoints

By the AKIN criteria, CA-AKI was defined as an increase in serum creatinine ≥ 0.3 mg/dL or 50% from baseline within 48 h [6], and by the CIN criteria, CA-AKI was defined as an increase in serum creatinine ≥ 0.5 mg/dL or 25% from baseline within 72 h after contrast media exposure [7]. The primary endpoint of this study was 1-year MAKE, which was defined as a composite of all-cause mortality and persistent renal dysfunction [19]. Persistent renal dysfunction was defined as a sustained elevation ≥ 1.5 times in serum creatinine at 1-year follow-up compared with baseline level or progression to ESRD requiring dialysis.

Statistical analysis

Continuous variables were expressed as means and standard deviation for normally distributed variables or as median and interquartile range for nonnormally distributed variables. The categorical variables were presented as numbers and percentages (%). The independent t-test for normally distributed values and the Mann–Whitney test for nonnormally distributed values were used to compare continuous variables across trial groups. The χ² was used to compare proportions, and if the expected frequency was < 5, the Fisher’s exact test was conducted. The association between CA-AKI and 1-year MAKE was estimated by fitting a multivariable Cox regression model adjusted for risk factors based on univariate Cox regression. The results were shown as hazard ratios (HR) and 95% confidence intervals (CI). To compare the predictive accuracy between the AKIN and CIN definitions, the receiver operating characteristic (ROC) curve analysis and area under the curve (AUC) were assessed. P < 0.05 was considered significant. SPSS 23.0 software (SPSS Inc., Chicago, Illinois, USA) was used to perform all analyzes.

Results

Participant characteristics

During the study phase, 2 patients were excluded due to death within 24 h after admission. Another 11 patients were excluded for cardiogenic shock requiring IABP. We further excluded 2 patients with ESRD on maintenance dialysis, 7 patients with recent exposure to contrast media or concomitant use of nephrotoxic agents, and 17 patients who were unwilling to participate in the study. Therefore, a total of 402 participants were finally included in the study.

Table 1 shows the baseline characteristics of all patients. Briefly, the mean age was 60 ± 14.1 years. Among them, 72 (17.9%) patients were older than 75 years. The mean door-to-balloon time was 77.5 ± 16.9 min. Baseline renal dysfunction (eGFR < 60 mL/min/1.73 m²) was detected in 35 (8.7%) patients. The mean volume of administered contrast medium was 92.7 ± 32.9 mL.

Table 1.

Baseline characteristics of all patients.

Characteristics	All patients (n = 402)	CA-AKI according to AKIN definition		P	CA-AKI according to CIN definition		P
		Yes (n = 41)	No (n = 361)		Yes (n = 80)	No (n = 322)
Age [years]	60 ± 14.1	69.4 ± 14.4	59.0 ± 13.7	0.000**	64.7 ± 14.3	59.0 ± 13.9	0.001**
Aged over 75 years	72 (17.9%)	17 (41.5%)	55 (15.2%)	0.000**	23 (28.8%)	49 (15.2%)	0.005**
Male gender	349 (86.8%)	31 (75.6%)	318 (87.6%)	0.025*	62 (77.5%)	287 (89.1%)	0.006**
Hypertension	195 (48.5%)	24 (58.5%)	171 (47.4%)	0.175	46 (57.5%)	149 (46.3%)	0.072
Diabetes mellitus	70 (17.4%)	9 (22.0%)	61 (16.9%)	0.419	16 (20.0%)	54 (16.8%)	0.495
Door-to-balloon time [min]	77.5 ± 16.9	85.3 ± 21.2	76.7 ± 16.1	0.015*	82.3 ± 19.5	76.3 ± 16.0	0.013*
Anterior MI	256 (63.7%)	30 (73.2%)	226 (62.6%)	0.182	60 (75.0%)	196 (60.9%)	0.019*
Killip class 2 or 3	47 (11.7%)	10 (24.4%)	37 (10.2%)	0.012*	15 (18.8%)	32 (9.9%)	0.028*
Multivessel coronary disease	158 (39.3%)	22 (53.7%)	136 (37.7%)	0.047*	41 (51.3%)	117 (36.3%)	0.015*
Post-procedural TIMI grade flow 3	360 (89.6%)	35 (85.4%)	325 (90.0%)	0.415	72 (90.0%)	288 (89.4%)	0.884
Contrast volume [mL]	92.7 ± 32.9	97.3 ± 39.6	92.2 ± 32.1	0.347	98.6 ± 41.2	90.7 ± 30.3	0.148
Creatinine [μmol/L]	78.4 ± 23.2	80.4 ± 23.6	78.2 ± 23.2	0.562	74.0 ± 19.9	80.6 ± 23.5	0.190
eGFR [mL/min/1.73 m2]	89.7 ± 19.8	80.4 ± 22.0	90.8 ± 19.2	0.001**	92.0 ± 20.5	89.2 ± 19.6	0.254
eGFR < 60 mL/min/1.73 m2	35 (8.7%)	7 (17.1%)	28 (7.8%)	0.045*	6 (7.5%)	29 (9.0%)	0.669
LVEF < 40%	22 (5.5%)	6 (14.6%)	16 (4.4%)	0.017*	9 (11.3%)	13 (4.0%)	0.017*
Statins	400 (99.5%)	40 (97.6%)	360 (99.7%)	0.194	79 (98.8%)	321 (99.7%)	0.359
ACEI/ARB	362 (90.0%)	36 (87.8%)	326 (90.3%)	0.784	71 (88.8%)	291 (90.4%)	0.664
MRA	19 (4.7%)	4 (9.8%)	15 (4.2%)	0.117	6 (7.5%)	13 (4.0%)	0.234

