Abstract

Harkness, J. R., Sabatine, M. S., Braunwald, E., Morrow, D. A., Sloan, S., Wiviott, S. D., Giugliano, R. P., Antman, E. M., Cannon, C. P., Scirica, B. M. Extent of ST-segment resolution after fibrinolysis adds improved risk stratification to clinical risk score for ST-segment elevation myocardial infarction Am Heart J. 2010;159(1):55-62.

BACKGROUND: The TIMI risk score (TRS) for ST-segment elevation myocardial infarction (STEMI) is a convenient validated clinical risk score for predicting mortality. Although not part of the risk score, ST-segment resolution (STRes) may provide a simple method of risk stratification based on the response to reperfusion. We sought to determine whether STRes provides incremental risk stratification to the TIMI risk score.

METHODS: The Clopidogrel as Adjunctive Reperfusion Therapy--Thrombolysis in Myocardial Infraction (CLARITY-TIMI 28) trial randomized STEMI patients receiving fibrinolysis to clopidogrel or placebo. A total of 2,340 patients had electrocardiograms (ECGs) valid to calculate STRes at 90 minutes, which was defined as complete (>70%), partial (30%-70%), or no resolution (30%). TRS was defined as low (0-2), medium (3-4), and high (> or =5). Clinical follow-up was through 30 days. Results were validated in 2,743 patients from the ExTRACT-TIMI 25 study.

RESULTS: The degree of STRes at 90 minutes after fibrinolysis correlated in a stepwise fashion with death or heart failure (5.1% complete STRes, 8.9% partial STRes, 13.4% no STRes, P < .001). Furthermore, the degree of STRes provided a consistent and significant gradient of risk across all risk score categories (low, medium, or high) and significantly improved the discriminatory ability of TIMI risk score to predict death or heart failure (c-statistic 0.69 for TIMI risk score alone and 0.74 with STRes added to the model, P < .001). With the inclusion of STRes to the TIMI risk score, 913 patients (39%) were reclassified to higher or lower risk groups, and the net reclassification improvement (NRI) was highly significant (P < .001). In the ExTRACT-TIMI 25 trial, addition of the STRes improved also the c-statistic (P = .012) and NRI (P < .001).

CONCLUSIONS: The extent of STRes based on routinely obtained ECGs is an independent predictor of death and heart failure when used together with the TIMI risk score and significantly improves the ability to risk stratify patients after fibrinolysis.

Trial: CLARITY-TIMI 28