Comparison of prognostic risk scores after successful primary percutaneous coronary intervention

IJCA-24256; No of Pages 6 International Journal of Cardiology xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard

Comparison of prognostic risk scores after successful primary percutaneous coronary intervention☆ Andreas Synetos 1, George Georgiopoulos ⁎,1, Voula Pylarinou 1, Konstantinos Toutouzas, Katerina Maniou, Maria Drakopoulou, Panagiotis Tolis, Antonios Karanasos, Aggelos Papanikolaou, George Latsios, Eleftherios Tsiamis, Dimitrios Tousoulis First Department of Cardiology, Hippokration Hospital, University of Athens, Athens, Greece

a r t i c l e

i n f o

Article history: Received 13 July 2016 Received in revised form 11 November 2016 Accepted 16 December 2016 Available online xxxx Keywords: Primary percutaneous coronary intervention MACE SYNTAX score clinical SYNTAX score ACEF GRS EuroSCORE EuroSCORE II

a b s t r a c t Background: The aim of this study was to compare the predictive ability of clinical risk scores (ACEF, EuroSCORE and EuroSCORE II) to angiographic (SYNTAX score) and combined risk scores (Global Risk Score and Clinical SXscore) towards cardiovascular death and/or major adverse cardiac events (MACE) in patients with STsegment elevation acute myocardial infarction (STEMI) managed with primary percutaneous coronary intervention (pPCI). Methods: A total of 685 patients successfully treated with pPCI were evaluated and the risk scores were calculated. The primary endpoint was the 2-year incidence of fatal cardiac events. Secondary end points were target lesion failure (TLF), repeat revascularization (RR) and MACE. Results: Patients distributed in the highest tertile of EuroSCORE II presented increased rates of CV death (CVD), all-cause mortality and MACE (p b 0.001 for all). EuroSCORE II was associated with increased C-statistics (0.873, 95% CIs: 0.784–0.962 and 0.825, 95% CIs: 0.752–0.898 respectively) for predicting CVD and MACE over competing risk scores (p b 0.05). EuroSCORE II conferred incremental discrimination (Harrell's C, p b 0.05 for all, apart from CSS for predicting CVD) and reclassification value (Net Reclassification Index, p b 0.05 for all, apart from CSS for reclassifying MACE) over alternative risk scores for study's main endpoints. EuroSCORE II independently predicted CVD (HR = 1.06, 95% CIs: 1.03–1.09, p b 0.001) and MACE (HR = 1.07, 95% CIs: 1.04–1.10, p b 0.001). Conclusion: EuroSCORE II has the best predictive ability of CVD and/or MACE after successful pPCI for the treatment of STEMI. © 2016 Elsevier Ireland Ltd. All rights reserved.

1. Introduction Primary percutaneous coronary intervention (PCI, pPCI) has been proven to be the most effective treatment for ST-segment elevation myocardial infarction (STEMI). There are several risk scores for the prediction of the outcome of patients with chronic stable angina or acute coronary syndromes (ACS) undergoing revascularization. SYNTAX (Synergy between Percutaneous Coronary Intervention with TAXUS and Cardiac Surgery) score (SXscore) is a visual angiographic prognostic model that evaluates lesion complexity, the extent and distribution of coronary atheromatosis and stratifies individual risk [1,2]. However, the absence of clinical factors has

☆ This author takes responsibility for all aspects of the reliability and freedom from bias of the data presented and their discussed interpretation. ⁎ Corresponding author at: 21 Orfanidou Street, 11142 Athens, Greece. E-mail address: [email protected] (G. Georgiopoulos). 1 These authors contributed equally to this article.

led to the creation of a pure clinical model ACEF (age, creatinine, left ventricular ejection fraction (LVEF)). There is a renewed interest in combining clinical and angiographic information to define the risk of patients undergoing revascularization in ACS. So, two combined risk models, the Global Risk Classification (GRS) and the Clinical SYNTAX score (CSS) [3,4] have incorporated clinical variables into the SXscore. The performance of these models has been validated and compared in patients with left main disease undergoing PCI or CABG [5]. The majority of prognostic models that have been applied in STEMI were studied before the widespread application of pPCI [6–11]. Studies evaluating the impact of SXscore on the outcome of patients undergoing pPCI have shown that SXscore predicts better the overall mortality and the incidence of major adverse events in patients with STEMI, but failed to establish an association with cardiovascular (CV) mortality [12]. Several risk scores have been proposed for the outcome after pPCI. Moreover, CCS has been identified as the best combined risk score for the prediction of outcome in patients with successful pPCI [13]. However,

http://dx.doi.org/10.1016/j.ijcard.2016.12.078 0167-5273/© 2016 Elsevier Ireland Ltd. All rights reserved.

Please cite this article as: A. Synetos, et al., Comparison of prognostic risk scores after successful primary percutaneous coronary intervention, Int J Cardiol (2016), http://dx.doi.org/10.1016/j.ijcard.2016.12.078

