Coronary CT angiography (CCTA) has been recognized as the first-line diagnostic test for most patients with suspected coronary syndrome, often acting as a gatekeeper for invasive coronary angiography. It is therefore pivotal to understand instances of discrepancies that are encountered in clinical practice. Moreover, most of the literature on this topic relies on obsolete machines or definitions of coronary artery stenosis that cannot be defined as severe.

The investigators aim 1) to report the real word data on the performance of last-generation CCTA in identifying obstructive coronary artery disease (also considering different thresholds of stenosis, i.e., moderate or severe) and 2) to identify predictors of discrepancies.

Most updated international guidelines recommend Coronary Computed Tomography Angiography (CCTA) as the initial test to rule out coronary artery disease (CAD). CCTA should also be considered an alternative to invasive coronary angiography (ICA) for non-diagnostic or indeterminate results of other noninvasive tests. Thanks to spatial and temporal resolution increase, CCTA is now considered in an extensive range of pre-test probability (PTP), from 5% to 90%. Indeed, the accuracy of CCTA for identifying patients with at least one significant coronary arterial stenosis, defined as moderate (≥50%) by ICA, has reached almost 90%. Furthermore, CCTA and anatomical evaluation seem superior to stress testing for risk prediction among patients with at least moderate ischemia. As a result, CCTA has been recognized as the first-line diagnostic test for most patients with suspected chronic coronary syndrome and even in some acute chest pain presentation.

Suppose CCTA serves as a gatekeeper for ICA because of its high negative predictive value and eventually will replace ICA in its diagnostic role, as hypothesized. In that case, it is pivotal to understand instances of discrepancies that are encountered in clinical practice. In addition, prior studies have primarily evaluated the performance of CCTA in identifying a ≥moderate coronary stenosis (i.e., ≥50% lumen narrowing) as compared with ICA. Instead, there is much less evidence of its ability to rule out severe coronary stenosis (i.e., ≥70% lumen narrowing). This is noteworthy because recent studies have shown that the anatomic severity of CAD has a strong prognostic impact, even more than ischemia. Finally, new techniques such as dynamic stress CT perfusion (stress-CTP) and fractional flow reserve CT derived (FFR-CT) emerged as potential strategies to combine anatomical and functional evaluation providing additional diagnostic accuracy.

Against this background, the investigators aim 1) to report the real word data on the performance of last-generation CCTA in identifying obstructive CAD (also considering different thresholds of stenosis, i.e., moderate or severe) and 2) to identify predictors of discrepancies. The investigators hope this study will help interpret CCTA findings in clinical practice and eventually refine the diagnostic algorithm for patients with obstructive CAD.

Inclusion Criteria:

• CCTA with >64 rows
• ICA performed within one month from CCTA

Exclusion Criteria:

- age<18 years

Real-world patients who underwent a last-generation CCTA and an ICA within one month from 2010 until 2023