Comparison of Anatomical and Clinical Scores in Predicting Outcomes in Primary Percutaneous Coronary Intervention

Mailing Address: Fernando Pivatto Júnior Rua Ramiro Barcelos, 2.350, sala 700, PostaL Code 90.035-903, Porto Alegre, RS Brazil. E-mail: fpivatto@gmail.com Comparison of Anatomical and Clinical Scores in Predicting Outcomes in Primary Percutaneous Coronary Intervention Fernando Pivatto Júnior, Gustavo Neves de Araújo, Felipe Homem Valle, Luiz Carlos Corsetti Bergoli, Guilherme Pinheiro Machado, Bruno Führ, Elvis Pellin Cassol, Ana Maria Rocha Krepsky, Rodrigo Vugman Wainstein, Marco Vugman Wainstein


Introduction
Different scores based on anatomical and/or clinical features have been developed for risk stratification of patients undergoing percutaneous coronary intervention (PCI).However, studies comparing the ability of these different models to predict cardiac events in patients submitted to primary PCI are limited.
the SYNTAX 2 trial with the objective of characterizing and objectively quantifying the severity and extent of coronary artery disease. 3SS is broadly used to stratify the outcomes of elective PCI in left main and multivessel coronary intervention when compared with coronary artery bypass grafting (CABG).[6][7][8][9][10][11][12] Pivatto Júnior et al.

Original Article
Age, creatinine and ejection fraction (ACEF) 13 score appears to be equivalent to more complex scores in predicting mortality in patients undergoing elective CABG.5][16] Modified ACEF score (ACEF Mod ) considers creatinine clearance (CrCl) as a semicontinuous variable, representing a better estimate of the underlying renal function compared with serum creatinine.This modification improves the predictive accuracy of ACEF in patients undergoing PCI. 17 Combining clinical and anatomical variables in the same score provides a better performance in risk stratification. 3he Clinical SYNTAX score (CSS) incorporates ACEF Mod to SS and is able to predict major adverse cardiac and cerebrovascular events (MACCE) in patients with complex coronary artery disease. 15However, limited number of studies have evaluated the role of this score in patients undergoing PCI. 11,12,15sk stratification is a relevant issue in patients undergoing PCI after MI. 16 The aim of this study was to compare SS, CSS, ACEF, and ACEF Mod in predicting MACCE in patients with ST-elevation MI (STEMI) undergoing primary PCI.

Methods
This cohort study included consecutive patients with STEMI undergoing primary PCI between April/2011 and December/2015 in a tertiary university hospital in southern Brazil.STEMI was defined as a typical chest pain at rest associated with ST-segment elevation of at least 1 mm in two contiguous leads in the frontal plane or 2 mm in the horizontal plane, or typical pain at rest in patients with a new, or presumably new, left bundle-branch block.Exclusion criteria were previous CABG (excluded in the SYNTAX trial 2 ), absence of admission laboratory testing or echocardiogram, and lack of 30-day follow-up.The study was approved by the institution's Research and Ethics Committee and informed consent was obtained from all patients.
All patients were pretreated with a loading dose of acetylsalicylic acid (300 mg) and clopidogrel (600 mg).Unfractioned heparin was used during the procedure (70-100 IU/kg).Use of IIb/IIIa glycoprotein, aspirative thrombectomy, and PCI technical strategies (i.e., predilation, direct stent placement, postdilation) were performed according to the operator's choice.Coronary flow before and after the procedure was assessed and described according to the Thrombolysis in Myocardial Infarction (TIMI) criteria. 18Anticoagulants were suspended after the end of the procedure, and double antiplatelet therapy was recommended for 12 months after the event.
SS was derived from the sum of individual scores for each separate lesion (defined as > 50% stenosis in vessels > 1.5 mm).Full details on the SS calculation are reported elsewhere. 1ACEF was computed as follows: (age/left ventricle ejection fraction) + 1 if serum creatinine value was > 2 mg/dL. 13In ACEF Mod , 1 point was added for every 10 mL/min reduction in CrCl < 60 mL/min/1.73m 2 (up to a maximum of 6 points). 15Therefore, a CrCl between 50 to 59 mL/min/1.73m 2 , 40 to 49 mL/min/1.73m 2 , and 30 to 39 mL/min/1.73m 2 would receive 1, 2, and 3 points, respectively.CSS was calculated retrospectively for each patient using the following formula: CSS = SS x ACEF Mod . 15We determined cutoff values for the scores above to define them as low or high risk.These cutoff values were obtained by multiplying sensitivity and specificity of each value within the receiver operator characteristic (ROC) curve of the different scores; the value with the highest product (sensitivity X specificity) was established as the cutoff point.
Blood samples were collected by venipuncture before the procedure, as part of routine patient care.CrCl was estimated according to the Modification of Diet in Renal Disease (MDRD) equation.Left ventricular ejection fraction (LVEF) was determined before patient discharge using transthoracic echocardiography and applying either Simpson (in the presence of segmental dysfunction) or Teicholz method.Clinical follow-up was performed with either outpatient visit or telephone contact.MACCEs were defined as death from all cause, new MI, stroke, Canadian Cardiovascular Society (CCS) class III/IV angina, or rehospitalization for congestive heart failure 30 days after the primary PCI.New MI was defined as recurrent chest pain with ST-segment elevation or new Q waves and increase in serum biomarkers after their initial decrease.Stroke was defined as a new, sudden-onset focal neurological deficit of presumably cerebrovascular cause, irreversible (or resulting in death), and not caused by other readily identifiable causes.

