FAME 3: PCI does not meet noninferiority to CABG in patients with three-vessel disease.

By Wally Omar on

Key Points:

    • All-comers with three vessel disease were randomized to undergo FFR-guided PCI or Coronary Artery Bypass Grafting in this multicenter, international randomized control trial.
    • FFR-guided PCI failed to meet noninferiority for the primary endpoint of MACCE (death, MI, stroke or repeat revascularization) at 1 year. There was no difference in the secondary endpoints, including those of procedural MI and stroke.
    • A subgroup analysis of patients with very low SYNTAX scores showed this to be the only group in which PCI is non-inferior to CABG.

Prior studies assessing patients with three-vessel disease have found coronary artery bypass grafting (CABG) to be superior to percutaneous coronary intervention (PCI) for long-term outcomes. The SYNTAX trial, for example, showed significantly higher rates of MACCE in patients undergoing CABG. Many of these studies, however, were performed when bare-metal and first-generation drug-eluting stents were in the market. The investigators

for FAME 3, therefore, sought to answer whether current generation drug eluting stents would hold up against CABG in patients with three-vessel disease, with the added benefit of using FFR to help guide necessity of PCI. The results of the trial were presented at a late-breaking session of the 33rd annual TCT scientific symposium by Dr. William Fearon (Stanford University). In this multi-center, international study, all comers with three-vessel disease, not involving the left main, were randomized to either receive FFR-guided PCI, in which all lesions with an FFR <0.8 were stented, or CABG guided by the coronary angiogram. The primary endpoint was MACCE at 1 year, including all cause death, myocardial infarction stroke or repeat revascularization. Importantly, and to avoid controversy, the definition of procedural MI was constant for both CABG and PCI. Key exclusion criteria were left main involvement, cardiogenic shock, recent STEMI within five days, and LVEF <30%. The study enrolled 1500 patients, approximately 750 to each arm. Patients were mostly white males with an average age of 65 years old. Less than one-third had diabetes, and approximately 40% in each arm presented with acute coronary syndrome. In terms of coronary anatomy, 0% had intermediate SYNTAX scores of 23-32, and more than two-thirds of each arm had bifurcation lesions.

After 1 year, the intention to treat analysis showed a MACCE rate of 10.6% in the PCI arm and 6.9% in the CABG arm (HR 1.5, 95% CI 1.1-2.2, p=0.35 for non-inferiority). This was mostly driven by repeat revascularization (5.9% in the PCI group vs. 3.9% in the CABG group). There was no significant difference in death (1.6% vs. 0.9%), MI (5.2% vs. 3.2%) or stroke (0.9% vs. 1.1%). All studied safety endpoints favored PCI, including BARC 3-5 Bleedin

g, AKI, Atrial fibrillation, and rehospitalization. In a subgroup analysis, participants with low SYNTAX scores (<22) were more likely to have better outcomes with PCI.  “The one-year rate of death, MI, or stroke was not significantly different between the two strategies. However, FFR-guided PCI with a current generation drug-eluting stent performed favorably in comparison with CABG in three-vessel coronary artery disease with less complex disease according to the SYNTAX score,” said Dr. Fearon, before adding “In patients with more complex three-vessel coronary artery disease, CABG remains the treatment of choice.”

In a panel discussion after the presentation, Dr. Roxana Mehran of Icahn School of Medicine Mt. Sinai questioned whether that interpretation could indeed hold up, as any subgroup analyses would have to be considered hypothesis generating with the primary outcome for non-inferiority was not met. The panel agreed that randomized-control trials needed to take place in order to come to definitive conclusions.

For many, this study was the last nail in the coffin for PCI in patients with three-vessel disease. Others are bothered by the low rates of imaging-guided PCI (12%), and question whether imaging-guided PCI rather than physiology-guided PCI would have resulted in better outcomes. These hypotheses will likely be tested in the coming years.

The study was simultaneously published in the New England Journal of Medicine.

Leave a Reply