Advertisement

Muscle Wasting Is Associated With Mortality in Patients With Cirrhosis

Published:September 05, 2011DOI:https://doi.org/10.1016/j.cgh.2011.08.028

      Background & Aims

      Sarcopenia, defined as a low level of muscle mass, occurs in patients with cirrhosis. We assessed its incidence among cirrhotic patients undergoing evaluation for liver transplantation to investigate associations between sarcopenia and mortality and prognosis.

      Methods

      We studied 112 patients with cirrhosis (78 men; mean age, 54 ± 1 years) who were consecutively evaluated for liver transplantation and had a computed tomography scan at the level of the third lumbar (L3) vertebrae to determine the L3 skeletal muscle index; sarcopenia was defined by using previously published, sex-specific cutoffs.

      Results

      Of the patients studied, 45 (40%) had sarcopenia. Univariate Cox analysis associated mortality with ascites (hazard ratio [HR], 2.12; P = .04), encephalopathy (HR, 1.99; P = .04), level of bilirubin (HR, 1.007; P < .01), international normalized ratio (HR, 7.69; P < .001), level of creatinine (HR, 1.01; P = .005), level of albumin (HR, 94; P = .008), serum level of sodium (HR, 89; P < .001), Model for End-Stage Liver Disease (MELD) score (HR, 1.14; P < .01), Child–Pugh score (HR, 2.84; P < .001), and sarcopenia (HR, 2.18; P = .006). By multivariate Cox analysis, only Child–Pugh (HR, 1.85; P = .04) and MELD scores (HR, 1.08; P = .001) and sarcopenia (HR, 2.21; P = .008) were independently associated with mortality. The median survival time for patients with sarcopenia was 19 ± 6 months, compared with 34 ± 11 months among nonsarcopenia patients (P = .005). There was a low level of correlation between L3 skeletal muscle index and MELD (r = −0.07; P = .5) and Child–Pugh scores (r = −0.14; P = .1).

      Conclusions

      Sarcopenia is associated with mortality in patients with cirrhosis. It does not correlate with the degree of liver dysfunction evaluated by using conventional scoring systems. Scoring systems should include evaluation of sarcopenia to better assess mortality among patients with cirrhosis.

      Keywords

      Abbreviations used in this paper:

