Bridging the PCI-CABG Complex
The SYNTAX trial cracked open the door for the use of PCI as a treatment option for select patients with complex coronary artery disease (CAD). Since the trial's enrollment closed, refinements in technologies and techniques continue to nudge the door wider for PCI at the same time that longer-term data from SYNTAX shine favorably on CABG for higher-risk patients. But the verdict is still out on how best to treat patients with multivessel or left main coronary disease if they are not clear-cut candidates for either procedure.PCI ventures into left main

"The results of the SYNTAX trial were more positive for left main disease and somewhat less positive for complex triple-vessel disease than were anticipated [for PCI]," says Gregg W. Stone, MD, director of the medical research and education division at the Cardiovascular Research Foundation and director of cardiovascular research and education at the Center for Interventional Vascular Therapy at New York-Presbyterian Hospital/Columbia University Medical Center in New York City. "From SYNTAX, left main PCI in particular got out of Class III 'stent jail' and went to Class IIA/IIB in the guidelines."

SYNTAX compared PCI using paclitaxel-eluting stents (Taxus, Boston Scientific) with CABG surgery in 1,800 patients with untreated left main CAD, three-vessel disease or a combination of both diseases in a randomized trial (N Engl J Med 2009;360:961-972). First-year results showed less invasive PCI with a higher rate of major adverse cardiac or cerebrovascular events (MACCE) of 17.8 percent vs. 12.4 percent for CABG, prompting investigators to conclude that CABG should remain the standard of care. But each method or approach had its drawbacks: in the case of PCI, a revascularization rate of 13.5 percent compared with 5.9 percent for CABG; and in CABG, a higher rate of stroke, at 2.2 percent vs. 0.6 percent for PCI.

The picture has changed over time though, particularly when the SYNTAX score entered the equation. The SYNTAX score assesses CAD in a calculation that includes the location, extent and severity of coronary stenosis, with scores ranked as low (22 or below), intermediate (between 23 and 32) and high (33 and above). Three-year SYNTAX results presented at the 2010 European Association of Cardiothoracic Surgery meeting in Geneva gave a nod to PCI, showing that while overall MACCE rates remained higher in the PCI group, cumulative MACCE rates for patients with a low score were nearly identical in both the PCI and CABG groups.

Four-year results, unveiled at TCT.11 in San Francisco, showed a cumulative MACCE rate for the low SYNTAX score left main subset of 26 percent for PCI and 28.4 percent for CABG. When broken down by complexity of the disease, though, PCI appeared to benefit patients with left main only and left main plus single-vessel diseases but not patients with more complex disease states.

"In isolated left main disease and low SYNTAX score left main disease, the outcomes appear equivalent with CABG and PCI at four years," says Michael J. Mack, MD, medical director of cardiovascular surgery at Baylor Health Care System in Dallas, and a member of the writing panel for the American College of Cardiology (ACC) and American Heart Association (AHA) 2011 CABG guidelines. In a collaboration with the 2011 PCI guideline writers, the authors of each publication offered identical recommendations in sections dealing with PCI and CABG to ensure interventional cardiologists and cardiac surgeons receive the same guidance.

"That is what drove the guidelines to change in terms of upgrading PCI in left main disease," says Mack, who also is president of the Society of Thoracic Surgeons (STS). CABG to improve survival in significant left main CAD received a Class I level recommendation in the 2011 guidelines. PCI was elevated from a Class III "no benefit/harm" designation to a Class II "may be considered" ranking and was deemed an acceptable alternative to CABG for left main disease patients with a low SYNTAX score (J Am Coll Cardiol 2011;58;e44-e122).

"In left main with multivessel disease or SYNTAX scores higher than 33, there is a mortality benefit to surgery," Mack adds. "It is not appropriate to use PCI in those patients." CABG also was associated with better outcomes for patients with diabetes mellitus.

The multivessel milieu

The use of PCI has outpaced CABG in recent years. An analysis of coronary revascularizations in the U.S. concluded that the annual CABG surgery rate had declined 38 percent between 2001 and 2008, while the PCI annual rate remained unchanged (JAMA 2011;305:1769-1776). In Canada, the Cardiac Care Network of Ontario's Variations in Revascularization Practice in Ontario Working Group reported that the ratio of PCI to CABG had increased from 1.6 in 2001 to 2.7 in 2007 in a study on index cardiac catheterization practices in Ontario (CMAJ December 2011, online).

The Ontario authors noted that improvements in stents and greater operator experience led some to expand from treating single-vessel disease to more complex diseases with PCI. "As interventionalists become more confident in their abilities to handle complex anatomies, they are increasingly advocating for angioplasty in patients who 10 years ago almost certainly would have undergone bypass surgery," says Jack V. Tu, MD, PhD, of the Institute for Clinical Evaluative Sciences in Toronto, and the study's lead author.

The researchers found that the majority of patients with single-vessel disease underwent PCI and patients with left main disease most often underwent CABG. Based on SYNTAX, CABG was a better option for certain patients with multivessel disease because it offered better longer-term outcomes.

Instead, there was a great deal of variability among hospitals, the authors found, particularly in treatment for patients with non-emergent multivessel disease. For instance, for three-vessel disease, the percentage of patients who underwent PCI ranged between 22.3 percent and 49.6 percent.

"Some variation is expected," Tu points out, due to the varied skill levels among interventionalists and surgeons. But the ratio across individual hospitals varied fivefold, he noted, from 1.3 to 6.1. "In our study, there was a huge amount of variation, and you wouldn't expect that within a universal healthcare system where, in theory, everyone should be using the same guidelines and the care provided across centers should be somewhat similar."

