By Chris Berrie
MUNICH, Germany -- September 3, 2008 -- Percutaneous coronary intervention (PCI) appears to be as effective as coronary artery bypass grafting (CABG) in patients with diabetes with multivessel disease or complex single vessel disease in terms of 1-year outcomes of mortality, nonfatal myocardial infarction (MI), and nonfatal stroke, according to results of a multicentre, randomised trial.
Akhil Kapur, MD, London Chest Hospital, Barts and the London NHS Trust, London, United Kingdom, presented the study on behalf of the Coronary Artery Revascularisation in Diabetes (CARDIA) investigators here on September 1 at the European Society of Cardiology 2008 Congress (ESC).
"When it comes to revascularisation in diabetic patients with multivessel disease, surgery has long been thought of as superior to [PCI]," Dr. Kapur said. And there have been no specific randomised comparisons of CABG and PCI in this setting.
The CARDIA investigators hypothesised that optimal PCI is not inferior to up-to-date CABG in these patients who are amenable to either procedure.
The study excluded patients who were more than 80 years old, and those who had previous CABG or PCI, left main stem disease, cardiogenic shock, recent ST-segment elevation MI, or contradictions to abciximab, aspirin, and clopidogrel use.
The primary endpoint was the composite of death, nonfatal MI, or nonfatal stroke at 1 year. The major secondary endpoint of further revascularisation at 1 year was accompanied by additional secondary endpoints.
The researchers randomised 254 patients to CABG and 256 patients to PCI. Patients in the 2 groups were a mean of 63.7 and 64.3 years old, respectively, and males made up 77.8% and 70.7% of each group, respectively.
Dr. Kapur noted that PCI was performed initially with bare metal stents (n = 29%) prior to availability and use of drug-eluting stents (71%; cypher).
Baseline clinical characteristics were similar across the groups, with three-quarters of patients in each group under elective admission and two-thirds of each group not receiving insulin treatment; the 2 groups had an average 10-year history of diabetes.
For CABG, 58% of patients had 3-vessel disease and 31% had off-pump surgery, with a mean of 2.8 grafts per patient. For PCI, 65% had 3-vessel disease, with means per patient of 3.5 stents implanted and 71 mm stent length.
As an intention-to-treat analysis, the primary outcome showed no significant difference between CABG and PCI (10.2% vs 11.6%; odds ratio [OR], 1.15; P = .63). Within this primary endpoint composite, there was a trend for fewer nonfatal strokes with PCI (2.5% vs 0.4%; OR, 0.16; P = .09).
For the secondary endpoint of further revascularisation, there was significant benefit with CABG (2.0% vs 9.9%; OR, 5.31; P = .001). When this was combined with the primary endpoint, this CABG benefit remained (11.0% vs 17.5%; OR, 1.72; P = .04), although compared with the DES subgroup of PCI, this significance was lost (11.0% vs 15.1%; P = .217).
Dr. Kapur also noted that this study was underpowered for formal noninferiority parameters, due to its early termination as a result of slow recruitment.
"Based on these results, we may be able to say that [PCI] could be considered a reasonable strategy in diabetic patients with multivessel disease, with the proviso that longer-term follow-up is needed," Dr. Kaput noted.
Funding for this study was provided by Eli Lilly and Company, Cordis Corporation, and Bristol-Myers Squibb Company/Sanofi-Aventis, with further support from Boston Scientific, Guidant Corporation, Jomed, and Medtronic, Inc.
[Presentation title: Coronary Artery Revascularisation in Diabetes. The CARDIA trial. Abstract 1694]
E-Mail this DGDispatch to a colleague
