Coronary CT Fractional Flow Reserve before Transcatheter Aortic Valve Replacement: Clinical Outcomes

Area of Science:

• Cardiovascular Imaging
• Interventional Cardiology
• Medical Artificial Intelligence

Background:

• The clinical utility of CT angiography-derived fractional flow reserve (CT-FFR) in guiding pre-transcatheter aortic valve replacement (TAVR) decisions remains unclear.
• Accurate risk stratification is crucial for optimizing outcomes in patients with severe aortic stenosis undergoing TAVR.

Purpose of the Study:

• To assess the predictive capability of machine learning-based CT-FFR for adverse clinical outcomes in patients referred for TAVR.
• To determine if CT-FFR enhances the prognostic value of current noninvasive assessment methods.

Main Methods:

• Retrospective analysis of 196 patients with severe aortic stenosis undergoing TAVR, with CT-FFR derived using an on-site machine learning algorithm.
• Clinical endpoints included major adverse cardiac events (MACE) and all-cause mortality, assessed via survival analysis.
• Model performance was evaluated using the C-index, comparing models with and without CT-FFR.

Main Results:

• CT-FFR (abnormal ≤0.75) was independently associated with MACE (hazard ratio [HR] 4.0; 95% CI: 1.5, 10.5; P = .01).
• Incorporating CT-FFR into existing predictive models significantly improved their ability to predict MACE (P = .002), with a C-index of 0.71.
• CT-FFR did not significantly predict all-cause mortality or improve models for this outcome.

Conclusions:

• CT angiography-derived fractional flow reserve is a significant predictor of major adverse cardiac events in TAVR candidates.
• CT-FFR integration enhances the predictive power of coronary CT angiography for MACE in this patient population.
• Machine learning-based CT-FFR shows promise for improving risk stratification prior to TAVR.