2

A. Synetos et al. / International Journal of Cardiology xxx (2016) xxx–xxx

more recent scores such as EuroSCORE II have not been tested as an outcome predictor in such populations. EuroSCORE II incorporates 10 patient related factors (age, gender, renal impairment, extracardiac arteriopathy, poor mobility, previous cardiac surgery, chronic lung disease, active endocarditis, critical perioperative state and diabetes on insulin) and 5 cardiac related factors (presence of pulmonary hypertension, left ventricular dysfunction, recent myocardial infarction within 90 days, NYHA and CCS classification). This clinical prognostic score meticulously outlines the most common comorbidities aggregating in patients presenting with STEMI along with haemodynamics (preoperative ventricular tachycardia, ventricular fibrillation/sudden cardiac death, cardiopulmonary resuscitation or intubation) and echocardiographic variables (Supplementary Figure1). The aim of this study was 1) to validate and compare the performance of EuroSCORE II in patients undergoing successful pPCI towards prediction of CV death and/or major adverse cardiac events (MACE), and 2) to evaluate whether the combined risk models provide additive prognostic information to the ACEF score, the EuroSCORE, EuroSCORE II and the SXscore. 2. Methods 2.1. Study design and sample This is an observational study; from October 2008 to December 2013, 685 consecutive patients that underwent successful primary PCI due to STEMI in our hospital were recruited out of 703 primary PCIs. Exclusion criteria included: post-arrest primary PCI with or without spontaneous recovery of circulation, administration of thrombolytic agents in the previous 30 days, history of bleeding, major surgery within 15 days, active bleeding or previous stroke within the last 6 months and previous CABG. Before the procedure, all patients enrolled into the study received 500 mg of acetylsalicylic acid, whereas the 600-mg loading dose of clopidogrel or 60 mg of prasugrel (after coronary anatomy was known) or 180 mg of ticagrelor was only given if no clopidogrel had been administered in the previous 7 days. The use of clopidogrel, prasugrel or ticagrelor was left to the discretion of the operator. 596 (87%) patients received clopidogrel as the second anti-platelet agent while in 21 (3.06%) and 68 (9.9%) patients prasugrel (available since 3/2010 in our Hospital) and ticagrelor (available since 5/ 2011 in our Hospital) were administered respectively. The study protocol was approved by the ethics committee of our institution and all participants signed informed consent. Study has been carried out in accordance with the ethical guidelines of the 1975 Declaration of Helsinki. All patients were discharged on 100 mg of acetylsalicylic acid indefinitely and clopidogrel 75 mg for at least 1 year or prasugrel 10 mg or ticagrelor 90 mg b.i.d. for the same period. Baseline clinical characteristics and procedural characteristics were recorded in a dedicated electronic database. Successful primary PCI was defined as the presence of TIMI flow 3 after the procedure [14,15]. Out of the 685 patients of the study, 9 patients were in cardiogenic shock and required temporary circulation support (Intra Aortic Balloon Pump). All 9 patients were successfully weaned from Intra Aortic Balloon Pump within the first 48 h post-PCI. One or more 2nd generation Drug Eluting (DES) Stents (Everolimus and Zotarolimus eluting stents) were implanted during the primary PCI. No Bare Metal Stent (BMS) was used in the study population. During the primary PCI, the therapeutic choice for treating only the culprit vessel or all angiographically significant lesions was left on the operator discretion. In cases where only the culprit lesion was stented, complete revascularization for significant coronary stenoses was performed during the indexed hospitalization. 2.2. Scoring systems The SXscore for each patient was calculated by a team of 2 interventional and experienced cardiologists. All coronary lesions with a diameter stenosis ≥50% in vessels ≥1.5 mm were scored using the SXscore algorithm, which is available on the Web site (www. syntaxscore.com). SYNTAX scoring was performed after wiring or after the use of a small balloon or thrombectomy. The application of predilatation with a balloon, thrombectomy or direct stenting was left to the discretion of the operator. If TIMI flow improved with these measures, this allowed assessment of lesion severity as well as additional disease downstream. Persistence of TIMI 0/1 that did not allow adequate visualization of the lesion was scored as in SXscore I (total occlusion with thrombus) [16]. The investigators that calculated the SXscore were blinded to the patients' clinical characteristics. The EuroSCORE and EuroSCORE II were calculated on the basis of the original methodology [17,18]. The online interactive calculator of EuroSCORE II is available in http://www. euroscore.org/calc.htm The ACEF score was calculated on the basis of the modified formula proposed by Ranucci et al. [19] (i.e., ACEF = [age / left ventricular ejection fraction (LVEF)] + 1 if serum creatinine N2 mg/dl).

The GRS and the CSS were derived as previously described [4,5]. Briefly, the GRS score is a combination of EuroSCORE and SXscore and CSS is a combination of ACEF and Syntax score. Three ΕuroSCOREII categories were identified by tertiles: LOW ≤ 1.98, MID 1.98 to 3.14 and HIGH ≥ 3.14. Three classes of risk were also grouped by tertiles for the ACEF score (LOW ≤1.28, MID 1.28 to 1.55, HIGH ≥ 1.55), for SXscore (LOW ≤8, MID 8 to 14, HIGH ≥ 14), CSS (LOW ≤11, MID 11 to 20.27, HIGH ≥ 20.27) and for EuroSCORE (HIGH ≤ 5, MID 5 to 8 and HIGH ≥ 8).

2.3. Follow-up The investigated outcomes included: adverse events (see below) that were assessed during hospitalization, as well as at 1, 6, 12 and 24 months after hospital discharge. The follow-up was performed in our outpatient department or by telephone. The primary endpoint was the 2-year incidence of fatal cardiac events (i.e., sudden death, myocardial infarction, or death secondary to heart failure). Deaths were considered cardiac following the ICD-10 definitions. In particular, myocardial infarction was defined according to an extended historical protocol definition and according to Academic Research Consortium (ARC) definitions (18, 19). Secondary end points were: target lesion failure (TLF), repeat revascularization (RR), stent thrombosis (ST) and major adverse cardiac events (MACE). Specifically, target lesion failure (TLF) was defined as heart attack attributed to the target vessel (target vessel MI), and ischemia-driven target lesion revascularization [20]. Repeat revascularization (RR) was considered as any kind or revascularization (PCI or CABG) in any coronary artery. Percutaneous revascularizations for significant coronary stenoses other than the culprit lesion during the initial hospitalization by definition were not included in the endpoint of “repeat revascularizations”. Stent thrombosis (ST) was defined according to the ARC definitions [21]. Major adverse cardiac events (MACE) were defined as the composite of CVD, nonfatal myocardial infarction, or target vessel revascularization.

