A novel risk score model for prediction of contrast-induced nephropathy after emergent percutaneous coronary intervention

    A novel risk score model for prediction of contrast-induced nephropathy after emergent percutaneous coronary intervention Kai-yang Lin, Wei-ping Zheng, Wei-jie Bei, Shi-qun Chen, Sheikh Mohammed Shariful Islam, Yong Liu, Lin Xue, Ning Tan, Ji-yan Chen PII: DOI: Reference:

S0167-5273(16)34586-7 doi:10.1016/j.ijcard.2016.12.095 IJCA 24273

To appear in:

International Journal of Cardiology

Received date: Revised date: Accepted date:

30 July 2016 11 November 2016 17 December 2016

Please cite this article as: Lin Kai-yang, Zheng Wei-ping, Bei Wei-jie, Chen Shiqun, Islam Sheikh Mohammed Shariful, Liu Yong, Xue Lin, Tan Ning, Chen Jiyan, A novel risk score model for prediction of contrast-induced nephropathy after emergent percutaneous coronary intervention, International Journal of Cardiology (2016), doi:10.1016/j.ijcard.2016.12.095

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT 1

A novel risk score model for prediction of contrast-induced

PT

nephropathy after emergent percutaneous coronary intervention

RI

Kai-yang Lin, MDa,b,c,§, Wei-ping Zheng, MDc,§, Wei-jie Bei, MDb,§, Shi-qun Chen,

SC

MSb,§, Sheikh Mohammed Shariful Islam, MBBS, MPH, PhDd, and Yong Liu, MDb¶, Lin Xue, MDb¶, Ning Tan, MD, FACC, FESC, FAPSICb¶, Ji-yan Chen, MD, FACC,

NU

FESCb¶

Southern Medical University, Guangzhou 510515, China

b

Department of Cardiology, Provincial Key Laboratory of Coronary Heart Disease,

MA

a

D

Guangdong Cardiovascular Institute, Guangdong General Hospital, Guangdong

Department of Geriatric Medicine, Fujian Provincial Hospital, Fujian Provincial

AC CE P

c

TE

Academy of Medical Sciences, Guangzhou 510100, China

Institute of Clinical Geriatrics, Fujian Provincial Key Laboratory of Geriatric Disease, Fujian Medical University, Fuzhou, 350001, China d

The George Institute for Global Health, University of Sydney, Camperdown NSW,

2050, AUSTRALIA §

These authors contributed equally to this work

¶

Corresponding authors

Addresses for correspondence: Ji-yan Chen, MD, FACC, FESC, Ning Tan, MD, FACC, FESC, FAPSIC, Yong Liu, MD, and Lin Xue, MD Department of Cardiology, Guangdong Cardiovascular Institute, Guangdong

ACCEPTED MANUSCRIPT 2

Provincial Key Laboratory of Coronary Disease, Guangdong General Hospital, Guangdong Academy of Medical Sciences, Guangzhou 510100, China. Tel:

PT

+86-15920172292, Fax: +86-20-83824369, E-mail: [email protected] (JC),

AC CE P

TE

D

MA

NU

SC

Subject Codes: Percutaneous Coronary Intervention

RI

[email protected] (NT), [email protected] (YL), [email protected] (LX).

ACCEPTED MANUSCRIPT 3

Background: A few studies developed simple risk model for predicting CIN with

PT

poor prognosis after emergent PCI. The study aimed to develop and validate a novel

RI

tool for predicting the risk of contrast-induced nephropathy (CIN) in patients

SC

undergoing emergent percutaneous coronary intervention (PCI).

Methods: 692 consecutive patients undergoing emergent PCI between January 2010

NU

and December 2013 were randomly (2:1) assigned to a development dataset (n = 461)

MA

and a validation dataset (n = 231). Multivariate logistic regression was applied to identify independent predictors of CIN, and established CIN predicting model, whose

D

prognostic accuracy was assessed using the c-statistic for discrimination and the

TE

Hosmere Lemeshow test for calibration.

AC CE P

Results: The overall incidence of CIN was 55(7.9%). A total of 11 variables were analyzed, including age>75 years old, baseline serum creatinine (SCr)>1.5 mg/dl, hypotension and the use of intra-aortic balloon pump(IABP), which were identified to enter risk score model (Chen). The incidence of CIN was 32(6.9%) in the development dataset (in low risk (score=0),1.0%, moderate risk (score:1-2),13.4%, high risk (score≥3),90.0%). Compared to the classical Mehran’s and ACEF CIN risk score models, the risk score (Chen) across the subgroup of the study population exhibited similar discrimination and predictive ability on CIN (c-statistic:0.828, 0.776, 0.853, respectively), in-hospital mortality, 2, 3-years mortality (c-statistic:0.738,0.750, 0.845, respectively) in the validation population. Conclusions: Our data showed that this simple risk model exhibited good

ACCEPTED MANUSCRIPT 4

discrimination and predictive ability on CIN, similar to Mehran’s and ACEF score, and even on long-term mortality after emergent PCI.

AC CE P

TE

D

MA

NU

SC

RI

intervention, risk score model, acute coronary syndromes

PT

Keywords: contrast-induced nephropathy, emergent percutaneous coronary

ACCEPTED MANUSCRIPT 5

1. Introduction

PT

Contrast-induced nephropathy (CIN) develops as a complication after coronary

RI

diagnostic and interventional procedures, and is associated with increased mortality,

SC

longer hospitalization stays, increased health care costs, and higher risk of adverse clinical outcomes, especially among patients undergoing emergent or primary

NU

percutaneous coronary intervention (PCI) [1-5]. The risk of CIN after emergent PCI is

MA

significantly higher than after elective PCI [6-8].

Current clinical guidelines suggest that patients should be assessed for the risk of

D

CIN after PCI [9-10] with CIN predictive model included patient-related risk factors

TE

(advanced age, renal insufficiency, hypotension, intra-aortic balloon pump, diabetes

AC CE P

mellitus) or procedure-related risk factors (large contrast volume, insufficient hydration and high osmolality agents) [11-13]. The hemodynamic unstable markers, including reduced ejection fraction, cardiogenic shock, hypotension were reported as the most import risk factors of CIN after emergent PCI [14-17]. Although the combination of above risk factors is common, a cumulative risk score model for prediction for CIN and prognosis among patients undergoing emergent PCI is limited in clinical practice. The most commonly used Mehran’s risk score model comprised 8 variables, which was established and validated in patients without acute myocardial infarction (AMI) [11]. And majority of other models use variables that are not known at the time of an emergent PPCI [12-13]. There is a need to develop a simple and rapid risk prediction tool for patient with

ACCEPTED MANUSCRIPT 6

ST-elevation-myocardial infarction (STEMI) or non-ST-elevation acute coronary syndromes (NSTE-ACS), who are at higher risk of undergoing emergent PCI.

PT

Therefore, in this prospective study, the purpose was to develop a simple risk score

RI

model that could be rapidly applied to evaluate the risk of CIN following emergent

SC

PCI.

2. Methods

NU

2.1. Study population

MA

Data for this study were obtained from a prospective observational study that included all consecutive patients who underwent coronary angiography or PCI

D

between January 2010 and December 2013 in Guangdong general hospital. Patients

TE

[14] were included if they were diagnosed with STEMI and presented quite high risk

AC CE P

in those with NSTE-ACS (ie, those with refractory angina, hemodynamic instability) in the analysis. The exclusion criteria were pregnancy, lactation, intravascular administration of contrast medium within the last 7 or 3 days postoperatively (n=34), lack of use of low-osmolality contrast agents (n = 21), cardiovascular surgery or endovascular repair (n=2), end-stage renal disease or renal replacement (n=2), missing preoperative creatinine data (n=3), lack of use of isotonic saline for hydration (n=5). The study was approved by the ethic committee of our institution, and written informed consent was obtained from all patients. Finally, 692 patients were included in the analysis. 2.2. Study protocol

ACCEPTED MANUSCRIPT 7

Emergent PCI, as primary PCI for patients with STEMI and immediate PCI for patients with NSTE-ACS who were at very high risk, was used with standard guide

PT

catheters, guide wires, balloon catheters, and stents via the femoral or radial approach.

