| Background The prevalence of acute myocardial infarction (AMI) in elder patient is elevating as the social aging. According to the GRACE and CRUSADE study, one third of the acute coronary syndrome (ACS) patients are the elder patients above 75 years old. Reperfusion therapy could markedly reduce the mortality of ST-segment elevation myocardial infarction (STEMI) and improve the clinical prognosis. The safety and efficacy of emergent percutaneous coronary intervention (PCI) are gradually increased as the devices updating and technology maturing. However, there is a little data from large scale randomized control trials revealing the clinical prognosis in elder STEMI patients. Thus it is a difficulty for doctors to choose the intervention therapy for elder patients. Multiple risk prediction models could qualify the risk of cardiovascular events and analyze the prognosis. Therefore, according to the models, doctors could screen the high risk patients, choose the appropriate therapeutic strategy and strengthen the management of outpatient. As a result, it will reduce the incidence of adverse events even further.Objective In the present study, we aimed to evaluate the clinical value of the multiple risk prediction models, including the SYNTAX (Synergy between Percutaneous Coronary Intervention with Taxus and Cardiac Surgery) score, GRACE (Global Registry of Acute Coronary Events)score, ACEF (age, creatinine, ejection fraction) score, EuroScore (European System for Cardiac Operative Risk Evaluation), GRC (global risk classification) score and CSS (clinical SYNTAX) score in elderly STEMI patients referring to emergent PCI.Methods It was a prospective study. We calculated the SYNTAX score, GRACE score, ACEF score, EuroScore, clinical SYNTAX score, GRC score in a consecutive series of 104 patients≥75 years presenting with STEMI and treated with emergent PCI from Nov. 2011 to Jan.2014. Patients were stratified into two groups (lower/mid risk group and high risk group) according to those models and analyzed retrospectively the differences between the groups in baseline and endpoint. The primary endpoint was 30-day and 1 year all-cause mortality. Survive curve, log-rank test and univariate COX regression model were used to analyzed the regressive relationship between the 2 groups in different score models. We used Receiver Operating Characteristic (ROC) to assess the predict power of 30-day and 1 year all-cause mortality in different models.Results1.104 patients, including 65 male patients and 39 female patients, were consecutively enrolled in the study. The mean age was 78.7±3.6.The average BMI was 24±3.9kg/m2. According to the baseline data, there were 41 patients diagnosed as the anterior MI (39.42%),13 lateral MI (12.5%) and 53 inferior and posterior MI (48.1%). The average left ventricular ejection fraction was 53.2±9.8%. 2. The different scores using different models are as follows:ACEF was 1.56± 0.46, Euroscore was 10.0±1.57, GRACE was 146.1±18.3, SYNTAX was 18.5±8.7, CSS was 30.0±22.0. In the ACEF model, low/mid risk group contained 18 patients, whereas the high risk group got 86 patients, in which were more elder, female, free of history of cerebral infarction and low LVEF patients. Short-term death of GRACE score showed that low/mid risk group contained 72 patients, whereas the high risk group got 32 patients, in which were more elder, free of history of hypertension, with the history of cardiac shock and low LVEF patients. However, the long-term death of GRACE score revealed that low/mid risk group contained 46 patients, whereas the high risk group got 58 patients, in which were more elder, free of history of hypertension and low LVEF patients. Almost all the patients using the EuroScore model were in high risk group. The SYNTAX score of 104 patients was 18.5±8.7 according to the results of their coronary angiography. The SYNTAX score showed that low/mid risk group contained 97 patients, whereas the high risk group got 7 patients, in which were more cardiac shock, left main and 3-vessel lesions, and low LVEF patients. The GRC score showed that low/mid risk group contained 97 patients, whereas the high risk group got 7 patients, in which were more cardiac shock and low LVEF patients. The CSS score showed that low/mid risk group contained 53 patients, whereas the high risk group got 51 patients, in which were more high BMI and low LVEF patients.3. All the models were evaluated by the ROC test as the sensitive analysis for estimating the short and long term prognosis. The AUC values of each models were as follows:SYNTAX 0.503, ACEF 0.723, EuroScore 0.832, GRACE 0.858, CSS 0.555. The AUC values of 30-day morbidity in those models were 0.646,0.669, 0.786,0.882 and 0.675 respectively; and the values of 1 year morbidity were 0.661, 0.68,0.811,0.788 and 0.693 separately.4 In the present study,5 patients died in hospital (4.8%),7 died in 30-day post PCI, and 10 died in 1 year. The COX regression analysis of thirty-day mortality in SYNTAX score showed that more patients died in the high risk group than in the low/mid risk group (42.9% vs.4.1%, log-rank test p<0.0001). It was the same situation in the short-term GRACE model (18.7% vs.1.4%, log-rank test p=0.0011). The COX regression analysis of 1-year mortality in SYNTAX score showed the similar trend that more patients died in the high risk group than in the low/mid risk group (42.9% vs.7.2%, log-rank test p=0.0008). It was the same situation in the long-term GRACE model (21.9% vs.4.2%, log-rank test p=0.0039). There was no significant difference in the other models.5. The univariate COX regression analysis of SYNTAX model showed that the 30-day and 1 year morbidity in the high risk group was much higher than in the low/mid risk group (the 30-day hazard ratio [HR]=11.199,95% confidence interval [CI]:2.502-50.134, p=0.0016; and lyear HR=7.221,95% CI:1.855-28.099, p=0.0044). It was the same in the short-term GRACE model (in hospitalization HR=10.43,95% CI:1.086-94.573, p=0.0421; 30-day HR=14.288,95% CI: 1.712-118.211,p=0.014; and 1 year HR=5.75,95% CI:1.486-22.256, p=0.0133). There was no significant difference in ACEF, CSS and GRC models.Conclusion1.The thirty-day and one-year mortality in elder patients (≥=75 years) received primary PCI due to STEMI were higher than in all-age patients. 2.The SYNTAX score and GRACE score are more valuable in predicting the thirty-day and one-year mortality in elder STEMI patients (≥75 years) underwent primary PCI.3.There was less predictive value of ACEF score, EuroScore and CSS score in elder STEMI patients (≥75 years) received primary PCI. 4.Short-term GRACE score had more predictive value in screening the high risk patients and predicting the thirty-day and one-year mortality in elder STEMI patients.5.The GRC and CSS score systems had less predictive value in risk classification and prognosis prediction.6. We could try to establish the suitable risk models for elder STEMI patients (≥75 years) underwent emergent PCI. |
PDF Full Text Request