Sarcopenia as a prognostic index of nutritional status in concurrent cirrhosis and hepatocellular carcinoma.
BACKGROUND AND AIMSAbnormal body composition such as severe skeletal muscle depletion or sarcopenia has emerged as an independent predictor of clinical outcomes in a variety of clinical conditions. This study is the first study to report the frequency and prognostic significance of sarcopenia as a marker of nutritional status in patients with hepatocellular carcinoma (HCC).
METHODSWe analyzed 116 patients with HCC who were consecutively evaluated for liver transplant. Skeletal muscle cross-sectional area was measured by CT. Sarcopenia was defined using previously established cutpoints.
RESULTSNinety-eight patients were males (85%), and the mean age was 58±6 years. Sarcopenia was present in 35 patients (30%). By univariate Cox analysis, male sex (HR, 3.84; P=0.02), lumbar skeletal muscle index (HR, 0.97; P=0.04), INR (HR, 8.18; P<0.001), MELD score (HR, 1.19; P<0.001), Child-Pugh (HR, 3.95; P<0.001), serum sodium (HR, 0.84; P<0.001), TNM stage (HR, 2.59; P<0.001), treatment type (HR, 0.53; P<0.001), and sarcopenia (HR, 2.27; P=0.004) were associated with increased risks of mortality. By multivariate Cox regression analysis, only MELD score (HR, 1.08; P=0.04), Child-Pugh (HR, 2.14; P=0.005), sodium (HR, 0.89; P=0.01), TNM stage (HR, 1.92; P<0.001), and sarcopenia (HR, 2.04; P=0.02) were independently associated with mortality. Median survival for sarcopenic patients was 16±6 versus 28±3 months in nonsarcopenic (P=0.003).
CONCLUSIONSSarcopenia is present in almost one third of patients with HCC, and constitutes a strong and independent risk factor for mortality. Our results highlight the importance of body composition assessment in clinical practice.
*Department of Oncology, Cross Cancer Institute †Division of Gastroenterology & Liver Unit, University of Alberta Hospital, Edmonton, AB, Canada ‡Department of Nutrition, Food and Exercise Sciences, The Florida State University, Tallahassee, FL., , , , , , ,
Aged, 80 and over
Proportional Hazards Models
Pub Type(s)Journal Article
Research Support, Non-U.S. Gov't