RI

The contrast volume and types (nonionic, low-osmolality [either Iopamiron or

SC

Ultravist]), medication, as well as the indication to intra-aortic balloon pump (IABP) support, were left to the discretion of interventional cardiologist and patient’s

NU

condition according to clinical guidelines [9,18-19]. Serum creatinine (SCr)

MA

concentrations were measured on hospital admission before the procedure, every day for the following 3 days, and at discharge from the CCU.

D

The Cockcroft-Gault and Modified Diet in Renal Disease formula were used to

TE

calculate creatinine clearance (CrCl) and estimated glomerular filtration rate (eGFR)

AC CE P

respectively [20]. We administrated hydration with normal saline at a rate of 1 mL/kg/h (0.5 mL/kg/h in cases of severe left ventricular <40% or congestive heart failure) at the start of the procedure or just before the procedure, which was continued 6～24 hours after the procedure. 2.3. Definitions and follow-up CIN0.5 was defined as an absolute increase in the serum creatinine concentration by 0.5 mg/dl compared to the baseline value within 72 hours of contrast exposure [21]. Advanced Congestive Heart Failure (CHF): New York Heart Association (NYHA) class >2, pulmonary edema or Killip class>1 [11]. “Anemia” was defined using World Health Organization criteria, namely, baseline hematocrit value <39% for men and<36% for women [11]. “Hypotension” was systolic blood pressure (SBP) <80 mm

ACCEPTED MANUSCRIPT 8

Hg for at least 1 h requiring inotropic support with medications or IABP within 24 h peri-procedurally [11].

PT

Follow-up events were carefully monitored and recorded by trained nurses

RI

through office visits and telephone interviews at 1, 6, 12, 24 and 36 months after the

SC

coronary angiography. Major adverse clinical events (MACEs) was defined as mortality, re-non-fatal acute myocardial infarction (Re-AMI), target vessel

NU

revascularization (TVR), contrast-induced acute kidney injury (CI-AKI) requiring

hospitalization.

D

2.4. Development of Risk Score

MA

renal replacement therapy (RRT), stroke, and re-hospitalization after index

TE

692 eligible patients from the entire data base were randomized in a 2:1 ratio to

AC CE P

established a development dataset (n = 461) and a validation dataset (n = 231). The univariate associations between baseline clinical and key procedural characteristics and CIN was identified in the risk score development dataset. Independent predictors of CIN were performed through estimating odds ratios (ORs) obtained from multivariate logistic regression analysis. Risk factors that were significant in the univariate analysis were used for selection in the final model, this average of the measure was then subtracted to select the best subset of risk factors by bootstrap method in development dataset to avoid overfitting the data. We established and validated sub-new risk model (as Chen Sub Score) on patients with CKD or renal insufficiency, defined as CrCl<60ml/min undergoing emergent PCI. The variables with P<0.05 in final multivariate model were assigned a weighted

ACCEPTED MANUSCRIPT 9

integer coefficient value based on its β value for the prediction of CIN. Therefore, the final risk score for each patient represented the sum of integer coefficients. The risk

PT

score was applied to test the incidence of CIN and the short- and long-term outcomes

RI

in the validation dataset. The predictive accuracy for the incidence of CIN, in-hospital,

SC

two-year and three-year all-cause mortality and MACEs in our model were compared with Mehran’s score, ACEF score [22] and Chen plus score, including additional

NU

variables: LVEF<40%, diabetes mellitus and anemia, by receiver-operating

MA

characteristic (ROC) curves and the Hosmere Lemeshow test for calibration. The prognostic significance of risk score on follow-up mortality was estimated using the

TE

3. Results

D

Kaplan-Meier method.

AC CE P

3.1. Baseline Characteristics

The cumulative incidence of CIN was 55(7.9%) in the whole study population (n=692), of which, 32(6.9%) in the development dataset. The comparison of the baseline clinical and procedural characteristics of subgroups defined by CIN0.5 were listed in Table 1. Overall (table1-A), the mean age was 62.2 years (SD 12.4 years), and there were 132 (16.8%) females. The mean baseline serum creatinine level was 1 mg/dl (SD 0.47mg/dl), whereas 80 (11.6%) of patients presented creatinine levels>1.5 mg/dl. 73 (10.5%) and 70 (10.1%) patients suffered from hypotension and IABP during the procedure respectively. The average left ventricular ejection fraction (LVEF) showed in the table 1-A was 53.91% (SD 10.65%). Laboratory measurements such as B-type natriuretic peptide (BNP), serum urea nitrogen, uric acid were

ACCEPTED MANUSCRIPT 10

significantly higher in the CIN group when compared with non-CIN group, as well as procedural characteristics contrast volume and hydration volume. There were no

RI

myocardial infarction (MI), and diabetes mellitus (DM).

PT

inter-group differences in terms of sex and a medical history of anemia, previous

SC

3.2. Univariable logistic regression models and multivariate model Univariable logistic regression models associated with CIN are shown in Table 2.

NU

A total of 11 variables were analyzed in the development of CIN. The significant

MA

correlates included demographics (age>75 years and heart rate), risk factors for coronary artery disease (hypertension), several laboratory findings (serum urea

D

nitrogen, serum creatinine), treatment modalities (β-blockers, diuretics, angiotensin

TE

converting enzyme inhibitors/angiotensin receptor blocker use), and several

AC CE P

angiographic and/or procedural characteristics (hypotension, IABP use and contrast amount).

The multivariate model of CIN predictors was obtained from all 461 patients in development dataset with no missing co-variate value. Age>75 years, hypotension, use of IABP, SCr>1.5 mg/dl were identified as independent predictors which demonstrated associated with CIN remarkably (Table 3). The Hosmer Lemeshow statistic of multivariable model did not suggest a lack of fit (x2 = 1.367, p = 0.505). 3.3. Development of risk score The incidence of CIN by risk score assignment is depicted in Fig. 1, with significant trends across increasing score values for predicting CIN (Cochran Armitage chi-square, p <0.001). Based on the obtained frequencies of CIN in relation

ACCEPTED MANUSCRIPT 11

to different risk score, 461 patients in the development dataset were further categorized into three groups in low risk (score=0), 1.0%, moderate risk (score: 1-2),

PT

13.4%, high risk (score≥3), 90.0% (Fig. 2).

RI

3.4. Validation and comparison of risk score

SC

CIN occurred in 23 (9.9%) of 231 patients in the validation dataset. The rates of CIN in the validation set presented in parallel to those in the development set inside

NU

each of the three risk groups (Fig. 2).The ability of the risk score to predict the

MA

outcomes of post emergent PCI was further evaluated. The developed CIN model demonstrated similar discriminative power with respect to CIN occurrence in the

D

validation population while compared with Mehran, ACEF score (c-statistic: 0.832,

TE

0.776, 0.853, respectively) (Fig. 3). The main results of this study are also

AC CE P

summarized schematically in Figure 4. 3.5. Subgroup Analysis (renal insufficiency,CrCl<60ml/min) Chen score model demonstrated good discriminative power with respect to CIN occurrence in the patients with CKD or renal insufficiency (CrCl<60ml/min) compared with Mehran and ACEF score (c-statistic: 0.763, 0.726, 0.757, respectively) (Supplement Fig. 1). In addition, according to univariable logistic regression models among subgroup patients with renal insufficiency (CrCl<60ml/min) undergoing emergent PCI (n=179), a total of 6 variables (age, heart rate, blood urea nitrogen, SCr>1.5mg/dl, peri-hypotension, use of IABP) were analyzed in the development set (Supplement Table 1). The multivariate model of CIN predictors were obtained from all 179 patients with no missing co-variate value, age, heart rate, SCr>1.5mg/dl and

ACCEPTED MANUSCRIPT 12

use of IABP, which were identified as independent predictors associated with CIN remarkably (Supplement Table 2). The Hosmer Lemeshow statistic of multivariable

PT

model did not suggest a lack of fit (x2 = 9.803, p = 0.279). The developed CIN model

RI

demonstrated similar discriminative power with respect to CIN occurrence in the

SC

validation population while compared with Chen score, Mehran, ACEF score

3.6. Follow-up of clinical outcomes

NU

(c-statistic: 0.796, 0.770, 0.683, 0.711, respectively) (Supplement Fig. 2).

