FFR-CT: How CT-Derived Functional Assessment Works, What the Evidence Shows

The Scene

The following scene is drawn from the composite of patients I have cared for in clinic and on the hospital floor. All identifying details are changed.

Susan is 58 years old, a family physician in Springfield, Illinois. She ordered her own coronary CTA after her father died of a heart attack at 64. Her CAC score was 247. Her coronary CTA showed a 55% stenosis in the proximal left anterior descending artery.

Her cardiologist tells her the stenosis is “intermediate” and the clinical decision whether to proceed to catheterization and possible stenting is not straightforward from the anatomy alone. She offers two options: invasive coronary angiography with FFR measurement (the reference standard) or HeartFlow FFRct analysis of the existing CTA images to derive a non-invasive functional estimate without another procedure.

Susan, who is a physician and understands what an invasive catheterization involves, chooses the FFRct analysis first. She will proceed to catheterization only if the FFRct shows physiological significance.

Three days later, her HeartFlow report arrives. Her FFRct value at the distal point of her LAD lesion is 0.89. She does not need a catheterization.

This article explains what happened with those numbers.

What It Is

FFRct (fractional flow reserve derived from coronary CTA) is a computational technology that applies fluid dynamics modeling to coronary CTA images to derive a non-invasive estimate of the hemodynamic significance of coronary stenoses.

Invasive FFR (the reference standard) is measured with a pressure wire advanced across a stenosis during adenosine-induced hyperemia: the ratio of mean distal coronary pressure to mean aortic pressure. FFR below 0.80 is hemodynamically significant. FFR above 0.80 is not flow-limiting.

FFRct derives mathematically equivalent information without a catheter, without a pressure wire, without adenosine. The computation requires:

1. A high-quality coronary CTA dataset (coronary CTA images obtained in standard clinical protocols)
2. Computational fluid dynamics modeling: the coronary anatomy is reconstructed in 3D, and blood flow is simulated using boundary conditions derived from physiological coronary flow models
3. Output: a color-coded FFRct value map overlaid on the coronary tree, with values at every point in the vessel

HeartFlow FFRCT (HeartFlow, Inc., Redwood City, CA) is the dominant commercial platform. It received FDA clearance in 2014 (De Novo, Class II). The analysis is performed off-site by HeartFlow analysts and returned within 24 hours.

The Mechanism

How Fluid Dynamics Models Coronary Flow

The coronary arterial tree is a fluid dynamics system: blood flows from the aorta through the coronaries to the microvascular bed based on pressure gradients. The principle underlying FFRct is that if the precise 3D geometry of the coronary arteries is known (from the CTA), the distribution of flow and pressure at every point can be solved mathematically using computational fluid dynamics (CFD), provided accurate boundary conditions.

The boundary conditions for each patient are derived from physiological scaling laws:

• Total coronary flow at rest is estimated from myocardial mass (computed from the CTA)
• Microvascular resistance is estimated using allometric scaling relationships between vessel diameter and flow (the Murray Law of bifurcation flows)
• Hyperemia (the increased flow state during which FFR is conventionally measured) is simulated mathematically by reducing microvascular resistance by a factor representing the response to adenosine

The resulting simulation produces pressure and flow values throughout the coronary tree without requiring adenosine infusion, a catheter, or a pressure wire 5 / Solid .

What FFRct Measures vs. What Invasive FFR Measures

Both values reflect the same physiological construct: the ratio of distal coronary pressure to aortic pressure during maximal hyperemia. They are not identical because:

1. Coronary CTA image quality affects the accuracy of the 3D reconstruction; small vessel artifacts reduce FFRct accuracy
2. The boundary conditions are estimates, not measurements (true microvascular resistance is not measured in the FFRct computation)
3. Invasive FFR is measured at the actual hemodynamic state; FFRct is modeled

These differences produce diagnostic error: FFRct slightly overestimates physiological significance (lower values) in vessels with moderate stenoses, a phenomenon called “false positive” drift toward significance 5 / Solid .

How It Is Used

Clinical Integration

FFRct is used as an intermediate step between coronary CTA and invasive angiography. The clinical pathway:

1. Patient with stable chest pain or positive stress test undergoes coronary CTA
2. CTA shows anatomically intermediate stenosis (40 to 69%) in one or more vessels
3. The anatomical finding is insufficient to distinguish a physiologically significant lesion from a benign moderate stenosis
4. FFRct analysis of the CTA images provides a functional estimate: FFRct below 0.80 suggests physiological significance; above 0.80 suggests non-significance
5. If FFRct is above 0.80, invasive angiography and PCI are deferred; medical therapy is optimized
6. If FFRct is below 0.80, the patient proceeds to invasive coronary angiography for confirmatory FFR measurement and possible revascularization

The PLATFORM trial demonstrated that FFRct guidance can safely reduce unnecessary invasive angiographies 5 / Solid , discussed below.