Statistical analysis
Continuous variables are expressed as mean (± standard deviation) or median (interquartile range).

Results
We included 311 (78.3%) of the 397 patients who underwent primary PCI for STEMI in the analyzed period.Mean age was 60.2 ± 12.0 years, 35.4% were women, and 22.5% had diabetes.LVEF was < 40% in 18.3%, and estimated CrCl was < 60 mL/min/1.73m 2 in 21.9% of the patients.Complete demographic data are described in Table 1.
Complete procedure-related data are shown in Table 2.The incidence of MACCE at 30 days was 23.8%, as detailed in Table 3.
ROC curves are presented in Figure 1.All curves were statistically significant, and the CSS curve had the largest area under the curve (AUC): CSS > ACEF Mod > SS > ACEF.However, when the AUCs were compared two-by-two with DeLong test, there was no statistically significant differences, except in the comparison of ACEF versus CSS (p = 0.02) (Figure 2).Univariate analysis of MACCE according to high or low risk score values (cutoff point determination previously described in the Methods section) showed that high-risk CCS, SS, ACEF, and ACEF Mod were significantly associated with higher MACCE rates (p < 0.001, p < 0.001, p < 0.002, and p < 0.040, respectively).Other clinical variables associated with MACCE in univariate analysis were age > 65 years (p = 0.007), female sex (p = 0.041), Killip 3 or 4 (p < 0.001), and postprocedural TIMI 0-2 (p = 0.006).When adjusted by these variables, only SS and CSS remained independent predictors of MACCE (Table 4).

Discussion
We assessed in the present study the ability of SS, CSS, ACEF, and ACEF Mod in predicting MACCE in STEMI patients undergoing primary PCI.Our data showed that CSS had the largest AUC; however, when compared two-by-two, the AUC for CSS was only statistically larger than that for ACEF.When we divided the scores between low and high risk, high-risk SS and CSS emerged as independent MACCE predictors; high-risk ACEF and ACEF Mod were predictors of MACCE in univariate analysis, but this association was lost after adjustment for clinical variables.
The prognostic value of the ACEF score in patients who underwent PCI after acute MI was assessed by Lee et al., 16 who analyzed 12,000 patients in this setting.The ACEF was significantly higher in nonsurvivors (1.95 ± 0.82 versus 1.28 ± 0.50, p < 0.001) and was an independent predictor of 1-year mortality (HR, p < 0.001).Capodanno et al. 17    PCI.In our analysis, despite having a larger area under the ROC curve compared with ACEF and SS, ACEF Mod was not an independent predictor MACCE.
6][7][8][9][10][11][12] In a study including 807 patients with STEMI, Garg et al. 5 identified SS as an independent predictor of mortality, MACCE, and stent thrombosis up to a 1-year follow-up.However, it is important to highlight that the study was not performed aiming to define cutoff points for the analyzed scores to predict MACCE, but only to define the relationship of SS with MACCE occurrence.The same author showed an improvement in the ability of the SS to predict MACCE and mortality in patients undergoing PCI by combining SS and ACEF Mod , (CSS). 15This improvement was also observed in the present study.
The use of CSS in patients with STEMI was evaluated in two studies, 11,12 which showed an improved outcome prediction accuracy compared with SS.Cetinkal et al. 11 recently evaluated 433 patients in this setting with the objective of validating CSS as a predictor of prognosis, and also evaluated SS and ACEF.The primary endpoint was a composite of all-cause mortality, MI, and cerebrovascular events, with a follow-up of 15 months.CSS > 26 was identified as an independent predictor of events.The AUC was 0.66 (p < 0.001), 0.59 (p = 0.01), and 0.64 (p < 0.001) for CSS, SS, and ACEF, respectively.However, all cases were performed by femoral access and there was an extremely low mortality in patients with low/moderate CSS (one death over 285 patients in a 15-month follow-up), which jeopardizes the external validity of the study.
Girasis et al. 12 analyzed 848 patients undergoing PCI with drug-eluting stents (only 25.3% were patients with STEMI) and demonstrated that both SS and CSS were able to stratify risk of very long-term adverse clinical outcomes.The AUC for the incidence of MACCE was 0.61 (95% CI: 0.56-0.65)and 0.62 (95% CI: 0.57-0.67),There are some limitations in our study.First, the retrospective design may have influenced the quality and consistency of the collected data.Second, the relatively small number of patients may have reduced the power of the study to detect some associations.Third, the fact that the study was conducted at a single center may also be considered a limitation.

Conclusion
SS and CSS were independent MACCE predictors in this study.In our cohort of primary PCI in patients with STEMI, pure anatomical SS calculated at baseline coronary angiography was a useful tool in predicting short-term MACCE.
have demonstrated that including CrCl (calculated either by MDRD or Cockcroft-Gault) in ACEF yields superior calibration compared with the original serum creatinine-based equation, and improves the predictive accuracy of ACEF in patients undergoing

Table 1 -Demographic data
Categorical variables are represented by relative and absolute frequencies.ROC curves were used to evaluate the discriminatory power of the different scores.Comparison of ROC curves was performed by DeLong test using the software R, version 3.1.2.Patients groups were compared using independent samples Student's t test (for normally distributed variable) or Mann-Whitney U test (for other variables) for continuous variables and χ 2 test or Fisher's exact tests for categorical variables.Multivariate analysis was performed by multiple logistic regression.P-values were considered significant at < 0.05.Data were analyzed using Statistical Package for the Social Sciences (SPSS), version 18.0.