MA

The mean follow-up time was 2.27±1.07 years (median, 1.95 years; interquartile range, 1.52-3.26 years). According to log-rank analysis (Fig. 5), patients with a high

D

risk score(≥3) presented with a higher rate of all-cause death than patients with a

TE

moderate (1-2) and low risk score (<1). Significant increases in follow-up mortality

AC CE P

rate was observed with increment of risk score(x2 = 9.949, p = 0.007). In addition, patients developing CIN presented with a higher rate of all-cause death than those without CIN during follow-up (37.5% vs. 5.4%, x2 = 22.98, p <0.001). The association of risk score and short- and long-term outcomes are shown in Fig. 6, and the comparison of risk scores on the short- and long- term outcomes are presented (table 4). The present risk score model as assessed in the validation population by the c-statistic demonstrated the same predictive accuracy compared to Mehran, ACEF risk scores (Fig. 7), as well as the calibration (Fig. 8). The Chen, Mehran, ACEF risk scores had similarly high prognostic value in-hospital, 2,3-years mortality (c-statistic:0.738,0.750, 0.845, respectively) and MACEs. We also add new Chen score (as Chen Plus Score) including additional variables:

ACCEPTED MANUSCRIPT 13

heart function (LVEF<40%), diabetes mellitus and anemia in assessment of the

RI

accuracy compared to Mehran, ACEF risk scores (Fig. 7)

PT

present risk score (Chen score) by the c-statistic demonstrated the same predictive

SC

4. Discussion

In the present study, we established and validated a simple, novel CIN risk score

NU

including four risk factors: age>75 years, SCr>1.5 mg/dl, hypotension and the use of

MA

IABP for developing CIN risk stratification tool among patients undergoing emergent PCI. Our simple, novel risk score (Chen) model demonstrated good discrimination

D

and predictive ability on CIN, and even for long-term outcomes after emergent PCI,

TE

which maybe a good alternative to other models.

AC CE P

Mehran’s risk score model，which is regarded as the classic CIN risk score, comprised 8 variables, covering the four risk factors in our model, and was established in patients without AMI or/and shock, that is difficult to extent to emergent PCI without enough time to assess several risk factors [11]. The predictive value of our risk score maybe similar to that of Mehran’s score (according to ROC analysis), but with simple application for risk stratification. Bartholomew’s risk score model also contained 8 variables with a range of 0 to 11 points for unselective patients PCI, which had similar predictive accuracy for CIN [23-24]. In another study, risk score model based on 5 variables demonstrated higher discriminative ability, while compared with previously published risk scores for CIN mentioned above [25]. Tsai et al. reported a risk score that uses pre-procedure variables, but this model

ACCEPTED MANUSCRIPT 14

has modest discrimination and may not be appropriate for patient-level decision making [26]. Hitinder S. Gurm et al obtained a computational tool to predict CIN

PT

applied for pre-procedural risk stratification, but the model used 46 baseline clinical

RI

variables and identified 15 most influential variables, which make it complicated and

SC

unsuitable to emergent PCI [27].

Giancarlo Marenzi’ s relative sample CIN risk score model based on AMI

NU

patients undergoing primary PCI, included advanced age, anterior infarction, long

MA

time-to-reperfusion, high contrast volume and use of IABP, which is similar to ours, but with small samples sizes (n=208) [17]. Besides, in Alberto Bouzas-Mosquera’s 5

D

factor CIN score including cardiogenic shock, diabetes mellitus, urea levels have

TE

shown predictive value for ARF after urgent cardiac catheterization. Unfortunately,

AC CE P

they failed to evaluate the risk score model’s predictive value for long-term prognosis with median follow-up time of only 1.3 years [28]. Our model using4 sample factors demonstrated similar accuracy compared to Mehran’s and ACEF score, in predicting CIN development and long-term prognosis. The mechanism of CI-AKI is hypoxia damage (i.e. the outer medulla) and altered renal haemo-dynamics and tubulo-dynamics caused by acute tubular necrosis because of increased adenosine, endothelin, free radical-induced vasoconstriction, and the direct toxic effect of contrast agent [1]. The incidence of CIN in recent reports showed to range from 12% to 26%, based on different definitions [29], and was 8.2% in our study following the definition of absolute increase of SCr. According to our study, Age > 75 years, use of IABP, hypotension and SCr>1.5 mg/dl were 4 key risk factors

ACCEPTED MANUSCRIPT 15

of CIN occurrence in identifying patients with emergent PCI. The advanced age was an independent predictor of CIN development after

PT

primary PCI for STEMI in many previous studies. Age in Giuseppe Ando's study was

RI

calculated as a continuous variable with 6 stages (each 10 years) [26]. Another study

SC

used categoricalvariable [30]. Both of them showed high risk related with CIN in the multivariate models.

NU

As with these previous studies, patients undergoing primary PCI are at higher

MA

risk of CIN, which is strongly associated with hemodynamic instability. IABP as a marker of significant hemodynamic disturbances during PCI may be linked with the

D

occurrence of CIN. In current practice, use of IABP is most often required to provide

TE

hemodynamic support in patients admitted with cardiogenic shock, in case of

AC CE P

unexpected hypotension and hemodynamic deterioration during complex and high-risk PCI. The presence of hypotension and IABP support without hypotension are powerful predictors of CIN [11], which were reported as predictors of CIN after emergent PCI in the present study. Moreover, the study suggested that incidence of Bleeding Academic Research Consortium (BARC) bleeding was significantly higher in the urgent group (12.8%) compared with the elective (4.1%) groups for patients who required an IABP support [31]. Hemodynamic instability is essentially a well-known cause of AKI and independent of contrast media, especially a higher percent of the CIN occurred in patients with AMI or STEMI, it is important to understand that the patients with CIN were more likely to have required IABP [5]. Accordingly, it indicated that hypotension and IABP support enrolled in the

ACCEPTED MANUSCRIPT 16

observation make more sense in patients undergoing emergent PCI. In recent studies, uric acid, anemia, BNP and C-reactive protein emerged as

PT

independent predictors of CIN in multiple logistic regression analysis in patients with

RI

STEMI after primary PCI [32-34]. Peripheral arterial disease (PAD) was

SC

independently associated with a doubling of the in-hospital mortality risk among patients undergoing primary PCI for AMI [35]. In previous studies, DM and

NU

metformin was assessed to be independent predictor for CIN as well [11, 25]. It had

MA

also been suggested that initiation of metformin shortly after primary PCI has no adverse effect on renal function in patients without DM or prior renal impairment,

D

further providing evidence of the safety of metformin use after MI and subsequent

TE

contrast exposure [36], though there was no difference in DM between CIN group and

AC CE P

non-CIN group in our study, the reason may be that patients with diabetic nephropathy undergoing PCI had a very high risk of developing CIN which was inapplicable to DM with normal renal function [29, 37]. Patients undergoing primary PCI are at high risk for CIN. Increasing evidence exists that several risk factors have been demonstrated to be associated with deteriorative outcomes. Mehran risk score is able to stratify patients for poor shortand long-term clinical outcomes in patients with STEMI [38]. Liu’s study of predicting CIN and its outcomes after primary PCI in patients with STEMI using comparison of 6 risk scores (Mehran; Gao; Chen; ACEF; AGEF; and Global Registry for Acute Coronary Events risk scores), reported that the risk scores above had high discriminatory ability with the majority also having good calibration for CIN0.5,

ACCEPTED MANUSCRIPT 17

in-hospital death and MACEs, and 3-year all-cause mortality in those patients [39]. Patients with CI-AKI showed higher rates of net adverse clinical events (NACEs), a

PT

combination of major bleeding or composite major adverse cardiac events (MACEs),

RI

consisting of death, re-infarction, target vessel revascularization for ischaemia, or

SC

stroke at 30 days and 3 years in Narula A’ study applied the Mehran and Marenzi scores [11,17]. They also had higher rates of mortality at 30 days and 3 years which

NU

showed contrast-induced acute kidney injury is associated with poor short- and

MA

long-term outcomes after primary PCI in STEMI [40]. The Logistic Clinical Syntax Score (log CSS) calculated according to the formula developed by FarooqV has been

D

found to be effective for the prediction of mortality in patients with STEMI, a log

TE

CSS > 9.5 was associated with in-hospital and long-term mortality, re-infarction,

AC CE P

revascularization, and in-hospital hemodialysis (p < 0.001 for each) [41]. In addition, intravenous saline hydration recognized effective prophylactic therapy could reduce the risk of CIN more than 50% during primary PCI. Preventive hydration should be mandatory to patients in emergent procedures in case of increased mortality and MACEs resulting from CIN [42-43]. Generally, it is difficult to take pre-procedure preventive measures such as hydration by saline and various other pre-treatments in emergent PCI, which are considered effective for the prevention of CIN. However, because circulatory insufficiency was strongly associated with the occurrence of CIN, we have made further efforts to make an early diagnosis and to minimize door-to-balloon time in those patients. The risk of CIN or risk factors may have potential ethnic differences for the

ACCEPTED MANUSCRIPT 18

patients undergoing emergent PCI, especial in subgroup with renal insufficiency. Dr Liu’s and colleagues reported in the U.S. Chronic Renal Insufficiency Cohort Study

PT

that non-Hispanic black with moderate and mild CKD (eGFR 20-70 ml/min/1.73m2)

RI

have a significantly higher CVD risk factor score than non-Hispanic white, and the

SC

race differences in cardiovascular disease risk maybe similar to that in CIN score [44]. However, our study did not report ethnic differences (such as Hang race or non-Hang

NU

race) as the data were from a single center among Chinese. In the future study, we

MA

will include the ethnic variables to control the potential confounders for evaluating CIN or cardiovascular disease risk, especially in patients with renal insufficiency.