Requirements for Accurate FFRct

CTA image quality requirements:

• Heart rate below 65 bpm for retrospective gating (or below 75 bpm with prospective gating and 270 ms rotation time)
• Motion artifact less than 0.5 mm of blur
• Contrast enhancement adequate (Hounsfield units above 250 in the coronary lumen)
• Calcium burden not excessive (severe calcification creates blooming artifact that misestimates lumen diameter)

If image quality is suboptimal (estimated 10 to 15% of submitted CTAs), the HeartFlow analysis fails or returns a “low quality” designation and cannot be interpreted reliably.

FDA Clearance Status

HeartFlow FFRCT received FDA De Novo clearance (Class II) in 2014 as a non-invasive diagnostic aid for evaluating suspected CAD in patients with intermediate stenoses (40 to 69%) on coronary CTA in stable symptomatic patients. It is not cleared for emergency settings, ACS presentations, or anatomical configurations outside the validated range 5 / Solid database, K140197).

The Evidence

NXT Trial: FFRct Accuracy vs. Invasive FFR

NXT (Norgaard BL, et al. J Am Coll Cardiol. 2014; doi:10.1016/j.jacc.2014.01.019) enrolled 254 patients (484 vessels) with coronary CTA showing 30 to 90% stenosis, comparing FFRct against invasive FFR as reference standard. Per-vessel diagnostic accuracy:

• AUC 0.93 (95% CI 0.90 to 0.96) for FFRct 5 / Solid
• Sensitivity 86%, specificity 79%, PPV 65%, NPV 93% at FFRct cutoff of 0.80

The high NPV (93%) means that an FFRct above 0.80 reliably rules out physiological significance. The moderate PPV (65%) means that an FFRct below 0.80 requires invasive confirmation before revascularization is pursued in stable patients.

PLATFORM Trial: FFRct-Guided Strategy vs. Usual Care

PLATFORM (Douglas PS, et al. J Am Coll Cardiol. 2016; doi:10.1016/j.jacc.2016.03.505) enrolled 584 patients planned for invasive coronary angiography, comparing FFRct-guided strategy versus usual care. In patients with stable symptoms planned for elective catheterization, the FFRct-guided strategy reduced invasive angiography showing no obstructive CAD by 61% (the “unnecessary” catheterizations). Over 90 days, there were no significant differences in MACE or quality of life between strategies 5 / Solid . Cost per patient was lower in the FFRct group 4 / Promising . What PLATFORM did not show: the trial was not powered for hard clinical endpoints; a longer follow-up outcomes trial was needed.

ADVANCE Registry: Real-World FFRct Outcomes

The ADVANCE Registry (Norgaard BL, et al. J Am Coll Cardiol. 2020; doi:10.1016/j.jacc.2020.06.074) prospectively enrolled 4,730 patients who underwent FFRct, with 12-month outcomes. Patients with all vessels FFRct above 0.80 who were deferred from revascularization had a 12-month MACE rate of 3.3%, equivalent to the deferral safety demonstrated by invasive FFR in the DEFER trial 5 / Solid . Patients with FFRct below 0.80 who underwent revascularization had 12-month MACE of 2.1%, consistent with revascularization benefit in physiologically significant disease 4 / Promising . The registry supports the safety of FFRct-guided deferral of revascularization in real-world practice.

FORECAST Trial: FFRct vs. Standard Care in UK NHS

FORECAST (Berry C, et al. Lancet. 2021; doi:10.1016/S0140-6736(21)01112-6) enrolled 1,400 patients with stable chest pain randomized to coronary CTA with FFRct analysis versus standard care (functional testing). At 9 months, the FFRct-guided strategy achieved equivalent MACE rates (3.1% vs. 2.6%; p = non-inferior) 5 / Solid , with lower utilization of invasive angiography (22% vs. 19%). FFRct integration into a standard stable chest pain pathway is feasible and safe. What FORECAST did not show: no long-term outcome data; no cost-effectiveness analysis from a US payer perspective.

Diagnostic Accuracy Compared with Invasive FFR: Pooled Analysis

A pooled analysis of 5 validation studies (1,375 patients, 2,230 vessels) showed:

• Per-vessel sensitivity: 0.86 (95% CI 0.83 to 0.89)
• Per-vessel specificity: 0.79 (95% CI 0.76 to 0.82)
• Per-vessel AUC: 0.93

The Patient Experience

Susan received her FFRct report electronically three days after her CTA data were submitted. The report showed a color-coded coronary tree map: green (FFRct above 0.80) throughout her vessels except at the level of the proximal LAD stenosis, where the value was 0.89.

The 0.89 value means her stenosis does not cause physiologically significant flow restriction at maximal hyperemia. It is above the 0.80 threshold. She does not require catheterization for this lesion at this time.

She does require:

• High-intensity statin (her LDL-C was 134 mg/dL)
• Aspirin (her cardiologist’s choice for a patient with CAC 247 and documented moderate LAD plaque)
• Blood pressure optimization (her office blood pressure was 138/86)
• Interval reassessment in 2 years with clinical evaluation

Her coronary CTA showed the anatomy. The FFRct answered the functional question. Together, they determined her management without a catheter.

