Diagnostic Performance of CCTA for Coronary Artery Disease
CCTA demonstrates excellent sensitivity (95-99%) for detecting obstructive coronary artery disease but moderate specificity (64-83%), making it highly reliable for ruling out disease but less reliable for confirming stenosis severity. 1
Sensitivity and Specificity by Analysis Level
Per-Patient Analysis
- Sensitivity: 95-99% for detecting ≥50% stenosis 1, 2
- Specificity: 64-83% depending on patient population and calcification burden 1, 2
- Positive Predictive Value (PPV): 64-86%, indicating moderate reliability for confirming disease 1, 3
- Negative Predictive Value (NPV): 94-99%, making CCTA excellent for excluding significant CAD 1, 2, 4
The CCTA ACCURACY trial specifically reported 95% sensitivity, 83% specificity, 64% PPV, and 99% NPV for detection of CAD at the 50% stenosis threshold. 1
Per-Vessel Analysis
- Sensitivity: 85-88% 1, 3
- Specificity: 82-90% 1, 3
- PPV: 47% (lower due to overestimation of stenosis) 3
- NPV: 99% 3
Per-Segment Analysis
Performance in Women vs Men
CCTA demonstrates similar diagnostic accuracy in women compared to men, with no significant sex-based differences in sensitivity or specificity. 2
- Women: Sensitivity 90-93%, Specificity 77-88% 2
- The specificity of any detectable coronary calcium (CAC >0) is significantly better in women (40-66%) than men (23-36%) 2
- Women in these cohorts tend to be older with lower prevalence of obstructive CAD 2
Factors That Reduce Diagnostic Accuracy
Heavy Coronary Calcification
- Significantly reduces specificity to as low as 32% at per-patient level 1, 4
- Causes overestimation of stenosis severity 4, 5
- Limits accurate assessment of plaque composition 4
Technical and Patient Factors
- High or irregular heart rates cause motion artifacts that falsely suggest stenosis 4, 5
- Obesity and poor body habitus reduce image quality 4
- Small vessels (<1.5-3.0mm) are difficult to assess accurately 4, 5
- In-stent restenosis evaluation has lower accuracy, particularly for stents <3.0mm diameter 5
Comparison with FFR-CT
Adding FFR-CT to CCTA significantly improves specificity while maintaining high sensitivity, addressing CCTA's main limitation of overestimating stenosis. 1
FFR-CT Performance
- Sensitivity: 85-93% (similar to CCTA alone) 1
- Specificity: 65-82% (significantly higher than CCTA's 32-46%) 1
- Overall accuracy: 84% vs 59% for CCTA alone 1
- FFR-CT correctly reclassified 68% of false-positive CCTA results as true negatives 1
Per-Patient vs Per-Vessel FFR-CT Analysis
Clinical Context and Pre-Test Probability
In patients with high pre-test probability of CAD (67%), CCTA demonstrates PPV of 82.7% and NPV of 85.0%, with diagnostic performance not influenced by angina type. 1
The CORE-64 study reported a receiver operating characteristic (ROC) area of 0.93 for CCTA using quantitative coronary angiography as reference standard. 1
Common Pitfalls to Avoid
- Do not rely on CCTA alone for quantifying stenosis severity in intermediate lesions—functional testing or FFR-CT is needed 1, 5
- Recognize that positive CCTA often overestimates obstruction severity and requires further testing to guide management 3
- Avoid CCTA in patients with extensive calcification, irregular heart rates, or significant obesity where image quality will be inadequate 6
- Remember that anatomic stenosis does not always correlate with hemodynamic significance—functional assessment may be needed 5
Prognostic Value Beyond Diagnosis
CCTA provides prognostic information beyond stenosis detection, with presence of non-obstructive CAD and high-risk plaque features predicting adverse cardiovascular events. 1, 4
- Patients with extensive non-obstructive CAD have higher adverse event rates (14.5%) than those with less extensive but obstructive disease (13.6%) 1
- Absence of atherosclerosis on CCTA is associated with very favorable prognosis 1
- High-risk plaque features are independent predictors of major acute coronary events even in non-obstructive CAD 1