Perioperative myocardial infarction in cardiac surgery : a diagnostic dilemma

Sammanfattning: Perioperative myocardial infarction remains a major cause of morbidity and mortality after cardiac surgery. In spite of this there is no consensus regarding diagnostic criteria and consequently the reported incidence varies widely. In this thesis risk factors for PMI and outcome after PMI were studied in a retrospective case control study on 42 patients fulfilling strict criteria for PM! collected from a cohort of 1147 adult cardiac surgical patients. Traditional diagnostic criteria for PMI, release characteristics of biochemical markers for myocardial injury and VCG were evaluated in a prospective study on 302 consecutive patients undergoing isolated frrst time CABG. PM! was found to be a problem mainly associated with surgery for ischaemic heart disease. Unstable angina and unfavourable conditions for revascularisation were found to be the most important risk factors for PMI. Patients with PM! had an impaired short-term and mid-term outcome compared with controls. In the prospective study a sustained elevation of troponin-T was used as a marker for permanent myocardial damage. It was demonstrated that Q-wave criteria, previously accepted as the gold standard for diagnosis of PM!, correlated poorly with biochemical markers of myocardial injury and clinical outcome. One fourth of the patients with new Q-waves after CABG had no evidence of permanent myocardial injury. The use of biochemical markers for early diagnosis of myocardial injury is interfered by unspecific release unrelated to permanent myocardial damage. However, little is known about the magnitude of this "diagnostic noise". To address this issue a subgroup with no or minimal permanent myocardial damage was identified by use of the unique release characteristics of troponin-T. The time frame of unspecific release and the plasma levels of CKMB and troponin-T caused by unspecific release were assessed. A substantial early release of both CKMB and troponin-T nnrelated to permanent myocardial injury was found. As the unspecific release can be expected to differ depending on type of cardiac intervention this type of knowledge may prevent inappropriate comparisons. Repeated early sampling for CKMB provided additional information of value for early identification of patients who would later show sustained elevation of troponin-T. VCG was found to correlate better with sustained levels oftroponin-T and clinical outcome than scalar ECG. To conclude, diagnostic pitfalls associated with both ECG and enzymatic diagnosis of PM! were addressed and novel approaches to improve detection of permanent myocardial damage are suggested.