What Your Cardiologist Will Not Have Time to Explain

• FFRct below 0.80 requires invasive confirmation before stenting. FFRct is a non-invasive screening tool; it does not replace invasive FFR measurement when revascularization is planned. A PPV of 65% means 35% of lesions with FFRct below 0.80 will have invasive FFR above 0.80 and not need a stent. This confirmation step protects against over-treatment.

• FFRct can fail if CTA image quality is insufficient. Image artifacts from calcium, motion, or poor contrast timing may produce an uninterpretable result or an inaccurate one. About 10 to 15% of submitted studies return as low quality. If this happens, the clinical decision returns to the pre-FFRct state.

• FFRct is not validated for all scenarios. It is not validated for in-stent restenosis, diffuse disease without focal stenosis, or bifurcation lesions with major side-branch involvement. Application in these scenarios is off-label.

• The cost is $1,500 to $2,000 before insurance. Medicare coverage for FFRct is available under CPT code 0623T at covered institutions. Commercial insurance coverage varies; prior authorization is often required. The cost comparison relevant to the patient: FFRct versus the $8,000 to $15,000 facility fee for elective diagnostic catheterization.

Decisions and Trade-Offs

FFRct vs. Direct Angiography with Invasive FFR

For a stable patient with an intermediate stenosis on CTA, the choice between FFRct and direct angiography with invasive FFR involves:

Favoring FFRct:

• Patient prefers to avoid invasive procedure if possible
• Intermediate-to-low clinical risk (stable symptoms, no high-risk features)
• Good CTA image quality and minimal calcium in the target vessel
• Insurance coverage for FFRct is confirmed

Favoring direct angiography:

• High pretest probability or high-risk stress test results suggesting highly significant disease
• Patient with known prior stent failure (in-stent restenosis is an FFRct limitation)
• Anatomical complexity (significant bifurcation disease, left main involvement, calcification that impairs CTA quality)
• Patient who has already consented to invasive evaluation and wishes to complete it in one session

The Downstream Use Decision

An FFRct above 0.80 safely defers revascularization when combined with guideline-directed medical therapy. The ADVANCE registry data confirm a 3.3% annual MACE rate with deferral, comparable to the DEFER trial’s invasive FFR deferral outcomes 5 / Solid . This does not mean the lesion requires no attention: the coronary CTA and FFRct together define the patient’s plaque burden and the intensity of medical therapy required.

The Three Questions Every Patient Should Ask

1. “Is my coronary CTA image quality sufficient for FFRct analysis?” A cardiologist who reviews the CTA before ordering FFRct is more likely to get an interpretable result than one who automatically submits every CTA.

2. “If my FFRct is below 0.80, what is the plan?” The plan should include confirmatory invasive FFR before stenting, not immediate stenting based on FFRct alone.

3. “Does my insurance cover FFRct, and what is my out-of-pocket cost?” This should be answered before the analysis is ordered, not after the report arrives.

FFRct and Non-Obstructive Disease

A critical interpretive limit: FFRct is designed to assess obstructive epicardial stenosis. It provides no information about microvascular disease, the small-vessel dysfunction that causes chest pain and ischemia without obstructive CAD.

Non-obstructive CAD with microvascular disease is the predominant cardiovascular phenotype in women presenting with stable chest pain. Up to 60% of women undergoing coronary CTA for chest pain evaluation have no obstructive stenosis on imaging. An FFRct above 0.80 in a patient with persistent exertional chest pain and non-obstructive anatomy does not rule out ischemia. It rules out hemodynamically significant epicardial stenosis. These are different diagnoses, and the distinction determines the next clinical step.

For patients with non-obstructive coronary anatomy and ongoing symptoms, coronary physiology assessment requires different tools: coronary flow reserve (CFR) measurement, index of microcirculatory resistance (IMR), or positron emission tomography myocardial perfusion imaging. FFRct and invasive FFR both measure epicardial physiology only. 5 / Solid

When a coronary CTA shows minimal or no obstructive disease, FFRct is not clinically indicated and will not be ordered. The diagnostic question shifts from “is this stenosis flow-limiting?” to “does this patient have microvascular disease? That requires an entirely different evaluation pathway, including provocative testing, coronary reactivity testing, or PET-based myocardial perfusion imaging.

Clinical Synthesis

Susan’s story is the prevention argument in compressed form: she found her CAC score, she found her coronary CTA, she found her intermediate LAD lesion, and she got an answer without a catheter. She is on appropriate therapy. She has a 2-year follow-up plan. She did not have an invasive procedure she did not need.

This is the future of stable coronary disease management: non-invasive anatomy (CTA), non-invasive function (FFRct), and aggressive primary prevention therapy for the plaque that is present but not yet causing flow limitation. The risk of deferral is real (3.3% annual MACE) but lower than the risk of undertreated atherosclerosis without appropriate medical therapy.

If you have had a coronary CTA showing an intermediate stenosis: the Signal Check reviews whether FFRct is appropriate for your anatomy and whether your medical therapy is calibrated to your plaque burden, not just to your LDL-C.

If you have been told you need a catheterization for a moderate stenosis on CTA: a structured cardiovascular assessment reviews whether FFRct is an appropriate intermediate step in your specific case, and what the PLATFORM trial data mean for your decision.