Log in to MyACC
Guest Editorial | The Expanding Role of CCTA Post ESC Congress 2018; Editors' Perspective
Click the image above for a larger view.
The recent European Society of Cardiology meeting was a pivotal event for the field of cardiac computed tomography (CT), which will have long-term implications for the field of cardiovascular imaging and interventional cardiology.
The first – and most important trial presented – was the five-year data from SCOT-HEART. The study showed the use of CT angiography (CTA), in addition to standard care, most commonly an exercise treadmill test (ETT), resulted in a 41 percent lower rate of nonfatal myocardial infarction (MI) or death from coronary heart disease (Figure 1).1
Editors' Perspective
Diagnosing Ischemic Coronary Disease: Is the Future Already Here?
This month's Cardiology: Interventions guest editorial, by Ron Blankstein, MD, FACC, gives us a peek into a future that's coming soon, if not already here. Dr. Blankstein describes in detail the presentations at the recent European Society of Cardiology Congress 2018 that we'll likely view as a turning point in how we should think about making the diagnosis of ischemic coronary disease.
In short, noninvasive testing using multiple imaging modalities can diagnose coronary stenosis and provide information as to whether a lesion is ischemia-producing and the likely prognosis of such a lesion. Surely the misnamed, badly-used, but shockingly popular so-called "oculo-stenotic" reflex within diagnostic angiography has left the stage. Perhaps diagnostic angiography will soon be asked to leave as well.
In the ADVANCE registry, the safety of showing an FFR-CT >0.8 was supported by an astonishing 0/1,592 patients having a cardiac event over a three-month follow-up. If FFR-CT was "positive" (<0.8), 19 patients had an event (hazard ratio, 19.75).
The CREDENCE study compared the diagnostic value of using either SPECT, PET or cardiac MR (an integrated functional assessment) vs. invasive FFR (an integrated anatomic approach). A comprehensive CTA assessed quantitative percent stenosis, plaque volume, plaque composition, high-risk plaque features (positive remodeling, low attenuation plaque, spotty calcifications) and FFR-CT.
Myocardial perfusion analysis included data on perfusion, ventricular function and high-risk markers. Interestingly, the final model only included stenosis severity, number of lesions with >30 percent diameter stenosis, volume of noncalcified plaque, high-risk plaque characteristics and lumen volume. Add to the above that FFR-CT was not included in the final model – perhaps because it did not improve discrimination. All this is a far cry from good-old-fashioned diagnostic angiography.
Dr. Blankstein points out that the derived CTA model used in CREDENCE is not a routinely used model. Analysis of total plaque volume currently requires advanced software and it's technically challenging and time consuming. Naysayers may point to this as well as to the challenges of cost. But in CREDENCE, the evaluation of the integrated functional assessment used currently available clinical technology.
That aside, CCTA easily beat myocardial perfusion imaging in the validation model. The conclusion is simple: comprehensive CCTA interpretation is superior to a comprehensive myocardial perfusion imaging evaluation for diagnosing coronary lesions that cause ischemia.
Looking ahead, there's little doubt that advances in image analysis using artificial intelligence will enable us to assess so much more about plaque composition and stability. Combine that with CCTA and think of the resulting information that will be available without using an invasive test.
For now, we can say that FFR-CT already has the ability to identify patients with stable chest pain who can safely defer invasive therapy (at least in the relatively short term). As noninvasive testing becomes more robust and sophisticated, many more studies will be needed to identify which modalities truly will be valuable for clinicians making therapeutic decisions for/with patients. There is a lot more to learn, and costs will also play a role. But if we do our jobs correctly, we may be out of a job – at least a diagnostic one!
Peter C. Block, MD, FACC, is a professor of medicine and cardiology at Emory University Hospital and School of Medicine in Atlanta, GA.
Alfred A. Bove, MD, PhD, MACC, is professor emeritus of medicine at Temple University School of Medicine in Philadelphia, and a former president of the ACC.
John Gordon Harold, MD, MACC, is clinical professor of medicine at Cedars-Sinai Smidt Heart Institute and David Geffen School of Medicine at UCLA.
