Aims
Risk stratification and individual risk prediction plays a key role in making treatment decisions in patients with complex coronary artery disease (CAD).
The aim of this study was to assess whether machine learning (ML) algorithms can improve discriminative ability and identify unsuspected, but potentially important, factors in the prediction of long-term mortality following percutaneous coronary intervention (PCI) or coronary artery bypass grafting (CABG) in patients with complex CAD.
Methods and results
To predict long-term mortality, the ML algorisms were applied to the SYNTAXES database with 75 pre-procedural variables including demographic and clinical factors, blood sampling, imaging, and patient-reported outcomes. The discriminative ability and feature importance of the ML model was assessed in the derivation cohort of the SYNTAXES trial using a 10-fold cross validation approach.
The ML model showed an acceptable discrimination (AUC = 0.76) in cross-validation. C-reactive protein, patient-reported pre-procedural mental status, Gamma-glutamyl transferase, and HbA1c were identified as important variables predicting 10-year mortality.
Conclusions
ML algorithms disclosed unsuspected, but potentially important prognostic factors of very long-term mortality among patients with CAD. A “mega-analysis” based on large randomized or non-randomized data, so called “BIG Data”, may be warranted to confirm these findings.
Coronary computed tomographic angiography (CCTA) is becoming the first-line investigation for establishing the presence of coronary artery disease and, with fractional flow reserve (FFR CT ), its haemodynamic significance. In patients without significant epicardial obstruction, its role is either to rule out atherosclerosis or to detect subclinical plaque that should be monitored for plaque progression/regression following prevention therapy and provide risk classification. Ischaemic non-obstructive coronary arteries are also expected to be assessed by non-invasive imaging, including CCTA. In patients with significant epicardial obstruction, CCTA can assist in planning revascularisation by determining the disease complexity, vessel size, lesion length and tissue composition of the atherosclerotic plaque, as well as the best fluoroscopic viewing angle; it may also help in selecting adjunctive percutaneous devices (e.g., rotational atherectomy) and in determining the best landing zone for stents or bypass grafts.
Background
In patients with complex CAD, the presence of left main (LM) disease is an important prognostic factor in assessing the risk balance between PCI and CABG. Functional assessment has become standard of care to evaluate the significance of coronary stenosis and to justify the performance of PCI in the contemporary practice. FFRCT is a well-established method based on 3D reconstruction of coronary artery derived from CCTA. The Murray law-based quantitative flow reserve (μQFR) is a novel computational method of invasive angiography relying on a single angiographic view that takes into account side branches diameters to compute fractal flow division. The aim of the current analysis is to evaluate in patients with complex CAD the feasibility of μQFR in LM bifurcation and its diagnostic concordance with FFRCT. The impact of the optimal viewing angle defined by CCTA on the physiological assessment of the LM bifurcation using a single angiographic view was also evaluated.
Methods
In 299 consecutive patients with 3-vessel disease with or without LM coronary artery disease, up to 3 analyzable fluoroscopic projections per patient were analysed with μQFR retrospectively. FFRCT and μQFR were measured at 3 fiducial landmark points: i) point of LM bifurcation (POB); ii) proximal LAD 10 mm distal to POB; ii) proximal LCX 10 mm distal to POB. CCTA-based “optimal viewing angle” of LM bifurcation are computed by creating a 3-point closed spline involving the LM, LAD, and LCX at 5mm from the POB and subsequently by reconstructing the “en face” fluoroscopic viewing angle of the spline. The en face viewing angle provides an optimal assessment of the bifurcation geometry [1]. In terms of Rx gantry angulation, the closest angiographic projection to the optimal viewing angle derived from CCTA was defined as the “best fluoroscopic projection” for each patient.
Results
In 299 patients, 793 projections were analysed with μQFR and compared to FFRCT. Single view μQFR was analyzable in 100%. Correlation and agreement between μQFR and FFRCT for 793 projections in 299 patients are shown in Figure 1A, 2A. The Spearman's correlation coefficient showed moderate correlations at POB (r=0.481, p<0.001) and LCX (r=0.584, p<0.001), and strong correlation at LAD (r=0.642, p<0.001). Correlation and agreement between μQFR and FFRCT for best projections from each patient are shown in Figure 1B, 2B. Correlations were improved in the best projections with the following Spearman's correlation coefficient: at POB (r=0.522, p<0.001), LCX (r=0.622, p<0.001), and LAD (r=0.695, p<0.001).
Conclusion
Computation of μQFR from a single angiographic view has a high feasibility. Tailored optimal fluoroscopic view is essential for the physiological assessment of the LM bifurcation using a single angiographic view. Evaluation of diagnostic accuracy of μQFR warrants further analysis of the LMCAD after prospective planning of the optimal fluoroscopic view based on the selection of the best CCTA 3D view.
Funding Acknowledgement
Type of funding sources: None.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.