SYNTAX Score Predicts Long-term Benefits of Left Main PCI and CABG: MAIN-COMPARE

For those with low-to-intermediate anatomic complexity, PCI and CABG fared equally at 10 years.

SYNTAX Score Predicts Long-term Benefits of Left Main PCI and CABG: MAIN-COMPARE

Coronary revascularization with PCI or CABG surgery results in comparable clinical outcomes at 10 years in patients with left main coronary artery disease and low-to-intermediate anatomic complexity, but there is a survival advantage with surgery in patients with the most complex anatomy, according to long-term follow-up of the MAIN-COMPARE registry.

Led by Yong-Hoon Yoon, MD (Chungnam National University Hospital, Daejeon, South Korea), and published today in JACC: Cardiovascular Interventions, the investigators reported that for individuals with a low-to-intermediate SYNTAX score, all-cause mortality and the combined endpoint of death, MI, or stroke were not statistically different between patients treated with PCI or CABG surgery, although rates of target vessel revascularization were higher with PCI.

In patients with complex anatomy, as indicated by a high SYNTAX score, the risk of all-cause mortality was significantly higher with PCI than with CABG surgery, as was the rate of target vessel revascularization. There was a nonsignificant trend toward a higher risk of death, MI, or stroke with PCI.

David Kandzari, MD (Piedmont Heart Institute, Atlanta, GA), the lead US interventional investigator for the EXCEL trial comparing CABG surgery versus PCI in patients with left main CAD, said extended follow-up from MAIN-COMPARE is a welcome addition to the field as well as a reflection of current clinical practice. For patients with left main CAD and low-to-intermediate lesion complexity, there is a discussion with the heart team about treatment options with PCI or CABG surgery, he said. As disease complexity increases, the choice for surgery is clear.

“I did a patient last night, my last case of the day, where he had left main disease but the disease was so complex there was no decision other than to recommend bypass surgery for him,” Kandzari told TCTMD. “Having said that, these data, along with EXCEL, should be interpreted as very reassuring of the long-term safety and durability of left main stenting compared with surgery in patients with low-to-intermediate complexity.”

Kevin Bainey, MD (Mazankowski Alberta Heart Institute, Edmonton, Canada), said that at his institution patients with left main CAD are reviewed by interventional cardiologists and surgeons as part of the heart-team approach to care. In those discussions, the SYNTAX score does come into play when making treatment decisions, although a “gestalt-type approach” is often employed, he added.

Like Kandzari, Bainey cited the 10-year follow-up from the SYNTAX trial showing clinical equipoise between CABG and PCI in patients with left main CAD but also said this latest analysis is helpful in that justifies how physicians are currently practicing as a heart team. Most patients with low-to-intermediate anatomic complexity choose PCI over CABG, he told TCTMD. Even so, “it’s quite clear that in the high-SYNTAX-score patients, we do think the cases need to be peer reviewed and we do still also believe that CABG is a better approach for those patients.”


The MAIN-COMPARE registry included 2,240 patients with unprotected left main CAD treated between 2000 and 2006, of whom 660 were excluded from analysis because they did not have a SYNTAX score. Of the remaining 1,580 patients, 897 had a low-to-intermediate SYNTAX score (≤ 32) and 683 had a high SYNTAX score (≥ 33). In the low-to-intermediate and high groups, 71% and 27% of patients were treated with PCI, respectively.

In addition to differences in clinical events, the researchers found that the SYNTAX score had incremental prognostic value on the 10-year risk of outcomes in the PCI-treated patients, but not in those treated with surgery. In the PCI group, a high SYNTAX score was associated with increased risks of death (HR 1.59; 95% CI 1.06-2.38), the composite of death, Q-wave MI, or stroke (HR 1.59; 95% CI 1.08-2.32), and target lesion revascularization (HR 1.57; 95% CI 1.01-2.44) compared with a low score.

Adjusted Risks for 10-Year Outcomes




P Value


    SYNTAX Score ≤ 32

    SYNTAX Score ≥ 33










Death/Q-wave MI/Stroke

    SYNTAX Score ≤ 32

    SYNTAX Score ≥ 33














    SYNTAX Score ≤ 32

    SYNTAX Score ≥ 33








< 0.001

< 0.001

In their paper, Yoon and colleagues note that their results are at odds with EXCEL and NOBLE, the other trial comparing CABG versus PCI in patients with left main CAD. In those trials, “the anatomic SYNTAX score did not have any impact on relative clinical outcomes after PCI and CABG, limiting the clinical utility of the SYNTAX score in left main coronary artery revascularization.” However, as they and Kandzari point out, patients with a large degree of anatomic complexity were excluded from EXCEL and NOBLE.