2.4. Statistical analysis Data are presented as mean ± SD or absolute and relative frequencies. Continuous variables not following normal distribution are summarized as median (interquartile range). In all patients EuroSCORE II, SXscore, clinical SXscore, EuroSCORE, Euroadditive, ACEF and GRS were calculated. Analyses were stratified according to EuroSCORE II tertiles as the main score of interest. Patients' characteristics pertaining to the primary PCI were compared across EuroSCORE II groups using the analysis of variance (ANOVA) for the normally distributed continuous variables and the chi-square or the Z tests, for categorical variables. The Kolmogorov–Smirnov test as well as P\ \P plots was used to assess normality. 5-year incidence rates of cardiac mortality were estimated by the Kaplan–Meier method and the log-rank test was used to evaluate differences between tertiles of EuroSCORE II. Data were censored at the time of the last visit. For patients lost during follow-up, their survival data were censored at the last date they were known to be alive. Subsequently, Cox proportional hazard models were fitted to evaluate the predictive ability of the scores on the studied outcomes. The proportional hazard assumption of Cox model was assessed using the appropriate graph and statistical test (Schoenfeld residuals). Associations are presented as hazard ratio (HR) with 95% confidence intervals (CI). Multivariable survival models for main endpoints were built under a bootstrap resampling procedure as previously described [22]. Hundred repeats with forward selection (P b 0.05 for selection) and 100 repeats with backward selection [23] (P b 0.1 for selection) were performed and variables selected in 80% of all repeats were included in the final multivariable model. Certain variables of biological interest (i.e. gender) were forced to be included in the final models. To avoid overfitting of the multivariable Cox models a ratio of ten events per one confounder incorporated was used as a rule of thumb. The scores' performances were evaluated in terms of calibration, discrimination and reclassification [24]. Calibration of the multivariable survival models was performed by comparing predicted probabilities and actual observed risk. Improvement in goodness of fit after adding each score to established risk factors was assessed by the likelihood ratio test and the Hosmer–Lemeshow statistic [25]. In terms of discrimination, Receiver Operating Characteristic (ROC) curves were plotted and the corresponding C-statistics were calculated in order to evaluate scores' performance in predicting the outcomes. Comparisons between C-statistic values were performed using the Z-test, while the Bonferroni rule for multiple comparisons was applied to control for the inflation of type I error. The incremental predictive value of EuroSCORE II over established risk factors and alternative risk scores was assessed by the Harrell's C-index [26] for censored time-to event data [22] (measure for model discrimination with larger values indicating better discrimination). Harrell's c of inverse hazard ratio was used as a measure of the predictive power of survival regression models estimates and statistics derived with the STATA procedures “somers d” and “lincom” [22]. Finally, to evaluate EuroSCORE's II performance and classification ability we calculated the continuous NRI (cNRI), a category-free version of the NRI [22], and the integrated discrimination index (IDI), which integrates the NRI over all possible cutoffs and is equivalent to the difference in discrimination slopes. For survival analysis, the final sample size of 685 subjects with survival follow-up data provided over 85% power to establish two-fold alteration in HR (two-sided) for Cox proportional hazards models towards primary endpoint. Type I error was predefined at 0.05. Statistical analysis was performed by STATA package, version 11.1 (StataCorp, College Station, Texas USA). We deemed statistical significance at a = 0.05.

Please cite this article as: A. Synetos, et al., Comparison of prognostic risk scores after successful primary percutaneous coronary intervention, Int J Cardiol (2016), http://dx.doi.org/10.1016/j.ijcard.2016.12.078

