GRACE, TIMI, Zwolle and CADILLAC risk scores — Do they predict 5-year outcomes after ST-elevation myocardial infarction treated invasively?

https://doi.org/10.1016/j.ijcard.2009.10.026Get rights and content

Abstract

Background

GRACE, TIMI, Zwolle, and CADILLAC are risk scores designed for predicting short-term outcome after acute coronary syndromes. The aim of our study was to test their utility for a prognosis of 5-year survival in a “real-life” population of patients with ST-elevation myocardial infarction (STEMI) treated with primary percutaneous coronary interventions (pPCI).

Methods

Our registry consisted of consecutive patients with STEMI treated with pPCI. Five-year follow-up was performed with all-cause mortality as the end-point.

Results

Out of 505 patients (mean age 58.6 ± 11.3 years) 32 died during the first 30 days (6.3%) and an additional 74 within 5 years (15.6%). PCI was successful in 95.2% (n = 481). Prognostic values (c statistics) for predicting 5-year mortality equaled: 0.742 (CI 0.69–0.79) for the GRACE risk score, 0.727 (CI 0.67–0.78) for TIMI, 0.72 (CI 0.67–0.77) for Zwolle, and 0.687 (CI 0.63–0.74) for CADILLAC. In a univariate analysis all the scores were associated with the 5-year outcome.

Conclusions

GRACE, TIMI, and Zwolle risk scores predicted well 5-year all-cause mortality in patients with STEMI treated with pPCI. Our data show that the usefulness of initial bedside risk assessment can be further extended for long-term follow-up.

Introduction

There are few reports concerning 5-year outcomes in patients with ST-elevation myocardial infarction (STEMI) treated with primary percutaneous coronary interventions (pPCI). In the randomized study PRAGUE-2, the overall mortality rate was 19%, including 6.8% within the first 30 days and 14.9% during the next 5 years among 30-day survivors [1]. In the Zwolle trial, 22% of patients died within 8 years [2]. Next, the Italian single-center experience with a mean follow-up of 51 months showed the overall mortality of 20% (including cardiac mortality—16%) [3]. Independent predictors of time of survival (a multivariate Cox analysis) were age, cardiogenic shock, multivessel coronary artery disease (CAD), and previous myocardial infarction [3].

Identifying high-risk-individuals with STEMI remains a challenge, especially in the case of long-term prognosis. There are several risk scores used for prediction of short-term survival (GRACE, TIMI, Zwolle, or CADILLAC risk scores — Table 1), but none of them has been tested for follow-up for as long a period as 5 years.

The most widespread tool is the GRACE risk score, recently recommended by The European Society of Cardiology for risk stratification in patients with non-ST-elevation acute coronary syndrome [4]. It includes basic, but very detailed, data from physical examination and laboratory tests. It was developed from a large unselected population of patients with all forms of acute coronary syndromes and was primarily used for prognosis of in-hospital all-cause mortality [5]. The TIMI risk score for STEMI is a very simple alternative, derived from highly selected patients treated with fibrinolytics from the InTIME II trial [6]. The Zwolle risk score is also valuable, because it takes into account additional data from the coronary intervention [7]. Finally, there is a more sophisticated tool, the CADILLAC risk score that also analyzes biochemical data and ejection fraction [8]. Each of the scores has different characteristics and advantages. Both TIMI and Zwolle are based on logistic regression of the 30-day follow-up, whereas CADILLAC was constructed with the use of a 1-year analysis. In the case of TIMI and CADILLAC, patients with the highest risk of death were not included in the model, as only highly selected patients participated in the clinical trials that served as derivation sets. On the contrary, GRACE and Zwolle were developed from “real-life” registries. Finally, only Zwolle and CADILLAC scores were based entirely on individuals treated invasively.

As listed above, there are several factors that might influence performance of specific risk scores in patients with myocardial infarction. Further application of the risk scores for long-term follow-up would also bring a very interesting perspective.

The aim of our analysis was to assess the usefulness of GRACE, TIMI, Zwolle, and CADILLAC risk scores for predicting 5-year mortality in our registry of unselected patients with STEMI treated with pPCI.

