CT derived FFR: "The Game Changer?" revisited.
J Cardiovasc Comput Tomogr. 2018 Nov - Dec;12(6):447-449. doi: 10.1016/j.jcct.2018.08.006.
The practice of cardiology has been revolutionized by the emergence of coronary computed tomographic angiography (CTA) as a major addition to the armamentarium for the evaluation of coronary artery disease. In an extraordinarily short time span, the CTA literature has been populated with studies documenting accuracy superior to the most commonly utilized functional testing modalities of treadmill testing, stress echocardiography (SE) and SPECT myocardial perfusion imaging (MPI) for the identification of invasive coronary angiographic (ICA) stenoses >50%.1, 2, 3 Recognizing that ICA stenosis correlates inadequately with invasive fractional flow reserve (IFFR) and is not the appropriate gold standard, CTA has also been correlated with IFFR, with similar superiority in sensitivity to the standard modalities but lower specificity. The less often utilized and more expensive positron emission tomography (PET) and magnetic resonance imaging (MRI) stress testing have sensitivities similar to CTA but superior specificities.4, 5, 6, 7
In addition, studies have documented the better prognostic value of CTA compared to SE and MPI,8 as well as superior outcomes in chest pain patients without established CAD randomized to CTA versus the standard modalities,9, 10, 11, 12 despite the relatively poor specificity using the IFFR gold standard. The reasons for the low specificity of CTA are multiple: Firstly, as with ICA, there is an imperfect correlation of percent stenosis with IFFR. Secondly, the classification by intention to diagnose of heavily calcified or motion artifact degraded segments as “positive” will decrease specificity. Thirdly, visual overestimation of % DS on CTA is undoubtedly as common as it is in invasive angiography.