Coronary revascularization guided by instantaneous wave-free ratio compared with fractional flow reserve: pooled 5-year mortality in the DEFINE-FLAIR and iFR-SWEDEHEART trials

Abstract

The DEFINE-FLAIR1 and iFR-SWEDEHEART2 clinical trials compared non-hyperaemic (resting) instantaneous wave-free ratio (iFR™)–guided management vs. hyperaemic fractional flow reserve (FFR)-guided management of patients being considered for coronary revascularization. The primary composite outcome of death, myocardial infarction, and urgent revascularization at 12 months and the non-inferiority designs were consistent across both trials. Overall, an iFR-guided strategy reduced revascularization, and the primary endpoint results of both trials met the non-inferiority criteria. Based on these results, the use of iFR or FFR was recommended in European3 and North American4 clinical practice guidelines (Class I, Level of Evidence A). The guidelines of the European Society of Cardiology noted the limited follow-up period of only 1 year, identifying an evidence gap for the prognostic role of iFR that required clarification.3 In a pooled analysis5 of death and myocardial infarction in the DEFINE-FLAIR and iFR-SWEDEHEART trials, there were 160 deaths or myocardial infarction events in 4345 participants during the 12 months after randomization. Of these, 90 events occurred in the iFR group (n = 2159), and 70 events occurred in the FFR group (n = 2186) {hazard ratio [HR] [95% confidence interval (CI)] 1.30 [0.96, 1.77]; P = .09}. This excess occurred despite fewer participants with primary outcome data in the iFR groups due to more withdrawals. Furthermore, at 2 years, the HR for mortality was two-fold higher in the iFR group [2.02 (1.24, 3.31); P = .01].6 The 5-year results of the iFR-SWEDEHEART trial have been published,7 and the 5-year outcomes of the DEFINE-FLAIR trial were reported at the EuroPCR.2023 conference.8 Therefore, we performed a pooled analysis at 5 years of these randomized controlled trials. The methods conformed to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (2009). The principal summary measure was the risk ratio (95% CI) and P-value, which were calculated for each individual study. Meta-analysis estimates were calculated from a random effects model using the restricted maximum likelihood method. Fixed effects analyses using the Cochrane–Mantel–Haenzel method produced near identical results (not shown). I2 was used to measure the consistency of the meta-analysis. The analysis was conducted using R (version 4.2.0) using the metaphor (https://CRAN.R-project.org/package=metafor) and rmeta (https://CRAN.R-project.org/package=rmeta) packages. The SWEDEHEART trial randomized 2037 participants (mean age 67 years, 25% women, and 22% with non-ST-elevation myocardial infarction); 1019 were assigned to the iFR group, and 1018 were assigned to the FFR group. Follow-up data were available for 1012 individuals in the iFR group and 1007 individuals in the FFR group (Figure 1). The DEFINE-FLAIR trial randomized 2492 participants (mean age 65 years, 24% women, and 15% with an acute coronary syndrome); 1242 were assigned to the iFR group and 1250 were assigned to the FFR group. Follow-up data were available for 1000 individuals in the iFR group and 1019 individuals in the FFR group. Forest plots of clinical outcomes at 5 years in the iFR-SWEDEHEART7 and DEFINE-FLAIR8 clinical trials. In the combined trials, after a follow-up period of 5 years, iFR-guided management was associated with a significant excess of death with a risk ratio of 1.34 (95% CIs 1.08, 1.65; P = .007) (Figure 1). There was directional consistency for the excess in deaths in the iFR groups in both trials. Considering the subgroups of death by cause in relation to the risk ratios (95% CI) for iFR- vs. FFR-guided management (Figure 1), there was also directional consistency in the increased risks for cardiovascular death [1.24 (0.61, 2.50); P = .551], confirmed non-cardiovascular death [1.27 (0.95, 1.68); P = .103], and undetermined cause of death [1.62 (1.06, 2.48); P = .025], with the latter being statistically significant. There were no differences between groups in the proportions of patients with unplanned coronary revascularization or non-fatal myocardial infarction (Figure 1). In this pooled analysis of two international clinical trials of iFR-guided vs. FFR-guided invasive management in patients with coronary heart disease, iFR-guided management was associated with an increase in the risk of death at 5 years, consistent with the increased risk of death previously reported at 2 years.6 In contemporary guidelines, the use of iFR or FFR is currently recommended (Class I, Level of Evidence A) to guide myocardial revascularization decisions.3,4 This recommendation is based on non-inferiority criteria being met for the primary composite outcome of death, myocardial infarction, and urgent revascularization at 12 months. However, the numerically dominant component of the primary outcome was unplanned revascularization, which was an investigator decision. The excess in spontaneous death and myocardial infarction at 12 months underpinned the case for further research in line with the guideline committee statement.3 The observed excess in deaths at 5 years with iFR-guided management is unlikely to be a chance finding as the trials were independently conducted, the signal was apparent at 1 and 2 years, and the 5-year results are consistent between the trials. Considering the subgroups of death by cause in relation to the risk ratios (95% CI) for iFR- vs. FFR-guided management (Figure 1), there was a directional consistency in the increased risk for death in each of the subgroups of death by cause. Given the limited number of deaths within these subgroup analyses, CIs around the point estimate of each risk ratio are wide. On the other hand, the increased risk of all-cause death is clear in terms of the magnitude (+34%) and the statistical significance. Taking all points together, the increased risk of death at 5 years associated with iFR-guided management is likely to be a true effect. In ∼20% of patients, a discrepant result occurs between iFR and FFR, leading to different treatment decisions. Mechanisms of harm may include incomplete revascularization based on deferred management in patients with multi-vessel coronary artery disease or revascularization of coronary arteries with non-flow limiting atherosclerosis. The proportions of myocardial infarction events were low in both studies, and ∼20% of the participants had missing data for death at 5 years, more so in the iFR groups; hence the HR may under-estimate the true effect.9 In conclusion, compared with FFR-guided management, iFR-guided management is associated with an increased risk of death at 5 years. A patient-level pooled analysis is warranted to better understand the mechanisms. These new results are relevant to patients, clinicians, and guideline committees. C.B. is employed by the University of Glasgow, which holds consultancy and research agreements for his work with AstraZeneca, Abbott Vascular, Boehringer Ingelheim, Coroventis, HeartFlow, Menarini, Merck, Novartis, Servier, Siemens Healthcare, TherOx, Inc, and Valo Health. These companies were not involved in any aspect of this manuscript. Research funding—British Heart Foundation (reference RE/18/634217). J.D.M.: Research grant—none; other research support—none; honoraria—none; expert witness—none; ownership interest—none. K.G.O.: Research grant—none; other research support—none; honoraria—none; expert witness—none; ownership interest—none; stock options—Biosensors International; full-time employee of Biosensors International. No data were generated or analysed for or in support of this paper. Research funding—British Heart Foundation (reference RE/18/634217). Ethical approval was not required. Not applicable—this is a meta-analysis of previously registered trials.

Topics & Concepts