Study objective: To evaluate the performance of cardiac specific markers, cardiac troponin I (cTnI) and CK-MB by mass assay (CK-MB mass), for the early diagnosis of myocardial ischemia and/or infarction after coronary bypass surgery. Methods: Prospective clinical, electrocardiograpic and biologic follow-up of 117 patients undergoing isolated corollary surgery with the use of intermittent anterograde normothermic blued cardioplegia. Blood samples for biochemical analysis were drawn before surgery (T-0) and at 2 (T-1), 6 (T-2), 10 (T-3) and 20 h (T-4) after aortic cross-clamp release. Without knowledge of the biochemical data, patients were classified according to the electrocardiographic evolution into two groups: group I, uneventful recovery and group 2. evidence of ischemia/infarction based on continuous ST-T segment monitoring and 12-lead EGG. Results: No patients had abnormal markers at T-0. At T-1, although both markers were elevated, no difference was noted between the two groups. At T-2 h after surgery, cTnl and CK-MB mass levels were significantly higher in group 2 than in group 1 (median = 17 mu g/l Interquartile Range (IR): 14.7-27.3 vs. 3.1 mu g/l, IR 1.9-5.3 for cTnI and median 42.5 mu g/l, IR: 27.1-95.7 vs. 13.6 mu g/l, IR: 9.5-18.5 for CK-MB mass). A receiver operating characteristic (ROC) curve analysis shows that a cTnI value of. 13.1 mu g/l has 100% specificity and 90% sensitivity to separate both groups, Whereas a value of 33.2 mu g/l for CEC-MB mass has a specificity of 100% and a sensitivity of 73%. At T-3 and T-4, the same difference was noted between the groups. cTnl values in all six patients with a Q-wave infarction were greater than or equal to 20 ng/l, whereas only one of five patients with prolonged ischemia bad cTnI level > 20 ng/ml. Conclusion: As soon as 6 h postoperatively, cTnI and CK-MB by mass assay were able to separate those patients with an uneventful recovery from those with significant ischemia. This is particularly useful in frequent cases when the ECG is difficult to interpret. (C) 1998 Elsevier Science B.V. All rights reserved.