Angiography-Based Physiology to Guide Coronary Revascularization (FAST III)

Authors, Journal, Affiliations, Type, DOI

• Authors: Joost Daemen, Jari A. van der Eijk, Marco Barbierato, Robert A. Byrne, et al.; FAST III Investigators
• Journal: New England Journal of Medicine
• Lead institution: Erasmus University Medical Center, Rotterdam (co-ordinating); 37 sites across 7 European countries
• Type: Investigator-initiated, international, open-label, randomized, controlled, noninferiority trial
• Regulatory sponsor: European Cardiovascular Research Institute (unrestricted grants from Pie Medical Imaging and Siemens Healthineers)
• DOI: https://doi.org/10.1056/NEJMoa2601841
• Published: March 29, 2026

Overview

The FAST III trial randomized 2,235 patients with intermediate coronary-artery lesions (30–80% diameter stenosis) across 37 European centers to revascularization guided by either vessel fractional flow reserve (vFFR) — derived from 3D quantitative coronary angiography without a pressure wire or adenosine — or standard pressure-wire–based FFR. At 1 year, the composite of death, myocardial infarction, or revascularization occurred in 7.5% of patients in both groups (risk difference −0.02 pp; 95% CI −2.25 to 2.21; P=0.004 for noninferiority), establishing vFFR-guided revascularization as noninferior. vFFR identified functionally significant lesions at a higher rate (40.9% vs 31.3%) and resulted in more revascularizations (45.0% vs 36.0%), yet outcomes were equivalent; procedure time was ~5 minutes shorter, and intraprocedural complications were lower with vFFR (3.7% vs 6.0%).

Keywords

Vessel fractional flow reserve, vFFR, fractional flow reserve, FFR, angiography-based physiology, quantitative coronary angiography, 3D-QCA, intermediate coronary lesions, percutaneous coronary intervention, noninferiority trial, coronary revascularization

Key Takeaways

Background and Rationale

• Guidelines recommend physiological assessment of intermediate coronary lesions (30–80% stenosis) to guide revascularization; FFR (pressure wire + adenosine) is the established gold standard
• Angiography-based physiological indices (vFFR, QFR) were developed to increase adoption and reduce procedure complexity by eliminating wire placement and pharmacologic hyperemia
• vFFR (Pie Medical Imaging) derives FFR from two angiographic projections ≥30° apart and invasive aortic-root pressure, applying physical flow laws (viscous resistance + separation loss; Gould and Kirkeeide models); prior multicenter validation studies (FAST, FAST II, FAST EXTEND) showed diagnostic accuracy equivalent to pressure-wire FFR
• Prior trials comparing angiography-based physiology vs angiographic guidance (FAVOR III China: QFR superior to angiography; FAVOR III Europe: QFR inferior to FFR) had not established the head-to-head comparison of angiography-based vs wire-based FFR

Trial Design and Patient Population

• 1:1 randomization: vFFR-guided vs FFR-guided revascularization of all intermediate study lesions
• Eligibility: age ≥18; chronic coronary syndrome, unstable angina, or NSTEMI; ≥1 native epicardial intermediate lesion (30–80% stenosis by visual estimation or QCA); reference vessel diameter ≥2.5 mm; TIMI flow grade 3
• Key exclusions: STEMI within 72h, cardiogenic shock, left main disease, severe valvular disease, aorto-ostial stenosis >50%, severe tortuosity, adenosine contraindications, bypass graft lesions
• Full analysis set: 1,116 (vFFR) vs 1,095 (FFR); mean age 67 years; 24.3% women; 18.7% ACS; 26.6% diabetes
• Mean number of study lesions per patient: 1.27 (vFFR) vs 1.28 (FFR)

Physiological Assessment

• vFFR: computed directly on-site after certification from two angiographic projections + intracoronary nitrates + invasive aortic pressure; performed in 96.7% of patients
• FFR: pressure wire + IV or intracoronary adenosine per standard practice; performed in 99.1% of patients
• Median vFFR 0.83 (IQR 0.74–0.89) vs median FFR 0.85 (IQR 0.79–0.91)
• Cut-off for revascularization: ≤0.80 for both strategies

