Continuous 12-lead ST-segment recovery analysis in the TAMI 7 study. Performance of a noninvasive method for real-time detection of failed myocardial reperfusion.

BACKGROUND

If a practical, reliable, noninvasive marker of failed reperfusion was available in real time, the benefits of further therapy in this patient subgroup could be tested. We developed a method of 12-lead ST-segment recovery analysis using continuously updated reference points to provide such a marker.

METHODS AND RESULTS

In this study, our method was prospectively tested in 144 patients given thrombolytic therapy early in myocardial infarction. All patients had 12-lead continuous ST-segment monitoring and acute angiography, each analyzed in an independent, blinded core laboratory. ST-segment recovery and re-elevation were analyzed up to the moment of angiography, at which time patency was predicted. Predictions were correlated to angiographic infarct artery flow, with TIMI flow 0 to 1 as occluded and TIMI flow 2 to 3 as patent. Infarct artery occlusion was seen on first injection in 27% of patients. The positive predictive value of incomplete ST recovery or ST re-elevation by our method was 71%, negative predictive value 87%, with 90% specificity and 64% sensitivity for coronary occlusion. ST recovery analysis predicted patency in 94% of patients with TIMI 3 flow versus 81% of patients with TIMI 2 flow and predicted occlusion in 57% of patients with collateralized occlusion versus 72% of patients with non-collateralized occlusion. In a regression model including other noninvasive clinical descriptors, ST recovery alone contained the vast majority of predictive information about patency.

CONCLUSIONS

In a blinded, prospective, angiographically correlated study design, 12-lead continuous ST-segment recovery analysis shows promise as a practical noninvasive marker of failed reperfusion that may contribute substantially to currently available bedside assessment. Our data also suggest that patients with TIMI 2 flow or with collateralized occlusions may represent a physiological spectrum definable with ST-segment recovery analysis.