FREEDOM Sets Us Free—for the Sickest Diabetics

In November 2009, Cardiovascular Business asked if the much-anticipated FREEDOM trial would “set us free” by better informing decisions to treat high-risk diabetics with multivessel disease using either PCI or CABG. The findings demonstrated resounding support for surgery, but some physician experts are concerned that the decisions may overlook the patient’s perspective on risks, such as stroke, and that these results may get translated to lower-risk diabetics.

'Practice-changing trial'

 Source: Elizabeth A. Magnuson, PhD

In real-life clinical practice, one-third of the patients who undergo cardiac catheterization and PCI have diabetes with multivessel disease—similar to the FREEDOM population, according to senior author Valentin Fuster, MD, PhD, director of Mount Sinai Heart in New York City and past-president of the American Heart Association.

In the trial, Fuster et al enrolled 1,900 patients (mean age, 63.1 years, 29 percent women and 83 percent had three-vessel disease) at 140 international centers from 2005 through 2010 (N Engl J Med 2012;367:2375-2384). After five years, the CABG group had a lower combined rate of strokes, MIs and deaths (18.7 percent) than the PCI group (26.6 percent).

“FREEDOM, a well-conducted, practice-changing trial, demonstrated that CABG is the preferred strategy for most diabetics with multivessel disease, particularly if there aren’t simple, straightforward lesions, and a mammary artery can be used to bypass the left anterior descending (LAD) coronary artery,” says interventional cardiologist Peter B. Berger, MD, chairman of cardiology at Geisinger Health System and co-director of the Geisinger Heart & Vascular Institute in Danville, Pa.

“Until this study, there were no clear guidelines to direct physicians for these patients, and it was left to a case-by-case basis,” says Fuster. “Any diabetic patient with significant coronary disease who is scheduled for a cardiac catheterization should be informed about the results favoring CABG, if this is what is found at the time of catheterization.” In the evolving drug-eluting stent (DES) era, the difference between adverse events for CABG vs. DES will not be a greater than 8 percent, he adds.

Stroke, a well-known risk of bypass surgery, occurred more often in the CABG group (5.2 percent) than in the PCI group (2.4 percent). However, more people died from any cause in the PCI group (16.3 percent) than in the CABG group (10.9 percent). The survival advantage of CABG over PCI was consistent regardless of race, gender, number of blocked vessels or disease severity, according to the study authors.

“If you were to look solely at death and MI, CABG is clearly superior in this trial,” says Ajay J. Kirtane, MD, chief academic officer and director of the interventional cardiology fellowship program at NewYork-Presbyterian Hospital and Columbia University Medical Center in New York City. “But the two procedures are clearly different from a patient perspective, particularly with regards to the invasive nature of the two procedures and the increased stroke risk, especially within 30 days.”

Conversely, Fuster calls the risk of stroke “relative,” if a physician considers the number of events with heart attack and mortality that are avoided. “The risk of stroke is minimal, compared with what you gain. Regardless, all the risks and benefits of both procedures need to be addressed with the patient.”

To better assess these risks, this “study reaffirms the heart team strategy,” Kirtane adds. “For the past five years in our practice, if we are presented with a diabetic patient with multivessel disease, interventionalists take the patient off the cath lab table to consult with a surgeon.”

While repeat revascularization drove many of the outcomes, particularly the economic outcomes, of FREEDOM, Berger says “death and stroke are much more important to physicians and patients.” Kirtane adds that the risk of stroke may be even more compelling to some patients than mortality.

Mark A. Hlatky, MD, of Stanford University School of Medicine in Stanford, Calif., wrote in the accompanying editorial that mortality has been consistently reduced by CABG compared with PCI, in more than 4,000 patients with diabetes who have been evaluated in 13 clinical trials (N Engl J Med 2012;367:2437-2438). “The controversy should finally be settled,” he wrote.

However, Hlatky concurred with Kirtane that informed decision-making is necessary for patients and specialists. He wrote there is “little time for informed discussion about alternative treatment options … Well-informed patients might choose any of those options on the basis of their concerns about the various outcomes of treatments, such as survival, stroke, MI, angina and recovery time.”

Hlatky also wrote that these discussions should begin before coronary angiography “to provide enough time for the patient to digest the information, discuss it with family members and members of the heart team and come to an informed decision.”

Yet, Kirtane stresses that interventionalists need to get comfortable taking patients off the cath lab table after diagnostic catheterization to have these conversations, even though that has not been the traditional pathway for interventionalists. Fuster adds that this conversation is a physician’s “ethical responsibility.”

“Interventionalists are there to make a diagnosis first, not necessarily intervene on a patient,” Kirtane says. “To truly assess the patient, you need to perform a diagnostic catheterization.”

