Restricting percutaneous interventions (PCI) to only those stenotic lesions that are ischemic by fractional flow reserve (FFR) thresholds is associated with better 5-year outcomes whether or not PCI is deployed, according to a cohort study presented at the American Heart Association scientific sessions.
For those that met the FFR threshold for ischemia, defined as up to 0.80, PCI reduced the risk of a major adverse cardiac event (MACE) at 5 years by 23% (hazard ratio, 0.77) relative to no PCI. Conversely, those not indicated for PCI because of a higher FFR had a 37% higher risk of MACE (HR, 1.37) at 5 years if treated with PCI relative to those who were not.
“The story of overuse of PCI is important,” reported the senior author Dennis Ko, MD, a scientist affiliated with the Schulich Heart Research Program, Sunnybrook Research Institute, University of Toronto, Canada. “We as interventionalists often think that putting in a stent is not harmful, and that turned out not to be the case.”
The FFR threshold for intervening with PCI is evidence based. Several trials, including one published in 2014, have associated PCI with better outcomes relative to medical therapy when FFR is 0.80 or lower. Other evidence suggests no advantage and possible harm for PCI performed if FFR is higher. Multiple guidelines, including those from the AHA, recommend against PCI if FFR is more than 0.80.
“As FRR is gaining in popularity, we were interested in whether physicians follow the thresholds in routine clinical practice and what happens to patient outcomes [if they are or are not followed],” Dr. Ko explained.
In this retrospective study by Dr. Ko’s trainee, Maneesh Sud, MD, and simultaneously published in JAMA, the answer was that there is deviation, and deviation leads to bad outcomes.
The 9,106 coronary artery disease patients included in the study underwent single-vessel FFR assessment within a 5-year period in Canada. The two cohorts evaluated were those with a lesional FFR of 0.80 or less, defined as ischemic, and those with a lesion with higher FFR, defined as nonischemic. The primary MACE outcome comprised death, myocardial infarction, unstable angina, or urgent coronary revascularization.
Of the 2,693 patients who met the FFR threshold of ischemia, 75.3% received PCI, and 24.7% were treated with medical therapy only. Of the 6,413 patients with nonischemic FFR, 87.4% were treated with medical therapy and 12.6% received PCI.
In those with ischemic FFR, event curves for MACE separated rapidly. At 30 days, the risk of MACE was 53% lower (HR, 0.47) in those receiving PCI. By 1 year, the advantage was less (HR, 0.76), but it was steady thereafter and remained about the same at 5 years (HR, 0.77; 95% confidence interval, 0.63-0.94). Relative advantages for each component of MACE went in the same direction. At 5 years, PCI exerted its greatest numerical advantage for the outcome or urgent coronary revascularization (HR, 0.71) and its least numerical advantage for MI (HR, 0.92), but none of these differences reached statistical significance.
In those with nonischemic coronary lesions on FFR, PCI was associated with more than twice the risk for MACE at 30 days (HR, 2.11), but the increase in risk relative to medical management fell at 1 year (HR 1.67) and 5 years (HR, 1.37). All of the individual components of MACE were numerically increased at all time points except for death, which was numerically lower at 30 days (HR, 0.41) and 5 years (HR, 0.94), even though these differences were not significant.
It could not be ascertained from these data why PCI was not performed when there was an indication or why it was performed when there was not. The investigators speculated that some clinicians may decide against PCI for ischemic lesions in the absence of symptoms or when concerned about comorbidities. They might offer PCI in nonischemic lesions because of symptoms, positive tests other than FFR, or FFR values near the threshold.
“I think the main message of our paper is that adherence of the FFR threshold as established by clinical trials is important,” Dr. Ko said in an interview. This not only means performing PCI when it is indicated, but refraining from PCI when it is not.
Basically, this study confirms that the guideline thresholds are valid, according to Jared M. O’Leary, MD, who is experienced with FFR and is Medical Director for Quality at the Vanderbilt Heart and Vascular Institute, Nashville, Tenn.
“It confirms the utility of FFR in the real world,” he said, adding that the results are “totally consistent with our practice.” He called FFR “an important tool in the cardiac cath lab” not only for determining when revascularization will benefit the patient but the opposite.
“The flip side is also true: Stenting should be avoided if a negative FFR is obtained,” he said, calling this technique “particularly useful for lesions that appear borderline by visual estimation alone.”
SOURCE: Sud M et al. AHA 2020. JAMA. 2020 Nov 13. doi: 10.1001/jama.2020.22708.