D

Finally, the prediction model of CIN risk score described offers a great

TE

investigational tool in future studies regarding CIN prevention in patients undergoing

AC CE P

emergent PCI. We considered that certain measures may be very effective in the prevention of CIN only in certain risk score-based patient subsets. For example, debate already exists on whether statins exert the protective effect on renal function for patients with STEMI or AMI undergoing primary PCI, whereas data are more supportive of its utility in patients [45-47]. On one hand, it would be detrimental to entirely dismiss a preventive measure because it may not prevent CIN after emergent PCI [48], but it would also be inappropriate to apply universally a protective measure to these patients, if it is only effective in a certain subgroup (patients at different risk score classification). The application of CIN risk score might help clarify such controversial issues and potentially lead to patient subset-oriented recommendations. 4.1 Study limitations

ACCEPTED MANUSCRIPT 19

The present study had several limitations. First, as with other observational studies, our study findings must be evaluated with certain caveats, which was conducted with

PT

small sample size in a single center. Second, the CrCl was computed using the

RI

Cockcroft-Gault formula, rather than measured directly. Third, variations in our

SC

measurement times might have given rise to missing post-procedure peak creatinine levels. Furthermore, 50% of the patients were discharged 3 days after the coronary

NU

angiography, so SCr concentrations were not measured on day 3 in these patients.

MA

Variation in the measurement times may have led to overlook peak levels of the SCr post procedure, which may have also led to an underestimation of the true incidence

D

of nephropathy in the current study population. Moreover, poor patient compliance

TE

caused a high lost to follow-up rate, which may have affected the results about

AC CE P

clinical adverse outcomes, and influenced the significance of the analysis. Lastly, our study did not report ethnic differences (such as Hang race or non-Hang race) as the data were from a single center among Chinese. Future studies on ethic difference and including patients with renal insufficiency are needed.

5. Conclusion There is a need for a simple and robust risk tool for prediction of CIN and prognosis in patients undergoing emergent PCI. Our simple, novel risk score (Chen) model demonstrated good discrimination and predictive ability on CIN with similar predictive value for long-term outcomes after emergent PCI, which maybe a good alternative to other models. However, the accuracy of our Chen model needs to be

ACCEPTED MANUSCRIPT 20

validated in further large scan of cohorts. Funding Sources: This work was supported by the Guangdong Provincial

PT

Cardiovascular Clinical Medicine Research Fund (grant number 2009X41 to Y.L. and

RI

N.T.); Science and Technology Planning Project of Guangdong Province

SC

(PRECOMIN study by Y.L. in 2011 and study grant number 2008A030201002 to JY.C.); and the Guangdong Cardiovascular Institute, the Youth project of Fujian

NU

provincial health and Family Planning Commission (grant number 2015-1-9).

AC CE P

TE

D

MA

Disclosures: None

ACCEPTED MANUSCRIPT 21

Reference

PT

[1] Seeliger E, Sendeski M, Rihal CS, Persson PB. Contrast-induced kidney injury:

RI

mechanisms, risk factors, and prevention. Eur Heart J. 33 (16) (2012) 2007-15.

SC

[2] Liu Y, Chen JY, Tan N, et al. Safe limits of contrast vary with hydration volume for prevention of contrast-induced nephropathy after coronary angiography among

NU

patients with a relatively low risk of contrast-induced nephropathy. Circ Cardiovasc

MA

Interv. 8 (6) (2015) pii: e001859.

[3] Aslanger E, Uslu B, Akdeniz C, Polat N, Cizgici Y, Oflaz H. Intrarenal

D

application of N-acetylcysteine for the prevention of contrast medium-induced

TE

nephropathy in primary angioplasty. Coron Artery Dis. 23 (4) (2012) 265-70.

AC CE P

[4] Wi J, Ko YG, Shin DH, et al. Prediction of Contrast-Induced Nephropathy With Persistent Renal Dysfunction and Adverse Long-term Outcomes in Patients With Acute Myocardial Infarction Using the Mehran Risk Score. Clin Cardiol. 36 (1) (2013) 46-53.

[5] Watabe H, Sato A, Hoshi T, et al. Association of contrast-induced acute kidney injury with long-term cardiovascular events in acute coronary syndrome patients with chronic kidney disease undergoing emergent percutaneous coronary intervention. Int J Cardiol. 174 (1) (2014) 57-63. [6] Ueda H, Yamada T, Masuda M, et al. Prevention of contrast-induced nephropathy by bolus injection of sodium bicarbonate in patients with chronic kidney disease undergoing emergent coronary procedures. Am J Cardiol. 107 (8) (2011) 1163-7.

ACCEPTED MANUSCRIPT 22

[7] Thayssen P, Lassen JF, Jensen SE, et al. Prevention of contrast-induced nephropathy with N-acetylcysteine or sodium bicarbonate in patients with

PT

ST-segment-myocardial infarction: a prospective, randomized, open-labeled trial. Circ

RI

Cardiovasc Interv. 7 (2) (2014) 216-24.

SC

[8] Marenzi G, Assanelli E, Campodonico J, et al. Contrast volume during primary percutaneous coronary intervention and subsequent contrast-induced nephropathy and

NU

mortality. Ann Intern Med. 150 (3) (2009) 170-7.

MA

[9] Levine GN, Bates ER, Blankenship JC, et al; American College of Cardiology Foundation; American Heart Association Task Force on Practice Guidelines; Society

D

for Cardiovascular Angiography and Interventions. 2011 ACCF/AHA/SCAI

TE

Guideline for Percutaneous Coronary Intervention. A report of the American College

AC CE P

of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines and the Society for Cardiovascular Angiography and Interventions. J Am Coll Cardiol 58 (2011) e44–122. [10] American College of Emergency Physicians; Society for Cardiovascular Angiography and Interventions, O'Gara PT, et al. 2013 ACCF/AHA guideline for the management of ST-elevation myocardial infarction: a report of theAmerican College of Cardiology Foundation/American Heart Association Task Forceon Practice Guidelines. J Am Coll Cardiol. 61 (4) (2013) e78-140. [11] Mehran R, Aymong ED, Nikolsky E, et al. A simple risk score for prediction of contrast-induced nephropathy after percutaneous coronary intervention: development and initial validation. J Am Coll Cardiol. 44 (7) (2004) 1393-9.

ACCEPTED MANUSCRIPT 23

[12] Yong Liu, Yuan-Hui Liu, Ning Tan, et al. Novel risk scoring for pre-procedural prediction of contrast-induced nephropathy and poor long-term outcomes among

PT

patients with chronic total occlusion undergoing percutaneous coronary intervention.

RI

European Heart Journal Supplements (2015) 17 (Supplement C), C34–C41.

SC

[13] Ji L, Su X, Qin W, et al. Novel risk score of contrast-induced nephropathy after percutaneous coronary intervention. Nephrology(Carlton). 20 (8) (2015) 544-51.

NU

[14] Liu Y, Lin L, Li Y, et al. Relationship Between the Urine Flow Rate and Risk of

MA

Contrast-Induced Nephropathy After Emergent Percutaneous Coronary Intervention. Medicine (Baltimore). 94 (50) (2015) e2258.