While the absolute event rates were low (2.3 vs. 3.9 percent over 4.8 years for standard care vs. CTA, respectively) and the mechanism related to the lower event rates observed in SCOT-HEART was a topic of intense discussion, the undeniable take-home point is this: the way in which we act on the results of cardiovascular imaging tests can have a significant impact on patient outcomes.
The SCOT-HEART results – together with other recent studies showing the benefit of coronary CTA (CCTA)2,3 – suggest there should be a greater role for CCTA among patients with stable chest pain. However, previous studies have shown that CCTA, by detecting more disease, may lead to a higher rate of invasive angiography and coronary revascularization – a finding that was not observed during the five-year follow-up of SCOT-HEART.
Since the benefit of performing coronary revascularization in the absence of objective evidence of ischemia is uncertain, many critics have appropriately pointed out that CCTA alone may be insufficient for selecting patients who may benefit from coronary revascularization.
While the noninvasive identification of ischemia in stable chest pain patients has traditionally been performed with functional testing, such as nuclear myocardial perfusion imaging (MPI) or stress echocardiography, the availability of fractional flow reserve via CT (FFR-CT) – a method by which computational fluid dynamic modeling is applied to rest CCTA data sets to estimate the hemodynamic consequence of coronary plaque – provides us with a new noninvasive tool. Supporting the diagnostic capabilities of this technique, there have been several studies examining the accuracy of CTA with or without FFR-CT against invasive angiography with FFR.4,5
The available data suggest a good accuracy of FFR-CT, although concerns have been raised about the diagnostic performance of this test when the FFR-CT values range between 0.7 to 0.8, as well as the fact that FFR-CT may be more sensitive but less specific when compared with other techniques. However, up until recently, there have been little to no data on how FFR-CT performs in the real world, whether it is safe to defer revascularization in patients with normal FFR-CT, and how this technique compares with other noninvasive functional tests. Several studies now shed light on some of these issues.
The ADVANCE registry, presented at ESC Congress 2018 by Timothy A. Fairbairn, MBChB, PhD, aimed to determine how FFR-CT would change clinical decision-making, patient management and clinical outcomes.6 This study – which included 5,083 patients from 38 international sites – found the addition of FFR-CT resulted in a modification to the anticipated treatment plan based on CTA results alone in two-thirds of patients.
Nevertheless, since the decision to revascularize patients in practice is not made by a core lab reviewing angiographic findings alone, but is based on multiple clinical parameters, the true magnitude of how often FFR-CT impacts care is likely over-estimated in this study.
Perhaps the more important data from this large ADVANCE registry is the safety of deferring revascularization when FFR-CT is greater than 0.8. Specifically, over the 90-day follow-up of this study, none of the 1,592 patients with negative FFR-CT experienced death, MI or unplanned hospitalization for an acute coronary syndrome (ACS) and urgent revascularization. In contrast, there were 19 adverse events (10 deaths, four MIs, five hospitalizations and urgent revascularization) in patients with positive FFR-CT (hazard ratio 19.75; p<0.001).
Further detailing the utility of FFR-CT following CCTA, Bjarne L. Norgaard, DO, PhD, presented an observational study of 3,674 consecutive patients with stable chest pain who were evaluated with CTA followed by selective FFR-CT for those with intermediate stenosis (30-70 percent).7 FFR-CT was performed for 697 patients (18 percent of cohort), reflecting the fact this test is only needed for a minority of CTA cases, when there is stenosis of uncertain hemodynamic significance.
Notably, patients with intermediate stenosis who had a negative FFR-CT (>0.80) had similar long-term outcomes when compared with patients with no to minimal stenosis (0-30 percent) by CCTA. On the other hand, adverse events were higher among patients with a positive FFR-CT who were not referred for invasive angiography. While the latter findings may be influenced by selection bias (i.e., patients who were not treated could have been higher risk), the overall findings from this trial support the safety of using FFR-CT to defer coronary revascularization in patients who have intermediate lesions deemed non-flow limiting by FFR-CT.
