Rationale and development of the computed tomography simulated pressure loss index (CTSPLI) for non-invasive phenotyping of coronary atherosclerosis

Abstract Introduction Combining coronary CT angiography (CCTA) and computational fluid dynamics offers promising non-invasive alternatives for functional assessment of coronary artery disease (CAD). Purpose We describe the rationale and development of the Computer Tomography Simulated Pressure Loss Index (CTSPLI), a non-invasive computational metric based on CT-FFR that can objectively phenotype CAD at the coronary artery level. Materials CTSPLI was developed in 15 healthy individuals who underwent CCTA and PET-myocardial perfusion imaging and were characterized by lack of atherosclerotic plaques, normal endothelium-dependent-(cold pressor test-induced hyperemia), and endothelium independent-(adenosine-induced hyperemia) vasodilatation and hence normal epicardial coronaries and microvasculature. In these individuals, CT-FFR analysis identified functional disease as regions with an FFR drop of ≥0.0015/mm. CTSPLI was then calculated by combining the length of functional disease with the magnitude of pressure loss along the artery. CTSPLI provides a quantitative assessment of the distribution of flow resistance in the epicardial coronaries, where lower values (approaching 0) indicate diffuse CAD and higher values (approaching 1) focal CAD. A second cohort of 29 patients undergoing CCTA and CT-FFR due to suspected obstructive CAD was retrospectively enrolled for the clinical investigation of CTSPLI. Coronaries were phenotyped as having focal, diffuse of mixed CAD pattern based on a. visual assessment of the distribution of plaques and CT-FFR pullback patterns and b. using CTSPLI. Results A total of 76 arteries (LAD: 36, LCX: 19, RCA: 21) from 29 patients were analyzed. Expert visual assessment based on anatomy and CT-FFR classified 21 arteries as focally diseased, 46 as diffusely diseased and 9 with mixed pattern. CTSPLI-derived physiology led to reclassification of a total of 46 arteries (57.8%) with 42% of focal disease reclassified to a diffuse or mixed pattern, whereas 24% of diffuse disease was reclassified as functional focal CAD. Conclusion CTSPLI can effectively and objectively phenotype CAD patterns, potentially enabling non-invasive personalization of management and revascularization strategies.