New HORIZONS-AMI analysis links stent thrombosis with pharmaco strategies
Orlando, Fla.—A new analysis of the HORIZONS-AMI trial found no increased stent thrombosis with drug-eluting stents (DES) compared with bare-metal stents (BMS)—both shortly after PCI and at one year. It also found that adjunct pharmacologic therapy influenced the risk of stent thrombosis, according to research presented Sunday during the i2 Summit at the American College of Cardiology’s 58th annual scientific session. 

There has been some concern that late stent thrombosis might be more common with DES than BMS, said lead investigator George Dangas, MD, an associate professor of medicine at Columbia University Medical Center and director of academic affairs and investigational pharmacology at the Cardiovascular Research Foundation, both in New York City. 

For the study, researchers analyzed data from 3,202 patients who were treated with DES (any type) and BMS in the main HORIZONS-AMI trial. Within one year, stent thrombosis occurred in 107 patients (3.4 percent). Of these, 0.9 percent of cases were acute, 1.6 percent were subacute, and 1 percent were late.  

The one-year rate of stent thrombosis was identical among patients treated with either DES or BMS alone. Among patients who received the anti-thrombin bivalirudin (Angiomax), the rate of one-year stent thrombosis was 3.6 percent, compared to 3.2 percent among patients who received unfractionated heparin plus glycoprotein IIb/IIIa inhibitors (UFH+GPI). The difference was not significant and was mainly driven by the higher rate of acute stent thrombosis (within the first five hours) associated with bivalirudin alone. Between 24 hours and one year, there tended to be less stent thrombosis associated with bivalirudin compared to UFH+GPI, Dangas said. 

Researchers also found that acute, sub-acute and late stent thrombosis appear to be related to different factors. Pharmacologic therapy, vessel flow, lesion characteristic, and the number and length of stents were the most important independent predictors of acute and sub-acute stent thrombosis. The most important independent predictors of late stent thrombosis were cigarette smoking and prior MI. Insulin-treated diabetes was a predictor for sub-acute and late stent thrombosis.   

A pre-randomization bolus of UFA and a high loading dose of clopidogrel (600 mg) were protective factors against acute and subacute stent thrombosis, respectively, Dangas reported. 

In his commentary to the study, E. Magnus Ohman, MD, a professor of medicine and director of the Program for Advanced Coronary Disease at Duke University Medical Center, Durham, N.C., praised the study for being well conducted.  

He commented on the significant mortality reduction for bivalirudin monotherapy at one year versus UFH+GPI in the original study. “Even though HORIZONS-AMI is not a blinded trial, it’s very hard to bias mortality,” he said. But the early higher rate of stent thrombosis associated with bivalirudin, along with the drug’s improved rates of sub-acute and late stent thrombosis “makes it complicated to understand the overall effect,” Ohman said.  

The small number of events (107) has limited power to define nonrandomized associations of stent thrombosis in STEMI patients. “We’d probably need between 400 and 500 events to conduct adequate statistical models. Nevertheless this study is important,” he said.  

The analysis suggests a number of things, Ohman said. “Getting upstream antiplatelet or anti-thrombin therapy is very important. My take home message from this analysis is we have to give a loading dose of 600 mg of clopidogrel, and we have to give a heparin bolus in the emergency department. That’s an easy strategy for most emergency departments to remember.” 

Another point of discussion that arose from the panel concerned the use of more rapid and effective ADP antagonists such as prasugrel (Effient), which has shown a 51 percent lower rate of early stent thrombosis in the TRITON-STEMI trial. “We need to recognize, however, that the dose of clopidogrel in that trial was 300 mg, Ohman said.  

Panelist John Douglas, MD, director of the Interventional Cardiology and Cardiac Catheterization Laboratory at the Emory Clinic in Atlanta, Ga., said the issue of whether to prolong anti-thrombin therapy with bivalirudin should be tested before being adopted. “The literature that extends anti-thrombin therapy in this field is associated with more bleeding, so it may actually lose the value of a simplified strategy as was tested in HORIZONS-AMI. Nevertheless, this is an important study and will influence practice,” he said.