Author + information
- Spencer B. King III, MD⁎ ()
- ↵⁎Address for correspondence to:
Spencer B. King III, MD, Editor-in-Chief, JACC: Cardiovascular Interventions, Saint Joseph's Heart and Vascular Institute, 5665 Peachtree Dunwoody Road, NE, Atlanta, Georgia 30342
Those of us involved in the performance and interpretation of BARI 2D (Bypass Angioplasty Revascularization Investigation Type 2 Diabetes) have been trying to explain the findings to patients and the public. It is a difficult job which involves a lot of nuance and is not suited to 1-line conclusions. I will try to give my observations of what we showed because I think some remain confused. Even the National Institute of Health website headlines falls into the word-conserving trap of trying to simplify the message: “Optimal Medical Therapy As Beneficial As Elective Revascularization Procedures in Patients with Type 2 Diabetes and Stable Coronary Heart Disease” (1). What was the trial setting out to show (2)? I will not discuss the insulin-providing versus the insulin-sensitizing arm of the study, but the revascularization versus continued medical therapy arm. The question here was: Among patients with diabetes and ischemic heart disease with clinical and anatomic presentations that allowed equipoise between routine revascularization or no routine revascularization, will the early revascularization be associated with improved clinical outcomes? There was no plan to compare percutaneous coronary intervention (PCI) and coronary artery bypass grafting (CABG) but to select the 2 revascularization strategies at the discretion of the investigators. Guidance was given to the investigative sites to consider the results of the BARI trial and others that showed better outcomes in diabetic patients with surgery for more complex multivessel disease and to use PCI for less extensive disease. The greater extent of disease in the CABG stratum was reflected by the 5-year mortality for patients randomized to medical therapy (16.4%) compared with the PCI stratum patients randomized to medical therapy (10.2%). Almost all of the CABG was done in 2- and 3-vessel patients; however, about 40% of the PCI was done in 1-vessel patients. This study of the strategy of revascularization with the tool the experienced investigators felt best for their patients at their sites was perfectly reasonable. The result was that the strategy of routine revascularization did not result in an improved 5-year outcome compared with a deferred strategy of performing intervention in the medical group only if needed for future symptoms or ischemia. Of the medically treated patients, 40% did receive this deferred revascularization during follow-up.
The comparison of the tools used to achieve this revascularization result is inappropriate. Either PCI or CABG was chosen to perform the revascularization in very different patients. Analysis of the patients who ended up in the surgical group showed that they were significantly more ischemic and had more extensive coronary artery disease. These are patients well-known to have adverse outcomes and to have benefit from relief of that ischemia. The patients treated with PCI had less ischemia, a lesser extent of coronary artery disease, a better prognosis, and a decreased opportunity to improve that prognosis with intervention. It is conceivable that patients who had improvement with revascularization may have benefited disproportionately by having CABG rather than PCI, and it is conceivable that those with less disease may have been more safely and effectively treated with PCI than undergoing CABG, but neither of these questions were studied in BARI 2D.
In talking to patients I try to help them understand the gradation of risk that exists among patients with various extents of coronary artery disease and ischemia and the concept that relief of extensive ischemia is of greater benefit than lesser degrees of ischemia relief. This approach should allow many patients similar to those in the trial to continue aggressive medical therapy without incurring increased risk compared with routine revascularization. There are, however, many diabetic patients who can benefit significantly from relief of major ischemia, most of whom were not similar to the BARI 2D patients; in such patients revascularization (the best procedure at each local institution, i.e., CABG, PCI, or hybrid procedures) should not be inappropriately denied. Some have confused the patient selection in BARI 2D with that in the ongoing FREEDOM (Future REvascularization Evaluation in patients with Diabetes mellitus: Optimal management of Multivessel disease) trial, which is limited to patients judged to be requiring revascularization. BARI 2D would by definition exclude such patients. Additional analyses of BARI 2D will expand our knowledge on how best to manage revascularization. It was a very well done trial that helps aim us in the right direction—but, it is not so well suited for a single headline.
- American College of Cardiology Foundation
- U.S. Department of Health and Human Services