Long-term FREEDOM: At Nearly 8 years, CABG Maintains Mortality Benefit Over PCI

CHICAGO, IL—Extended follow-up of patients with type 2 diabetes and multivessel disease who participated in the FREEDOM trial shows that coronary artery bypass graft surgery retains a significant mortality advantage when compared with PCI.

After a median follow-up of 7.5 years, the data showed that CABG remained superior to PCI with a drug-eluting stent, with those undergoing PCI at a 36% increased risk of all-cause mortality when compared with surgical patients, reported senior investigator Valentin Fuster, MD, PhD (Icahn School of Medicine at Mount Sinai, New York, NY), here at the American Heart Association 2018 Scientific Sessions.

The landmark FREEDOM trial—first presented 6 years ago—showed that revascularization with surgery for patients with diabetes and multivessel coronary artery disease was associated with a reduced risk of all-cause mortality, MI, and stroke when compared with PCI. The median time to follow-up in FREEDOM was just 3.8 years, however.

“Follow-up beyond 5 years in coronary revascularization trials is unusual owing to a lack of funding and logistical obstacles,” said Fuster during a presentation at the late-breaking clinical trial session. Considering the long-term nature of atherosclerotic cardiovascular disease, the initial FREEDOM trial “was a relatively short-term study and longer-term follow-up was needed for a better understanding of the comparative benefit of CABG surgery, specifically on mortality,” he said.

Roughly Half of Patients Included in Follow-up

After the study was completed in 2012, patients and centers were invited to participate in the FREEDOM Follow-On Study. Twenty-five of the original 140 hospitals agreed to participate, leading to long-term follow-up of 943 patients. Fuster stressed the inclusion of just 49.6% of the original 1,900 patients in the trial does not imply that more than 50% of patients were lost to follow-up but instead reflects the number of patients included from centers who participated in the extension.

After 3.8 years, all-cause mortality was 16.3% and 10.9% in the PCI and CABG arms, respectively (P = 0.05). After 7.5 years of follow-up in the patients enrolled in the extension study, all-cause mortality in the two arms was 23.7% and 18.7%, respectively (HR 1.32; 95% CI 0.97-1.78). In the entire FREEDOM cohort, which reflects patients followed for just 3.8 years and those in extended follow-up, mortality was 24.3% in the PCI group and 18.3% in the surgical arm (HR 1.36; 95% CI 1.07-1.74).

There were trends favoring CABG surgery in nearly every subgroup. “One thing that was interesting was that the younger patients did much better [with CABG] than the older ones,” said Fuster. “This was really very significant.” For patients 63.3 years and younger, all-cause mortality at 7.5 years was 10.2% with CABG surgery and 20.7% with PCI. Comparatively for those 63.3 years and older, the rate of all-cause mortality was 27.6% with CABG and 26.3% with PCI (P = 0.01 for interaction by age).

Fuster noted that there are limitations to the analysis, which was published simultaneously online in Journal of the American College of Cardiology. In addition to studying just half of the patients in extended follow-up, Fuster said it also doesn’t involve the newer drug-eluting stents that have been developed since the FREEDOM trial. However, Fuster said he is unaware of any new stent technology capable of bridging the gap between PCI and CABG at the present time given the 6% absolute difference in all-cause mortality. The medical therapy landscape for diabetes has also changed since FREEDOM, with the introduction of several agents capable of reducing cardiovascular outcomes. How this medical therapy advance might impact differences between PCI and surgery is unknown, said Fuster. 

‘Fragility’ of Mortality Endpoint 

Speaking during the late-breaking session, Alice Jacobs, MD (Boston University School of Medicine, MA), noted that 1 million coronary revascularization procedures are performed each year, with 35% done in patients with diabetes. Jacobs praised the FREEDOM investigators for adding new longer-term follow-up data to the evidence base comparing PCI versus CABG. She noted, however, that despite the randomized nature of the trial, bias is possibly introduced with just under half of patients followed. The follow-up study is also likely underpowered.

Nonetheless, the long-term data are consistent with the overall findings, said Jacobs. In terms of applicability, the results support clinical guidelines, she added, referring specifically to the 2014 focused update from several US societies that stated CABG is preferred to PCI in patients with diabetes and multivessel coronary disease if the patient is a good candidate for surgery (class I recommendation, level of evidence B). In Europe, CABG is also a class I recommendation (level of evidence A) in diabetic patients with three-vessel disease regardless of SYNTAX score.    

In an editorial accompanying the study in the Journal of the American College of Cardiology, Sripal Bangalore, MD (New York University School of Medicine, NY), and Marco Zenati, MD (Brigham and Women’s Hospital, Boston, MA), questioned the “fragility" or robustness of the mortality benefit in FREEDOM. For example, the point out that just 16 more deaths in the CABG arm would have rendered the between-group difference statistically nonsignificant.

Additionally, Bangalore and Zenati highlight improvements in guideline-directed medical therapy and stent technology, both of which have the potential to affect the outcomes in the trial. They suggest that data from clinical trials and registries have shown reductions in death and/or MI with new-generation drug-eluting stents, mainly driven by decreases in the risk of stent thrombosis and restenosis.

“It is foolish to ignore the results from the FREEDOM and FREEDOM follow-on trials,” write Bangalore and Zenati. “Equally foolish is to ignore the wealth of data supporting clear impact on death and/or MI with improvements in medical therapies and stent technology.”   

The editorialists state the optimal management of patients with diabetes and multivessel disease should be individualized using the heart team, which would include an assessment of surgical risk and late benefits, ability to completely revascularize with PCI, compliance with medical therapy, and patient preference.