乌鸦传媒

WASHINGTON -- Pitting coronary artery bypass grafting (CABG) against percutaneous coronary intervention (PCI) for the treatment of unprotected left main coronary artery disease, two randomized trials found conflicting winners -- and left researchers searching for yet more answers.

In the long run, between CABG and PCI recipients (14.7% versus 15.4%, HR 1.00, 95% 0.79-1.26), according to , of New York-Presbyterian Hospital/Columbia University Medical Center, and co-investigators of the EXCEL trial. Patients were followed for a median of 3 years.

Ischemia-driven revascularization was more common after PCI (7.5% for CABG versus 12.6%, HR 1.72, 95% CI 1.27-2.33). However, definite stent thrombosis was less likely than graft occlusion (5.4% versus 0.7%, HR 0.12, 95% CI 0.05-0.28), giving PCI the upper hand, Stone suggested during his presentation at the annual Transcatheter Cardiovascular Therapeutics meeting. The data were simultaneously published online in the New England Journal of Medicine.

The EXCEL investigators randomized patients with unprotected left main disease to PCI with the Xience everolimus-eluting stent (n=948) or CABG (n=957).

Importantly, results were flipped in the short-term: at 4.9%, adverse events were reduced in the PCI group within the first 30 days of intervention (versus 7.9% for CABG, HR 0.61, 95% CI 0.42-0.88). Especially favoring PCI in this period were rates of MI -- especially those of ST-elevation myocardial infarction (0.7% versus 2.3%, HR 0.32, 95% CI 0.14-0.74).

"It should be noted that the PCI group exhibited a greater increase in these events between 30 days and 3 years than did the CABG group (11.5% vs. 7.9%, P=0.02). Therefore, it is reassuring that the EXCEL investigators plan further follow-up of these patients," wrote , of Brigham and Women's Hospital in Boston, in an .

Braunwald concluded that "the majority of patients with unprotected left main coronary artery disease, which was a very serious, life-shortening, and disabling condition early in my professional lifetime, can now be managed equally well by means of two strategies of revascularization if carried out by expert experienced teams such as those participating in the EXCEL trial."

The decision between the two procedures may ultimately rest on each patient's characteristics and require the operator to weigh the short-term advantage of PCI against the surgery's (albeit eroding) upper edge in the long run, Stone suggested to the audience. He emphasized the role of the heart team during the decision-making process.

Also presented at TCT were the results of the NOBLE trial, which suggested that CABG was actually the better choice for left main disease.

Over 5 years, all-cause mortality rates were no different between PCI and CABG groups (11.6% versus 9.5%, HR 1.07, 95% CI 0.67-1.72), as were those for stroke (4.9% versus 1.7%, HR 2.25, 95% CI 0.93-5.48). Symptomatic graft occlusion from CABG was just as likely as definite stent thrombosis (4% versus 3%, HR 0.59, 95% CI 0.26-1.36).

However, major adverse cardiac or cerebrovascular (28.9% versus 19.1%, HR 1.48, 95% CI 1.11-1.96), which landed CABG the superior position (P=0.0066), reported , of Denmark's Aarhus University Hospital Skejby. The NOBLE trial was simultaneously published online in The Lancet.

"This was a surprise for us and was the opposite of the SYNTAX trial," Christiansen told the audience at TCT.

Specifically, nonprocedural MI (6.9% versus 1.9%, HR 2.88, 95% CI 1.40-5.90) and repeat revascularization (16.2% versus 10.4%, HR 1.50, 95% CI 1.04-2.17) were the weaknesses of endovascular therapy.

Stone emphasized that unlike EXCEL, NOBLE did not include procedural MI in its analysis. With different patient populations and different endpoints, he emphasized that trials were not exactly comparable.

"When we designed the trial in 2008 we did not have SCAI criteria for procedural MI," defended Christiansen, arguing that his results remain valid: event curves "shift to favor CABG" in his longer-term study and that was a trend that may not have yet been statistically significant in Stone's study.

"If you look at the endpoints, the curves are diverging late through 3 years in EXCEL," agreed , of Saint Luke's Mid America Heart Institute in Kansas City, Mo. In a TCT press conference, Cohen pointed out the "catch-up" phenomenon that perhaps made EXCEL and NOBLE not so different after all.

Paradoxically, Christiansen's group also found that lowest-risk patients -- those with SYNTAX scores under below 23 -- appeared to benefit the most from CABG over PCI (HR 1.88, 95% CI 1.23-2.89). On stratification, their higher-risk peers had statistically similar event rates between groups.

"We saw a slight difference in patients refusing the allocated treatment in favour of PCI and in some patients' view, the need for surgery, the long stay in hospital, the risk of reoperation for bleeding and infection, and a longer recovery time might not be worth the lower risk of repeat revascularization and myocardial infarction because no difference in all-cause mortality was found," the NOBLE trialists noted.

In NOBLE, patients were randomized to PCI (n=598) or CABG (n=603). The first tenth of patients got a first-generation Cypher stent; the rest got the Biomatrix Flex. Investigators did not use SYNTAX criteria for trial inclusion.