The trial’s results have directly impacted the formal policy at Geisinger, where “physicians now—as a matter of policy—take diabetic patients with multivessel [disease] off the cath lab table to consult with a surgeon and discuss the relative risks and benefits of CABG with patients,” Berger says. Also, all diabetic patients are warned before a diagnostic catheterization that if multiple blockages are discovered, they will be taken off the table for additional consultation.

Unfortunately, there isn’t a precise way to mitigate for stroke risk currently. “There are physicians who would perform off-pump surgery or screen patients ahead of time with carotid Doppler, but those methods are exploratory,” Kirtane points out. “It’s not just for stroke risk, but also all the neurocognitive complications that accompany bypass surgery, which lead many patients to opt for stenting first.”  

To better understand which patients are most at risk for stroke, Fuster and his colleagues are further assessing those patients who experienced a stroke in the FREEDOM trial.

*Intention-to-treat population (includes planned staged procedures)
Source: Elizabeth A. Magnuson, PhD

Cost outcomes

Based on the data from the FREEDOM trial, Magnuson et al conducted a cost-effectiveness analysis. The researchers assessed costs from the perspective of the U.S. healthcare system and quality of life, using the EuroQOL. To measure the long-term impact of CABG vs. PCI on life expectancy, quality-adjusted life expectancy and costs, they developed a patient-level microsimulation model. Their goal was to estimate the monthly risk of death based on U.S. lifetables calibrated to the PCI trial population and the observed impact of CABG vs. PCI on five-year mortality. 

Lead author Elizabeth A. Magnuson, PhD, director of health economics and technology assessment at Saint Luke’s Mid America Heart Institute in Kansas City, Mo., acknowledges that both procedures are expensive, especially given the use of DES in the PCI arm, which were assigned a unit cost of $1,500 per stent. The average number of stents used in patients enrolled in FREEDOM and assigned to PCI was 4.1.

Although initial procedural costs were lower for CABG by approximately $3,000, the total costs for the index hospitalization were $8,622 per patient higher for the CABG patients. 

Over the next five years, follow-up costs were higher with PCI, driven by the need for more frequent repeat revascularization and higher outpatient medication costs. Nonetheless, cumulative five-year costs remained $3,641 per patient higher with CABG. Although there were “only modest gains” in survival with CABG during the trial period, when the in-trial results were extended to a lifetime horizon, a relatively large gain in life expectancy was found.

 “With an estimated gain in quality-adjusted life years (QALYs) of 0.663, the cost-effectiveness ratio was around $8,100, which is highly favorable,” says Magnuson. “Though CABG was not less expensive over the long term, the cost per incremental gain in QALY was very reasonable, well below the commonly used U.S. benchmark for cost effectiveness of $50,000, not only in our base case analysis, but across a broad range of assumptions regarding the effect of CABG on post-trial survival and costs.”

Based on their economic findings, the researchers reported that CABG was projected to be “economically attractive” relative to PCI with DES.

“Our findings reaffirm the clinical findings from FREEDOM, and reinforce the guidelines that recommend CABG as the preferred revascularization strategy for diabetic patients with multivessel disease,” Magnuson says.

Personalizing care

In the 2009 Cardiovascular Business article, Berger had expressed concern that the results of this trial might get translated into less sick diabetic patients in real-life practice, even though these patients were not evaluated in FREEDOM. “Now that the results have emerged, my concerns are doubly reinforced.”

However, Fuster suggests that the majority of diabetics who are appropriate to undergo CABG or PCI have multivessel disease, and diabetics with single-vessel disease are “exceptions.”

Berger highlights a few cases of diabetics with multivessel disease when the benefits of CABG are “markedly reduced,” such as when a left internal mammary artery bypass graft cannot be placed to the LAD artery.

“If a diabetic or nondiabetic has lesions in the circumflex and/or right coronary artery and/or the left main artery does not require bypass, the evidence does not support the FREEDOM strategy,” he adds. “Furthermore, even if a patient’s disease is in the LAD artery, but the distal vessel is not suitable for bypass with the mammary artery, I wouldn’t apply the results of FREEDOM in such a patient.”

Another area of uncertainty in the post-FREEDOM era is how diabetic patients with multivessel disease fared if their diseased lesions were straightforward and simple. The FREEDOM researchers are now reviewing these patients in the trial.

To address these issues, Fuster speaks to the need to individualize revascularization care. “If I have a patient in his or her 80s or 90s, and his or her main symptom is angina, [then] stenting would likely produce a better quality of life,” Fuster says. “Although the study’s results were remarkable, it doesn’t mean that all diabetic patients should undergo surgery.”

Berger sees these types of personalized care discussions as a way to shape a more informed and transparent patient consent process, where physicians and patients can come to a mutual decision on the best treatment option for the individual.