D

[15] Ivanes F, Isorni MA, Halimi JM, et al. Predictive factors of contrast-induced

TE

nephropathy in patients undergoing primary coronary angioplasty. Arch Cardiovasc

AC CE P

Dis. 107 (8-9) (2014) 424-32.

[16] Maioli M, Toso A, Leoncini M, Micheletti C, Bellandi F. Effects of hydration in contrast-induced acute kidney injury after primary angioplasty: a randomized, controlled trial. Circ Cardiovasc Interv. 4 (5) (2011) 456-62. [17] Marenzi G, Lauri G, Assanelli E, et al. Contrast-induced nephropathy in patients undergoing primary angioplasty for acute myocardial infarction. J Am Coll Cardiol. 44 (9) (2004) 1780-5. [18] Canadian Cardiovascular Society; American Academy of Family Physicians; American College of Cardiology, et al. 2007 focused update of the ACC/AHA 2004 guidelines for the management of patients with ST-elevation myocardial infarction: a

ACCEPTED MANUSCRIPT 24

report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol. 51 (2008) 210-47.

PT

[19] Kushner FG, Hand M, Smith SC Jr, et al. 2009 focused updates: ACC/AHA

RI

guidelines for the management of patients with ST-elevationmyocardial infarction

SC

(updating the 2004 guideline and 2007 focusedupdate) and ACC/AHA/SCAI guidelines on percutaneous coronaryintervention (updating the 2005 guideline and

NU

2007 focused update): areport of the American College of Cardiology

MA

Foundation/AmericanHeart Association Task Force on Practice Guidelines. J Am CollCardiol. 2009;54:2205-41. Erratum in: J Am Coll Cardiol. 54 (2010) 2464.

D

[20] Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum

TE

creatinine. Nephron 16 (1976) 3-41.

AC CE P

[21] Slocum NK, Grossman PM, Moscucci M, et al. The changing definition of contrast-induced nephropathy and its clinical implications: insights from the Blue Cross Blue Shield of Michigan Cardiovascular Consortium (BMC2). Am Heart J. 163 (5) (2012) 829-34.

[22] Andò G, Morabito G, de Gregorio C, Trio O, Saporito F, Oreto G. The ACEF scoreas predictor of acute kidney injury in patients undergoing primary percutaneouscoronary intervention. Int J Cardiol. 168 (4) (2013) 4386-7. [23] Bartholomew BA, Harjai KJ, Dukkipati S, et al. Impact of nephropathy after percutaneous coronary intervention and a method for risk stratification. Am J Cardiol. 93 (12) (2004) 1515-9.

ACCEPTED MANUSCRIPT 25

[24] Tziakas D, Chalikias G, Stakos D, et al. Development of an easily applicable risk score model for contrast-induced nephropathy prediction after percutaneous coronary

PT

intervention: a novel approach tailored to current practice. Int J Cardiol. 163 (1) (2013)

RI

46-55.

SC

[25] Tziakas D, Chalikias G, Stakos D, et al. Validation of a new risk score to predict contrast-induced nephropathy after percutaneous coronary intervention. Am J Cardiol.

NU

113 (9) (2014) 1487-93.

MA

[26] Tsai TT, Patel UD, Chang TI, et al. Validated contemporary risk model of acute kidney injury in patients undergoing percutaneous coronary interventions: insights

TE

Assoc. 3 (6) (2014) e001380.

D

from the National Cardiovascular Data Registry Cath-PCI Registry. J Am Heart

AC CE P

[27] Gurm HS, Seth M, Kooiman J, Share D. A novel tool for reliable and accurate prediction of renal complications in patients undergoing percutaneous coronary intervention. J Am Coll Cardiol. 61 (22) (2013) 2242-8. [28] Bouzas-Mosquera A, Vázquez-Rodríguez JM, Calviño-Santos R, et al. Contrast-induced nephropathy and acute renal failure following emergent cardiac catheterization: incidence, risk factors and prognosis. Rev Esp Cardiol. 60 (10) (2007) 1026-34. [29] Sany D, Refaat H, Elshahawy Y, Mohab A, Ezzat H. Frequency and risk factors of contrast-induced nephropathy after cardiac catheterization in type II diabetic patients: a study among Egyptian patients. Ren Fail. 36 (2) (2014) 191-7.

ACCEPTED MANUSCRIPT 26

[30] Andò G, Morabito G, de Gregorio C, Trio O, Saporito F, Oreto G. Age, glomerular filtration rate, ejection fraction, and the AGEF score predict

PT

contrast-induced nephropathy in patients with acute myocardial infarction undergoing

RI

primary percutaneous coronary intervention. Catheter Cardiovasc Interv. 82 (6) (2013)

SC

878-85.

[31] Davidavicius G, Godino C, Shannon J, et al. Incidence of overall bleeding in

NU

patients treated with intra-aortic balloon pump during percutaneous coronary

MA

intervention: 12-year Milan experience. JACC Cardiovasc Interv. 5 (3) (2012) 350-7. [32] Fu N, Li X, Yang S, et al. Risk score for the prediction of contrast-induced

D

nephropathy in elderly patients undergoing percutaneous coronary intervention.

TE

Angiology. 64 (3) (2013) 188-94.

AC CE P

[33] Jarai R, Dangas G, Huber K, et al. B-type natriuretic peptide and risk of contrast-induced acute kidney injury in acute ST-segment-elevation myocardial infarction: a substudy from the HORIZONS-AMI trial. Circ Cardiovasc Interv. 5 (6) (2012) 813-20.

[34] Mendi MA, Afsar B, Oksuz F, et al. Uric Acid is a Useful Tool to Predict Contrast-Induced Nephropathy. Angiology. (2016) Mar 22. [35] Jeremias A, Gruberg L, Patel J, Connors G, Brown DL. Effect of peripheral arterial disease on in-hospital outcomes after primary percutaneous coronary intervention for acute myocardial infarction. Am J Cardiol. 105 (9) (2010) 1268-71. [36] Posma RA, Lexis CP, Lipsic E, et al. Effect of Metformin on Renal Function After Primary Percutaneous Coronary Intervention in Patients Without Diabetes

ACCEPTED MANUSCRIPT 27

Presenting with ST-elevation Myocardial Infarction: Data from the GIPS-III Trial. Cardiovasc Drugs Ther. 29 (5) (2015) 451-9.

PT

[37] Chong E, Poh KK, Liang S, Tan HC. Risk factors and clinical outcomes for

RI

contrast-induced nephropathy after percutaneous coronary intervention in patients

SC

with normal serum creatinine. Ann Acad Med Singapore. 39 (5) (2010) 374-80. [38] FA sgura, L Bertelli,D Monopoli,et al. Mehran contrast-induced nephropathy risk

NU

score predicts short- and long-term clinical outcomes in patients with

MA

ST-elevation-myocardial infarction. Circ Cardiovasc Interv. 2010 ;3(5):491-8. [39] Liu YH, Liu Y, Zhou YL, et al. Comparison of Different Risk Scores for

D

Predicting Contrast Induced Nephropathy and Outcomes After Primary Percutaneous

AC CE P

Cardiol. (2016) Apr 6.

TE

Coronary Intervention in Patients With ST Elevation Myocardial Infarction. Am J

[40] Narula A, Mehran R, Weisz G, et al. Contrast-induced acute kidney injury after primary percutaneous coronary intervention: results from the HORIZONS-AMI substudy. Eur Heart J. 35 (23) (2014) 1533-40. [41] Ozturk D, Celik O, Erturk M, et al. Utility of the Logistic Clinical Syntax Score in the Prediction of Contrast-Induced Nephropathy After Primary Percutaneous Coronary Intervention. Can J Cardiol. 32 (2) (2016) 240-6. [42] Jurado-Román A, Hernández-Hernández F, García-Tejada J, et al. Role of hydration in contrast-induced nephropathy in patients who underwent primary percutaneous coronary intervention. Am J Cardiol. 115 (9) (2015) 1174-8.

ACCEPTED MANUSCRIPT 28

[43] Manari A, Magnavacchi P, Puggioni E, et al. Acute kidney injury after primary angioplasty: effect of different hydration treatments. J Cardiovasc Med (Hagerstown).

PT

15 (1) (2014) 60-7.

RI

[44]Liu, L. Using Multivariate Quantile Regression Analysis to Explore

SC

Cardiovascular Risk Differences in Subjects with Chronic Kidney Disease by Race and Ethnicity: Findings from the US Chronic Renal Insufficiency Cohort Study.

