Evaluation of patients with symptoms suggestive of an acute myocardial infarction (AMI) and a nondiagnostic electrocardiogram can be difficult. A variety of specific myocardial markers, including creatine kinase (CK-MB [mass]), troponin T and myoglobin, have been used to increase diagnostic accuracy. The value of most studies assessing the appropriate use of these markers has been diminished by small or inappropriate populations. Jernberg and colleagues compared the early diagnostic performance of new biochemical markers to evaluate their role in accurately appraising patients with chest pain and a non-diagnostic electrocardiogram.

Patients with chest pain or other symptoms suggestive of an acute myocardial infarction who were admitted to the coronary care unit of a university hospital were included in the study. Of the 738 patients, 237 (32 percent) had normal electrocardiograms on admission, 251 (34 percent) had ST-segment depressions and 250 (34 percent) had other pathologic electrocardiographic results. A diagnosis of acute myocardial infarction required one of the following features: (1) pathologic Q waves developing in at least two leads; (2) symptoms suggestive of acute myocardial infarction or nondiagnostic electrocardiographic changes, and typically elevated plasma levels of biochemical markers with a CK-MB level of at least 10 ng per mL (10 μg per L); or (3) signs of acute myocardial infarction at autopsy. Plasma samples were analyzed on admission and after three, six and 12 hours for CK-MB, with troponin T at three hours and myoglobin at 12 hours. For each tested marker, the upper reference level provided by the laboratory was used as a cutoff value to discriminate between acute myocardial infarction and no acute myocardial infarction.

On admission, CK-MB and myoglobin were equally sensitive, while troponin T had a significantly lower sensitivity. At six hours after admission, CK-MB was significantly better than myoglobin in terms of sensitivity and specificity. The CK-MB demonstrated a high sensitivity with a negative predictive value of 99.8 percent at 6 hours after admission, whereas the specificity decreased, resulting in a positive predictive value of 64 percent. To confirm acute myocardial infarction with a higher specificity while maintaining a high sensitivity, the combination of CK-MB and troponin T showed the best diagnostic performance. The myoglobin level, generally acknowledged to increase before CK-MB in acute myocardial infarction, did not add value when obtained with the CK-MB in some patients with small acute myocardial infarction and early peak values because of its fast clearance.

The authors conclude that by using a combination of CK-MB and troponin T levels, it is possible to correctly confirm acute myocardial infarction in 98 percent of patients at six hours after admission, and this combination would be useful in evaluating patients with chest pain and a nondiagnostic electrocardiogram.

editor's note: Identifying acute myocardial infarction in patients with a normal electrocardiogram requires biochemical testing. Ebell and colleagues performed a systematic review of data on the accuracy of troponin T and I values for the diagnosis of acute myocardial infarction. The researchers noted that the sensitivity of these troponin markers increases from 10 to 45 percent within one hour of the onset of chest pain to more than 90 percent at eight or more hours. Specificity declines from 87 to 80 percent from one to 12 hours after the onset of chest pain for troponin T and is approximately 95 percent for troponin I. These results demonstrate that troponin levels are particularly useful for ruling out acute myocardial infarction at eight or more hours after onset of chest pain. The combination of CK-MB and troponin levels obtained in a serial manner, ideally three measurements within a 16-hour period, is best to accurately determine the presence or absence of an acute myocardial infarction when diagnostic electrocardiogram changes are absent.—r.s.