A. Synetos et al. / International Journal of Cardiology xxx (2016) xxx–xxx

3. Results 3.1. A. Baseline clinical and procedural characteristics Baseline clinical and procedural characteristics of the population are depicted in Supplementary Table 1. In total, 685 subjects (age: 61.4 ± 11.8 years, 83.4% (n = 571) males) were included in the study. The clinical and procedural characteristics, stratified by tertiles of the EuroSCORE II, are summarized in Table 1. 3.2. B. Clinical outcomes The median follow-up period was 556 days (interquartile range: 321–836). Thirty three subjects (4.8%) died from cardiovascular causes, 32 (4.7%) patients had TLF, ST was present in 8 (1.2%) patients, 36 (5.3%) patients required RR and 61 (8.9%) patients had MACE during the follow-up. The event rates of interest are summarized by EuroSCORE II tertiles in Table 2. Patients distributed in the highest EuroSCORE II score presented increased rates of CV death as compared to patients in the two others groups (12.0% vs 1.1% for middle and 1.1% for low tertile respectively, p-value b 0.001). In addition, the incidence of all cause mortality was increased in patients in the upper tertile of EuroSCORE II (20.2% vs 3.5% and 2.2% for middle and low tertile respectively, pvalue b 0.0001). The occurrence of MACE was also increased in high risk group patients according to EuroSCORE II (21.1% in highest tertile versus 2.2% and 3.5% in middle and lower tertile, p-value b 0.0001) (Table 2). Similar results were observed for alternative outcomes of interest, including TLF, ST and RR (p b 0.05 for increased rate of occurrence across the tertiles of EuroSCORE II). When we allocated post hoc our study's population into two subgroups (primary PCI performed during 2008–2011, 408 patients and primary CI performed during 2012–2013, 277 patients), no difference in incidence of major outcomes was established between groups (p N 0.05 for all pairwise comparisons) (Supplementary Table 2). For both subgroups, patients distributed in the highest EuroSCORE II score tertile presented increased rates of CV death, MACE and TLF as compared to patients in lower tertiles (p-value b 0.05 for all three). 3.3. C. Comparison of scores 3.3.1. 1. Discrimination The ROC curves for CVD and MACE are presented in Fig. 1 respectively. The corresponding C-statistics for the EuroSCORE II, euroadditive, EuroSCORE, SX Score, GRC, ACEF and CSS CVD mortality were 0.873, 0.746, 0.745, 0.651, 0.690, 0.778 and 0.741 and for MACE 0.825, 0.645, 0.708, 0.622, 0.659, 0.673 and 0.686 respectively. Pairwise comparisons between C-statistics indicated that the EuroSCORE II classified better event from event-free patients, as compared with all other scores in terms of MACE prediction (Bonferronni corrected p's b 0.05). EuroSCORE II was associated with marginally increased AUC for MACE in comparison to CSS (p = 0.653). In terms of CVD, EUROSCORE II was superior (p b 0.05) to other risk scores or marginally different from Euroadditive score (p = 0.051) and CSS score (p = 0.0593). For ACEF score, ROC analysis did not establish a significant difference (p = 0.1205) with Euroscore II for predicting CVD in our population (Table 3). When time to event was taken into consideration for evaluating the discrimination ability of alternative predictors for CVD and MACE, EuroSCORE II was consistently associated with increased Harrell's cstatistics over alternative competing risk scores (Table 4). For the incidence of MACE in detail, EuroSCORE II offer significant incremental discriminative value over traditional risk factors and ACEF but was only marginally superior from clinical SS (p = 0.07, Table 4). In addition, EuroSCOREII conferred incremental discriminative value for predicting CVD and MACE over ACEF (p = 0.015 and p = 0.001, respectively for CVD and MACE) only in the subgroup of early (2008–2011) enrolled patients and over CSS in both subgroups (p b 0.05 for all).

3

Table 1 Baseline demographic and angiographic characteristics stratified by the tertiles of EuroSCORE II.

Sex (male), n (%) Age (years), mean Hypertension, n (%) Diabetes mellitus, n (%) Dyslipidemia, n (%) Smoking, n (%) Family history for CAD, n (%) Left main, n (%) 1VD, n (%) 2VD, n (%) 3VD, n (%) Culprit vessel LAD, n (%) LCX, n (%) RCA, n (%)

EuroSCORE II LOW (n = 269)

EuroSCORE II MID (n = 183)

EuroSCORE II HIGH (n = 233)

P-value

258 (95.9%) 52.43 (7.73) 127 (47.2%) 43 (16%)

145 (79.2%) 62.29 (7.13) 105 (57.4%) 39 (21.3%)

168 (72.1%) 71.00 (10.63) 132 (56.7%) 60 (25.8%)

b0.001 b0.001 0.044 0.026

166 (61.7%) 180 (66.9%) 90 (33.5%)

106 (57.9%) 121 (66.1%) 46 (25.1%)

156 (67%) 122 (52.3%) 67 (28.8%)

0.475 0.001 0.282

2 (0.9%) 114 (50%) 64 (28.1%) 50 (21.9%)

10 (4.4%) 101 (44.1%) 73 (31.9%) 56 (24.5%)

11 (4.8%) 113 (49.6%) 58 (25.4%) 57 (25%)

0.007 0.497 0.463 0.645

105 (46.1%) 31 (13.6%) 91 (39.9%)

110 (48%) 33 (14.4%) 86 (37.6%)

138 (60.5%) 23 (10.1%) 68 (29.7%)

b0.001

Data are presented as mean ± SD or absolute and relative frequencies. P-values are derived by ANOVA or Pearson's chi-square test. CAD: coronary artery disease; LM: left main coronary artery; VD: vessel disease, LAD: left anterior descending artery; LCX: left circumflex artery; RCA: right coronary artery.

3.3.2. 2. Calibration and reclassification Subsequently, Euroscore II was directly compared to individual traditional risk factors or competing risk scores in logistic regression nested models for CVD. For an indicative model of traditional risk factors not included in prognostic models or for alternative risk scores, incorporation of EuroSCORE II improved both calibration and conferred incremental reclassification value. However, EuroSCORE II did not confer significant reclassification improvement over the CSS for predicting CVD (Table 4). In contrary, EuroSCORE II improved calibration and reclassification over all alternative risk scores and traditional risk factors in logistic regression models of MACE incidence in our study population (Table 4). When stratified analysis per chronological group was performed, EuroSCOREII conferred incremental calibration and reclassification value for predicting CVD and MACE over alternative risk scores (i.e. CSS and ACEF) in both subgroups (p b 0.05 for all).

3.4. D. Survival analysis A survival analysis was performed in order to assess the predictive ability of EuroSCORE II, the score that showed the highest discriminating ability (i.e., higher C-statistic), on MACE and CVD outcomes. Log-rank test indicated that EuroSCORE II predicted both CV death (p b 0.001) and MACE (p b 0.001). The Kaplan Meier curves

Table 2 Clinical outcome of the patients stratified by the tertiles of EuroSCORE ΙΙ.

