CT-FFR: functional assessment

Fractional Flow Reserve from CT adds physiology to the anatomic scan — when to use it, how to read the numbers, and where it falls short.

Standard CCTA is an excellent purely morphological imaging modality — but it is a poor predictor of reversible myocardial ischemia. Because the anatomic severity of a plaque doesn’t always correlate with its hemodynamic relevance, clinicians traditionally had to rely on separate stress tests to determine whether a lesion actually restricted blood flow.

Fractional Flow Reserve from CT (CT-FFR) bridges that gap — it produces both anatomic and functional/physiological assessments from a single noninvasive scan, with no additional pharmacological stress agent.

1 How the technology works

CT-FFR takes standard anatomical CCTA data and uses cloud-based supercomputers to run Computational Fluid Dynamics (CFD) modeling. The system creates a patient-specific 3D mesh model of the aorta and epicardial coronaries and simulates hyperemia to calculate pressure and flow.

The algorithms rely on three core physical principles:

  • Allometric scaling law — baseline coronary blood flow is proportional to left ventricular mass
  • Murray’s law & Poiseuille solution — flow rate is proportional to vessel diameter and shear stress, and inversely proportional to blood viscosity
  • Hyperemic boundary conditions — model parameters tuned to simulate adenosine-induced maximal vasodilation

Processing is currently done off-site via FDA-approved platforms (e.g. HeartFlow). The CCTA dataset is uploaded; CT-FFR results return in 2–3 hours.

2 Interpreting CT-FFR values

Invasive FFR — measured with a pressure wire in the cath lab — uses a threshold of < 0.80 to indicate a hemodynamically significant lesion requiring revascularization.

CT-FFR provides pressure values throughout the entire coronary tree. Rather than a strict binary, interpret as a continuous variable:

FFRCTInterpretation
> 0.80Not significant for lesion-specific ischemia
0.75 – 0.80Borderline for lesion-specific ischemia
< 0.75Significant for lesion-specific ischemia

Interpretation pearls:

  • A sharp pressure drop across a focal stenosis resulting in FFRCT < 0.75 is highly indicative of lesion-specific ischemia.
  • A gradual pressure drop is less specific. Normal vessels experience a natural, gradual pressure drop along their length even in the absence of CAD — e.g. the distal LAD may naturally reach FFRCT > 0.90 in a clean coronary.
  • Don’t refer for ICA based solely on a low end-vessel FFRCT — the location and gradient of the pressure drop matter more than the terminal number.

3 Clinical evidence & impact

CT-FFR is most clinically valuable for adjudicating moderate (50–69%) stenoses. Standard CCTA struggles with specificity in this range — particularly when calcification creates blooming artifacts that artificially inflate the appearance of stenosis.

TrialFinding
NXT & PACIFIC86–87% diagnostic accuracy vs. invasive catheterization gold standard
PLATFORMCombining CT-FFR with CCTA cancelled 60% of planned ICAs and reduced invasive procedures on non-obstructive lesions
RIPCORDCT-FFR data changed clinical management (medical therapy vs. PCI vs. CABG) in 36% of patients

4 Limitations

  • Image-quality requirements. Demands an exceptionally high-quality CCTA dataset — slice thickness < 1 mm, strict heart-rate control, and proper nitroglycerin administration. Real-world rejection rates due to motion or noise are around 10%.
  • Diagnostic blind spots. Accuracy drops for lesions that fall into the borderline range (0.70–0.80).
  • Contraindications. Cannot be reliably used in patients with extensive heavy calcification, prior stents, or surgical bypass grafts — all distort the CFD model.
References
1 — Nørgaard BL et al. NXT trial — CT-FFR diagnostic performance. JACC 2014;63(12):1145–1155.
2 — Driessen RS et al. PACIFIC trial — Comparison of CT-derived FFR, perfusion CT, SPECT, PET, and ICA. JACC 2019;73(2):161–173.
3 — Douglas PS et al. PLATFORM trial — Outcomes of FFRCT-guided diagnostic strategies. JACC 2015;65(15):1437–1450.
4 — Curzen NP et al. RIPCORD study — Routine pressure-wire assessment changes management. Circulation 2014;130(18):1626–1635.
← previous
Plaque burden (P-modifier)
next →
Stents