［Original Article］
Yamashita M, Kamiya K, Matsunaga A, Kitamura T, Hamazaki N, Matsuzawa R, Nozaki K, Ichikawa T, Nakamura T, Yamamoto S, Kariya H, Maekawa E, Meguro K, Ogura M, Yamaoka-Tojo M, Ako J, Miyaji K. Interact Cardiovasc Thorac Surg. 2019 Dec 30. pii: ivz307.
OBJECTIVES:
Although skeletal muscle density (SMD) is useful for predicting mortality, the cut-off in an acute clinical setting is unclear, especially in patients with cardiovascular disease (CVD). This study was performed to determine the preoperative SMD cut-off using the psoas muscle and to investigate the effect on postoperative outcomes, including sarcopaenia, in CVD patients.
METHODS:
Preoperative psoas SMD was measured by abdominal computed tomography in CVD patients. Postoperative sarcopaenia was defined according to the criteria of the Asia Working Group for Sarcopaenia. The Youden index was used to test the predictive accuracy of survival models. The prognostic capability was evaluated using multivariable survival and receiver operating characteristic curve analyses.
RESULTS:
Continuous data were available for 1068 patients (mean age 65.5 years; 63.6% male). A total of 105 (9.8%) deaths occurred during the 1.99-year median follow-up period (interquartile range 0.71-4.15). The psoas SMD cut-off estimated by the Youden index was 45 Hounsfield units with high sensitivity and moderate specificity for all-cause mortality and was consistent in various stratified analyses. After adjusting for the existing prognostic model, EuroSCORE II, preoperative and postoperative physical status, psoas SMD cut-off was predicted for mortality (hazard ratio 2.42, 95% confidence interval 1.32-4.45). The psoas SMD cut-off was also significantly associated with postoperative sarcopaenia and provided additional prognostic information to EuroSCORE II on receiver operating characteristic curve analysis (area under the curve 0.627 vs 0.678, P = 0.011).
CONCLUSIONS:
Reduced psoas SMD was associated with postoperative mortality and added information prognostic for mortality to the existing prognostic model in CVD patients.
PMID: 31886866
Comments