Deep Learning-Derived Body Composition Analysis Predicts Long-Term Mortality After Transcatheter Aortic Valve...

Area of Science:

• Cardiovascular Imaging
• Medical Physics
• Geriatric Medicine

Background:

• Transcatheter aortic valve replacement (TAVR) is a key treatment for aortic stenosis.
• Patient outcomes after TAVR can be influenced by various factors, including body composition.
• Preprocedural assessment is crucial for optimizing TAVR success.

Purpose of the Study:

• To investigate the association between body composition metrics from computed tomography (CT) angiography and all-cause mortality post-TAVR.
• To determine if specific body composition parameters can predict long-term survival after TAVR.

Main Methods:

• Utilized CT angiography data from 2642 TAVR patients (2011-2023).
• Quantified skeletal muscle (SM), subcutaneous adipose tissue (SAT), visceral adipose tissue (VAT), and SM index (SMI) using a deep learning model.
• Assessed 3-year all-cause mortality using multivariable Cox proportional hazards models.

Main Results:

• Lower SM, SAT, VAT, and SMI were independently associated with increased 3-year all-cause mortality.
• Specific thresholds for these metrics indicated higher mortality risk (e.g., SM <128 cm², SMI <41 cm²/m²).
• Adjusted hazard ratios indicated significant associations for all analyzed body composition parameters.

Conclusions:

• Diminished skeletal muscle and adipose tissue reserves are significant independent predictors of mortality after TAVR.
• Automated CT-derived body composition analysis offers a promising tool for preoperative risk stratification in TAVR candidates.
• These findings can aid in refining clinical decision-making for patients undergoing TAVR.