Atlanta, GA - Two trials examining the role of sirolimus- and paclitaxel-eluting stents, respectively, in acute myocardial infarction (AMI) have produced disparate results, prompting investigators to underscore the pitfalls of comparing outcomes between trials that randomized different types of patients, with different control arms.

Results of the 712-patient Trial to Assess the Use of the Cypher Stent in Acute Myocardial Infarction Treated with Balloon Angioplasty (TYPHOON) showed the sirolimus-eluting stent to significantly reduce target vessel failure in patients undergoing primary PCI for AMI, compared with a bare-metal stent. By contrast, the 619-patient Paclitaxel-Eluting Stent versus Conventional Stent for STEMI (PASSION) trial found no differences in the primary end point of MACE at one year and no differences in target lesion revascularization (TLR) rates between the bare-metal-stent and paclitaxel-eluting-stent groups. PASSION, however, included a number of more complicated lesions/patients that were specifically excluded in TYPHOON.

Dr Christian Spaulding

Both trials were presented during an i2 Summit late-breaking clinical-trials session here at the American College of Cardiology 2006 Scientific Sessions.

Speaking with heartwire following the presentation of both trials, lead investigator for TYPHOON, Dr Christian Spaulding (Cochin Hospital, René Descartes University, Paris, France), said, "I was a little surprised [that trial outcomes were different], but as soon as you include more complex patients, it's a little more difficult to show a significant difference, and that was the main difference between their study and ours. You should never try to compare two randomized trials that are done with two different designs using two different stents. In TYPHOON, there was a significant difference; in the PASSION trial, there was no significant difference. That's the only answer."

TYPHOON: Sirolimus beats bare-metal stents in AMI

TYPHOON randomized 712 patients AMI patients from 48 centers in Europe, Australia, and Israel to either a sirolimus-eluting or any kind of bare-metal stent. As Spaulding showed today, target vessel failure—a composite of target vessel revascularization (TVR), MI, or cardiac death—occurred nearly twice as often in the bare-metal-stent-treated patients over one year of follow-up than in those who received a Cypher stent. This difference was driven largely by a higher rate of TVR. Angiographic follow-up, performed at eight months in a subset of 210 patients, showed significantly less late loss, less diameter stenosis, and a lower rate of in-stent restenosis in the sirolimus-eluting stent group.

TYPHOON: Clinical and angiographic results

End point	Cypher	Bare-metal stent	p
Target vessel failure* (%)	7.3	14.3	<0.0036
MACE (%)	5.9	14.6	<0.001
Target vessel revascularization (%)	5.6	13.4	<0.001
Target lesion revascularization (%)	3.7	12.6	<0.0001
Binary restenosis (%)	3.5	20.3	0.001
Late loss (mm)	0.13	0.83	<0.0001
% diameter stenosis	16.4	37.1	<0.0001

"[Sirolimus-eluting stents] will further increase the long-term clinical benefit of primary PTCA, which has been demonstrated as the most efficient therapy for AMI when performed rapidly, by experienced teams," Spaulding concluded.

No differences in PASSION

Dr Maurits T Dirksen

Dr Maurits T Dirksen (Onze Lieve Vrouwe Gasthuis, the Netherlands) presented the results from PASSION immediately following TYPHOON. Surprising many in the audience, who had expected the results to mirror those of TYPHOON, Dirksen showed one-year MACE and TLR that were not statistically different between the paclitaxel-eluting-stent and bare-metal-stent groups.

"We did not observe a statistically significant difference in events after AMI in patients treated either with a Taxus or a bare-metal stent," he said.

A major factor was the lower-than-expected MACE rate in the bare-metal-stent group, he added, a finding that might be partially related to the fact that PASSION did not bring patients back for angiographic follow-up. The so-called "oculostenotic reflex" is frequently fingered as leading to higher TLR rates when restenosis is seen on angiogram.

"Overall, the use of a paclitaxel-eluting stent in AMI was shown to be safe and at least as effective as a bare-metal stent in this randomized trial," Dirksen concluded. "However, outcome differences were less pronounced than in other indications, and this has definite implications for cost/benefit profiles."

PASSION: Clinical results

End point	Taxus	Bare-metal stent	Hazard ratio (95% CI)	p
MACE	8.7	12.6	0.68 (0.41-1.10)	0.12
Death/MI	4.8	6.5	0.74 (0.38-1.45)	0.39
TLR*	6.2	7.4	0.68 (0.36-1.28)	0.23

In an interview with heartwire, Dirksen emphasized that there were important differences between the two studies. For one, the MACE rate in the bare-metal-stent group in TYPHOON was considerably higher than that seen in the bare-metal-stent-treated patients in PASSION. One reason, Dirksen noted, was that PASSION compared the Taxus paclitaxel-coated Express stent with a bare Express stent, while TYPHOON compared the Cypher with any bare stent. PASSION also enrolled patients with left main disease, bifurcation lesions, and patients with large thrombus, while TYPHOON specifically excluded these patients. Time from symptom onset to balloon inflation was also slightly longer in TYPHOON.

The major point, said Dirksen, was that the TLR rates seen in the drug-eluting-stent arms of both trials were similar. "So I don't think the sirolimus stent is more favorable than the Taxus stent in this setting."

Session moderators and panelists unanimously agreed that comparisons of the two trials were treacherous. To heartwire, Dr Nicolas Chronos (American Cardiovascular Research Institute, Atlanta, GA) commented: "These types of trials are extraordinarily dangerous when people start to compare one with the other. These are very small studies, and the definitions in some of the end points were different. One study used all-comers in the bare-metal-stent arm, whereas in PASSION, they used only the bare Express stent, and that's probably what drove the statistical difference in TYPHOON, but not in PASSION."