PREDICTORS OF UNDERLYING CORONARY-ARTERY DISEASE IN COCAINE-ASSOCIATED MYOCARDIAL-INFARCTION - A METAANALYSIS OF CASE-REPORTS

Age, number of cardiac risk factors, inferior infarction and bradydysr rhythmias predicted underlying coronary artery disease (CAD) in patien ts with cocaine associated myocardial infarction. We sought to validat e these criteria through a meta-analysis of published case reports. Co caine associated myocardial infarctions reported in the English langua ge literature were retrospectively identified. Data abstracted include d patient demographics; cardiac risk factors, history of myocardial in farctions, last use of cocaine and complications during hospital cours e, and the presence or absence of CAD. Patients with CAD (>50% stenosi s on cardiac catheterization or autopsy) were compared to patients wit hout CAD (normal coronaries of cardiac catheterization). In all cases of Incomplete information. the authors were contacted for missing resu lts. We identified 66 patients with cocaine associated myocardial infa rction in whom the presence or absence of CAD was known: 41% had CAD. Of the 27 patients with CAD, 14 (52%) had 1 vessel disease, 7 (26%) ha d 2 vessel and 6 (22%) had 3 vessel disease. CAD was associated with t he presence of 2 or more cardiac risk factors (18% vs 0%, p = 0.03), h owever the sensitivity for detection of CAD was only 18%. Bradydysrrhy thmias were uncommon but only occurred in patients with CAD (8% vs 0%, p = 0.17), however the sensitivity for detection of CAD was only 8%. Patients with and without CAD did not differ in terms of the other cri teria (age > 40 y and infarct location) previously found associated wi th a higher likelihood of CAD. In this meta-analysis of published case reports, patients with cocaine associated myocardial infarction and 2 or more traditional cardiac risk factors were more likely to have und erlying CAD, but only 18% of patients with underlying CAD would be det ected through this screening criteria. Larger prospective trials are n eeded to better determine criteria predictive of underlying CAD in pat ients with cocaine associated myocardial infarction.