      BMI (body mass index), CI (confidence interval), CT (computed tomography), HCC (hepatocellular carcinoma), HCV (hepatitis C virus), HR (hazard ratio), HU (Hounsfield unit), INR (international normalized ratio), MELD (Model for End-Stage Liver Disease), ROC (receiver operating characteristic), SMI (skeletal muscle index), TIPS (transjugular intrahepatic portosystemic shunt)
      See editorial on page 100.
      Prognostic assessment of patients with liver cirrhosis remains a difficult task; the natural history of these patients is particularly variable because of several factors, including hepatic synthetic function, presence and degree of portal hypertension, the cause of cirrhosis, the possibility of resolution of the underlying damaging process, and the occurrence of liver cancer.
      • Durand F.
      • Valla D.
      Assessment of prognosis of cirrhosis.
      Currently, Child–Pugh
      • Pugh R.N.
      • Murray-Lyon I.M.
      • Dawson J.L.
      • et al.
      Transection of the oesophagus for bleeding oesophageal varices.
      and Model for End-Stage Liver Disease (MELD)
      • Kamath P.S.
      • Kim W.R.
      Advanced Liver Disease Study Group
      The model for end-stage liver disease (MELD).
      scores constitute the best tools to predict mortality in patients with cirrhosis. Child–Pugh score (initially termed Child–Turcotte) was originally designed to predict mortality during surgery,
      • Child C.G.
      Surgery and portal hypertension.
      and it has been shown to be useful in determining prognosis, treatment response, and necessity for liver transplant.
      • Planas R.
      • Ballesté B.
      • Alvarez M.A.
      • et al.
      Natural history of decompensated hepatitis C virus-related cirrhosis: a study of 200 patients.
      • Albers I.
      • Hartmann H.
      • Bircher J.
      • et al.
      Superiority of the Child-Pugh classification to quantitative liver function tests for assessing prognosis of liver cirrhosis.
      MELD was originally developed as a prognostic model of early mortality in patients with cirrhosis who received a transjugular intrahepatic portosystemic shunt (TIPS)
      • Malinchoc M.
      • Kamath P.S.
      • Gordon F.D.
      • et al.
      A model to predict poor survival in patients undergoing transjugular intrahepatic portosystemic shunts.
      . The original MELD has subsequently been simplified,
      • Kamath P.S.
      • Wiesner R.H.
      • Malinchoc M.
      • et al.
      A model to predict survival in patients with end-stage liver disease.
      and currently it is widely used to predict short-term mortality in different patient populations with cirrhosis.
      • Salerno F.
      • Merli M.
      • Cazzaniga M.
      • et al.
      MELD score is better than Child-Pugh score in predicting 3-month survival of patients undergoing transjugular intrahepatic portosystemic shunt.
      • Angermayr B.
      • Cejna M.
      • Karnel F.
      • et al.
      Child-Pugh versus MELD score in predicting survival in patients undergoing transjugular intrahepatic portosystemic shunt.
      • Said A.
      • Williams J.
      • Holden J.
      • et al.
      Model for end stage liver disease score predicts mortality across a broad spectrum of liver disease.
      Moreover, in many liver transplant centers, MELD score has replaced the Child–Pugh score for priority organ allocation.
      • Freeman R.B.
      • Wiesner R.H.
      • Harper A.
      • et al.
      The new liver allocation system: moving toward evidence-based transplantation policy.
      The MELD score has the advantage over Child–Pugh score of being based on objective variables (serum bilirubin, international normalized ratio of prothrombin time [INR], and serum creatinine) rather than on subjective evaluation of the severity of clinical findings (ascites and encephalopathy). Nevertheless, the MELD score has limitations,
      • Merion R.M.
      • Wolfe R.A.
      • Dykstra D.M.
      • et al.
      Longitudinal assessment of mortality risk among candidates for liver transplantation.
      • Heuman D.M.
      • Abou-assi S.G.
      • Habib A.
      • et al.
      Persistent ascites and low serum sodium identify patients with cirrhosis and low MELD scores who are at high for early death.
      the most important of which are the variability of biochemical parameters and lack of evaluation of the nutritional and functional status.
      Previously, a variety of indexes to evaluate nutritional and functional capacity status in patients with cirrhosis have been tested with diverse results.
      • Merli M.
      • Riggio O.
      • Dally L.
      Does malnutrition affect survival in cirrhosis? PINC (Policentrica Italiana Nutrizione Cirrosi).
      • Gunsar F.
      • Raimondo M.L.
      • Jones S.
      • et al.
      Nutritional status and prognosis in cirrhotic patients.
      • Abad-Lacruz A.
      • Cabré E.
      • González-Huix F.
      • et al.
      Routine tests of renal function, alcoholism, and nutrition improve the prognostic accuracy of Child-Pugh score in nonbleeding advanced cirrhotics.
      • Alvares-da-Silva M.R.
      • Reverbel da Silveira T.
      Comparison between handgrip strength, subjective global assessment, and prognostic nutritional index in assessing malnutrition and predicting clinical outcome in cirrhotic outpatients.
      • Carey E.J.
      • Steidley D.E.
      • Aqel B.A.
      • et al.
      Six-minute walk distance predicts mortality in liver transplant candidates.
      Consequently, an optimal index for nutritional status in terms of availability, reproducibility, practicality, and prognostic performance is needed.
      Sarcopenia is defined as a muscle mass 2 standard deviations below the healthy young adult mean.
      • Baumgartner R.N.
      • Koehler K.M.
      • Gallagher D.
      • et al.
      Epidemiology of sarcopenia among the elderly in New Mexico.
      Although sarcopenia is associated with aging, it can also be present as a result of chronic diseases and malignancy,
      • Lang T.
      • Streeper T.
      • Cawthon P.
      • et al.
      Sarcopenia: etiology, clinical consequences, intervention, and assessment.
      and it ultimately leads to decreased functional capacity and higher risk of mortality in different groups.
      • Janssen I.
      Evolution of sarcopenia research.
      In this analysis of prospectively acquired data in patients with cirrhosis, we aimed to establish the frequency of sarcopenia, to evaluate correlations of sarcopenia with conventional prognostic scores for patients with cirrhosis (Child–Pugh, MELD), and to determine the relationship between sarcopenia and mortality.