NU

International Cardiovascular Forum Journal. 2015; 2:20-26.

MA

[45] Kaya A, Kurt M, Tanboğa IH, et al. Rosuvastatin versus atorvastatin to prevent contrast induced nephropathy in patients undergoing primary percutaneous coronary

D

intervention (ROSA-cIN trial). Acta Cardiol. 68 (5) (2013) 489-94.

TE

[46] Zhao JL, Yang YJ, Zhang YH, You SJ, Wu YJ, Gao RL. Effect of statins on

AC CE P

contrast-induced nephropathy in patients with acute myocardial infarction treated with primary angioplasty. Int J Cardiol. 126 (3) (2008) 435-6. [47] Jo SH, Hahn JY, Lee SY, et al. High-dose atorvastatin for preventing contrast-induced nephropathy in primary percutaneous coronary intervention. J Cardiovasc Med (Hagerstown). 16 (3) (2015) 213-9. [48] Bouzas-Mosquera A, Vázquez-Rodríguez JM, Calviño-Santos R, Vázquez-González N, Castro-Beiras A. Statin therapy and contrast-induced nephropathy after primary angioplasty. Int J Cardiol. 134 (3) (2009) 430-1.

ACCEPTED MANUSCRIPT 29

Figure Legends

PT

Fig. 1. Incidence of contrast induced nephropathy according to the risk score. Increasing risk of CIN with increasing risk score is evident, Cochran Armitage

SC

RI

chi-square, p<0.001.

NU

Fig. 2. Incidence of contrast induced nephropathy in the development (n=461) and validation (n=231) datasets according to risk strata. Low risk (n =3 [1.0%]), moderate

MA

risk (n=20 [13.4%]), high risk (n = 9 [90.0%] in development dataset, and low risk (n

D

=3 [1.9%]), moderate risk (n=17 [23.9%]), high risk (n = 3 [50.0%] in validation

TE

dataset.

AC CE P

Fig. 3. Comparison of predictive accuracy of CIN risk score models between Chen score, Mehran score and ACEF score in CIN0.5 (validation dataset).

Fig. 4. Scheme to define contrast-induced nephropathy (CIN) risk score.

Fig. 5. Cumulative mortality as a function of time for patients with low, medium, and high present risk score. Chi-Square = 9.949, p = 0.007.

Fig. 6. Risk score and short-and long-term outcomes. Rates of in-hospital death and MACEs, 2-year and 3-year all-cause mortality and MACEs, in the low-, moderate-, and high-risk groups were showed according to the Chen, Mehran, ACEF risk scores.

ACCEPTED MANUSCRIPT 30

Fig. 7. Predictive ability of the risk scores for in-hospital death and MACEs, 2-year

PT

and 3-year all-cause mortality and MACEs by Chen plus, Chen, Mehran, ACEF risk

SC

RI

scores (validation dataset).

Fig. 8. Discrimination (c-statistic) and calibration (Hosmere Lemeshow test) for Chen

NU

score, Mehran score and ACEF score. a = Chen score; b = Mehran score; c = ACEF

D

Supplement Figure Legends

MA

score (validation dataset).

TE

Supplement Fig. 1. Comparison of predictive accuracy of CIN risk score models

AC CE P

between Chen score, Mehran score and ACEF score in CIN0.5 (whole dataset).

Supplement Fig. 2. Comparison of predictive accuracy of CIN risk score models between Chen subgroup score, Chen score, Mehran score and ACEF score in CIN0.5 (validation dataset).

ACCEPTED MANUSCRIPT

AC CE P

Figure 1

TE

D

MA

NU

SC

RI

PT

31

ACCEPTED MANUSCRIPT

AC CE P

Figure 2

TE

D

MA

NU

SC

RI

PT

32

ACCEPTED MANUSCRIPT

AC CE P

TE

D

MA

NU

SC

RI

PT

33

Figure 3

ACCEPTED MANUSCRIPT

MA

NU

SC

RI

PT

34

AC CE P

TE

D

Figure 4

ACCEPTED MANUSCRIPT

AC CE P

TE

D

MA

NU

SC

RI

PT

35

Figure 5

ACCEPTED MANUSCRIPT

D

MA

NU

SC

RI

PT

36

AC CE P

TE

Figure 6

ACCEPTED MANUSCRIPT

AC CE P

Figure 7

TE

D

MA

NU

SC

RI

PT

37

ACCEPTED MANUSCRIPT

NU

SC

RI

PT

38

AC CE P

TE

D

MA

Figure 8

ACCEPTED MANUSCRIPT

PT

39

RI

Table 1

Total(n=461)

CIN0.5

CIN(n=32)

Non- CIN(n=429)

p-Value

MA

Characteristic

NU

SC

Baseline Clinical, Biochemical, and Procedural Characteristics in the Development Dataset

Demographics 62.17±12.417

71.78±12.500

61.46±12.124

<0.001

Age ＞ 75 y, n (%)

80(17.4)

16(50.0)

64(14.9)

<0.001

Female sex, n (%)

87(18.9)

6(18.8)

81(18.9)

1.000

Weight, kg

63.962±10.417

61.984±11.375

64.109±10.341

0.266

122.73±22.448

116.72±26.479

123.17±22.089

0.117

DBP, mm Hg,

73.43±12.420

72.41±12.210

73.51±12.446

0.629

HR, b.p.m,

78.36±16.314

84.00±17.124

77.94±16.194

0.043

Medical history, n (%)

CE

AC

SBP, mm Hg,

PT ED

Age, y

ACCEPTED MANUSCRIPT

86(18.7)

9(28.1)

0.154

Hypertension, n (%)

227(49.2)

24(75.0)

203(47.3)

0.003

Hyperlipidemia, n (%)

74(16.1)

1(3.1)

73(17.0)

0.039

Anemia, n (%)

129(28.0)

13(40.6)

116(27.0)

0.099

History of smoking, n (%)

209(45.3)

11(34.4)

198(46.2)

0.197

Previous MI, n (%)

24(5.2)

4(12.5)

20(4.7)

0.054

AMI, n (%)

438(95.0)

32(100.0)

406(94.6)

0.179

LVEF, %

53.62±10.605

48.14±12.711

54.03±10.332

0.003

LVEF<40%, n (%)

45(10.5)

37(9.3)

0.003

8(26.7)

Killip class＞1, n (%)

106(23.0)

16(59.3)

90(22.9)

<0.001

NYHA class >1, n (%)

50(71.4)

5(83.3)

45(70.3)

0.500

NYHA class >2, n (%)

13(18.6)

3(50.0)

10(15.6)

0.038

Heart Function

AC

CE

PT ED

MA

NU

RI

77(17.9)

SC

Diabetes mellitus, n (%)

PT

40

ACCEPTED MANUSCRIPT

PT

41

134.060±19.119

129.177±27.311

134.418±18.409

0.276

BNP, pg/ml

3.089±0.592

3.482±0.619

SC

3.051±0.577

0.009

Serum urea nitrogen, mg/dl

5.546±2.840

8.370±5.526

5.336±2.411

<0.001

Serum albumin, g/L

33.542±4.809

29.813±4.256

33.801±4.746

<0.001

Uric acid, mmol/L

375.279±116.539

479.492±137.613

366.982±110.804

<0.001

LDL-C, mmol/L

3.290±1.039

2.754±1.167

3.324±1.024

0.033

HDL-C, mmol/L

0.938±0.270

1.285±0.332

0.926±0.264

0.064

Total cholesterol, mmol/L

4.995±1.149

4.536±1.368

5.024±1.130

0.099

1.004±0.304

1.180

0.991±0.312

0.570

HbA1c, %

6.637±1.526

6.595±1.209

6.640±1.549

0.897

SCr, mg/dl

1.083±0.450

1.648±0.775

1.041±0.386

<0.001

SCr> 1.5 mg/dl, n(%)

50(10.8)

15(46.9)

35(8.2)

<0.001

MA

PT ED

AC

Serum cystatin C, ng/ml

NU

Hemoglobin, g/L

CE

RI

Laboratory measurements

ACCEPTED MANUSCRIPT

<0.001

SC

70.799±30.407

42.652±23.476

≤30, n(%)

26(5.6)

11(34.4)

15(3.5)

