Coronary CT angiography and CT-derived FFR

What it is and how it works

Coronary CT angiography (CCTA) is a CT scan of the heart with iodinated contrast injected through a vein, timed to the heartbeat so the coronary arteries can be seen. It shows the extent and severity of both nonobstructive and obstructive coronary artery disease (CAD), the makeup of the plaque, and high-risk features such as positive remodeling and low-attenuation plaque Gulati et al., 2021, Section 3.1.1. It can also identify noncardiac causes for some symptoms Winchester et al., 2023, Table A.

The 2021 chest pain guideline describes current radiation dosimetry for CCTA as low, with effective doses for most patients of 3 to 5 mSv Gulati et al., 2021, Section 3.1.1. For comparison, the same guideline gives about 3 mSv for rest/stress rubidium PET, about 10 mSv for technetium SPECT, and an average of 4 to 10 mSv for an interventional angiographic procedureGulati et al., 2021, Section 3.2.3 Gulati et al., 2021, Section 3.1.2.

What the result tells you

Reports use CAD-RADS, a standardized system for describing the degree of stenosis seen on CCTA, which the SCCT consensus recommends for reporting Winchester et al., 2023, Definitions for Table 1.2 Narula et al., 2021, Table 1. In the AUC scenarios the categories read as follows Winchester et al., 2023, Table 1.2.

The newer CAD-RADS 2.0 also attaches modifiers for plaque burden (P) and for ischemia (I), as in the example “CAD-RADS 4A/P3/HRP/I+” in the 2026 FFR-CT consensus Weir-McCall et al., 2026, Table 4. Outcomes step down with each category: 5-year event-free survival was 95% at CAD-RADS 0 and 69.3% at CAD-RADS 5 Narula et al., 2021, Section 2.2.

A finding of nonobstructive plaque matters too. It excludes the need for revascularization, but SCCT says it should prompt preventive therapy Narula et al., 2021, Section 5.1.1. The chronic coronary disease guideline notes that a negative functional test did not uniformly predict a low annual risk of cardiovascular death or MI, with a normal CCTA the lone exception Virani et al., 2023, Section 3.2.

CT-derived FFR: what it adds

Stenosis on CCTA and flow limitation do not always agree, which is why CCTA alone has lower specificity for ischemia Weir-McCall et al., 2026, Section 1. CT-derived fractional flow reserve (FFR-CT) closes part of that gap. Software uses the CCTA images and a computational model to estimate the pressure ratio that an invasive pressure wire would measure Weir-McCall et al., 2026, Section 3. Values are read 2 cm beyond the stenosis: above 0.80 means a low probability of flow limitation, 0.70 to 0.80 an intermediate probability, and below 0.70 a high probability Weir-McCall et al., 2026, Section 9.2.

The 2026 SCCT/SCAI consensus recommends CT-FFR in stable chest pain, provided the scan has a stenosis of 50% to 90% outside the left main and the result could change management. It should be done only on a high-quality scan, with heart rate control where required and nitroglycerin Weir-McCall et al., 2026, Table 2 Weir-McCall et al., 2026, Section 8. In trials and registries, adding FFR-CT meant fewer invasive angiograms, fewer of them without significant disease, more of those done leading to revascularization, and no excess in death or MI Weir-McCall et al., 2026, Section 5.4.

Strengths and limitations

Practical points

• Who struggles. The chest pain guideline lists inability to cooperate with breath-holds, clinical instability, a contraindication to beta-blockade with a raised heart rate and no alternative, heart rate variability and arrhythmia, and a contraindication to nitroglycerin (if indicated) Gulati et al., 2021, Table 5.
• Contrast and kidneys. Allergy to iodinated contrast and renal impairment, defined by local protocols, are on the same list Gulati et al., 2021, Table 5. The sources give no single kidney-function cutoff.
• Heart rate. For stent imaging, SCCT sets a goal of under 60 bpm Narula et al., 2021, Table 1. Motion artifact, which depends on heart rate and scanner speed, was the most common reason a CCTA could not be used for CT-FFR (6.7% of more than 13,000 datasets) Weir-McCall et al., 2026, Section 4.
• Calcium. Heavy calcification causes blooming artifact. The 2026 consensus sets no absolute calcium score above which CT-FFR is barred, but accuracy against invasive FFR is lower in heavily calcified vesselsWeir-McCall et al., 2026, Section 4.1.
• Stents and grafts. SCCT considers CCTA appropriate for symptomatic patients with stents of 3.0 mm or more, and possibly appropriate for smaller stents. CCTA is also appropriate after bypass surgery, particularly when graft patency is the main questionNarula et al., 2021, Table 1. CT-FFR is not validated in stents or bypass grafts Weir-McCall et al., 2026, Section 5.7.
• Pregnancy and breastfeeding. Radiation exposure should generally be avoided in pregnancy. Iodinated contrast reaches the fetal circulation and calls for caution, while breastfeeding may continue after contrast because less than 1% is excreted in breast milk Gulati et al., 2021, Section 3.3.
• Reading the report. Look for the CAD-RADS category with its plaque and ischemia modifiers, the post-stenosis CT-FFR value if one was done, and whether the study was rated inconclusive (N).

An example from the tool

CCTA is often considered after another test has not settled the question. The 2023 AUC assume that after an equivocal or inconclusive noninvasive test, a different modality should be usedWinchester et al., 2023, Assumption 7. Open the scenario below to see how the criteria rate each test after an inconclusive stress imaging result.

Try the scenario builder → for a patient with a stent, a high heart rate or a contrast concern.