      Patients and Methods

       Study Population

      One hundred twelve patients with cirrhosis who were consecutively evaluated for liver transplant at the University of Alberta Hospital and who had a computed tomography (CT) scan at the third lumbar (L3) vertebrae were selected. In our center routine diagnostic CT scans are regularly done for diagnosis of hepatocellular carcinoma (HCC) as well as for evaluation of biliary and vascular anatomy in all patients under consideration for liver transplantation.

       Clinical and Laboratory Assessments

      Data recovered from medical charts included gender, age, weight, height, liver cirrhosis etiology, presence of ascites or hepatic encephalopathy, liver biochemistries, serum albumin, serum creatinine, INR, and Child–Pugh and MELD scores. The clinical and laboratory data used for regression and correlation analysis and to calculate MELD and Child–Pugh scores were obtained within 1 week from the index CT used to determine the skeletal muscle index (SMI).

       Muscularity and Sarcopenia Assessment

      CT scans used for analysis were done solely as part of the liver transplant evaluation. A transverse CT image from L3 in the inferior direction was assessed from each scan. Images were analyzed with SliceOmatic V4.3 software (Tomovision, Montreal, Quebec, Canada), which enables specific tissue demarcation by using previously reported Hounsfield unit (HU) thresholds. Skeletal muscle is identified and quantified by HU thresholds of −29 to +150.
      • Mitsiopoulos N.
      • Baumgartner R.N.
      • Heymsfield S.B.
      • et al.
      Cadaver validation of skeletal muscle measurement by magnetic resonance imaging and computerized tomography.
      Muscles in the L3 region encompass psoas, erector spinae, quadratus lumborum, transversus abdominis, external and internal obliques, and rectus abdominis. The following HU thresholds were used for adipose tissues: −190 to −30 for subcutaneous and intermuscular adipose tissues,
      • Kvist H.
      • Sjöström L.
      • Tylén U.
      Adipose tissue volume determinations in women by computed tomography: technical considerations.
      and −150 to −50 for visceral adipose tissues.
      • Vehmas T.
      • Kairemo K.J.
      • Taavitsainen M.J.
      Measuring visceral adipose tissue content from contrast enhanced computed tomography.
      With these specific HU thresholds, measurements of the SMI are not influenced by the presence of ascites in patients with cirrhosis. Cross-sectional areas (cm2) were automatically computed by summing tissue pixels and multiplying by pixel surface area. All CT images were analyzed by 2 trained observers. Cross-sectional area of muscle and adipose tissue was normalized for stature (cm2/m2) as reported elsewhere.
      • Mourtzakis M.
      • Prado C.M.
      • Lieffers J.R.
      • et al.
      A practical and precise approach to quantification of body composition in cancer patients using computed tomography images acquired during routine care.
      The L3 SMI was expressed as cross-sectional muscle area/height,
      • Pugh R.N.
      • Murray-Lyon I.M.
      • Dawson J.L.
      • et al.
      Transection of the oesophagus for bleeding oesophageal varices.
      and cutoffs for sarcopenia were based on a CT-based sarcopenic study (L3 SMI, ≤38.5 cm2/m2 for women and ≤52.4 cm2/m2 for men).
      • Prado C.M.
      • Lieffers J.R.
      • McCargar L.J.
      • et al.
      Prevalence and clinical implications of sarcopenic obesity in patients with solid tumours of the respiratory and gastrointestinal tracts: a population-based study.

       Statistical Analyses

      The Fisher exact probability test was used to compare categorical variables, and the unpaired t test was used to compare differences in means of continuous variables. Nonparametric variables in independent samples were compared by the Mann–Whitney test. Survival was calculated by using methods of Kaplan–Meier, and they were compared by using the log-rank (Mantel–Cox) test.
      • Heinze G.
      • Gnant M.
      • Schemper M.
      Exact log-rank tests for unequal follow-up.
      Patients were followed from the date of the index CT (time zero) done as part of the liver transplant assessment and used to determine the L3 SMI until the date of death, liver transplantation, or the last visit. Prognostic factors for mortality were analyzed by Cox regression univariate and multivariate analysis.
      • Cox D.R.
      Regression models and life tables (with discussion).
      Biochemical parameters such as serum bilirubin, creatinine, and sodium were included in the regression analysis as dimensional variables, whereas clinical variables such as ascites, encephalopathy, and sarcopenia were entered as categorical variables (present or absent). Correlations between sarcopenia and MELD and Child–Pugh scores and between sarcopenia and biochemical parameters were determined by Pearson correlation coefficient (r) analysis. A receiver operating characteristic (ROC) curve was plotted to measure how the L3 SMI and other variables such as MELD score, Child–Pugh score, and serum sodium performed in predicting mortality at 3 and 6 months. The validity of the model was measured by the area under the curve or C statistic, because for binary outcomes C is identical to the area under the curve.
      • Hanley J.A.
      • McNeil B.J.
      The meaning and use of the area under a receiver operating characteristic (ROC) curve.
      Data are presented as the means ± standard error of the mean, absolute frequencies, percentages, or ranges in tables and text.

