Impact of computed-tomography defined sarcopenia on outcomes of older adults undergoing transcatheter aortic valve implantation.

Background: The adoption of Computed tomography (CT)-defined sarcopenia to risk stratify transcatheter aortic valve implantation (TAVI) candidates remains limited by a lack of both standardized definition and evidence of independent value over currently adopted mortality prediction tools.

Methods: 391 consecutive TAVI patients with pre-procedural CT scan were included (81 ​± ​6 years, 57.5% male, STS-PROM score 4.4 ​± ​3.6%) and abdominal muscle retrospectively quantified. The two definitions of radiologic sarcopenia previously adopted in TAVI studies were compared (psoas muscle area [PMA] at the L4 vertebra level: "PMA-sarcopenia"; indexed skeletal muscle area at the L3 vertebra level: "SMI-sarcopenia"). The primary endpoint was longer available-term all-cause mortality. Secondary endpoints were Valve Academic Research Consortium-2-defined in-hospital and 30-day outcomes.

Results: SMI- and PMA-sarcopenia were present in 192 (49.1%) and 117 (29.9%) patients, respectively. After a median of 24 (12-30) months follow-up, 83 (21.2%) patients died. PMA-(adj-HR 1.81, 95%CI 1.12-2.93, p ​= ​0.015), but not SMI-sarcopenia (adj-HR 1.23, 95%CI 0.76-2.00, p ​= ​0.391), was associated with all-cause mortality independently of age, sex and in-study outcome predictors (atrial fibrillation, hemoglobin, history of peripheral artery disease, cancer and subcutaneous adipose tissue). PMA-defined sarcopenia provided additive prognostic value over current post-TAVI mortality risk estimators including STS-PROM (p ​= ​0.001), Euroscore II (p ​= ​0.025), Charlson index (p ​= ​0.025) and TAVI2-score (p ​= ​0.020). Device success, early safety, clinical efficacy and 30-day all-cause death were unaffected by sarcopenia status regardless of definition.

Conclusions: PMA-sarcopenia (but not SMI-sarcopenia) is predictive of 2 year mortality among TAVI patients. The prognostic information provided by PMA-sarcopenia is independent of the tools currently adopted to predict post-TAVI mortality in clinical practice.