Total death, n (%) CVD, n (%) TLF, n (%) ST, n (%) RR, n (%) MACE, n (%)

EuroSCORE II LOW (n = 269)

EuroSCORE II MID (n = 183)

EuroSCORE II HIGH (n = 233)

P-value

6 (2.2%) 3 (1.1%) 3 (1.1%) 2 (0.7%) 6 (2.2%) 9 (3.3%)

6 (3.3%) 2 (1.1%) 0 (0.0%) 0 (0%) 2 (1.1%) 4 (2.2%)

47 (20.2%) 28 (12%) 29 (12.45%) 6 (2.6%) 28 (12%) 48 (20.6%)

b0.001 b0.001 b0.001 0.037 b0.001 b0.001

Data are presented as absolute and relative frequencies. P-values are derived by chi-square test. CVD: cardiovascular death; TLF: target lesion failure; ST: stent thrombosis; RR: repeat revascularization; MACE: major adverse cardiac events.

Please cite this article as: A. Synetos, et al., Comparison of prognostic risk scores after successful primary percutaneous coronary intervention, Int J Cardiol (2016), http://dx.doi.org/10.1016/j.ijcard.2016.12.078

4

A. Synetos et al. / International Journal of Cardiology xxx (2016) xxx–xxx Table 3 Corresponding area(s) under ROC curve to various risk scores (euroadditive, EuroSCORE, SYNTAXScore, GRC, ACEF and CSS) for predicting cardiovascular death and MACE. Area under the curve CVD Risk scores

Area

⁎P-value

EuroSCOREII (reference) Euroadditive EuroSCORE SYNTAXScore GRC ACEF CSS

0.8726 0.7457 0.7448 0.6505 0.6896 0.7782 0.7412

0.051 0.036 0.001 0.003 0.121 0.0593

Area under the curve MACE EuroSCOREII (reference) Euroadditive EuroSCORE SYNTAXScore GRC ACEF CSS

0.8249 0.6448 0.7077 0.6223 0.6586 0.6732 0.6864

b0.001 0.015 0.003 b0.001 0.003 0.065

CVD: cardiovascular death; MACE: major adverse cardiac events; GRC: global risk score; ACEF: age-creatinine-ejection fraction score; SS: syntax score. ⁎ P-values correspond to pairwise comparisons of EuroSCORE II to risk alternative scores for predicting CVD.

Fig. 1. Different risk scores and corresponding ROC Curves for predicting A. CVD. B. MACE. CVD: cardiovascular death, MACE: major adverse cardiac events.

for CVD, stratified according to the tertiles of the EuroSCORE II, are shown in Fig. 2A and for MACE are shown in Fig. 2B. The results of the un-adjusted and the multi-adjusted Cox regression analysis are shown in Tables 5 and 6. Un-adjusted analysis showed that EuroSCORE II, renal dysfunction, LVEF and cholesterol were predictors of CVD and MACE; whereas, multi-adjusted analysis, controlling for the above confounders revealed that EuroSCORE II was an independent predictor of our study's main outcomes during follow-up. Additional adjustment for variables of biological interest (i.e. age, gender, DM) already incorporated in EuroSCORE II did not alter our results (data not shown). Relevant results for the independent predictive role of EuroSCOREII were yielded when a stratified survival analysis was performed on the basis of the time of enrolment of the patients of our study (p b 0.05 for both groups). 4. Discussion The main findings of the current study are that in patients with successful pPCI, 1) EuroSCORE II was an independent predictor of CVD and MACE after adjusting for plausible confounders and 2). EuroSCORE II predicts better—compared to CSS, GRS, SxScore, ACEF and EuroSCORE—the incidence of CVD and/or MACE. The prognosis of patients after pPCI is affected by coronary angiographic characteristics such as the culprit lesion location, the TIMI flow at presentation and the presence of multivessel disease and/or chronic total occlusion [10–12]. Moreover, some clinical factors such as age, renal dysfunction, diabetes mellitus [12] and LVEF have been

recognized as independent predictors of MACE in the same population [27,28]. Regarding patients with successful pPCI recent data have shown that combined clinical and angiographic scoring systems improve the prognostic accuracy of clinical or angiographic stand-alone tools and that CSS was the best model for the prediction of the incidence of CVD and/or MACE after successful pPCI for the treatment of STEMI [13]. These results were obtained in an era when EuroSCORE II that is known to substantially improve the calibration of the older models and at least maintain and perhaps further improve the discriminatory power of the previous models was not available. Our study for the first time validated EuroSCORE II in patients with STEMI that underwent successful pPCI and compared it with the combined (clinical and angiographic) CSS and GRS. EuroSCORE II had the best predictive value and was an independent predictor for both CVD and MACE in this population. This can be partially explained by the fact that although EuroSCORE II is a pure clinical score that lacks any angiographic variables incorporates the most important clinical variables. EuroSCORE II includes ten patient-related risk factors. Nine of them remained the same as in EuroSCORE (with changed definitions for some of them) and diabetes on insulin therapy was added. The accuracy of EuroSCORE II compared to that of EuroSCORE has been tested in patients undergoing coronary artery bypass grafting or other cardiac surgery with conflicting results [29]. In patients undergoing transcatheter aortic valve replacement, EuroSCORE II had better accuracy compared to EuroSCORE, however none of the risk models was optimal in predicting 30-day mortality in unselected, real-life population with aortic stenosis [30]. The data provided in these studies are not comparable to our population as we included only patients with successful pPCI. However the inclusion criteria of the current study justify the results regarding the outcome of MACE and of stent thrombosis that are comparable to previous published studies. The clinical characteristics of the studied population and primary outcomes of the present study were similar to studies previously published. The data provided in other studies are not comparable to our population, as we included only patients with successful pPCI. However, the inclusion criteria of the current study justify the better results regarding the outcome of MACE and of ST. In other studies the range for CV mortality and MACE was 13–36% and 8.9–11.9%, respectively, versus 5.5 and 8.9% of the studied population [12,16,27,28,31,32]. The small study sample and the absence of data concerning the door-to-balloon and pain-to-balloon time are a limitation of the study. The recording of the LVEF after the PCI which is correlated with the

