Predictors and impact of target vessel revascularization after stent implantation for acute ST-segment elevation myocardial infarction: lessons from HORIZONS-AMI.

BACKGROUND

Target vessel revascularization (TVR) may compromise the benefits of primary percutaneous coronary intervention in ST-segment elevation myocardial infarction (STEMI) We set out to identify the predictors and examine the impact of TVR after STEMI in patients receiving a coronary stent.

METHODS

In HORIZONS-AMI, 3,602 patients with STEMI were randomized to bivalirudin versus heparin and a glycoprotein IIb/IIIa inhibitor. Stents were implanted in 3,202 patients (2,982 were randomized to bare-metal stents versus paclitaxel-eluting stents, and 220 received nonrandomized stents).

RESULTS

Target vessel revascularization occurred in 219 patients (6.9%) at 1 year and in 437 patients (14.4%) at 3 years. Target vessel revascularization was ischemia-driven in 418 cases (95.7%). Target vessel revascularization was due to restenosis in 219 patients (50.1%), definite stent thrombosis in 124 (28.4%), and disease progression in 94 (21.5%). Independent predictors of TVR were more extensive coronary artery disease, smaller vessel size, longer lesion length and the number of stents implanted, post-percutaneous coronary intervention diameter stenosis, symptom onset to balloon time, treatment with bare-metal stents rather than paclitaxel-eluting stents, and scheduled angiographic follow-up. Target vessel revascularization was an independent predictor of subsequent myocardial infarction (hazard ratio [HR] 5.25, P < .0001), ST (HR 5.98, P < .0001), and major bleeding (HR 5.25, P < .0001) but not mortality (HR 0.88, P = .61).

CONCLUSIONS

In HORIZONS-AMI, TVR within 3 years after stent implantation was performed in ~1 of every 7 patients and was associated with more extensive coronary disease, more complex procedures, and bare metal stents. Target vessel revascularization was often due to stent thrombosis and disease progression as well as restenosis and was strongly associated with adverse outcomes but not mortality.