Angiography-Derived Fractional Flow Reserve to Guide PCI (ALL-RISE)

Authors, Journal, Affiliations, Type, DOI

• Authors: William F. Fearon, Allen Jeremias, Guy Witberg, Rasha Al-Lamee, David J. Cohen, Amir Kaki, Rahul P. Sharma, Robert W. Yeh, Bassem M. Chehab, Michael C. Kim, Hiromasa Otake, Rajiv Tayal, Hitoshi Matsuo, Margaret McEntegart, Ankitkumar K. Patel, Yader Sandoval, Karim M. Al-Azizi, Kazuhiro Dan, Louai Razzouk, Stephane Fournier, Alexandra Popma, Jeffrey J. Popma, Duzhi Zhao, Bjorn Redfors, Mahesh V. Madhavan, Ran Kornowski, Martin B. Leon, Ajay J. Kirtane; for the ALL-RISE Investigators
• Journal: New England Journal of Medicine, published online March 29, 2026
• Affiliations: Stanford University; Cardiovascular Research Foundation, New York; and 59 sites across North America, Europe, Asia and the Middle East
• Type: International, multicentre, randomized, open-label, noninferiority trial
• DOI: 10.1056/NEJMoa2600949

Overview

ALL-RISE tested whether FFRangio — a software-derived FFR computed from ordinary coronary angiograms, without a pressure wire or hyperaemic agent — could replace pressure-wire physiology to guide PCI decisions without worsening clinical outcomes. In 1,930 patients with at least one intermediate stenosis (50–90%), 1-year death, myocardial infarction or unplanned clinically indicated revascularization occurred in 6.9% with FFRangio and 7.1% with pressure-wire assessment (HR 0.98; difference −0.2 percentage points; upper 97.5% CI boundary 2.1 points, below the 3.5-point margin; P<0.001 for noninferiority). FFRangio shortened procedural and fluoroscopy time and reduced contrast use, but classified more lesions as significant and led to more PCI (44.3% vs 35.4% of lesions). The trial is industry-sponsored and open-label.

Keywords

Fractional flow reserve; FFRangio; Angiography-derived physiology; Pressure wire; Intermediate coronary stenosis; Percutaneous coronary intervention; Noninferiority

Key Takeaways

Background

• Pressure-wire FFR and non-hyperaemic pressure ratios have Class I guideline recommendations for assessing intermediate stenoses, yet are used in <20% of cases, partly because of workflow burden: a guide catheter, a dedicated wire and often a hyperaemic agent are required.
• FFRangio (CathWorks) reconstructs a three-dimensional coronary tree from angiographic images and applies a resistance model to estimate pressure gradients from vessel contours. It correlates well with wire FFR but needs neither wire nor adenosine.

Methods

• 59 sites in North America, Europe, Asia and the Middle East; 4,704 screened, 1,930 randomized 1:1 (965 FFRangio, 965 pressure wire).
• Inclusion: age ≥18, invasive angiography with ≥1 stenosis of 50–90% by visual assessment requiring haemodynamic evaluation.
• Key exclusions: recent STEMI; culprit vessel in NSTE-ACS; CABG candidacy; prior CABG to the study vessel; severe aortic or mitral valve disease; severe LV dysfunction.
• Randomization was stratified by presentation (stable angina vs NSTE-ACS) and by whether the investigator planned FFR or a non-hyperaemic pressure ratio.
• FFRangio: three angiographic images ≥30° apart; automated lumen tracing with manual correction; 3D reconstruction; FFR ≤0.80 = significant.
• Pressure wire: FFR (intracoronary or intravenous adenosine/ATP, or intracoronary papaverine) or any non-hyperaemic pressure ratio; FFR ≤0.80 or NHPR ≤0.89 = significant.
• PCI was per local standard; post-PCI physiology was not permitted to guide PCI; angiograms and physiology were sent to an independent core laboratory.
• Primary endpoint: death from any cause, myocardial infarction, or unplanned clinically indicated revascularization at 1 year.
• Noninferiority margin: 3.5 percentage points (one-sided 97.5% CI); assumed 1-year event rate 7.5% in both arms.

Results

Population and measurements

• Mean age 68.4 ± 10.2 years; 25.0% women; 90.6% without elevated cardiac biomarkers; 37.8% diabetic; LAD was the study vessel in 49.6%.
• FFRangio succeeded in 98.5% of lesions (mean 0.81 ± 0.12); 43.0% were abnormal (≤0.80).
• Pressure-wire assessment succeeded in 98.6%; FFR in 55.5%, NHPR alone in 44.5%; mean FFR 0.84 ± 0.10, mean NHPR 0.89 ± 0.11; 37.0% abnormal.

Primary endpoint

• 64 patients (6.9%) vs 65 patients (7.1%); HR 0.98 (95% CI 0.70–1.39); difference −0.2 percentage points; upper boundary of the one-sided 97.5% CI 2.1 points; P<0.001 for noninferiority.
• No apparent heterogeneity across geographic regions.
• Individual components were similar: death 2.1% vs 2.3%; MI 1.6% vs 2.5%; clinically indicated repeat revascularization 4.1% vs 4.6%.