Please cite this article as: A. Synetos, et al., Comparison of prognostic risk scores after successful primary percutaneous coronary intervention, Int J Cardiol (2016), http://dx.doi.org/10.1016/j.ijcard.2016.12.078

A. Synetos et al. / International Journal of Cardiology xxx (2016) xxx–xxx

5

Table 4 Improvement in model calibration and reclassification of EuroSCORE II over established risk scores and the best predictive model of traditional risk factors for study main endpoints. Regression parameters

Discrimination parameters

Reclassification parameters Continuous NRI

HR (95%CI)

P-value

IDI

Harrell's c

P-value

Among event subjects

Among non-event subjects

Overall (SE)

P-value

Coefficient (SE)

Primary endpoint (cardiovascular mortality, n = 33) Model1 Model1 + EUROSCOREII 1.11 (1.06–1.15) b0.001 ACEF 7.04 (4.83–10.3) b0.001 ACEF + EUROSCOREII 1.07 (1.04–1.1) b0.001 CSS 1.02 (1.01–1.02) b0.001 CSS + EUROSCOREII 1.11 (1.07–1.15) b0.001

0.838 (0.758–0.917) 0.89 (0.832–0.947) 0.811 (0.721–0.9) 0.886 (0.818–0.954) 0.771 (0.691–0.852) 0.908 (0.845–0.97)

0.03

51.52%

26.16%

77.7% (17.8)

b0.001

⁎14.4 (4.4)

0.002

45.46%

38.18%

83.6% (17.8)

b0.001

⁎⁎10.3 (3.9)

10.52%

7.46%

17.98% (10.6)

0.359

4.4 (3.9)

Major adverse cardiac events (n = 61) Model1 Model1 + EUROSCOREII 1.11 (1.07–1.14) ACEF 4.89 (3.54–6.75) ACEF + EUROSCOREII 1.07 (1.05–1.1) CSS 1.02 (1.01–1.02) CSS + EUROSCOREII 1.11 (1.08–1.14)

0.739 (0.672–0.806) 0.817 (0.754–0.88) 0.689 (0.61–0.768) 0.814 (0.75–0.877) 0.694 (0.622–0.765) 0.796 (0.715–0.877)

0.008

31.14%

63.72%

94.9% (13.4)

b0.001

⁎22.1(3.7)

b0.001

37.7%

76.6%

114% (14.4)

b0.001

⁎15.7 (3.3)

0.07

34.42%

72.76%

107% (13.4)

b0.001

⁎21.6(3.7)

b0.001 b0.001 b0.001 b0.001 b0.001

0.005

Model 1: age, gender, ejection fraction, diabetes mellitus, dyslipidemia, smoking. HR: hazard ratio; NRI: Net Reclassification Index; SE: Standard Error; IDI: integrated discrimination index; ACEF: Age-Creatinine-Ejection Fraction Score; CSS: clinical Syntax Score. ⁎ Indicates level of statistical significance b 0.001. ⁎⁎ Indicates level of statistical significance b 0.05.

time interval from symptoms to intervention, overcomes the absence of data regarding the time delay of the patients. The current study demonstrated that the use of EuroSCORE II provides the highest predictive value for cardiovascular mortality and MACE. In patients with STEMI undergoing successful pPCI, EuroSCORE II improves the prognostic accuracy of combined, clinical or angiographic stand-alone tools. The use of this score for risk stratification after the successful pPCI may improve our treatment strategies in patients with high EuroSCORE II. An aggressive treatment of this population

combined with a closer clinical follow-up of these patients needs to be further evaluated in prospective clinical studies. Conflict(s) of interest None. Funding This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.ijcard.2016.12.078.

Table 5 Unadjusted and multi-adjusted predictors of CVD. Unadjusted analysis for CVD

Cholesterol (mg/dl) Creatinine (mg/dl) EF (%) EuroSCORE II (%)

Multi-adjusted⁎ analysis for CVD

HR

95% CI

P-value HR

0.464 40.9 0.82 1.1

0.231–0.934 17.7–95 0.785–0.864 1.095–1.14

0.031 b0.001 b0.001 b0.001

95% CI

P-value

– 6.54 2.25–19 0.001 0.9 0.846–0.956 0.001 1.061 1.028–1.095 b0.001

EF: ejection fraction; CVD: cardiovascular death; HR: hazard ratio; CI: confidence intervals.

Table 6 Unadjusted and multi-adjusted predictors of MACE. Unadjusted analysis for MACE

Cholesterol (mg/dl) Creatinine (mg/dl) EF (%) EuroSCORE II (%) Fig. 2. Kaplan–Meier Curve for A. CVD. B. MACE according to EuroSCORE II tertiles. CVD: cardiovascular death, MACE: major adverse cardiac events.

Multi-adjusted⁎ analysis for MACE

HR

95% CI

P-value HR

0.594 23.6 0.877 1.1

0.358–0.985 11.3–49.3 0.847–0.909 1.09–1.12

0.045 b0.001 b0.001 b0.001

95% CI

P-value

– 5.11 2.1–12.4 b0.001 0.943 0.904–0.984 0.007 1.07 1.043–1.1 b0.001

EF: ejection fraction; MACE: major adverse cardiac events; HR: hazard ratio; CI: confidence intervals.

