Author + information
- James K. Min, MD∗ ()
- Departments of Radiology and Medicine, Weill Cornell Medical College, New York, New York
- Dalio Institute of Cardiovascular Imaging, New York-Presbyterian Hospital, New York, New York
- ↵∗Reprint requests and correspondence:
Dr. James K. Min, Department of Radiology, New York-Presbyterian Hospital, Weill Cornell Medical College, 413 East 69th Street, Suite 108, New York, New York 10021.
Just 4 years ago, Choi et al. (1) reported a novel metric for discrimination of the hemodynamic significance of coronary artery lesions, a measure termed transluminal attenuation gradient (TAG). This gauge of functionally important coronary artery disease (CAD) is based upon typically acquired coronary computed tomography angiograms (CTA), and is defined as the linear regression coefficient between luminal attenuation and axial distance across the length of a coronary artery possessing a stenosis. In their first report, Choi et al. (1) observed a consistent reduction in contrast opacification, as measured by Hounsfield unit gradients, in arteries with progressively more severe stenoses reported as a percent of normal vessel diameter. Interestingly, TAG was able to improve the diagnostic accuracy of coronary computed tomography stenosis alone to discriminate lesions of anatomically obstructive quality by invasive coronary angiography. On the basis of these findings, more than a dozen reports have followed that demonstrated similar results against not only invasive angiography, but also fractional flow reserve and myocardial perfusion tools. Yet more recently, Stuijfzand et al. (2) recently evaluated TAG compared with an invasive fractional flow reserve (FFR) reference standard. In 34 patients with hemodynamically significant CAD, as defined by an FFR ≤0.8, TAG—both including and excluding calcified coronary segments—did not effectively diagnose hemodynamically significant lesions.
More than 40 years ago, Gould et al. (3) reported in 12 dogs the flow responses and regional distributions of coronary arteries at rest and during hyperemia, noting that hemodynamically significant reductions in flow were more observed in the hyperemic state. This study was supported by that of Wilson et al. (4), who examined coronary flow reserve as a function of arterial stenosis. Using a papaverine-based approach to induce hyperemia, these investigators compared individuals with stenoses versus those with normal coronary arteries. At maximal hyperemia, the changes in coronary blood flow velocity were evident in patients with higher grade coronary stenoses, with a curvilinear decline occurring at ∼50% area stenosis and steeply declining for lesions greater than 70%. Conversely, they noted a coronary flow reserve cutpoint of 3.5 to be effective in identifying lesions smaller and greater than 70% stenosis. These and an array of other studies have established the importance of increased coronary flow—either by exercise or pharmacologic vasodilation—to excellently identify hemodynamically significant lesions, with rest-related translesional flow or pressure differences less effective in their discriminatory capability to detect or exclude functionally important coronary lesions.
How, then, can the results of Choi et al. (1), Stuijfzand et al. (2), Gould et al. (3), and Wilson et al. (4) be reconciled? Some have posited that TAG is sensitive enough to allow the generally small differences in coronary flow at rest. Others have argued that the nitroglycerin often administered before coronary CTA performance—combined with the vasodilatory effect of iodinated contrast itself—induces an effective hyperemic state that allows the changes in contrast opacification to reflect hemodynamic significance. In their study, Stuijfzand et al. (2) commented on the large overlap between the groups of individuals with versus without >50% anatomic stenosis and reduced Thrombolysis in Myocardial Infarction (TIMI) flow in the study by Choi et al. (1). Thus, it also remains possible that, although intergroup differences exist for TAG positivity versus negative in stenoses >50%, only highly stenotic vessels (e.g., >90%) may result in reduced TAG, thus limiting its utility in those with hemodynamically positive significance but lower grade stenoses. For whichever reason, it seems unlikely for an at-rest coronary CTA to be able to offer a measure of functional ischemia generally proven only to be diagnostic at the hyperemic state.
