An electrocardiographic algorithm for the prediction of the culprit lesionsite in acute anterior myocardial infarction

Background: Although the 12-lead electrocardiogram (ECG) has been found use ful in identifying the left anterior descending (LAD) coronary artery as th e infarct-related artery in acute myocardial infarction (MI), it has tradit ionally been felt to be incapable of localizing the culprit lesion within t he LAD itself. Such a capability would be important, because anterior MI du e to proximal LAD lesions carry a much worse prognosis than those due to mo re distal or branch vessel lesions. Hypothesis: This study investigated whether certain ECG variables-especiall y an ST-segment injury pattern in leads aVL and/or V-1-would correlate with culprit lesion site, and an ECG algorithm was developed to predict culprit lesion site. Methods: The initial ECGs of 55 patients who had undergone cardiac catheter ization after an anterior or lateral MI were reviewed to identify the leads with an ST-segment injury pattern; the corresponding catheterization films were then reviewed to identify the location of the culprit lesion; and the se separate findings were then compared. Results: The sensitivity and specificity of an ST-injury pat tern in aVL in predicting a culprit lesion before the first diagonal branch were 91 and 9 0%, respectively; the same values in predicting a lesion prior to the first septal branch were 85 and 78%. ST-segment elevation in VI, on the other ha nd, was a much less sensitive and specific predictor of a preseptal lesion. Overall, our algorithm correctly identified the culprit lesion location in 82% of our patients. Conclusion: Based on our findings, we conclude that an ST-segment injury pa ttern in aVL during an anterior myocardial infarction predominantly reflect s a proximal LAD lesion and therefore constitutes a high-risk finding.