The CREDENCE study, also presented at ESC Congress 2018, aimed to compare the diagnostic value of an integrated functional – using either SPECT, PET or CMR – vs. an integrated anatomic approach for diagnosis of vessel specific ischemia, as defined by invasive FFR. This was a prospective, multicenter, diagnostic performance study that included 612 patients who were referred for nonemergent invasive angiography across 13 international sites. Per design, patients were assigned to either a derivation or a validation cohort.
The comprehensive CTA analysis included an assessment of CCTA for stenosis severity, and a quantitative assessment of percent stenosis, plaque volume, plaque composition, high-risk plaque features (i.e., positive remodeling, low attenuation plaque, spotty calcifications) and FFR-CT. Myocardial perfusion analysis included data on perfusion, ventricular functional and high-risk markers. The reference standard was invasive FFR ≤0.8 or ≥90 percent diameter stenosis by invasive angiography.
The final derived CTA model included the following variables: stenosis severity, number of lesions with ≥30 percent diameter stenosis, volume of noncalcified plaque, high-risk plaque characteristics and lumen volume. FFR-CT was not included in the final model, suggesting it did not add further improved discrimination once accounting for the aforementioned variables.
Notably, the derived CTA model used in CREDENCE was incredibly comprehensive and robust. For instance, the use of a quantitative plaque analysis is not currently performed as part of routine clinical CCTA interpretation, as an analysis of total plaque volume is technically challenging (often requiring a separate dedicated software) and time consuming.
On the other hand, the interpretation of functional images employed techniques currently used in routine clinical interpretation. Notwithstanding these differences, there was a highly significant improvement in the area under the curve (AUC), which in the validation model increased from 0.68 for MPI to 0.83 for OCTA. The investigators concluded that a comprehensive CCTA interpretation was superior to a comprehensive MPI evaluation for diagnosis of coronary arteries that cause ischemia.
These finding support the increasingly recognized concept that the identification of ischemia requires us to look beyond just stenosis severity, as adding information on plaque volume8 and plaque characteristics9 provides a more robust discrimination between lesions that cause ischemia from those that do not.
The CREDENCE study further extends these observations. Specifically, when evaluating the derivation cohort of CREDENCE, there was an increase in AUC from 0.82 to 0.88 (p=0.003) when adding atherosclerotic plaque features to stenosis severity. Currently, clinical measures of plaque extent often involve simpler measures – such as counting the number of segments with plaque10 – but ongoing advances in image analysis and machine learning will undoubtedly make measurements of plaque type and volume more readily available so it can be incorporated as part of the routine interpretation of CCTA.
In summary, due to the ability of CCTA to improve clinical outcomes by accurately diagnosing both nonobstructive and obstructive disease, this test is likely to have an increasing role in the evaluation of patients with stable chest pain. In select cases where there is stenosis of uncertain hemodynamic significance, FFR-CT may help discriminate nonflow limiting lesions for which coronary revascularization can be safely deferred from those associated with higher risk. Nevertheless, further studies are needed to compare the clinical effectiveness of FFR-CT vs. other noninvasive functional tests.
Ultimately, the optimal use of CCTA and functional testing could improve outcomes by leading to identification and treatment of more patients who have coronary artery disease, but limiting the use of invasive angiography to those who stand to benefit from coronary revascularization.
This invited guest editorial was authored by Ron Blankstein, MD, FACC. He is the associate director of the Cardiovascular Imaging Program, director of Cardiac Computed Tomography and associate physician in preventive cardiology at Brigham and Women's Hospital in Boston, MA. He's also associate professor in medicine and radiology at Harvard Medical School.
Keywords: ACC Publications, Cardiology Interventions, Acute Coronary Syndrome, Coronary Angiography, Artificial Intelligence, Chest Pain, Constriction, Pathologic, Coronary Stenosis, Echocardiography, Stress, Exercise Test, Follow-Up Studies, Hemodynamics, Hospitalization, Hydrodynamics, Myocardial Infarction, Myocardial Ischemia, Plaque, Atherosclerotic, Myocardial Perfusion Imaging, Prognosis, Registries, Selection Bias, Tomography, Emission-Computed, Single-Photon, Tomography, X-Ray Computed, Ventricular Function
< Back to Listings