These concepts were evaluated in a very elegant study by Park et al. (5), as reported in this issue of iJACC. To ascertain the mechanism of TAG, the authors performed a 3-step investigation. First, they studied phantoms with idealized vessels of varying diameter and at different contrast opacifications. Second, they performed an experimental in vivo study in a canine model to examine intraluminal attenuations by TAG using a prescribed stenosis model. Third, they evaluated 62 patients and 152 arteries that underwent 320-detector row coronary CTA and invasive angiography. In the phantom experiments, the investigators observed a reliable reduction in TAG as a function of decreasing vessel diameter, as expressed by a transluminal diameter gradient. These in vitro findings were confirmed in the canine model, in which the maximal Hounsfield unit opacification was consistently smaller for smaller arteries than larger arteries, a finding independent of stenosis severity. Indeed, no differences in TAG were noted for stenotic versus nonstenotic vessels. Unsurprisingly, the in vivo patient-based studies substantiated the pre-clinical findings. These results should be considered generalizable, given the in vitro, pre-clinical in vivo, and clinical in vivo assessments.
Fundamental to our evaluations of the physiologic significance of CAD has been the provocation of an enhanced coronary flow state through exercise or pharmacologic hyperemia, and the initial observations of Choi et al. (1) seemed contrary to the very basic tenets that guide our approach. Importantly, the findings of Park et al. (5) do not obviate the results observed by Choi et al. (1) Instead, the changes in TAG as a function of stenosis severity and TIMI frame count simply reflect a confounding variable not adequately accounted for in the initial studies. Of interest, the study by Stuijfzand et al. (2) was a careful prospective evaluation of consecutive patients of intermediate pre-test likelihood and without a history of manifest CAD. In their study, no relation of TAG was observed in relation to invasive fractional flow reserve, a widely accepted “gold standard” for determining lesion functionality. Their seeming dissonance to the TAG studies that preceded it may lie in the choice of patients, with many prior studies evaluating patients with highly stenotic vessels or chronic total occlusions, where TIMI frame count might be expectedly lower. Although TIMI frame count may be considered a method of functional CAD evaluation, it does not offer the same measure of physiologic provocation of ischemia that FFR offers.
Perhaps the most convincing argument to be made to support the study by Park et al. (5)—and to negate the contention of TAG to identify hemodynamically significant CAD—may be the symptomatic state of the patients themselves. Allowance of TAG from resting coronary CTA as a measure of functional CAD implies impaired basal coronary flow that, without an adequate collateral circulation, should coexist with symptomatic rest angina. In the absence of symptoms, prior TAG studies may have offered a novel finding—related to vessel size as proven by the study by Park et al. (5)—but existed in discord with the physiologic principles of CAD that we have proven and come to rely upon. As appropriately stated by the poet David Law Proudfit: “Nature abhors imperfect work and on it lays her ban” (6). The imperfection in TAG is not in the veracity of its observation, but simply in the reason for its existence. The study by Park et al. (5) has determined this, and the authors should be commended for their careful evaluation and ascertainment of the reason of this imaging finding. With TAG, what we see is true and merits belief, but on the basis of a mechanism that explains it, the proven methods of determining perturbations in coronary flow through hyperemic provocation remain warranted and needed.
↵∗ Editorials published in JACC: Cardiovascular Imaging reflect the views of the authors and do not necessarily represent the views of JACC: Cardiovascular Imaging or the American College of Cardiology.
This work was supported by grants from the National Institutes of Health (R01 HL111141, R01 HL 115150, and R01 HL 118019) and a gift from the Dalio Foundation.
Dr. Min is a consultant to GE Healthcare, HeartFlow, Cardiovascular Research Foundation, and Abbott Vascular; and serves on an advisory board for Arineta.
- 2016 American College of Cardiology Foundation
- Wilson R.F.,
- Marcus M.L.,
- White C.W.
- Park E.-A.,
- Lee W.,
- Park S.J.,
- Kim Y.K.,
- Hwang H.Y.
- Proudfit D.L.