^*p < 0.05;

^**p < 0.01;

CA-AKI — contrast-associated acute kidney injury; AKIN — Acute Kidney Injury Network; CIN — contrast induced nephropathy; MI — myocardial infarction; TIMI — Thrombolysis In Myocardial Infarction; eGFR — estimated glomerular filtration rate; LVEF — left ventricular ejection fraction; ACEI — angiotensin converting enzyme inhibitor; ARB — angiotensin II receptor blocker; MRA — mineralocorticoid receptor antagonists

Incidence of CA-AKI by the AKIN and the CIN criteria

By the CIN criteria, CA-AKI was detected in 80 (19.9%) patients; while according to the AKIN criteria, only 41 (10.1%) patients were diagnosed as CA-AKI. Among 41 patients with CA-AKI by the AKIN criteria, 40 (97.6%) patients met the CIN criteria; on the contrary, only 50.0% (40/80) of them fulfilled the AKIN definition among patients with CA-AKI by the CIN definition.

Irrespective of the definition adopted, patients developing CA-AKI were older, more often female, had a higher proportion of Killip class 2 or 3, experienced prolonged door-to-balloon time, and had a reduced LVEF. The AKIN definition was more likely to distinguish patients with impaired baseline renal function, while patients with CA-AKI by the CIN definition tended to have an anterior STEMI.

Predictive value of the AKIN and the CIN criteria on long-term MAKE

During the 1-year follow-up period, the primary endpoint was documented in 29 (7.2%) patients, including all-cause mortality in 18 patients and persistent renal dysfunction in 11 patients (with 1 patient progressed to ESRD). Patients developing CA-AKI were more likely to suffer 1-year MAKE than those without CA-AKI, either by the AKIN (39.0% vs. 3.6%, p = 0.000) or by the CIN (22.5% vs. 3.4%, p = 0.000) definition (Fig. 1A). A similar trend was also found in the frequency of persistent renal dysfunction, which was higher in patients with CA-AKI by both the AKIN (9.8% vs. 1.9%, p = 0.018) and the CIN (6.3% vs. 1.9%, p = 0.047) criteria than that in those without CA-AKI (Fig. 1B).

Figure 1.

The incidence of 1-year major adverse kidney events (A) and persistent renal dysfunction (B) was significantly higher in patients with contrast-associated acute kidney injury (CA-AKI) by both the Acute Kidney Injury Network (AKIN) and the contrast-induced nephropathy (CIN) definitions; *p < 0.05; **p < 0.01.

