BARI 2D: Revascularization no more beneficial than intensive medical therapy in diabetes, CAD

The much anticipated results of the Bypass Angioplasty Revascularization Investigation in Type 2 Diabetes — BARI 2D — trial indicate no significant advantage of prompt revascularization alone over intensive medical therapy on mortality in patients with type 2 diabetes and stable coronary artery disease.

At five years, there was no difference in mortality, the primary outcome, between patients randomly assigned to revascularization (88.3%) — percutaneous coronary intervention or coronary artery bypass graft surgery — compared with intensive medical therapy alone (87.8%; P=.97). Rates of survival were also similar between patients assigned to insulin-sensitization (88.2%) vs. insulin-provision therapy (87.9%; P=.89).

Prompt revascularization also had no benefit on freedom from major CV events (composite of death, myocardial infarction or death), the principal secondary outcome, compared with intensive therapy (77.2% vs. 75.9%; P=.70). Similarly, there was no significant difference among patients assigned insulin-sensitizers (77.7%) vs. insulin-providers (75.4%; P=13).

When examined separately, however, although PCI vs. medical therapy showed no difference in rate of death or CV events, prompt CABG, when compared with medical therapy alone, yielded significantly better outcomes when major CV events were considered in addition to death (22.4% vs. 30.5%; P=.01). Much of the benefit with CABG was an observed reduction in nonfatal MI (7.4% vs. 14.6%).

“This is the first demonstration in a properly conducted randomized trial that CABG reduces nonfatal MI,” said study chair Robert L. Frye, MD, of the Mayo Clinic. The results were presented in June at the American Diabetes Association’s 69th Scientific Sessions in New Orleans.

Robert L. Frye, MD, said BARI 2D was not a negative trial; a select, high-risk group of patients may benefit from revascularization. Photo by: Joe Kane

Taken together, the new data provide evidence that treatment can safely begin with an initial program of intensive medical therapy, the researchers concluded.

However, “one really important issue is to understand the BARI 2D outcomes and interpret the results,” Frye said. “Nothing stays fixed for five years, clinically, in a patient with diabetes or CAD.”

“The leading cause of death in diabetes is heart disease. Unfortunately, we do not know the best way to treat either the diabetes or the heart disease in patients. BARI 2D was designed to add to the basic clinical question of how to treat the typical 60-year-old patient with diabetes who presents for evaluation of potential heart disease to a cardiac catheterization facility,” said researcher Trevor Orchard, MD, professor of epidemiology, University of Pittsburgh Graduate School of Public Health.

The BARI 2D data build on results of the first BARI trial, released in 1996, which compared CABG and PCI using balloon angioplasty in patients with CAD. Results demonstrated no difference in long-term mortality rate and MI; however, results indicated slightly better survival after CABG among patients with diabetes.

The researchers cautioned that this trial did not compare CABG and PCI; however, “the differences in secondary endpoints render indirect comparisons likely,”

William E. Boden, MD, and David P. Taggart, MD, PhD, wrote in a New England Journal of Medicine editorial.

The latest BARI 2D findings have been compared to the Clinical Outcomes Utilizing Revascularization and Aggressive Drug Evaluation (COURAGE) trial. “The BARI 2D results replicate the principal finding of the COURAGE trial — that an initial strategy of PCI provides no incremental clinical benefit over intensive medical therapy,” Boden and Taggart wrote in the editorial.

Unique trial design

BARI 2D attempted “to marry the best cardiac care and the best diabetes care to see if either alone or together would provide the best outcome,” co-principal investigator Richard W. Nesto, MD, chairman of the department of cardiovascular medicine at Lahey Clinic Medical Center in Burlington, Mass., told Endocrine Today.

Richard W. Nesto

The trial compared both a CV treatment approach and diabetes control approach in 2,368 patients with type 2 diabetes and stable CAD. At baseline, 82.1% of patients had symptomatic myocardial ischemia; mean duration of diabetes was 10.4 years.

Patients were randomly assigned to prompt revascularization (CABG, n=763; PCI, n=1,605) or intensive medical therapy alone, per the referring physician’s recommendation. Patients selected for CABG had more extensive coronary disease, three-vessel disease, proximal disease of the left anterior descending artery, chronic coronary occlusions, and were more likely to have a history of MI than patients for whom PCI was intended.

The intensive medical therapy regimen followed clinical guidelines, with common use of aspirin, statins, beta-blockers, ACE inhibitors and/or angiotensin-receptor blockers.

Those assigned revascularization underwent the procedure within four weeks of randomization; patients assigned medical therapy underwent revascularization only following acute coronary syndrome or continuing/worsening angina or ischemia.

Coronary revascularization was performed within six months in 95.4% of patients in the prompt revascularization group vs. 13% of patients in the medical therapy group. At five years, 42.1% of patients in the medical therapy group (43.3% in the PCI stratum and 39.7% in the CABG stratum) had undergone clinically indicated revascularization.

Alan J. Garber

“A significant number of patients became unstable and required vascular procedures. But, medical management seems an appropriate choice for diabetic patients with mild to moderate CAD,” Alan J. Garber, MD, Endocrine Today Chief Medical Editor, said in an interview.