Please cite this article as: A. Synetos, et al., Comparison of prognostic risk scores after successful primary percutaneous coronary intervention, Int J Cardiol (2016), http://dx.doi.org/10.1016/j.ijcard.2016.12.078

6

A. Synetos et al. / International Journal of Cardiology xxx (2016) xxx–xxx

References [1] G. Sianos, M.A. Morel, A.P. Kappetein, M.C. Morice, A. Colombo, K. Dawkins, et al., The SYNTAX score: an angiographic tool grading the complexity of coronary artery disease, EuroIntervention 1 (2005) 219–227. [2] K.D. Dawkins, M.A. Morel, P.W. Serruys, Counting the score: the SYNTAX score and coronary risk, EuroIntervention 5 (2009) 33–35. [3] D. Capodanno, M. Miano, G. Cincotta, A. Caggegi, C. Ruperto, R. Bucalo, et al., EuroSCORE refines the predictive ability of SYNTAX score in patients undergoing left main percutaneous coronary intervention, Am. Heart J. 159 (2010) 103–109. [4] S. Garg, G. Sarno, H.M. Garcia-Garcia, C. Girasis, J. Wykrzykowska, K.D. Dawkins, et al., A new tool for the risk stratification of patients with complex coronary artery disease: the clinical SYNTAX score, Circ. Cardiovasc. Interv. 3 (2010) 317–326. [5] D. Capodanno, A. Caggegi, M. Miano, G. Cincotta, F. Dipasqua, G. Giacchi, et al., Global risk classification and clinical SYNTAX (synergy between percutaneous coronary intervention with TAXUS and cardiac surgery) score in patients undergoing percutaneous or surgical left main revascularization, JACC Cardiovasc. Interv. 4 (2011) 287–297. [6] S. Addala, C.L. Grines, S.R. Dixon, G.W. Stone, J.A. Boura, A.B. Ochoa, et al., Predicting mortality in patients with ST-elevation myocardial infarction treated with primary percutaneous coronary intervention (PAMI risk score), Am. J. Cardiol. 93 (2004) 629–632. [7] A. Halkin, M. Singh, E. Nikolsky, C.L. Grines, J.E. Tcheng, E. Garcia, et al., Prediction of mortality after primary percutaneous coronary intervention for acute myocardial infarction: the CADILLAC risk score, J. Am. Coll. Cardiol. 45 (2005) 1397–1405. [8] D.A. Morrow, E.M. Antman, A. Charlesworth, R. Cairns, S.A. Murphy, J.A. de Lemos, et al., TIMI risk score for ST-elevation myocardial infarction: a convenient, bedside, clinical score for risk assessment at presentation: an intravenous nPA for treatment of infarcting myocardium early II trial substudy, Circulation 102 (2000) 2031–2037. [9] G. De Luca, H. Suryapranata, A.W. van‘'t Hof, M.J. de Boer, J.C. Hoorntje, J.H. Dambrink, et al., Prognostic assessment of patients with acute myocardial infarction treated with primary angioplasty: implications for early discharge, Circulation 109 (2004) 2737–2743. [10] P. Sorajja, B.J. Gersh, D.A. Cox, M.G. McLaughlin, P. Zimetbaum, C. Costantini, et al., Impact of multivessel disease on reperfusion success and clinical outcomes in patients undergoing primary percutaneous coronary intervention for acute myocardial infarction, Eur. Heart J. 28 (2007) 1709–1716. [11] R.J. van der Schaaf, M.M. Vis, K.D. Sjauw, K.T. Koch, J. Baan Jr., J.G. Tijssen, et al., Impact of multivessel coronary disease on long-term mortality in patients with STelevation myocardial infarction is due to the presence of a chronic total occlusion, Am. J. Cardiol. 98 (2006) 1165–1169. [12] S. Garg, G. Sarno, P.W. Serruys, A.E. Rodriguez, L. Bolognese, M. Anselmi, et al., Prediction of 1-year clinical outcomes using the SYNTAX score in patients with acute ST-segment elevation myocardial infarction undergoing primary percutaneous coronary intervention: a substudy of the STRATEGY (single high-dose bolus Tirofiban and sirolimus-eluting stent versus Abciximab and bare-metal stent in acute myocardial infarction) and MULTISTRATEGY (multicenter evaluation of single high-dose bolus Tirofiban versus Abciximab with sirolimus-eluting stent or bare-metal stent in acute myocardial infarction study) trials, JACC Cardiovasc. Interv. 4 (2011) 66–75. [13] K. Toutouzas, A. Synetos, A. Karanasos, C. Nikolaou, A. Michelongona, D. Panagiotakos, et al., Prognostic models for cardiovascular events after successful primary percutaneous coronary intervention, Int. J. Cardiol. 158 (2012) 168–170. [14] S.C. Smith Jr., J.T. Dove, A.K. Jacobs, J.W. Kennedy, D. Kereiakes, M.J. Kern, et al., ACC/ AHA guidelines for percutaneous coronary intervention (revision of the 1993 PTCA guidelines)-executive summary: a report of the American College of Cardiology/ American Heart Association task force on practice guidelines (committee to revise the 1993 guidelines for percutaneous transluminal coronary angioplasty) endorsed by the Society for Cardiac Angiography and Interventions, Circulation 103 (2001) 3019–3041. [15] S. Hamada, T. Nishiue, S. Nakamura, T. Sugiura, H. Kamihata, H. Miyoshi, et al., TIMI frame count immediately after primary coronary angioplasty as a predictor of functional recovery in patients with TIMI 3 reperfused acute myocardial infarction, J. Am. Coll. Cardiol. 38 (2001) 666–671.

