Stenting Diabetics: Time to Get Personal
Safety concerns still linger about using drug-eluting stents (DES) in patients with diabetes, but complications could result from the individual’s condition and not from the stent itself.

Diabetes currently afflicts 171 million people globally, including 23.6 million Americans—and these figures are expected to double by 2030. The adverse vascular consequences of diabetes are well documented, along with the increased incidence of atherosclerosis. Research suggests that the representation of coronary artery disease in patients with diabetes varies widely and the decision whether to perform PCI or CABG is not so clear cut.

“That decision needs to be tailored to the individual patient,” says Suresh R. Mulukutla, MD, an interventional cardiologist from the Cardiovascular Institute at the University of Pittsburgh Medical Center.

Suresh R. Mulukutla, MD, interventional cardiologist at the University of Pittsburgh Medical Center
Peter B. Berger, MD, director of the Center for Clinical Studies at Geisinger Health System in Danville, Pa.
A diabetic patient who is typically a good surgical candidate has left main disease, or one or two bifurcation lesions, or mild to moderate left anterior descending artery disease with a single severe stenosis. PCI would be preferred if the atherosclerotic state is relatively straightforward—one or no bifurcation, no significant left main involvement and little diffused disease.

Recently, some confusion arose over the BARI 2D trial and its accompanying editorial (NEJM 2009;360(24):2503-2515/2570-2572). BARI 2D, which enrolled more than 2,000 diabetics, was misinterpreted as having been powered to compare CABG with PCI. “It was not a CABG versus PCI study, but rather a medical therapy versus revascularization study. The choice of revascularization, which included both angioplasty and CABG, was left up to the treating physician. It was not a randomized comparison of the two,” Mulukutla explains.

However, BARI 2D led to some over-reaching interpretations that revascularization (whether PCI or bypass) is in many ways equivalent to medical therapy. In the accompanying editorial, Boden et al wrote that BARI 2D “reinforces other current scientific evidence supporting the benefits of CABG over PCI, especially in patients with diabetes.” In response, Peter B. Berger, MD, director of the Center for Clinical Studies at Geisinger Health System in Danville, Pa., says, “I couldn’t disagree more strongly with that statement, as the BARI 2D trial does not only not support CABG over PCI—it doesn’t even address the issue.”


Recent clinical data have trended toward DES over bare-metal stents (BMS) for diabetics. At the 2008American Heart Association scientific sessions, Mauri et al presented results on 5,051 diabetics in the nonrandomized Mass-DAC Registry, which found that the risk-adjusted mortality at three years was 17.5 percent for DES versus 20.7 percent for BMS, with no excess adverse events.

Yet, some questions still remain about whether insulin-treated diabetics will respond differently to either a DES or BMS. Mulukutla et al sought to answer this concern in an analysis of the one-year outcomes in the NHLBI Dynamic Registry, which involved insulin-treated and noninsulin-treated diabetics who underwent PCI with either DES or BMS (JACC Intv 2008;1:139-147). They found that DES were associated with a lower risk for repeat revascularization compared with BMS in either group. Also, DES use was not associated with any significant increased safety risk compared with BMS.

The authors concluded that DES should be the “preferred strategy for diabetic patients.” Mulukutla adds that while “study after study has shown stronger efficacy data with DES over BMS, our study shows that they are at least as safe as BMS in this patient population. In fact, there appears to be no safety concerns with the use of DES in the diabetic population.”

Mulukutla says the only exceptions are patients with whom there will be dual-antiplatelet compliance issues, such as those who have scheduled surgery. He estimates that about 85 to 90 percent of the diabetic population are being treated with DES. In an accompanying editorial, Steven Marso wrote that despite the improved vessel patency and acceptable safety profile with DES, there is neither a neutralization of restenosis risk nor an optimal DES platform ideally suited for diabetics (JACC Intv 2008;1:148-149).

Likewise, Berger says he does not use DES in all diabetic patients, due to “concerns about the frequency of late thrombosis.” He tends to use BMS in diabetics with discreet lesions and large arteries. With the DES brands available on the U.S. market, Berger says that some DES allow for branch vessel occlusions and possibly very late stent thrombosis, based on randomized trials and pooled data, leading him to conclude that “all DES are not created equal.”

During his presentation at the 2009 Society for Cardiovascular Angiography and Interventions conference, Berger reported that based on trials enrolling diabetics, the Cypher DES (Cordis) has less late loss than Taxus (Boston Scientific) in the first year after placement, and appears to result in less target lesion revascularization (TLR); Cypher also outperformed Medtronic’s Endeavor in SORT OUT III. Yet, he said data indicate that Xience V (Abbott Vascular) outperforms Taxus but not in diabetics. Finally, he reported similar late 12-month TLR in diabetic patients treated with Endeavor and Taxus.

At the 2009 Transcatheter Cardiovascular Therapeutics meeting, researchers reported that the Xience V performed statistically superior to Taxus in every area, except with diabetics where the two stents performed relatively equally, with target lesion failure rates of 6.4 percent (Xience V) versus 6.9 percent (Taxus). The data were from SPIRIT IV, the first randomized trial stratified for the diabetic population.

Conversely, Mulukutla suggests that “so far” the data indicate that all DES are created equal. In fact, he and his colleagues conducted another analysis of the NHLBI Dynamic Registry, comparing Cypher to Taxus and the preliminary evidence, according to Mulukutla, shows that there is no difference between the two stents in the diabetic population. This study was presented in ?abstract form at the 2008 AHA scientific sessions.

“It’s not so much the stent that is associated with adverse safety events, but rather patient-specific considerations that affect safety outcomes,” Mulukutla says. “It’s no longer really a question of DES versus BMS. The most recent literature has leveled the playing field.” And SPIRIT IV supports this view.

While Berger acknowledges that the data are not yet conclusive either way, he says that physicians have to rely on the data, rather than new indications or marketing hype, as a means to make decisions.

Will FREEDOM free us?

While cardiologists eagerly await the results of the FREEDOM trial, Berger and Mulukutla question whether the study’s diabetic patient population with multivessel disease is too sick to apply to the general diabetic population. “The results of FREEDOM shouldn’t influence a physician’s decision on how to treat diabetic patients with single vessel disease or two lesions in a single vessel, because these patients are less sick than the trial’s population and CABG is rarely the most appropriate therapy,” Berger says.

However, Mulukutla focuses on FREEDOM’s novel design. Trials such as COURAGE and SYNTAX included death and MI as well as repeat revascularization as their primary endpoints. FREEDOM investigators have “wisely” chosen, according to Mulukutla, to include repeat revascularization as a secondary endpoint. “In previous interventional studies, outcomes tended to be mostly driven by repeat revascularization. If you only examine the hard endpoints of death and MI, there is very little difference between PCI and surgery, and outcomes are mainly driven by repeat revascularization,” he says. “The FREEDOM design appears to be a more level approach to assessing the best revascularization approach to these patients.”

There has been some miscommunication about both COURAGE and SYNTAX regarding outcomes. For instance, the results of SYNTAX did not indicate that CABG is superior to PCI because if you remove repeat revascularization from its hard endpoint, there was very little difference between the two strategies.

Ideally, FREEDOM will shed some light on how to properly manage this burgeoning patient population as individuals rather than an aggregate group.