The findings suggest that patients with diabetes and stable CAD “would benefit from prompt revascularization mainly because of a lower rate of nonfatal MI,” the researchers wrote. “However, for the many patients with type 2 diabetes who have less extensive CAD and for whom PCI is judged to be more appropriate, prompt revascularization did not reduce the risk of CV events, as compared with medical therapy.”

The cardiology community was not surprised at these results, according to cardiologist Carl J. Pepine, MD, at the University of Florida at Gainesville and a member of the Endocrine Today Editorial Board.

Pepine said these results should “add some reassurance for patients who decide to defer revascularization until their symptoms sufficiently impair their quality of life.” However, he said, there are no sufficient, published data on quality of life to make that judgment.

Insulin-sensitization vs. insulin-provision

A second component of the study compared whether controlling diabetes with insulin-sensitizing drugs (metformin and thiazolidinediones) had an advantage over insulin-providing drugs (insulin and sulfonylureas). The same patient population was randomly assigned to one of the two treatment arms.

All patients were treated to a target HbA1c of <7%, LDL <100 mg/dL and BP <130/80 mm Hg. Patients also received lifestyle counseling for smoking cessation, weight loss and exercise.

Insulin-providing and insulin-sensitizing drugs generally yielded similar results on mortality and CV events.

In contrast to previous some reports, the researchers observed no increase in MI among patients in the insulin-sensitizing group who were assigned rosiglitazone (Avandia, GlaxoSmithKline).

The insulin-provision group had a greater frequency of severe hypoglycemia (9.2%) compared with the insulin-sensitization group (5.9%; P=.003). Less weight gain and higher HDL were reported with insulin-sensitizing drugs.

There was a trend toward more heart failure (22.6% vs. 20%) and fractures (7.6% vs. 6.9%) in the insulin-sensitization group compared with the insulin-provision group. These adverse events have been reported previously with TZDs.

“As we get more exposure data of patients to TZDs we are finding that the HF and fracture risk remains, but the concern about risk for adverse ischemic heart disease events seems to be going away. [Recent] data are more powerful than original data from several meta-analyses that suggested a greater risk for CV events with TZDs,” Nesto said.

Researchers noted that the benefit of early CABG was largely seen in patients who were also treated primarily with insulin-sensitizing drugs. The effect of revascularization on the rate of CV events was particularly evident among patients in the CABG stratum who were assigned to the insulin-sensitizing strategy, with a rate of 18.7% among patients in the revascularization group vs. 32% among those in the medical therapy group (P=.002).

There was a mean difference of less than 0.5% in between the two glycemic control strategies.

The treatment regimens in BARI 2D reflect clinical possibilities for patients with established type 2 diabetes and stable CAD, according to the researchers.

During the press conference at ADA, the BARI 2D researchers offered two schools of thought on the study results.

Trevor Orchard

“From a diabetes viewpoint, the take-home message is the reassurance that previous concerns about TZDs have not been manifested in this trial and that good diabetes control can be achieved with either insulin-sensitization or insulin-provision. It is possible to have early and aggressive intervention on CV risk factors,” Orchard said.

“From the cardiology perspective, the most striking finding was the identification of a high-risk group of patients who were selected for CABG and who benefited from prompt CABG. That emphasizes the importance of continuing what has been a long-time effort, hopefully enhancing it with some of the new imaging modalities to identify the high-risk group who would benefit from prompt revascularization,” Frye said.

While the BARI 2D researchers did not officially recommend intensive medical therapy for all patients with diabetes and CAD, “optimal medical therapy rather than any intervention is an excellent first-line strategy, particularly for those with less severe disease,” Boden and Taggart wrote in the editorial.

The majority of experts that Endocrine Today interviewed recommended individualized treatment for each patient. However, many patients in BARI 2D had minimal or no symptoms of CAD, according to Nesto. “Although screening for CAD comes to mind, that is clearly not what we should do; we know there is no indication for that,” he said.

Whatever strategy health care professionals choose, there must be “a strong emphasis on coordinated care between diabetology, primary care physicians and cardiologists,” Frye said.

One limitation was that “patients in the BARI 2D study are highly selected and represent only a very small slice of diabetes and heart disease patients. More research is needed on the larger diabetic population,” such as data from the FREEDOM trial, Steven R. Bailey, MD, president of the Society for Cardiovascular Angiography and Intervention, said in a statement.

He also said: “The majority of the BARI 2D patients did not receive drug-eluting stents, which have been shown to be superior to bare metal stents for diabetic patients. The standard of care has evolved since BARI 2D was initiated and many of the early patients in the trial did not receive the care we now know to be most effective.”

Taken together, BARI 2D “provides a new, noninvasive roadmap for the management of diabetes and stable angina that is highly useful for clinicians struggling with this highly prevalent subpopulation of patients with diabetes,” Garber said. – by Katie Kalvaitis

Is it difficult to defend revascularization with stable CAD?

For more information:

• BARI 2D Study Group. N Engl J Med. 2009;360:2503-2515.