[16] M. Magro, S. Nauta, C. Simsek, Y. Onuma, S. Garg, E. van der Heide, et al., Value of the SYNTAX score in patients treated by primary percutaneous coronary intervention for acute ST-elevation myocardial infarction: the MI SYNTAXscore study, Am. Heart J. 161 (2011) 771–781. [17] M. Singh, C.S. Rihal, R.J. Lennon, K.N. Garratt, D.R. Holmes Jr., A critical appraisal of current models of risk stratification for percutaneous coronary interventions, Am. Heart J. 149 (2005) 753–760. [18] S.A. Nashef, F. Roques, L.D. Sharples, J. Nilsson, C. Smith, A.R. Goldstone, et al., EuroSCORE II, Eur. J. Cardiothorac. Surg. 41 (2012) 734–744 (discussion 44-5). [19] M. Ranucci, S. Castelvecchio, L. Menicanti, A. Frigiola, G. Pelissero, Risk of assessing mortality risk in elective cardiac operations: age, creatinine, ejection fraction, and the law of parsimony, Circulation 119 (2009) 3053–3061. [20] E. Nikolsky, A.J. Lansky, K. Sudhir, J. Doostzadeh, D.E. Cutlip, R. Piana, et al., SPIRIT IV trial design: a large-scale randomized comparison of everolimus-eluting stents and paclitaxel-eluting stents in patients with coronary artery disease, Am. Heart J. 158 (2009) 520–526, e2. [21] D.E. Cutlip, S. Windecker, R. Mehran, A. Boam, D.J. Cohen, G.A. van Es, et al., Clinical end points in coronary stent trials: a case for standardized definitions, Circulation 115 (2007) 2344–2351. [22] G. Montalescot, U. Zeymer, J. Silvain, B. Boulanger, M. Cohen, P. Goldstein, et al., Intravenous enoxaparin or unfractionated heparin in primary percutaneous coronary intervention for ST-elevation myocardial infarction: the international randomised open-label ATOLL trial, Lancet 378 (2011) 693–703. [23] G. Goliasch, M.E. Kleber, B. Richter, M. Plischke, M. Hoke, A. Haschemi, et al., Routinely available biomarkers improve prediction of long-term mortality in stable coronary artery disease: the Vienna and Ludwigshafen Coronary Artery Disease (VILCAD) risk score, Eur. Heart J. 33 (2012) 2282–2289. [24] N.R. Cook, Use and misuse of the receiver operating characteristic curve in risk prediction, Circulation 115 (2007) 928–935. [25] D.G. Altman, P. Royston, What do we mean by validating a prognostic model? Stat. Med. 19 (2000) 453–473. [26] F.E. Harrell Jr., K.L. Lee, D.B. Mark, Multivariable prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors, Stat. Med. 15 (1996) 361–387. [27] S. Pedersen, S. Galatius, R. Mogelvang, S.Z. Abildstrom, R. Sorensen, A. Galloe, et al., Long-term clinical outcome in STEMI patients treated with primary PCI and drugeluting or bare-metal stents: insights from a high-volume single-center registry, J. Invasive Cardiol. 23 (2011) 328–333. [28] C. Simsek, M. Magro, E. Boersma, Y. Onuma, S. Nauta, J. Daemen, et al., Comparison of six-year clinical outcome of sirolimus- and paclitaxel-eluting stents to bare-metal stents in patients with ST-segment elevation myocardial infarction: an analysis of the RESEARCH (rapamycin-eluting stent evaluated at Rotterdam cardiology hospital) and T-SEARCH (taxus stent evaluated at Rotterdam cardiology hospital) registries, J. Invasive Cardiol. 23 (2011) 336–341. [29] K. Janikowski, R. Morawiec, B. Jegier, R. Jaszewski, M. Lelonek, EuroSCORE II does not show better accuracy nor predictive power in comparison to original EuroSCORE: a single Centre study, Kardiol. Pol. (2015). [30] K. Zbronski, Z. Huczek, D. Puchta, K. Paczwa, J. Kochman, R. Wilimski, et al., Outcome prediction following transcatheter aortic valve implantation: multiple risk scores comparison, Cardiol. J. (2015). [31] M.T. Dirksen, M.A. Vink, M.J. Suttorp, J.G. Tijssen, M.S. Patterson, T. Slagboom, et al., Two year follow-up after primary PCI with a paclitaxel-eluting stent versus a baremetal stent for acute ST-elevation myocardial infarction (the PASSION trial): a follow-up study, EuroIntervention 4 (2008) 64–70. [32] M.A. Vink, M.T. Dirksen, M.J. Suttorp, J.G. Tijssen, J. van Etten, M.S. Patterson, et al., 5year follow-up after primary percutaneous coronary intervention with a paclitaxeleluting stent versus a bare-metal stent in acute ST-segment elevation myocardial infarction: a follow-up study of the PASSION (paclitaxel-eluting versus conventional stent in myocardial infarction with ST-segment elevation) trial, JACC Cardiovasc. Interv. 4 (2011) 24–29.

Please cite this article as: A. Synetos, et al., Comparison of prognostic risk scores after successful primary percutaneous coronary intervention, Int J Cardiol (2016), http://dx.doi.org/10.1016/j.ijcard.2016.12.078