Secondary and process outcomes

• Assessment time: median 6 vs 8 minutes (difference −2 min; 95% CI −3 to −2).
• Procedural time: 39 vs 42 minutes (difference −5 min; 95% CI −7 to −3).
• Fluoroscopy time: 9 vs 12 minutes (difference −3 min; 95% CI −3 to −2).
• Contrast volume: 100 vs 105 mL (difference −14 mL; 95% CI −20 to −10).
• PCI was performed in 44.3% of FFRangio-assessed lesions vs 35.4% of pressure-wire-assessed lesions (OR 1.45; 95% CI 1.24–1.70) — FFRangio identified more lesions as significant.

Safety

• BARC 3/4/5 bleeding: 0.7% vs 1.5% (HR 0.50; 95% CI 0.20–1.23).
• Acute kidney injury: 0 vs 1 patient.
• Periprocedural angiographic complications: 1.0% vs 0.5% (HR 2.00; 95% CI 0.68–5.85).

Discussion

• FFRangio was noninferior to pressure-wire assessment for clinical outcomes at 1 year, while reducing procedural time, radiation exposure and contrast use — supporting its use to increase adoption of physiological guidance.
• The higher PCI rate with FFRangio is attributed to more lesions falling near the revascularization threshold; the authors note that outcomes are similar whether such grey-zone lesions are treated or deferred (IRIS-FFR).
• The result contrasts with FAVOR III Europe, where QFR-guided revascularization was not noninferior to FFR. The authors attribute the divergence to different computational methods and to greater operator interaction and inter-observer variability with QFR — a FAVOR III Europe substudy (REPEAT-QFR) showed high site-to-core-laboratory variability, whereas FFRangio has shown strong core-laboratory agreement.
• Subgroup analyses raised the possibility of a differential effect by vessel: events were lower with FFRangio for LAD lesions and higher for non-LAD lesions, and there was a possible interaction with whether the comparator used FFR or NHPR — but the trial was underpowered for these analyses.
• The FAST III trial (vFFR vs FFR) is testing another angiography-derived method.

Limitations (as stated)

• Investigators first had to judge that the anatomy was amenable to both pressure-wire and FFRangio assessment — a selection step that may limit generalizability.
• Most patients had stable coronary disease or no biomarker elevation; the role in NSTE-ACS nonculprit lesions and STEMI is being tested in AIR-STEMI.
• The comparator was a mixture of FFR and NHPR, and more evidence supports FFR than NHPR — although this reflects real-world practice.
• Open-label design, which the authors argue would be expected to favour the established standard of care; the similar rates of unplanned revascularization are reassuring.
• The control-arm event rate was 7.1% versus the assumed 7.5%, giving achieved power of 87.3%.

Limitations of the document

• Industry-sponsored trial: CathWorks funded the trial and participated in site selection, data collection and monitoring; statistical analysis was performed independently by the Cardiovascular Research Foundation, but the sponsor's role warrants caution.
• Open-label, noninferiority design with a 3.5-percentage-point margin — a relatively permissive margin for a diagnostic strategy, and the trial cannot show superiority.
• 1-year follow-up is short for a diagnostic strategy intended to guide durable revascularization decisions, especially with a higher PCI rate in the FFRangio arm; longer-term stent-related outcomes are unknown.
• Composite primary endpoint includes unplanned clinically indicated revascularization, which is partly subjective and may dilute or inflate treatment differences.
• The higher revascularization rate (44.3% vs 35.4% of lesions) is a real difference in care — the trial shows noninferior outcomes despite more stenting, not equivalence of treatment decisions; whether the extra stents are beneficial, neutral or harmful beyond 1 year is unresolved.
• Subgroup signals (LAD vs non-LAD, FFR vs NHPR comparator) are exploratory and underpowered; they should not guide practice.
• FFRangio requires adequate angiographic technique (three or four projections ≥30° apart, no table movement, minimal overlap), which may not be achievable in all laboratories; the trial protocol recommended this, and real-world image quality may differ.
• Patients with severe LV dysfunction, valvular disease and recent STEMI were excluded, limiting applicability to high-risk ACS and structural populations.

Key Concepts Mentioned

• concepts/Angiography-Based-Coronary-Physiology — the class to which FFRangio belongs
• concepts/Fractional-Flow-Reserve — the pressure-wire reference standard
• concepts/Instantaneous-Wave-Free-Ratio — non-hyperaemic wire-based comparator used in 44.5%
• concepts/Complete-Revascularization — physiological lesion selection in multivessel disease
• concepts/Multivessel-PCI-STEMI-Timing — AIR-STEMI extends the same technology to STEMI nonculprit lesions

Key Entities Mentioned

• CathWorks — developer of FFRangio and trial sponsor
• Cardiovascular Research Foundation — independent statistical analysis
• ALL-RISE Investigators — trial group
• FAVOR III Europe, FAST III, AIR-STEMI — comparator and successor trials

Wiki Pages Updated

• Created: wiki/sources/ffrangio-allrise-nejm-2026.md
• Updated: wiki/concepts/Angiography-Based-Coronary-Physiology.md
• Updated: wiki/concepts/Fractional-Flow-Reserve.md
• Updated: wiki/sourceindex.md
• Updated: log.md