By multivariable regression analysis, there was a significant association between CA-AKI and 1-year MAKE assessed by both the AKIN (HR: 11.58, 95% CI: 4.29–31.24, p = 0.000) and the CIN (HR: 6.45, 95% CI: 2.56–16.25, p = 0.000) definitions (Table 2). However, the AKIN definition (HR: 4.95, 95% CI: 1.17–21.02, p = 0.030) was more reliable in the prediction of persistent renal dysfunction than the CIN definition (HR: 4.08, 95% CI: 0.99–16.87, p = 0.052) (Table 3). Moreover, the ROC curve analysis showed a modest increase in the AUC for predicting 1-year MAKE with the AKIN definition (AUC: 0.742, 95% CI: 0.629–0.856, p = 0.000) than that with the CIN definition (AUC: 0.727, 95% CI: 0.620–0.834, p = 0.000, Fig. 2), though failing to reach statistically significance (p = 0.55).

Table 2.

Regression model for long-term major adverse kidney events.

	Univariate analysis			Multivariate analysis
	HR	95% CI	P	CA-AKI by the AKIN definition			CA-AKI by the CIN definition
				HR	95% CI	P	HR	95% CI	P
Aged over 75 years	6.04	2.83–12.91	0.000**	2.80	1.07–7.31	0.035*	3.33	1.33–8.34	0.010*
Male gender	0.63	0.25–1.61	0.336
Hypertension	2.06	0.96–4.41	0.064
Diabetes mellitus	0.49	0.15–1.66	0.251
Anterior MI	1.73	0.75–3.97	0.197
Killip class 2 or 3	6.10	2.74–13.60	0.000**	4.31	1.54–12.06	0.005**	4.45	1.61–12.30	0.004**
Multivessel coronary disease	2.00	0.96–4.18	0.066
Post-procedural TIMI grade flow 3	0.21	0.09–0.48	0.000**	0.26	0.08–0.81	0.020*	0.28	0.09–0.85	0.024*
eGFR < 60 mL/min/1.73 m2	3.60	1.42–9.07	0.007**	2.18	0.68–7.05	0.193	3.19	0.99–10.27	0.051
LVEF < 40%	14.36	5.55–36.95	0.000**	11.26	3.21–39.46	0.000**	9.76	2.90–32.87	0.000**
CA-AKI by the AKIN definition	14.85	6.59–33.48	0.000**	11.58	4.29–31.24	0.000**
CA-AKI by the CIN definition	6.95	3.24–14.91	0.000**	6.45	2.56–16.25	0.000**

^*p < 0.05;

^**p < 0.01;

CA-AKI — contrast-associated acute kidney injury; AKIN — Acute Kidney Injury Network; CIN — contrast induced nephropathy; MI — myocardial infarction; TIMI — Thrombolysis In Myocardial Infarction; eGFR — estimated glomerular filtration rate; LVEF — left ventricular ejection fraction; HR — hazard ratio; CI — confidence interval

Table 3.

Regression model for persistent renal dysfunction.

	Univariate analysis			Multivariate analysis
				CA-AKI by the AKIN definition			CA-AKI by the CIN definition
	HR	95% CI	P	HR	95% CI	P	HR	95% CI	P
Aged over 75 years	9.75	2.77–34.35	0.000**	5.54	1.44–21.38	0.013*	6.00	1.57–22.92	0.009**
Male gender	0.40	0.10–1.57	0.190
Hypertension	11.40	0.95–89.97	0.071
Diabetes mellitus	0.47	0.06–3.72	0.474
Anterior MI	1.03	0.75–3.97	0.197
Killip class 2 or 3	5.31	1.49–19.00	0.010*	3.64	0.86–15.47	0.080	3.59	0.83–15.53	0.087
Multivessel coronary disease	0.93	0.27–3.23	0.909
Post-procedural TIMI grade flow 3	1.04	0.13–8.39	0.968
eGFR < 60 mL/min/1.73 m2	6.94	1.92–25.08	0.003**	4.43	1.06–18.52	0.041*	6.03	1.36–26.74	0.018*
LVEF < 40%	9.15	2.20–38.00	0.002**	7.57	1.51–37.94	0.014*	6.31	1.23–32.29	0.027*
CA-AKI by the AKIN definition	6.94	1.92–25.08	0.003**	4.95	1.17–21.02	0.030*
CA-AKI by the CIN definition	3.91	1.16–13.17	0.028*				4.08	0.99–16.87	0.052

^*p < 0.05;

^**p < 0.01;

CA-AKI — contrast-associated acute kidney injury; AKIN — Acute Kidney Injury Network; CIN — contrast induced nephropathy; MI — myocardial infarction; TIMI — Thrombolysis In Myocardial Infarction; eGFR — estimated glomerular filtration rate; LVEF — left ventricular ejection fraction; HR — hazard ratio; CI — confidence interval

Figure 2.

