Sarcopenia Defined by Psoas Muscle Thickness Predicts Mortality After Transjugular Intrahepatic Portosystemic Shunt

Purpose To assess and compare the value of psoas muscle thickness at the level of the third lumbar (L3) vertebra (TPML) or umbilicus (TPMU) and skeletal muscle index (SMI) for diagnosing sarcopenia and predicting mortality in patients undergoing transjugular intrahepatic portosystemic shunt (TIPS). Materials and Methods Two hundred forty-nine patients undergoing TIPS were included in this retrospective study. The cut-offs of L3-SMI for sarcopenia were 42.0 cm(2)/m(2) in men and 38.0 cm(2)/m(2) in women. The cut-offs for TPML/height and TPMU/height to predict mortality was established using a receiver-operating characteristic analysis. The Kaplan-Meier and Cox regression were used for survival analyses. Results Compared with TPMU/height, TPML/height was more consistent with L3-SM for the diagnosis of sarcopenia (Kappa coefficient: 0.63 vs. 0.36 in men; 0.61 vs. 0.45 in women). The Cox analysis showed that both TPML/height and TPMU/height were independent risk factors for mortality. The optimal cut-off values of TPML/height and TPMU/height for mortality in men and women were 11.2 mm/m, 9.4 mm/m, 18.4 mm/m, 15.1 mm/m, respectively. There were 119 (47.8%), 87 (34.9%), and 82 (32.9%) patients diagnosed with sarcopenia in the TPMU/height, TPML/height, and L3-SMI models, respectively. Kaplan-Meier analysis showed that the overall survival was significantly lower in the sarcopenia group in all three models. Conclusion TPMU/height and TPML/height have a similar survival prognostic value as L3-SMI. TPML/height has better consistency with L3-SMI in diagnosing sarcopenia and is a more stable alternative to L3-SMI for diagnosing sarcopenia in patients undergoing TIPS.