      Results

       Clinical and Biochemical Features of Patients With Cirrhosis

      Seventy-eight patients were men (70%), and the mean age was 54 ± 1 years (median, 54 years; range, 28–71 years). Etiology of liver cirrhosis was hepatitis C virus (HCV) (29%), alcohol (22%), autoimmune liver disease (19%), HCV and alcohol (16%), and others (13%). The remaining clinical and biochemical features are shown in Table 1.
      Table 1Patient Characteristics at Time of Muscularity Assessment
      Featuresn = 112Percentage or range
      Age (y)54 ± 128–71
      Men:women78:3470/30
      Weight (kg)81 ± 242–149
      Height (cm)170 ± 1147–198
      BMI (kg/m2)28 ± 618–53
      BMI category
       <18.5 kg/m211
       18.5–24.9 kg/m23632
       25–29.9 kg/m2140
       >30 kg/m23027
      Body surface area (m2)1.95 ± 0.021.33–2.75
      Ascites8374
      Encephalopathy1917
      Albumin (g/L) (normal, 35–50)31 ± 117–44
      Bilirubin (μmol/L) (normal, <20)69 ± 94–733
      ALT (U/L) (normal, <50)59 ± 511–400
      AST (U/L) (normal, <38)92 ± 616–322
      Creatinine (μmol/L) (normal, <115)86 ± 336–230
      Sodium (mmol/L) (normal, 133–146)137 ± 0.4125–148
      INR (normal, 0.8–1.2)1.4 ± 0.041.0–2.9
      Child–Pugh in points9 ± 0.25–13
      Child–Pugh score
       A1211
       B6659
       C3430
      MELD score13 ± 0.66–36
      Cirrhosis etiology
       Alcohol2522
       HCV3229
       Alcohol + HCV1816
       HBV22
       Autoimmune liver diseases
      Includes autoimmune hepatitis, primary biliary cirrhosis, and primary sclerosing cholangitis.
      2119
       Others
      Includes hemochromatosis, nonalcoholic fatty liver disease, and cryptogenic.
      1413
      Lumbar SMI (cm2/m2)51 ± 132–80
       Men53 ± 135–80
       Women45 ± 132–66
      Sarcopenia4540
       Men3950
       Women618
      Lumbar total adipose tissue index (cm2/m2)270 ± 1722–1147
       Men270 ± 2122–1147
       Women272 ± 2665–671
      NOTE. Numbers in the second column are means ± standard error or absolute frequencies.
      AP, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; HBV, hepatitis B virus.
      a Includes autoimmune hepatitis, primary biliary cirrhosis, and primary sclerosing cholangitis.
      b Includes hemochromatosis, nonalcoholic fatty liver disease, and cryptogenic.

       Lumbar Skeletal Muscle Index and Frequency of Sarcopenia in Patients With Cirrhosis