The receiver operating characteristic (ROC) curve and area under the curve showed a better diagnostic accuracy of the Acute Kidney Injury Network (AKIN) definition for 1-year major adverse kidney events than that of the contrast-induced nephropathy (CIN) definition (0.742 vs. 0.727).

Discussion

To the best of our knowledge, this was the first study to assess the association between CA-AKI and long-term MAKE in patients with STEMI undergoing primary PCI. Our data suggested that CA-AKI could be an independent predictor of 1-year MAKE, regardless of the definition used; moreover, the AKIN definition might be more reliable in the prediction of long-term adverse prognosis than the CIN definition.

Multiple large studies have demonstrated that patients who survive an onset of AKI are at potential risk of progression to CKD or ESRD. A cohort study by Wald et al. [20] enrolling 17,367 patients with a median follow-up of 3 years showed that the absolute risk of ESRD was 0.4% among those with normal baseline renal function without AKI, while it rose up to 4.6% (more than 10-fold) when complicated by AKI. Another cohort study of 36,980 patients by Chawla et al. [21] showed that patients with AKI were 2-fold as likely to develop CKD or ESRD (HR: 2.07, 95% CI: 1.99–2.16) compared to those without AKI. Based on these findings, the endpoint of MAKE has been endorsed by the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) clinical trial work-group to harmonize and encourage future clinical trials [22]. Hence, we systematically assessed pre-angiography renal function, prospectively tracked the development of CA-AKI and 1-year outcomes, seeking to evaluate the association between CA-AKI and long-term MAKE in patients with STEMI undergoing primary PCI. As shown in this study, a significant association was detected between CA-AKI and worse prognosis, irrespective of the definition used. The pathophysiologic mechanisms that CA-AKI would increase the risk of long-term MAKE remain unknown. Results from experimental animals suggest that AKI can induce tubulointerstitial fibrosis and a reduction in peritubular capillary density in the inner stripe of the outer medulla, thus resulting in persistent deterioration of renal function [23]. Although AKI has generally been considered reversible in nature, there may be subclinical damage that persists and mediates these adverse outcomes [24].

In this study, we further compared the difference between the AKIN and the CIN definitions. Over years, a debate has existed on a standardized definition of CA-AKI. The lack of consensus results in a significant variation in the incidence of CA-AKI in patients undergoing PCI, ranging from 3.3% to 10.5%, depending on the various definitions used [25]. In the setting of STEMI, the rate of CA-AKI was 16.1% according to the HORIZON-AMI trial, in which the CIN definition was adopted [26]. These results are quite similar to the findings in our study (19.8%). Meanwhile, the incidence of CA-AKI varied from 9.6% to 10.7% according to the previous observational study where the AKIN definition was applied [8, 27], which is also consistent with our data (10.1%). Besides, in this study, 97.6% (40/41) of the patients diagnosed with CA-AKI by AKIN definition fulfilled the CIN criteria; only 50.0% (40/80) of those with CA-AKI by CIN criteria met the AKIN definition. It seemed that the CIN definition was a more sensitive indicator, while the AKIN definition tended to be more rigorous and specific.

