Background: Transcatheter aortic valve replacement (TAVR) is often used in patients with high cardiovascular risk due to multiple comorbidities. Despite the role of obesity as a cardiovascular risk factor, improved long-term outcome in chronic diseases has been described in obese patients, leading to the medical hypothesis of an “obesity paradox”. In addition, an excess mortality occurs in sarcopenic patients. Here, we assessed the relevance of changes in body composition (fat, muscle) and related blood markers as predictors of TAVR outcome.

Methods: 403 TAVR patients (49 % male) were retrospectively analyzed. All-cause mortality was recorded over a median follow-up of 4.5 years. Subcutaneous, visceral and epicardial fat and psoas muscle area were quantified by computed tomography. Plasma levels of GDF15 (Growth differentiation factor 15) and leptin were analyzed by ELISA.

Results: Patients underwent TAVR via transapical (TA, 97 patients, 81.3±0.4 years) or transfemoral approach (TF, 306 patients, 79.8± 1.4years) and showed similar, overall 30-day or 1-year survival. Comparisons of patient outcome according to body mass index (BMI) subgroups demonstrated an increased 2- and 3-year mortality in severely obese patients (BMI>35kg/m²) compared to overweight or obese patients (BMI 25-34.9kg/m², p<0.05) but not compared to patients with a BMI<25kg/m². While all fat areas were positively correlated to BMI (epicardial fat: R²=0.05, p<0.01; visceral fat: R²=0.20, p<0.001; subcutaneous fat: R²=0.13, p<0.001), only an increase in epicardial fat area or visceral fat area but not subcutaneous fat area resulted in a higher short- (both p<0.01) and long-term mortality (both p<0.001). Severely obese patients (1781.3mm²±75.8, p<0.05) and lean patients (1729.4±52.8, p<0.01) showed significantly lower psoas muscle area compared to mildly obese patients (2055.2±91.7). Sarcopenia, determined as a reduction in psoas muscle area and increased visceral fat to psoas muscle ratio, was a strong, independent predictor of long-term mortality on multivariate logistic regression analysis (p<0.01). Serum levels of GDF15, a previously reported biomarker of muscle weakness, were highest in severely obese patients (2793.5pg/ml±123.2) compared to lean (2017.6±130.8), overweight (1881.8±127.4) or mildly obese patients (2054.2.5±124.1, all p<0.05) and showed a significant increase with mortality in all groups (GDF15 cutoff value: 2587pg/ml, AUC 0.94). Serum leptin increased with BMI but did not predict outcome.

Conclusions: Short- and long-term outcome in aged and chronically ill patients demonstrates a strong dependency on patient’s body composition. Morbidly increased visceral and epicardial fat, sarcopenia and increased serum GDF15 are strong predictors of adverse outcome. Patients with sarcopenic obesity represent a high-risk group.