Abstract 14608: Non-Invasive Derivation of Resting Coronary Artery Flow Directly From CT Angiography (CTA) via the Transluminal Attenuation Gradient (TAG) in Left- versus Right-Dominant Circulation
Introduction: The TAG characterizes the dropoff in CTA contrast enhancement along the length of a vessel. It can detect functionally significant CAD without additional CTA radiation or computation. However, its relationship to flow is not established.
Hypothesis: We sought to establish how TAG relates to previously-reported values of coronary flow.
Materials and Methods: TAG was measured in the 3 major coronary arteries (LAD,LCX,RCA) of 42 patients (35 right- [RD], 7 left/co-dominant [LD]) with <25% stenosis imaged with single heart-beat 320-row CT. TAG was adjusted for scan kVp based on the relationship of iodine concentration vs blood HU. First, the ratio of TAG in each artery for RD vs LD arteries was compared to that of invasively-measured flow reported in n=496 patients (PREDICTION trial, Sakamoto et al, Am J Cardiol 2013;111:1420). Second, in 34 patients with CTA bolus tracking images available, the rate at which contrast was increasing at CTA (ie, HU increase/sec) was measured as a linear fit of descending aortic attenuation in the tracking images vs time to calculate flow via the following theoretical relationship we derived for TAG and flow (Figure): Q[ml/sec]~Lumen Area[cm^2]*Contrast inflow slope[HU/sec]/TAG[HU/cm]. Lumen area was determined for the arterial length used for TAG measurement.
Results: TAG LD-to-RD ratios were within 2%, 6% and 30% of flow ratios in the LAD, LCX and RCA respectively (Table). LAD and LCX flow derived from TAG was within 4-16% of reported flow; RCA flow was over-/underestimated by 51%/20% in RD/LD patients, respectively (Table).
Conclusion: TAG in non-stenotic coronaries is related to resting coronary flow. Knowledge of the temporal change of inflowing contrast concentration enables derivation of reasonable values of coronary flow from CTA using the TAG. This is a critical step toward obtaining the flow boundary conditions for FFR-CT techniques to non-invasively detect functionally significant CAD.
Author Disclosures: D. Mitsouras: Research Grant; Modest; Toshiba America Medical Systems, Vital Images, a Toshiba Medical Systems Group Company. R. Sewatkar: None. A.A. Giannopoulos: None. M. Agrawal: None. E. George: None. M.K. Cheezum: None. S. Shanbhag: None. R. Blankstein: None. M.Y. Chen: None. F.J. Rybicki: Research Grant; Modest; Toshiba Medical Systems Corporation.
- © 2015 by American Heart Association, Inc.