      Mean L3 SMI was 51 ± 1 cm2/m2 (median, 50; range, 32–80). Sarcopenia was present in 45 patients (40%). In Figure 1 we compare 2 cirrhotic patients with identical body mass index (BMI) (32 kg/m2). Abdominal CT images in horizontal plane are images taken at third lumbar vertebrae, and the red color indicates skeletal muscles. The patient on the left is sarcopenic, with lumbar skeletal index of 50 cm2/m2, and the patient on the right is not sarcopenic, with lumbar skeletal index of 70 cm2/m2.
      Figure thumbnail gr1
      Figure 1Comparison of 2 cirrhotic patients with identical BMI (32 kg/m2). Abdominal CT images taken at third lumbar vertebrae. Red color indicates skeletal muscles: rectus abdominis, oblique and lateral abdominal muscles, psoas, and paraspinal muscles. The patient on the left is sarcopenic with L3 SMI of 49.82 cm2/m2; patient on the right is not sarcopenic with L3 SMI of 70.8 cm2/m2.
      Sarcopenia was more common in men compared with women (50% vs 18%, P < .001). Age was similar among sarcopenic and nonsarcopenic patients (55 ± 1 vs 54 ± 1 years, P = .6). Sarcopenic patients had lower BMI (26 ± 0.7 vs 29 ± 0.8 kg/m2, P = .003) compared with nonsarcopenic patients. Otherwise, there were no significant differences in clinical and biochemical features among sarcopenic and nonsarcopenic patients, as shown in Table 2.
      Table 2Features Associated With Sarcopenia in Patients With Cirrhosis
      Features associated with mortalitySarcopenia (n = 45)No sarcopenia (n = 67)P value
      Age (y)55 ± 154 ± 1.6
      Gender (M:F)39:639:28.001
      Weight (kg)80 ± 282 ± 3.5
      BMI (kg/m2)26 ± 0.729 ± 0.8.003
      Lumbar SMI (cm2/m2)45 ± 155 ± 1<.001
      Lumbar total adipose tissue index (cm2/m2)249 ± 28285 ± 20.3
      Ascites29 (64)54 (81).08
      Encephalopathy9 (20)10 (15).6
      Creatinine (μmol/L) (normal, 50–115)87 ± 684 ± 4.6
      INR (normal, 0.8–1.2)1.4 ± 0.061.3 ± 0.04.3
      Albumin (g/L) (normal, 35–50)30 ± 0.831 ± 0.7.8
      Bilirubin (μmol/L) (normal, <20)78 ± 1863 ± 8.4
      Sodium (mmol/L) (normal, 133–146)136 ± 0.6137 ± 0.5.2
      MELD13.0 ± 1.112.6 ± 0.6.7
      Child–Pugh (A/B/C)6/25/146/41/20.7
      Child–Pugh (points)8.6 ± 0.38.6 ± 0.2.9
      Etiology of cirrhosis.5
       Alcohol11 (33)14 (21)
       HCV15 (24)17 (25)
       Alcohol+ HCV7 (16)11 (16)
       HBV02 (3)
       Autoimmune liver diseases
      Includes autoimmune hepatitis, primary biliary cirrhosis, and primary sclerosing cholangitis.
      9 (20)12 (18)
       Others
      Includes hemochromatosis, nonalcoholic fatty liver disease, and cryptogenic.
      3 (7)11 (16)
      NOTE. Numbers are presented as means ± standard error or absolute frequencies; numbers in parentheses are percentages.
      AP, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase.
      a Includes autoimmune hepatitis, primary biliary cirrhosis, and primary sclerosing cholangitis.
      b Includes hemochromatosis, nonalcoholic fatty liver disease, and cryptogenic.

       Features Associated With Survival in Patients With Cirrhosis in Univariate and Multivariate Cox Analysis