<0.001

30-60, n(%)

153(33.2)

12(37.5)

141(32.9)

<0.001

60-90, n(%)

181(39.3)

8(25.0)

173(40.3)

<0.001

>90, n(%)

101(21.9)

1（3.1)

100(23.3)

<0.001

eGFR, ml/min/1.73mm2

80.710±28.827

52.604±25.256

82.806±27.990

<0.001

7(21.9)

5(1.2)

<0.001

87(18.9)

14(43.8)

73(17.0)

<0.001

60-90, n(%)

216(46.9)

9(28.1)

207(48.3)

<0.001

>90, n(%)

146(31.7)

2(6.3)

144(33.6)

<0.001

RI

72.896±29.835

CE

CrCl, ml/min

PT

42

Medication, n (%)

12(2.6)

AC

30-60, n(%)

MA

PT ED

eGFR class ≤30, n(%)

NU

CrCl class

ACCEPTED MANUSCRIPT

23(71.9)

0.008

β-blocker

365(79.2)

15(46.9)

350(81.6)

<0.001

CCB

44(9.5)

4(12.5)

40(9.3)

0.555

Diuretics

168(36.4)

18(56.3)

150(35.0)

0.016

Statin

454(98.5)

32(100.0)

422(98.4)

0.467

PT ED

MA

NU

RI

401(87.0)

Procedure performed

378(88.1)

SC

ACEI/ARB

PT

43

439(95.2)

30(93.8)

409(95.3)

0.659

Length of stents, mm

32.44±21.65

32.38±17.44

32.44±21.95

0.987

Number of stenting, n

1.37±0.84

1.37±0.84

0.805

1.41±0.84

Iopamidol, iso-osmia, n (%)

262(56.8)

17(53.1)

245(57.1)

0.661

Non-iopamidol, anisosmotic, n (%)

199(43.2)

15(46.9)

184(42.9)

0.661

Contrast volume, ml

127.484±47.662

146.719±55.250

126.049±46.805

0.018

Contrast type

AC

CE

Coronary lesion, n (%)

ACCEPTED MANUSCRIPT

Hydration volume, ml,

982.18±598.171

1598.00±1182.728

<0.001

Hypotention, n (%)

51(11.1)

13(40.6)

38(8.9)

<0.001

IABP, n( %)

47(10.2)

15(46.9)

32(7.5)

<0.001

Mehran risk score

6.17±4.717

13.34±5.683

5.63±4.175

<0.001

MA

NU

RI

936.24±502.766

SC

PT

44

236(51.2)

3(9.4)

233(54.3)

<0.001

Midd

142(30.8)

5(15.6)

137(31.9)

<0.001

High

58(12.6)

11(34.4)

47(11.0)

<0.001

Very high

25(5.4)

13(40.6)

12(2.8)

<0.001

AC

PT ED

low

CE

Mehran Risk Level

Abbreviations: MI, myocardial infarction; SBP , systolic blood pressure; LVEF, left ventricular ejection fraction; BNP, B-type natriuretic peptide; HbA1c, hemoglobin A1c; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; SCr, serum creatinine; eGFR, estimated glomerular filtration rate; CrCl, creatinine clearance; ACEI/ARB, angiotensin converting enzyme inhibitors/angiotensin receptor blocker; CCB, calcium channel blocker; IABP, intra-aortic balloon pump.

ACCEPTED MANUSCRIPT

PT

45

SC

RI

Table 1-A

Total(n=692)

CIN(n=55)

P

71.84±10.691

61.31±12.142

<0.001

24(43.6)

92(14.4)

<0.001

14(25.5)

118(18.5)

0.209

PT ED

Demographics

CIN0.5 Non- CIN(n=637)

MA

Characteristic

NU

Baseline Clinical, Biochemical, and Procedural Characteristics

62.14±12.359

Age ＞ 75 y, n (%)

116(16.8)

Female sex, n (%)

132(19.1)

Weight, kg

64.463±11.093

60.555±11.067

64.800±11.040

0.006

SBP, mm Hg,

122.54±22.118

121.64±29.380

122.62±21.405

0.752

DBP, mm Hg,

73.20±12.224

73.42±14.619

73.18±12.009

0.889

HR, b.p.m,

78.20±16.560

85.11±20.505

77.60±16.055

<0.001

AC

CE

Age, y

ACCEPTED MANUSCRIPT

PT

46

15(27.3)

131(20.6)

0.242

Hypertension, n (%)

344(49.7)

39(70.9)

305(47.9)

<0.001

Hyperlipidemia, n (%)

125(18.1)

5(9.1)

120(18.8)

0.071

Anemia, n (%)

198(28.6)

21(38.2)

177(27.8)

0.102

History of smoking, n (%)

316(45.7)

18(32.7)

298(46.8)

0.045

Previous MI, n (%)

39(5.6)

5(9.1)

34(5.3)

0.247

Previous CABG, n (%)

1(0.1)

0(0.0)

1(0.2)

0.769

AMI, n (%)

658(95.1)

55(100.0)

603(94.7)

0.079

LVEF, %

53.91±10.650

48.36±11.975

54.40±10.399

<0.001

LVEF<40%, n (%)

66(10.3)

14(27.5)

52(8.8)

<0.001

152(24.6)

28(63.6)

124(21.6)

<0.001

SC

146(21.1)

AC

CE

PT ED

MA

NU

Diabetes mellitus, n (%)

RI

Medical history, n (%)

Heart Function Killip class＞1, n (%)

ACCEPTED MANUSCRIPT

PT

47

79(65.8)

11(84.6)

68(63.6)

0.131

NYHA class >2, n (%)

20(16.7)

5(38.5)

15(14.0)

0.026

Hemoglobin, g/L

134.217±18.041

129.364±24.234

134.671±17.326

0.227

BNP, pg/ml

3.066±0.627

3.666±0.659

3.000±0.588

<0.001

Serum urea nitrogen, mg/dl

5.634±3.056

8.032±4.511

5.427±2.807

<0.001

Serum albumin, g/L

33.321±4.785

29.213±4.462

33.687±4.645

<0.001

Uric acid, mmol/L

371.328±120.929

448.941±140.565

363.843±116.322

<0.001

LDL-C, mmol/L

3.241±1.081

2.780±1.222

3.272±1.066

0.028

HDL-C, mmol/L

0.891±0.277

1.022±0.302

0.884±0.275

0.237

Total cholesterol, mmol/L

4.925±1.207

4.538±1.695

4.950±1.166

0.242

Serum cystatin C, ng/ml

1.007±0.353

1.180

1.000±0.358

0.626

HbA1c, %

6.700±1.581

6.679±1.264

6.702±1.606

0.937

SC

RI

NYHA class >1, n (%)

AC

CE

PT ED

MA

NU

Laboratory measurements

ACCEPTED MANUSCRIPT

PT

48

1.088±0.465

1.605±0.741

1.043±0.404

<0.001

SCr> 1.5 mg/dl, n(%)

80(11.6)

24(43.6)

56(8.8)

<0.001

CrCl, ml/min

71.540±31.376

42.174±22.954

74.075±30.722

<0.001

≤30, n(%)

45(6.5)

21(38.2)

24(3.8)

<0.001

30-60, n(%)

223(32.2)

20(36.4)

203(31.9)

<0.001

60-90, n(%)

258(37.3)

12(21.8)

246(38.6)

<0.001

>90, n(%)

166(24.0)

2(3.6)

164(25.7)

<0.001

eGFR, ml/min/1.73mm2

80.578±28.579

53.945±28.968

82.878±27.377

<0.001

NU

SC

RI

SCr, mg/dl

AC

CE

PT ED

MA

CrCl class

eGFR class ≤30, n(%)

21(3.0)

12(21.8)

9(1.4)

<0.001

30-60, n(%)

126(18.2)

23(41.8)

106(16.2)

<0.001

60-90, n(%)

319(46.1)

15(27.3)

304(47.7)

<0.001

ACCEPTED MANUSCRIPT

226(32.7)

5(9.1)

221(34.7)

<0.001

ACEI/ARB

610(88.2)

40(72.7)

570(89.5)

<0.001

β-blocker

550(79.5)

29(52.7)

521(81.8)

<0.001

CCB

69(10.0)

8(14.5)

61(9.6)

0.238

Diuretics

255(36.8)

31(56.4)

224(35.2)

