Chicago, IL - Question: When is one drug-eluting stent as effective as another? Answer: Two years after the procedure.

That's the case with respect to angiographic outcomes, at least in the SPIRIT-2 trial, where two-year results for the Xience-V everolimus-eluting stent (Abbott Vascular) and the paclitaxel-eluting Taxus stent (Boston Scientific) were reported here last week at the American College of Cardiology 2008 Scientific Sessions/i2 Summit-SCAI Annual Meeting [1].

As covered previously by heartwire, the randomized SPIRIT-2 comparison of drug-eluting stents (DES) for the treatment of de novo coronary lesions showed a superior late-loss result for Xience-V at six-month angiography, the study's primary end point.

The latest angiographic results suggest that by two years, late loss in vessels treated with Xience-V had all but caught up to late loss in those treated with Taxus.

Other SPIRIT-2 results were consistent with such a "late catch-up" effect for Xience-V. Angiographic recurrence, as defined by percent-diameter stenosis, trended better at six months in the Xience-V group, but the difference at even that statistical level had dissipated by two years. Intravascular ultrasound data showed less lesion regrowth with Xience-V at six months, but not during long-term follow-up.

Angiographic outcomes at six months and two years in SPIRIT-2*: Xience-V vs Taxus

End point	Xience-V, 97 lesions	Taxus, 35 lesions	p
In-stent late loss
6 mo (mm)	0.17	0.33	0.004
2 y (mm)	0.33	0.34	0.603
In-stent %-diameter stenosis
6 mo (%)	16.00	18.3	0.062
2 y (%)	19.20	18.8	0.959
In-stent binary restenosis rate
6 mo (%)	1.00	2.90	0.461
2 y (%)	2.10	2.90	1.000

Clinical outcomes may have followed the same pattern. What seemed to be an early trend favoring the everolimus-eluting stent for major adverse cardiac events (MACE) had become less distinct by two years; at neither point, however, did clinical-outcome differences reach statistical significance.

Dr Patrick W Serruys

When reporting the two-year SPIRIT-2 results, Dr Patrick W Serruys (Thoraxcenter, Erasmus Medical Center, Rotterdam, the Netherlands) observed that the trial was designed for both noninferiority and superiority in the angiographic primary end point but wasn't statistically powered to show significant clinical differences.

"It is fair to say that Xience-V remains numerically lower than Taxus in MACE and all its components at two-year follow-up, despite a modest increase in late loss and neointimal hyperplasia in the Xience-V arm over time," Serruys said.

SPIRIT-2 randomized 300 patients in a 3:1 ratio to receive either the Xience-V or the Taxus stent for de novo lesions <28 mm in length and with a target-vessel reference diameter of >2.5 and <4.25 mm; up to two such lesions could be treated per patient. The trial was conducted at 28 centers in Europe, New Zealand, and India.

Dr David J Cohen

There were originally 223 patients in the Xience-V group and 77 in the Taxus group. Importantly, the analyses of six-month and two-year data presented by Serruys included only lesions for which there were data at both six months and two years.

As the discussant assigned to comment on Serruys's presentation, Dr David J Cohen (Saint Luke's Mid America Heart Institute, Kansas City, MO) lauded the SPIRIT-2 trial for having an excellent early and late angiographic follow-up rate and said the focus on patients with available data at each time point was a strength of the analysis. It eliminates the introduction of "spurious results due to changing patient populations over time."

Intravascular ultrasound measurements at six months and two years in SPIRIT-2*: Xience-V vs Taxus

End point	Xience-V, 69 lesions	Taxus, 32 lesions	p
Neointimal hyperplasia
6 mo (mm3)	4.1	12.6	<0.001
2 y (mm3)	8.4	11.6	0.253
Stent obstruction volume
6 mo (%)	2.8	6.5	<0.001
2 y (%)	5.2	5.8	0.403

Another strength, Cohen said, was that the trial's definition of in-stent thrombosis followed the Academic Research Consortium (ARC) criteria, recently introduced to standardize reporting of the complication.

In SPIRIT-2, two Xience-V patients developed very-late stent thrombosis and one Taxus patient developed stent thrombosis twice, once in the subacute period and once late after PCI, Serruys reported.

Cohen also observed that "although there was a moderate degree of very late neointimal growth [associated with the Xience-V stent], it is reassuring—albeit somewhat confusing—that this did not lead to any apparent excess of late clinical events."

The data, Cohen said, show "angiographic and clinical dissociation," in that, despite the stent's late catch-up to Taxus with respect to neointimal hyperplasia, "if we look at the clinical outcomes and in particular the clinical outcomes that ought to be most closely associated with restenosis—that is, ischemia-driven target-lesion revascularization—there was no late catch-up; in fact, if anything, the curves pull slightly further apart."

Clinical outcomes at six months and two years in SPIRIT-2: Xience-V vs Taxus

End point	Xience-V, n=211 (%)	Taxus, n=73 (%)	p
Cardiac death
6 mo	0.0	1.3	0.257
2 y	0.5	1.4	0.449
MI
6 mo	0.9	3.9	0.110
2 y	2.8	5.5	0.286
TLR (ischemia driven)
6 mo	1.8	3.9	0.379
2 y	3.8	6.8	0.330
Total MACE
6 mo	2.7	6.5	0.158
2 y	6.6	11.0	0.308

The Xience-V stent has been available in Europe and elsewhere in the world since 2006 and is thought to be close to approval by the FDA in the US, having received a nearly unanimous positive recommendation from the agency's premarket advisory panel in November 2007.

As reported then by heartwire, the panel's decision was based largely on the strength of the SPIRIT-2 six-month angiographic outcomes plus data from the much larger US-based SPIRIT-3 trial, which at the time was available out to one year. In the latter, a 1002-patient comparison of the same two stents, Xience-V bested Taxus by significant margins for the primary end point, in-segment late loss at eight months, and was noninferior for the secondary composite end point of cardiac death, MI, TLR, and target-vessel revascularization at nine months. Nine-month MACE rates were significantly lower for Xience-V. All differences persisted at 12 months, heartwire reported from the TCT 2007 meeting a month before the FDA panel met.

At that time, SPIRIT-3's principal investigator, Dr Gregg W Stone (Columbia University, New York, NY), reported that according to ARC criteria, there were seven instances of stent thrombosis in the Xience-V group and two in the Taxus group by 12 months.

Enrollment to the SPIRIT-4 trial is well under way, with a target exceeding 3000 patients. Launched in August 2007, the study is expected to yield final data for its primary end point—ischemia-driven MACE—in August 2009 [2].

Taxus stent's manufacturer, Boston Scientific, plans to market a version of Xience-V called Promos, which would be automatically cleared in the event of Xience-V's FDA approval. Promos will be manufactured by Abbott, the company notes, along with Xience-V, and will be supplied to Boston Scientific "as part of a distribution agreement between the two companies." The unusual market situation stems from the 2006 sale of Guidant's vascular- and endovascular-therapy business to Abbott at the same time the rest of the company was swallowed up by Boston Scientific.

The SPIRIT trial series is sponsored by Abbott Vascular. Serruys reports no conflicts of interest related to his presentation of the trial. Cohen reports grant support for research related to drugs from The Medicines Company, Schering-Plough, CV Therapeutics, Eli Lilly, Bristol-Meyers Squibb/Sanofi, and Baxter and for research related to devices from Cordis/J&J, Medtronic, Boston Scientific, Angel Medical Systems, and Edwards Lifesciences.