Most importantly, our data revealed that the AKIN definition was superior to the CIN definition in the prediction of long-term MAKE. It has been suggested that the AKIN definition is the most discriminant definition to identify patients at higher risk of mortality. A cohort study of 402 patients undergoing primary PCI for STEMI by Centola et al. [8] demonstrated that the AKIN definition (HR: 9.70, 95% CI: 5.12–18.37, p = 0.000) provided better accuracy in the prediction of 1-year mortality than the CIN criteria (HR: 4.84, 95% CI: 2.56–9.16, p = 0.000) [8]. In another cohort which enrolled 1114 patients with STEMI undergoing primary PCI, Silvain et al. [28] confirmed the value of the AKIN definition in predicting long-term mortality; in addition, the AKIN definition (odds ratio [OR]: 4.60, 95% CI: 1.88–11.27, p = 0.001) was superior to the CIN definition (OR: 3.54, 95% CI: 1.49–8.39, p = 0.004) in predicting a hemodialysis requirement at 1-year follow-up. However, the conclusions remain to be debated as patients with preexisting renal failure, hemodynamic instability and resuscitated cardiac arrest were unselectedly included, which might also contribute to the tendency of hemodialysis treatment. Moreover, merely the hemodialysis requirement, rather than assessment of serum creatine was adopted as the renal endpoint, which might be biased in the overall estimation of renal impairment. In the present study, kidney function was assessed by an in- or outpatient clinic visit, and innovatively evaluated the predictive value of the two definitions of CA-AKI with 1-year MAKE and persistent renal dysfunction, based on either progression to ESRD requiring dialysis or a sustained elevation ≥ 1.5 times in serum creatinine compared with the baseline level, in accordance with the Risk, Injury, Failure, Loss of kidney function, and End-stage kidney disease (RIFLE) definition [29]. The adjusted HRs of CA-AKI for 1-year MAKE were 11.58 (95% CI: 4.29–31.24) and 6.45 (95% CI: 2.56–16.25) when the AKIN and the CIN definitions were, respectively, used. Nevertheless, the AKIN definition (HR: 4.95, 95% CI: 1.17–21.02, p = 0.030) seemed to be more reliable in the prediction of persistent renal dysfunction than the CIN definition (HR: 4.08, 95% CI: 0.99–16.87, p = 0.052). One plausible explanation for this phenomenon might be that the AKIN definition tended to distinguish patients with CKD in this study (17.1% vs. 7.8%, p = 0.045), who were more vulnerable to the deterioration of renal function than those without CKD [30]. Furthermore, the comparison of AUCs suggested that the AKIN definition was likely to provide better diagnostic accuracy for 1-year MAKE than the CIN definition (0.742 vs. 0.727). In a word, we complemented the previous results [8, 9, 28, 31], demonstrating that the AKIN definition might be a more reliable predictor of both long-term mortality and persistent renal dysfunction in patients with STEMI undergoing primary PCI, compared to the CIN definition.

Last but not least, the current study found that older age and reduced LVEF were also independent predictors of 1-year MAKE or persistent renal dysfunction after primary PCI for STEMI, apart from the episode of CA-AKI. The findings that elderly patients were prone to worse long-term prognosis were not surprising, since they generally had a high burden of medical comorbidity with physiologically reduced GFR [32]. On the other hand, our data also indicated that patients with cardiac insufficiency tended to develop long-term MAKE. The potential link between reduced LVEF and MAKE remains to be elucidated in the future study, while it is plausible that acute cardiorenal injury might induce a vicious cycle that persists long after the acute event [33].

Limitations of the study

There are several limitations to this study. First, this was a single-center study with a restricted sample size. A larger sample from multiple centers would make the results more reliable. Second, both the AKIN and CIN definitions refer to an increase in serum creatine, which could be affected in the setting of hemodynamic instability or decreased renovascular autoregulation [1, 34]. Therefore, patients with cardiogenic shock requiring IABP were excluded, which might have caused selection or surveillance bias. Third, we were unable to confirm whether CA-AKI was the cause of, or definitively contributed to, or was just a mediator to the development of long-term MAKE in this study. Fourth, the incidence of MAKE was comparatively low, as the follow-up data in the present study were obtained merely 1 year after the procedure. Future studies will be encouraged for a longer period of follow-up in this subset of patients. Last but not least, though a modest increase was observed in the AUC for predicting 1-year MAKE with the AKIN definition compared to that with the CIN one, the discrepancy failed to reach statistical significance (p = 0.55). The superior diagnostic accuracy of the AKIN definition might be validated in a larger cohort by using ROC analysis.

Conclusions

Irrespective of the definition used, CA-AKI was significantly associated with MAKE at 1-year in patients with STEMI undergoing primary PCI. Moreover, the AKIN definition might be more reliable in the prediction of long-term adverse prognosis than the CIN definition.