Usefulness of Psoas Muscle Area to Predict Mortality in Patients Undergoing Transcatheter Aortic Valve Replacement

Abstract

Frailty has become high-priority theme in cardiovascular diseases because of aging and increasingly complex nature of patients. Low muscle mass is characteristic of frailty, in which invasive interventions are avoided if possible because of decreased physiological reserve. This study aimed to determine if the psoas muscle area (PMA) could predict mortality and to investigate its utility in patients who underwent transcatheter aortic valve replacement (TAVR). We retrospectively reviewed 232 consecutive patients who underwent TAVR. Cross-sectional areas of the psoas muscles at the level of fourth lumbar vertebra were measured by computed tomography and normalized to body surface area. Patients were divided into tertiles according to the normalized PMA for each gender (men: tertile 1, 1,708 to 1,178 mm(2)/m(2); tertile 2, 1,176 to 1,011 mm(2)/m(2); and tertile 3, 1,009 to 587 mm(2)/m(2); women: tertile 1, 1,436 to 962 mm(2)/m(2); tertile 2, 952 to 807 mm(2)/m(2); and tertile 3, 806 to 527 mm(2)/m(2)). Smaller normalized PMA was independently correlated with women and higher New York Heart Association classification. After adjustment for multiple confounding factors, the normalized PMA tertile was independently associated with mortality at 6 months (adjusted hazard ratio 1.53, 95% confidence interval 1.06 to 2.21). Kaplan-Meier analysis showed that tertile 3 had higher mortality rates than tertile 1 at 6 months (14% and 31%, respectively, p = 0.029). Receiver-operating characteristic analysis showed that normalized PMA provided the increase of C-statistics for predicting mortality for a clinical model and gait speed. In conclusion, PMA is an independent predictor of mortality after TAVR and can complement a clinical model and gait speed.