SYNTAX MACCE to Four Years - 30.84 Kb
LM-left main; MACCE=major adverse cardiac or cerebrovascular events
Source: Patrick W. Serruys, MD, PhD, Erasmus Medical Center

Going beyond SYNTAX

Drawing conclusions about SYNTAX subgroups is problematic, Mack and Stone agree. The SYNTAX trial failed to reach its primary endpoint of noninferiority of PCI, making all findings on subanalyses hypothesis-generating. Additionally, SYNTAX results didn't reflect advancements used in current practice such as everolimus-eluting stents, which have a lower restenosis rate than the paclitaxel-eluting stents; assessment and imaging techniques such as fractional flow reserve (FFR) and intravascular ultrasound (IVUS), which have allowed for more targeted PCI; and pharmacological approaches such as bivalirudin (Angiomax, The Medicines Company) and antiplatelet agents to reduce stent thrombosis.

"We are doing smarter PCI and we are getting better procedural results because of improved devices, drugs, FFR and imaging," Stone says. "This begs the question: If the SYNTAX trial were done today, would the results be more positive for PCI?"

The SYNTAX trial was underpowered to definitively compare PCI and CABG solely for left main disease or three-vessel disease. In an effort to bolster evidence for PCI, Stone and colleagues conducted a meta-analysis to determine the safety and efficacy of PCI in left main disease. They reported no significant differences in one-year MACCE rates and endpoints of death or MI, but found PCI had a higher rate of target vessel revascularization and a lower rate of stroke compared with CABG. In their conclusion, they argued that the left main PCI guidelines should be revised (J Am Coll Cardiol 2011;58:1426-1432).

In the meantime, Stone is leading a trial that is seeking to settle the PCI-CABG debate in the left main disease population. The EXCEL trial, sponsored by Abbott Vascular, will evaluate use of Abbott's Xience stents vs. CABG for left main revascularization in 2,600 randomized patients with left main disease and a SYNTAX score of 32 or greater. The primary endpoint is a composite of all-cause mortality, MI or a stroke. The study was designed by both interventional cardiologists and cardiac surgeons.

Bigger, more contemporary

Surgeons in EXCEL will be encouraged to use off-pump surgery and bilateral internal mammary artery grafts, according to Mack. "Patients at risk for stroke and kidney failure have better outcomes with off-pump surgery, and a significant number of patients in EXCEL will be riskier patients," Mack says. "It is an attempt to have the best care approach from a surgical standpoint."

For interventionalists, Stone adds, EXCEL specifies the use of the Xience stent, FFR and IVUS protocols and pharmacological approaches. "Our goal is to determine whether contemporary PCI is really as good or better than CABG, for most patients other than those with the most complex coronary artery disease," he says. "Equal steps also have been taken to ensure the best contemporary results in surgery are obtained, so this should be a fair comparison."

To address multivessel disease, the ACC and the STS are collaborating on a study using their clinical registries combined with the Centers for Medicare & Medicaid Services (CMS) file data to compare PCI with CABG in patients with multivessel CAD. The National Heart, Lung and Blood Institute-funded ASCERT study will track outcomes in almost 190,000 patients who received CABG or PCI between 2004 and 2007, giving investigators sufficient power to evaluate subgroups (J Am Coll Cardiol Intv 2010;3:124-126).

"SYNTAX made a good attempt to look at real world patients," says Fred H. Edwards, MD, medical director of cardiothoracic surgery at University of Florida/Shands in Jacksonville, Fla., and a member of the ASCERT steering committee. "We can do that on a broader scale, with patients from across the U.S., and at a variety of hospitals with this trial."

The ASCERT study team plans to present results on long-term survival for patients with two- and three-vessel disease at the ACC.12 meeting in Chicago in March, according to Edwards.

Partnering in patient care

In their study of Ontario practices, Tu and colleagues found that fewer than 4 percent of patients were discussed in case conferences involving both interventionalists and cardiac surgeons in the participating 17 hospitals. They reported a strong association between the treatment recommended by the cardiologist and the treatment the patient received. An interventional cardiologist was more likely than a non-interventional cardiologist to proceed directly to PCI after the index catheterization, and a patient whose index catheterization was performed by an interventional cardiologist was 40 percent more likely to undergo PCI.

"There wasn't necessarily an informed discussion with all the parties involved," Tu said.

Tu and colleagues recommended that hospitals adopt a heart team strategy for patients who could undergo either PCI or CABG. The heart team approach, developed in the SYNTAX trial, brings interventional cardiologists and cardiac surgeons together to review patient cases, agree on a reasonable treatment and then discuss options with the patient. The heart team concept is endorsed in the PCI and CABG guidelines for treating patients with unprotected left main or complex CAD.

"It is having a meeting of the minds, looking at all the subtleties of a particular patient and a particular disease to determine the best treatment for this patient," says Mack. "Not all left main disease is the same, not all multivessel disease is the same and not all patients are the same."

The heart team approach helps minimize bias, Stone says, which gives patients a more accurate understanding of pros and cons for each procedure.

The outlier hospitals in the Ontario study have taken the advice to heart after getting their data, Tu says. Anecdotally, some have launched a heart team approach and are encouraging specialists to work together. The Ontario study group plans to continue their analyses to track any changes in practice.

"We expect the variation to shrink between the low and high centers over time as people become more aware of how they stand," Tu predicted. "It was certainly an eye opener for some centers."

The legacy of SYNTAX includes not only pivotal findings and a foundation for further research, but also validation of the heart team approach, interventional cardiologists and cardiac surgeons agree. The team-based collaboration, combined with current and future evidence about treatments for left main disease and multivessel disease, will help guide care for patients with complex coronary diseases.