Bernard Chevalier, MD (Hôpital Privé Jacques Cartier, Massy, France), who wrote an editorial accompanying the study, notes that PCI has evolved over time, with patients treated with better stents and improved dual antiplatelet therapy, among other developments. “Thus, enhanced stenting strategies seem to partly play a role of equalizer in terms of hard safety endpoints in patients with heavy coronary burden,” he writes.

To TCTMD, Kandzari said the 10-year MAIN-COMPARE results are in line with the SYNTAX trial showing that the greater the anatomic complexity, the greater the benefit of bypass surgery. He noted, however, that more and more data indicate that completeness of coronary revascularization is the driver of improved outcomes, and that if physicians can achieve complete revascularization with PCI akin to that achieved with CABG surgery, outcomes would be similar.

“I think that in this case the findings speak not so much to the baseline complexity as a predictor of adverse outcomes, but more to what is not reported in this study, which is residual disease,” he said. “In other words, the incompleteness of revascularization that is more common with PCI than with bypass surgery. This study unfortunately doesn’t address the residual SYNTAX score or the completeness of revascularization.”

Kandzari also noted that a paradox has been observed in interventional cardiology where physicians treat less coronary disease in patients with greater disease burden or anatomic complexity, such as in patients with chronic total occlusions, bifurcations, or calcified lesions. The driver of adverse clinical outcomes in these instances is not so much what gets treated, but what gets left behind, said Kandzari.

MAIN-COMPARE, EXCEL, and Other Trials

Compared with EXCEL, a trial that has been heavily criticized by one former investigator and several professional surgical societies, MAIN-COMPARE was nonrandomized and included a broader range of less selected patients. In EXCEL at 5 years, said Kandzari, the difference in repeat revascularization also favored surgery, but the absolute difference in event rates was approximately 7%. In contrast, the observed difference in target lesion revascularization between PCI and surgery in the low-to-intermediate SYNTAX score patients was roughly 15%.

“So, I think part of this highlights that these are probably more complex patients than were included in the EXCEL trial,” said Kandzari.

He pointed out that the prevalence of diabetes was similar in the PCI and surgery cohorts in MAIN-COMPARE, despite the large body of evidence showing an advantage with CABG surgery in these patients. The FREEDOM trial, a large-scale study showing a lower risk of mortality with surgery in diabetic patients and multivessel CAD, wasn’t yet available when the patients were treated and enrolled in the MAIN-COMPARE registry, however.

Bainey said it can be a challenge to convince patients they are more appropriate candidates for surgery. For those with complex anatomy, though, a discussion with surgeon and interventional cardiologist usually helps convince them that surgery is the better approach. In terms of potential for conflict between surgeons and interventionalists—some of which has been on full display on social media in light of the EXCEL controversy—long-term data from MAIN-COMPARE help reinforce the benefits of PCI in selected patients.

“It’s powerful data for us, and it actually compounds with other data showing that what we’re doing makes sense,” said Bainey.

In addition to anatomic complexity, patient comorbidities and age factor into treatment decisions, which can sometimes lead to differing opinions between surgeons and interventionalists on the best treatment option. “The comorbidities may be prohibitive for surgery and it falls upon us to treat,” said Bainey. “The SYNTAX score is great in patients where you’re not taking into consideration other clinical parameters. It’s great if the patient is 65 and doesn’t have much in the way of other comorbidities, but a high SYNTAX score in an 85-year-old with renal failure and bad lungs—that patient is not going to go for CABG.”

To TCTMD, Kandzari said rates of left main PCI have remained low and relatively stagnant over the last couple of years despite the positive data from EXCEL at 3 and 5 years. He added that he doesn’t think the controversy around the trial dampened enthusiasm for PCI in appropriately selected patients, noting that the debate around EXCEL emerged only in the last few months.   

Michael O’Riordan is the Managing Editor for TCTMD. He completed his undergraduate degrees at Queen’s University in Kingston, ON, and…

Read Full Bio
  • Yoon and Bainey report no relevant conflicts of interest.
  • Kandzari reports institutional grant support from Abbott Vascular, Boston Scientific, Medtronic, Biotronik, Medinol, Orbus Neich, and CSI; and honoraria from Medtronic, Biotronik, and CSI.
  • Chevalier reports consulting for Biotronik, Medtronic, and Terumo; and holding shares in CERC.