0.002

Statin

684(98.8)

55(100.0)

629(98.7)

0.403

53(96.4)

612(96.1)

>0.99

29.13±17.41

31.04±21.03

0.654

RI

>90, n(%)

PT

49

NU

MA

PT ED CE

Procedure performed

SC

Medication, n (%)

665(96.1)

Length of stents, mm

30.88±20.76

Number of stenting, n

1.29±0.79

1.25±0.78

1.30±0.79

0.382

393(56.8)

28(50.9)

365(57.3)

0.359

AC

Coronary lesion, n (%)

Contrast type Iopamidol, iso-osmia, n (%)

ACCEPTED MANUSCRIPT

299(43.2)

27(49.1)

272(42.7)

0.359

Contrast volume, ml

126.171±47.306

140.455±48.422

124.937±47.044

0.019

Vein HV, ml,

979.94±649.126

1515.98±1101.940

933.66±572.461

<0.001

Hypotention, n (%)

73(10.5)

19(34.5)

54(8.5)

<0.001

IABP, n (%)

70(10.1)

27(49.1)

43(6.8)

<0.001

Mehran risk score

5.72±4.727

11.78±5.620

5.19±4.259

<0.001

8(14.5)

372(58.4)

<0.001

15(27.3)

187(29.4)

<0.001

RI

Non-iopamidol, anisosmotic, n

PT

50

PT ED

MA

NU

SC

(%)

CE

Mehran Risk Level 380(54.9)

Midd

202(29.2)

High

75(10.8)

15(27.3)

60(9.4)

<0.001

Very high

35(5.1)

17(30.9)

18(2.8)

<0.001

AC

low

Abbreviations: MI, myocardial infarction; CABG, coronary artery bypass grafting; SBP, systolic blood pressure; LVEF, left ventricular ejection

ACCEPTED MANUSCRIPT

PT

51

RI

fraction; BNP, B-type natriuretic peptide; HbA1, hemoglobin A1c; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density

SC

lipoprotein cholesterol; SCr, serum creatinine; eGFR, estimated glomerular filtration rate; CrCl, creatinine clearance; ACEI/ARB, angiotensin

AC

CE

PT ED

MA

NU

converting enzyme inhibitors/angiotensin receptor blocker; CCB, calcium channel blocker; IABP, intra-aortic balloon pump.

ACCEPTED MANUSCRIPT

PT

52

RI

Table 2

OR

CI

p-Value

5.703

2.715-11.979

＜0.001

1.020

1.000-1.040

0.046

NU

Patients (%) Incidence of CIN (%) 17.4

20.00

Heart rate

n/a

n/a

Hypertension

49.2

10.57

3.340

1.468-7.601

0.004

Serum urea nitrogen

n/a

n/a

1.233

1.123-1.354

＜0.001

Hypotension

11.1

7.040

3.227-15.360

＜0.001

IABP

10.2

31.91

10.947

5.007-23.932

＜0.001

SCr＞1.5 mg/dl

10.8

30.00

9.933

4.573-21.573

＜0.001

β-bloctor

79.2

4.11

0.199

0.095-0.416

＜0.001

Diuretic

36.4

10.71

2.391

1.157-4.943

0.019

PT ED

MA

Age＞75 years

CE

Variable

SC

Association of baseline, clinical, pre-procedural characteristics and CIN in the Development Dataset （Univariate Analysis） ）

AC

25.49

ACCEPTED MANUSCRIPT

OR

87.0

5.74

0.345

Contrast volume

n/a

n/a

1.007

p-Value 0.011

1.001-1.014

0.021

0.151-0.786

NU

ACEI-ARB

CI

RI

Patients (%) Incidence of CIN (%)

SC

Variable

PT

53

MA

Abbreviations: CI, confidence interval; OR, odds ratio; CIN, contrast-induced nephropathy; ACEI/ARB, angiotensin converting enzyme

AC

CE

PT ED

inhibitors/angiotensin receptor blocker; IABP, intra-aortic balloon pump.

ACCEPTED MANUSCRIPT

PT

54

SC

RI

Table 3 Multivariate Predictors of CIN After emergent PCI in Development Dataset OR

CI

Age＞75 years

1.682

5.379

2.238-12.928

＜0.001

1

Hypotension

1.742

5.706

2.234-14.574

＜0.001

1

IABP

1.760

5.814

2.378-14.211

＜0.001

1

SCr＞1.5 mg/dl

2.107

8.222

3.353-20.162

＜0.001

1

MA

PT ED

CE AC

Abbreviations as in Table 2.

NU

Model Coefficient（β value）

Variable

p-Value

Integer Score

ACCEPTED MANUSCRIPT

PT

55

Predictive accuracy of Chen plus, Chen, Mehran and ACEF risk score

Chen score

NU

Chen plus score

MA

Events

SC

RI

Table 4

Mehran score

ACEF score

0.776(0.673-0.880)

0.853(0.779-0.927)

AUC(95%)

0.841(0.744-0.939)

0.828(0.737-0.920)

In-hospital mortability

0.840(0.714-0.966)

0.748(0.558-0.938)

0.805(0.653-0.956)

0.857(0.763-0.951)

In-hospital MACEs

0.749(0.669-0.830)

0.693(0.609-0.777)

0.677(0.585-0.770)

0.753(0.673-0.833)

2-year mortability

0.790(0.677-0.904)

0.732(0.591-0.872)

0.737(0.617-0.857)

0.849(0.763-0.935)

2-year MACEs

0.756(0.629-0.882)

0.706(0.566-0.846)

0.693(0.552-0.835)

0.830(0.740-0.923)

3-year mortability

0.793(0.686-0.899)

0.738(0.607-0.869)

0.750(0.636-0.863)

0.845(0.763-0.926)

3-year MACEs

0.760(0.641-0.879)

0.714(0.581-0.846)

0.708(0.573-0.844)

0.828(0.742-0.916)

AC

CE

PT ED

CIN0.5

CIN, contrast induced nephropathy. CIN0.5: Chen plus score vs. Chen score, p = 0.606; Chen plus score vs. Mehran score, p = 0.164; Chen plus

ACCEPTED MANUSCRIPT

PT

56

RI

score vs. ACEF, p = 0.796; Chen score vs. Mehran score, p = 0.132; Chen score vs. ACEF, p = 0.586; Mehran score vs. ACEF score, p = 0.204.

SC

In-hospital mortability: Chen plus score vs. Chen score, p = 0.024; Chen plus score vs. Mehran score, p = 0.492; Chen plus score vs. ACEF, p =

NU

0.779; Chen score vs. Mehran score, p = 0.253; Chen score vs. ACEF, p = 0.251; Mehran score vs. ACEF score, p = 0.537. In-hospital MACEs:

MA

Chen plus score vs. Chen score, p = 0.018; Chen plus score vs. Mehran score, p = 0.053; Chen plus score vs. ACEF, p = 0.932; Chen score vs. Mehran score, p = 0.635; Chen score vs. ACEF, p = 0.172; Mehran score vs. ACEF score, p = 0.135. 2-year mortability: Chen plus score vs.

PT ED

Chen score, p = 0.138; Chen plus score vs. Mehran score, p = 0.328; Chen plus score vs. ACEF, p = 0.323; Chen score vs. Mehran score, p = 0.916; Chen score vs. ACEF, p = 0.160; Mehran score vs. ACEF score, p = 0.157. 2-year MACEs: Chen plus score vs. Chen score, p = 0.192;

CE

Chen plus score vs. Mehran score, p = 0.229; Chen plus score vs. ACEF, p = 0.180; Chen score vs. Mehran score, p = 0.798; Chen score vs.

AC

ACEF, p = 0.106; Mehran score vs. ACEF score, p = 0.072. 3-year mortability: Chen plus score vs. Chen score, p = 0.146; Chen plus score vs. Mehran score, p = 0.411; Chen plus score vs. ACEF, p = 0.351; Chen score vs. Mehran score, p = 0.802; Chen score vs. ACEF, p = 0.174; Mehran score vs. ACEF score, p = 0.208. 3-year MACEs : Chen plus score vs. Chen score, p = 0.202; Chen plus score vs. Mehran score, p = 0.303; Chen plus score vs. ACEF, p = 0.194; Chen score vs. Mehran score, p = 0.909; Chen score vs. ACEF, p = 0.115; Mehran score vs. ACEF score, p = 0.102.