Section snippets

Materials and methods

Our registry consisted of consecutive unselected patients admitted to a single reference centre due to acute STEMI in the years 2000–2002 (including patients transported from local hospitals without invasive treatment facilities). They all underwent pPCI within 12 h of the onset of symptoms. STEMI was diagnosed on the basis of chest pain history, ST-elevation in ECG (at least 0.1 mV in 2 contiguous leads, in the case of leads V1–V3 — at least 0.2 mV) and a rise in cardiac necrosis markers above

Results

Our registry comprised 514 consecutive patients. Nine of them were lost to follow-up due to migration beyond the Voivodeship's border (the registry used for retrieving follow-up includes only Podlaskie Voivodship). All further analysis was performed in the remaining 505 individuals (98.2%) with a completed 5-year observation. Median time of follow-up was 63.6 months. All-cause mortality reached 20.99% (n = 106). Thirty-two patients died during the first 30 days (6.3%); 51 in the first year (10.09%).

Discussion

In our “real-life” single-center registry of patients with STEMI treated with pPCI, 5-year mortality was 21% — high, though comparable with previous publications [1], [2]. In these conditions the GRACE risk score had the highest prognostic accuracy both for 30-day and 5-year follow-up. The other risk scores had insignificantly lower performances.

The good prognostic value of the GRACE score was expected; the model is very detailed, designed for prognosis of all-cause mortality, and based on a

Acknowledgement

The project was supported by a grant No 4-53850. The authors of this manuscript have certified that they comply with the Principles of Ethical Publishing in the International Journal of Cardiology [22].

References (22)

  • G. Parodi et al.

    Five year outcome after primary coronary intervention for acute ST elevation myocardial infarction: results from a single centre experience

    Heart

    (2005)
  • Cited by (56)

    • Prediction of mortality in hospital survivors of STEMI: External validation of a novel acute myocardial infarction prognostic score

      2019, Cardiovascular Revascularization Medicine
      Citation Excerpt :

      Although some of these means were designed (PAMI and CADILLAC) or validated (TIMI) for patients treated with invasive procedures (i.e. primary PCI), they were all originally conceived for relatively short-term risk stratification; 30-days-1-year. Furthermore, although subsequent evaluations for longer periods were performed, it often yielded lower performance by these scores [17,18]. Importantly the dynamic TIMI risk model which is an upgrade of the classic TIMI risk score, was designed for reassessment of the risk of patients discharged from hospital [20].

    • Association between glycemic variability and major adverse cardiovascular and cerebrovascular events (MACCE) in patients with acute coronary syndrome during 30-day follow-up

      2017, Clinica Chimica Acta
      Citation Excerpt :

      This risk algorithm is recommended by contemporary guidelines and has been incorporated into clinical practice with potential improvements in decision-making. After the ability of GRACE scores to discriminate the risk of death and acute myocardial infarction for 6 months was confirmed, the same applies for heart failure admissions for 1 y [10] and more prolonged periods of up to 6 y [11]. The recently published updated version of GRACE score 2.0 demonstrated a better discrimination in the prediction of 1- and 3-y mortality and mortality/infarction [12].

    • Natriuretic peptides, nitrite/nitrate and superoxide dismutase have additional value on top of the GRACE score in prediction of one-year mortality and rehospitalisation for heart failure in STEMI patients - Multiple biomarkers prospective cohort study

      2016, International Journal of Cardiology
      Citation Excerpt :

      More importantly, our study showed that oxidant-antioxidant pathways, particularly plasma production of NOx and concentration of SOD, are independently powerful prognosticators of outcome. The GRACE score was designed to predict the 6-month mortality in patients with acute coronary syndrome [5] later, the ability of GRACE scores to discriminate the risk of death and acute myocardial infarction for 6 months was confirmed [5], the same applies for heart failure admissions for 1 year [7] and more prolonged periods of up to 6 years [37]. The recently published updated version of GRACE score 2.0 demonstrated a better discrimination in the prediction of one- and three-year mortality and mortality/infarction [38].

    View all citing articles on Scopus
    View full text