      By univariate Cox analysis, the presence of ascites (hazard ratio [HR], 2.12; P = .04), encephalopathy (HR, 1.99; P = .04), serum bilirubin level (HR, 1.007; P < .001), INR (HR, 7.69; P < .001), creatinine (HR, 1.01; P = .005), albumin (HR, 0.94; P = .008), sodium (0.89; P < .001), MELD (HR, 1.14; P < .001), Child–Pugh (HR, 2.84; P < .001), and sarcopenia (HR, 2.18; P = 0.006) were associated with mortality (Table 3). In the multivariate Cox analysis including the presence of sarcopenia, MELD score, Child–Pugh score, and serum sodium (excluding ascites, encephalopathy, albumin, INR, and creatinine to avoid colinearity), only Child–Pugh score (HR, 1.85; P = .04), MELD score (HR, 1.08; P = .001), and sarcopenia (HR, 2.21; P = .008) were independently associated with higher risks of mortality (Table 3).
      Table 3Features Associated With Mortality in Cirrhosis by Univariate and Multivariate Cox Analysis
      Death (n = 51)Alive (n = 61)HR95% CIP value
      Features associated with univariate analysis
       Age (y)55 ± 154 ± 11.020.98–1.06.4
       Gender (M:F)36:1542:191.060.58–1.93.9
       Weight (kg)80.7 ± 2.581.3 ± 2.60.790.99–1.02.8
       BMI (kg/m2)28.4 ± 0.925.5 ± 0.81.030.99–1.08.2
       Ascites42 (82)41 (67)2.121.03–4.37.04
       Encephalopathy12 (24)7 (12)1.991.04–3.83.04
       Creatinine (μmol/L) (normal, 50–115)93 ± 680 ± 31.011.00–1.02.005
       INR (normal, 0.8–1.2)1.5 ± 0.061.3 ± 0.037.693.83–15.47<.001
       Albumin (g/L) (normal, 35–50)29.7 ± 0.831.4 ± 0.70.940.90–0.98.008
       Bilirubin (μmol/L) (normal, <20)97.6 ± 17.744.9 ± 5.41.0071.004–1.009<.001
       Sodium (mmol/L) (normal, 133–146)136 ± 0.7138 ± 0.40.890.84–0.95.001
       Lumbar SMI (cm2/m2)
        Men51 ± 155 ± 10.950.91–0.99.02
        Women48 ± 344 ± 11.040.98–1.11.2
       Sarcopenia28 (55)17 (30)2.181.25–3.80.006
       MELD15.2 ± 0.910.7 ± 0.51.141.09–1.19<.001
       Child–Pugh (A/B/C)2/25/2410/41/102.841.76–4.56<.001
       Child–Pugh (points)9.3 ± 0.38.0 ± 0.21.561.32–1.85<.001
      Features associated with multivariate analysis
      Ascites, encephalopathy, creatinine, INR, bilirubin, and albumin were not included to avoid colinearity, because they are included in either Child–Pugh or MELD scores.
       Sodium (mmol/L) (normal, 133–146)136 ± 0.7138 ± 0.40.970.90–1.065.5
       Sarcopenia28 (55)17 (30)2.211.23–3.95.008
       MELD15.2 ± 0.910.7 ± 0.51.081.03–1.14.001
       Child–Pugh (A/B/C)2/25/2410/41/101.851.02–3.36.04
      NOTE. Numbers in parentheses are percentages.
      AP, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase.
      a Ascites, encephalopathy, creatinine, INR, bilirubin, and albumin were not included to avoid colinearity, because they are included in either Child–Pugh or MELD scores.
      We also performed a multivariate analysis including L3 SMI, serum sodium, and the individual components of MELD and Child–Pugh scores (creatinine, INR, bilirubin, albumin, and presence of ascites and hepatic encephalopathy), but without including the composite indexes. In this multivariate analysis only INR (HR, 3.67; P = .004) and the presence of sarcopenia (HR, 2.11; P = 0.02) were significantly and independently associated with higher risks of mortality (Supplementary Table 1).

       Correlations Between Sarcopenia and Biochemical Parameters, Prognostic Scores in Patients With Cirrhosis

      A poor correlation between sarcopenia and MELD score (r = 0.04, P = .7) and Child–Pugh score (r = 0.01, P = .9) was observed. Also, there was a poor correlation between sarcopenia and serum albumin (r = −0.02, P = .9) and serum sodium (r = −0.11, P = .2) (Figure 2) .
      Figure thumbnail gr2
      Figure 2Dispersion graphs depicting correlations between L3 SMI and MELD score (A), Child–Pugh score (B), serum albumin (C), and serum sodium (D).

       Survival Among Sarcopenic and Nonsarcopenic Patients

      After the index CT during a mean follow-up of 19 ± 2 months (median, 15 months; range, 1–82 months), 51 patients (46%) died, 11 patients (10%) received a liver transplant, and 50 patients (45%) were alive.
      Median survival for sarcopenic patients was 19 ± 6 months (95% confidence interval [CI], 7–30), compared with 34 ± 11 months (95% CI, 14–55) in nonsarcopenic patients (log-rank, P = .005) (Figure 3) ; 6-month probability of survival was 71% and 90%, respectively. The 1-year probability of survival was 53% and 83% in these same groups, but fewer patients were followed.
      Figure thumbnail gr3
      Figure 3Kaplan–Meier curve indicating the survival of patients with (- - -) and without (—) sarcopenia. The 6-month probability of survival was 71% and 90%, respectively (P = .005, log-rank test). One-year probability of survival was 53% and 83% in these same groups, but fewer patients were followed.
      Of the 51 patients who died, 25 patients (49%) died of liver failure, 15 (25%) died of sepsis, 3 (6%) died of HCC, 3 (6%) died of hepatorenal syndrome, 1 patient (2%) died of variceal bleeding, and 4 patients (8%) of other reasons (1 cholangiocarcinoma, 1 cerebral bleeding, 1 cardiac disease, and 1 unknown).
      The frequency of sepsis-related death was significantly higher in sarcopenic than nonsarcopenic patients (22% vs 8%, P = .02). Otherwise, there were no significant differences in the frequency of liver failure–related death (27% vs 19%, P = .5), hepatorenal syndrome (2% vs 3%, P = 1.0), and death related to HCC development (7% vs 0%, P = .06) among sarcopenic and nonsarcopenic patients.

