CABG Shows Survival Advantage Over PCI in Nondiabetic Patients With Multivessel Disease

The theory that contemporary PCI might diminish the advantage of CABG in patients with multivessel disease does not appear to be valid, at least based on pooled patient-level data from two large randomized trials.

“CABG, as compared with PCI with DES, significantly reduced the long-term risk of mortality in nondiabetic patients with multivessel CAD,” write Mineok Chang, MD (Asan Medical Center, Seoul, South Korea), and colleagues.

But another researcher who was not involved with the study told TCTMD the findings are based on underpowered, outdated data that do nothing to inform the decision-making process.

For their study published online June 27, 2016, ahead of print in the Journal of the American College of Cardiology, Chang and colleagues sought to compare long-term outcomes in 1,275 nondiabetic patients who had been enrolled in the SYNTAX and BEST clinical trials and randomized to PCI or CABG. While the SYNTAX trial used first-generation paclitaxel-eluting Taxus stents (Boston Scientific) in its PCI arm, BEST used newer everolimus-eluting stents. Patients enrolled in both trials had either 2- or 3-vessel CAD, or left-main disease.

At 5 years, more patients in the PCI group than in the CABG group had died from any cause (9.3% vs 6.0%; P = 0.04). This disparity was most pronounced in those with intermediate to high SYNTAX scores. Within this group, all-cause mortality was 11.6% at 5 years with PCI versus 7.1% with surgery (log-rank P = 0.02). The CABG group also had lower rates of the composite of death, MI, or stroke as well as less cardiac death and fewer individual MIs. However, the rate of stroke was similar between the two treatment groups. Not surprisingly, repeat revascularizations occurred less often among patients who had surgery compared with PCI.

In subgroup analyses, CABG maintained its advantage over PCI, with no interaction between BEST and SYNTAX for treatment effect or by DES type.

Some Answers, but Not Enough

According to Chang and colleagues, understanding long-term mortality in this population “may help physicians to decide the best treatment strategy for a particular patient.” Plus, they add, all-cause mortality may be “the most reliable endpoint” since it is not affected by bias in classifying cause of death.

“These findings strengthen the concept that CABG provides a mortality benefit for patients with a high burden of CAD, whereas in patients with a lower burden of CAD (SYNTAX score 0 to 22), CABG and PCI may perform similarly with respect to the long-term rates of major adverse cardiac and cerebrovascular events,” write Farouc A. Jaffer, MD, PhD (Massachusetts General Hospital, Boston, MA), and Patrick T. O’Gara, MD (Brigham and Women’s Hospital, Boston, MA), in an editorial accompanying the study.

However, “decision-making for nondiabetic patients with multivessel CAD should be individualized,” they point out, adding that in cases of clinical equipoise, a multidisciplinary CAD heart team can help improve the decision process.

Underpowered and Adds Nothing

While the findings are consistent with those of the large, observational ASCERT trial, other study data have shown comparable rates of death after CABG and EES PCI over 2.9 years of follow-up.

Sripal Bangalore, MD, MHA (NYU Langone Medical Center, New York, NY), who authored the latter study, told TCTMD that the pooled SYNTAX and BEST data add nothing new to the conversation on how best to treat nondiabetic patients with multivessel disease.

“We could have come to the same exact conclusion by looking at the SYNTAX trial alone,” he said. “BEST was underpowered for the composite endpoint. So, you can imagine that it is much more underpowered for individual outcomes. All of the results [of the pooled data] are mainly driven by the SYNTAX trial. The [BEST] everolimus group was pretty small, with only about 250 patients, so the question of whether CABG is superior to PCI with next-generation stents is still not answered.”

Bangalore said clinicians and patients are in need of more data on CABG versus contemporary PCI in multivessel disease, some of which likely will come from the ongoing FAME 3 and EXCEL trials.

“Until that time, it’s very difficult to rely on data from trials such as SYNTAX,” he added. “We know that Taxus is one of the worst possible stents, with very high rates of stent thrombosis even out to 6 years. Therefore, the decision between PCI and CABG should still be based on whether you can completely revascularize with PCI and if not you should consider bypass, but patient preference highly matters in this decision. Data from prior trials using outdated stents is not going to help us.”


  • Chang M, Ahn J-M, Lee CW, et al. Long-term mortality after coronary revascularization in nondiabetic patients with multivessel disease. J Am Coll Cardiol. 2016;68:29-36.
  • Jaffer FA, O’Gara PT. Multivessel CAD in nondiabetic patients: to operate or to dilate? J Am Coll Cardiol. 2016;68:37-39.


  • The study was supported by a research grant from the CardioVascular Research Foundation (CVRF), Seoul, South Korea.
  • Chang reports no relevant conflicts of interest.
  • Bangalore reports serving on the advisory board of Abbott Vascular.

Related Stories:

  • Guidelines Notwithstanding, More Diabetics with Multivessel CAD Receive PCI Than CABG
  • SYNTAX Score II Added to Modern PCI Translates Into Gains for Patients With Three-Vessel Disease
  • PCI, Surgery Show Similar Long-term Mortality in Diabetic, Nondiabetic Patients


Content retrieved from:


Recent Blog Posts

Perfusion Services Inc 5000

For the Third Consecutive Year, Keystone Perfusion Services Makes the Inc. 5000 List with Three-Year Revenue Growth of 410 Percent


Fifty Things a Cardiac Surgery Team Can Do for Blood Conservation


ECMO: Last-Resort Life Support Option Helped Majority of Critically Ill COVID-19 Patients Survive, Global Study Shows.


Keystone Perfusion Services, PC Makes Inc. Magazine’s List of 5000 Fastest-Growing Private Companies for Second Time.