Procedural Findings

• Functionally significant lesions (≤0.80): 40.9% of lesions in vFFR vs 31.3% in FFR group — 9 percentage point difference
• Revascularization of study lesions: 45.0% vFFR vs 36.0% FFR — driven by higher lesion detection rate with angiography-based tool
• Protocol adherence for revascularization decision: 96.0% vFFR vs 96.2% FFR
• Mean procedure time among PCI patients: 55.8 ± 26.8 min (vFFR) vs 60.9 ± 28.5 min (FFR); difference −5.13 min (95% CI −8.55 to −1.71) — shorter with vFFR due to absence of wire manipulation and adenosine

Primary End Point (1-Year MACE)

• Death/MI/revascularization: 80/1116 (KM 7.5%) vFFR vs 79/1095 (KM 7.5%) FFR
• Risk difference: −0.02 pp (95% CI −2.25 to 2.21); P=0.004 for noninferiority (margin 3.0 pp)
• Sensitivity analyses in per-protocol and intention-to-treat populations were concordant

Secondary End Points

• Study-vessel failure (cardiac death/study-vessel MI/clinically indicated study-vessel revascularization): 4.0% (vFFR) vs 4.6% (FFR); risk difference −0.62 pp (95% CI −2.35 to 1.10) — NS
• Death from any cause: 2.2% (vFFR) vs 2.3% (FFR) — NS
• Any myocardial infarction: 2.9% (vFFR) vs 2.4% (FFR) — NS
• Any revascularization: 4.2% vs 4.2% — identical
• Seattle Angina Questionnaire summary score at 1 year: 87.3 ± 14.4 (vFFR) vs 86.8 ± 15.1 (FFR) — equivalent quality of life

Safety

• Intraprocedural complications (at time of physiological assessment): 3.7% (vFFR) vs 6.0% (FFR) — fewer with vFFR, attributable to absence of wire-related complications (coronary dissection, spasm, perforation from wire manipulation)
• Overall serious adverse events appeared similar between groups

Why vFFR Detects More Significant Lesions Than FFR

• Angiography-based tools assume higher coronary flow (based on vessel size rather than actual measured hyperemic flow) → they do not capture actual microvascular resistance, demand, or patient-specific hyperemia response
• Higher assumed flow → higher computed pressure drop → lower vFFR value → more lesions appear hemodynamically significant
• Additionally, angiography-based tools may overestimate severity of longer or tapered lesions (diffuse disease) due to accumulated resistance assumptions
• These discordances were not apparent in initial validation studies but emerge at trial scale

Limitations of the Document

• Open-label design: Operators and patients knew group assignment; potential bias in clinical decision-making after initial assessment, though end-point events were adjudicated by an independent committee
• Small ACS proportion (18.7%): Limits generalizability to acute coronary syndrome presentations; trial primarily represents stable/CCS population
• Experience asymmetry: Operators had far greater experience with wire-based FFR than vFFR; may have disadvantaged vFFR (though vFFR training was mandated pre-enrollment)
• Operator and center selection: 37 high-volume European centers with physiological assessment expertise; may not generalize to lower-volume centers or settings with less angiographic computational experience
• Geographic limitation: Data skewed toward southern Europe and the Netherlands
• 1-year follow-up only: No long-term data; the higher revascularization burden in vFFR group has not been assessed beyond 12 months
• Prespecified cost-effectiveness analysis pending: Higher revascularization rate with vFFR vs shorter/simpler procedure creates an unresolved economic trade-off

Key Concepts Mentioned

• concepts/Angiography-Based-Coronary-Physiology — vFFR as the intervention tested; validated noninferior to FFR in this source
• concepts/Fractional-Flow-Reserve — comparator strategy; ≤0.80 threshold; adenosine/pressure-wire requirements
• concepts/Instantaneous-Wave-Free-Ratio — resting alternative to FFR (iFR); not directly tested but contextually related
• concepts/Multivessel-PCI-STEMI-Timing — ACS context for physiological assessment

Key Entities Mentioned

• Pie Medical Imaging — developer and software licensor of vFFR system; provided funding (unrestricted grant) and training
• Siemens Healthineers — co-funder (unrestricted grant)
• European Cardiovascular Research Institute (Rotterdam) — regulatory sponsor
• Cardialysis (Rotterdam) — contract academic research organization; data management, core laboratory, statistical reporting

Wiki Pages Updated

• Created wiki/sources/vffr-fastiii-nejm-2026.md
• Created wiki/concepts/Angiography-Based-Coronary-Physiology.md
• Updated wiki/concepts/Fractional-Flow-Reserve.md — vFFR section added; contradictions updated
• Updated wiki/sourceindex.md
• Updated wiki/wikiindex.md