       C-Statistics Analysis for Features Associated With Mortality in Cirrhosis

      In Supplementary Table 2 we show the ROC curve analysis for composite scores such as MELD score, Child–Pugh score, individual biochemical parameters, and the L3 SMI. C statistics were significant for MELD, Child–Pugh, bilirubin, albumin, INR, and sodium for predicting 3-month and 6-month mortality. Regarding the C statistics for the L3 SMI, it was significant for predicting mortality at 6 months, and in the subanalysis for male patients it was significant in predicting 3-month and 6-month mortality (Figure 4) .
      Figure thumbnail gr4
      Figure 4ROC curve for predicting 3- and 6-month mortality for MELD, Child–Pugh score, serum sodium, and L3 SMI.

      Discussion

      Our study indicates that sarcopenia is present in approximately 40% of patients with cirrhosis being evaluated for liver transplant, and sarcopenia constitutes a strong and independent predictive factor for mortality. This mortality risk is more than 2-fold higher in sarcopenic than nonsarcopenic patients. Moreover, sarcopenia does not correlate with degree of liver dysfunction evaluated with conventional scores such as Child–Pugh or MELD score and also does not correlate with other common biochemical parameters such as serum albumin and sodium.
      Interestingly, we found that the higher mortality risk in cirrhotic patients with sarcopenia seems to be related to a higher frequency of sepsis-related death and not to liver failure mortality. These findings are similar to a recent study that found correlation between protein malnutrition and sepsis in a cohort of hospitalized cirrhotic patients,
      • Merli M.
      • Lucidi C.
      • Giannelli V.
      • et al.
      Cirrhotic patients are at risk for health care-associated bacterial infections.
      and this might explain why conventional scores that reflect mainly liver function, such as MELD and Child–Pugh, do not detect mortality risks associated with low muscle mass. Also, in our analysis with ROC curves relating mortality at 3 and 6 months and comparison of C-statistics values, we found that L3 SMI including the whole group was useful to distinguish mortality at 6 months, and in the subanalysis on male patients it was useful to distinguish mortality at 3 and 6 months. Interestingly, the L3 SMI was not significant in the subanalysis of female patients, possibly because of the low number of female patients in our sample.
      Previously, the presence of sarcopenia has been demonstrated to increase risk of infections in other populations such as elderly patients
      • Cosquéric G.
      • Sebag A.
      • Ducolombier C.
      • et al.
      Sarcopenia is predictive of nosocomial infection in care of the elderly.
      ; therefore, sarcopenia might increase the risk of severe infections in cirrhotic patients, reflecting an impaired immunity.
      • Merli M.
      • Lucidi C.
      • Giannelli V.
      • et al.
      Cirrhotic patients are at risk for health care-associated bacterial infections.
      Also, sarcopenia has been associated with physical disability, injuries, and higher mortality in individuals with malignant
      • Mourtzakis M.
      • Prado C.M.
      • Lieffers J.R.
      • et al.
      A practical and precise approach to quantification of body composition in cancer patients using computed tomography images acquired during routine care.
      • Prado C.M.
      • Baracos V.E.
      • McCargar L.J.
      • et al.
      Sarcopenia as a determinant of chemotherapy toxicity and time to tumor progression in metastatic breast cancer patients receiving capecitabine treatment.
      and nonmalignant disease.
      • Cosquéric G.
      • Sebag A.
      • Ducolombier C.
      • et al.
      Sarcopenia is predictive of nosocomial infection in care of the elderly.
      • Morley J.E.
      • Baumgartner R.N.
      • Roubenoff R.
      • et al.
      Sarcopenia.
      This study analyzed the frequency and clinical impact of sarcopenia in cirrhotic patients.
      It is important to mention that a recent consensus recommends using the presence of both low muscle mass and low muscle function (strength or performance) for the diagnosis of sarcopenia.
      • Cruz-Jentoft A.J.
      • Baeyens J.P.
      • Bauer J.M.
      • et al.
      Sarcopenia: European consensus on definition and diagnosis—Report of the European Working Group on Sarcopenia in Older People.
      However, CT scan is a gold standard tool to quantify skeletal muscle mass
