FFR Versus IVUS with Angiography-Derived FFR for Clinical Outcomes in Patients with Coronary Artery Disease

1. Hypothesis: The IVUS-guided stent implantation after angiography-derived FFR-based decision-making will show superiority in terms of a lower rate of patients-oriented composite outcomes (POCO) at 24 months after randomization compared with the FFR-guided PCI strategy in patients with coronary stenosis.

2. Research materials and indication for revascularization:

   2.1 Experimental group: PCI will be performed if angiography-derived FFR ≤0.80 and will be deferred if angiography-derived FFR >0.80; If PCI is performed, PCI optimization using IVUS will be performed following the recommended criteria: ① Plaque burden at stent edge ≤55%; ② Minimal stent area ≥ 5.5 mm2, or minimal stent area ≥ distal reference lumen area.

   2.2 Control group: PCI will be performed if FFR ≤0.80 and will be deferred if FFR >0.80; If PCI is performed, PCI optimization using FFR will be performed following the recommended criteria: ① Post-PCI FFR ≥ 0.88, or ② Post-PCI ΔFFR ([FFR at stent distal edge] - [FFR at stent proximal edge]) < 0.05.

3. Sample size: In the post-hoc analysis of the FLAVOUR I study applying QFR analysis, the 2-year POCO rate was 13.0% in the PCI group with FFR ≤0.80 and undergoing FFR-based PCI optimization and 7.1% in the PCI group with QFR ≤0.80 and undergoing IVUS-based PCI optimization. Meanwhile, the 2-year POCO rate was 5.8% and 6.5% in the deferral of PCI group with FFR >0.80 and QFR >0.80, respectively. Assuming a PCI rate of 70% in patients with coronary artery lesions with 50-90% stenosis that is the inclusion criteria for the current study, and considering event rates from historical studies evaluating FFR- and QFR-guided PCI strategies, the cumulative incidence rate of POCO at 24 months was estimated to be 13.0% in the control group (FFR group) and 9.0% in the experimental group (QFR-IVUS group).

   • Primary endpoint: POCO, defined as a composite of death from any cause, MI, or any revascularization at 24 months after randomization.

   • Design: superiority

   • Sampling ratio: experimental group : control group = 1:1

   • Type I error (α): One-sided 2.5%

   • Accrual time: 24 months

   • Total time: 4 years (accrual 24 months + follow-up 24 months)

   • Assumption: POCO 13.0% vs. 9.0% in control or experimental group, respectively

   • Statistical power (1- β): 90%

   • Primary statistical method: Kaplan-Meier survival analysis with log-rank test

   • Estimated attrition rate: total 10%

   • Stratification in randomization: Presence of diabetes mellitus

Based on the above assumption, we would need total 1,942 patients (971 patients in each group) with consideration of an attrition rate.