      • Heymsfield S.B.
      Development of imaging methods to assess adiposity and metabolism.
      and hence constitutes a good resource for objective and detailed nutritional/metabolic assessment of patients and identification of sarcopenia. Importantly, muscle mass alone has been extensively used to define sarcopenia and to associate sarcopenia with prognosis in several conditions, and despite the newly published European consensus recommendation, the use of muscle mass versus muscle strength in defining sarcopenia is still disputed.
      Moreover, a CT scan is readily available for review in a patient's chart and hence is a feasible way to assess sarcopenia. Even though the BMI was lower in cirrhotic patients with sarcopenia (26 ± 0.7 vs 29 ± 0.8 kg/m2, P = .003), only 1 patient in our cohort would be considered underweight by commonly accepted criteria (BMI ≤18.5 kg/m2). Therefore, sarcopenia is not exclusively present in underweight patients and constitutes an occult condition that can be present in cirrhotic patients with any BMI. Furthermore, measurement of sarcopenia is independent of the fluid retention that plagues accurate measurement of body weight and BMI in cirrhotic patients.
      As current scores to evaluate prognosis in patients with cirrhosis need appropriate modifications, previous studies have evaluated the importance of other biochemical parameters. For example, serum sodium has been shown to be an independent risk factor for mortality in patients with cirrhosis
      • Heuman D.M.
      • Abou-assi S.G.
      • Habib A.
      • et al.
      Persistent ascites and low serum sodium identify patients with cirrhosis and low MELD scores who are at high for early death.
      • Kim W.R.
      • Biggins S.W.
      • Kremers W.K.
      • et al.
      Hyponatremia and mortality among patients on the liver-transplant waiting list.
      ; however, serum sodium is highly variable in patients who are using diuretics. Moreover, we did not find any correlation between serum sodium and sarcopenia, and hyponatremia (defined as sodium <133 mmol/L) was only found in 13% of our patients.
      Other tests such as the 6-minute walk distance
      • Carey E.J.
      • Steidley D.E.
      • Aqel B.A.
      • et al.
      Six-minute walk distance predicts mortality in liver transplant candidates.
      seem to identify cirrhotic patients with low functional capacity and increased risk of death that might be due to sarcopenia. Further studies evaluating this test and correlations with sarcopenia are warranted.
      • Huo T.I.
      • Lee S.D.
      • Lin H.C.
      Selecting an optimal prognostic system for liver cirrhosis: the model for end-stage liver disease and beyond.
      One drawback of our study is that we included patients being evaluated for liver transplant but not yet listed; consequently, we do not know whether addition of sarcopenia to conventional scores such as MELD score would help to better predict mortality risk in patients waiting for liver transplant. However, because only the presence of sarcopenia, Child–Pugh and MELD scores were associated with higher mortality risks in our multivariate analysis, subsequent analysis with equations including these variables (MELD-sarcopenia) could be used in a prospective fashion to determine the predictive value on mortality in cirrhotic patients. Importantly, the frequency and clinical significance of sarcopenia recognized in the present study suggest that that identification of these patients might permit early preventive strategies to maintain muscle mass to improve their outcome.
      Another limitation of our study is that we used a definition of sarcopenia based on cutoff values from a different population,
      • Prado C.M.
      • Lieffers J.R.
      • McCargar L.J.
      • et al.
      Prevalence and clinical implications of sarcopenic obesity in patients with solid tumours of the respiratory and gastrointestinal tracts: a population-based study.
      because predefined values for sarcopenia in patients with cirrhosis are lacking. However, the values we used were derived from optimum stratification of the L3 SMI, finding the most significant P values to define the gender-specific cutoffs associated with mortality in patients with malignancy and obesity.
      • Prado C.M.
      • Lieffers J.R.
      • McCargar L.J.
      • et al.
      Prevalence and clinical implications of sarcopenic obesity in patients with solid tumours of the respiratory and gastrointestinal tracts: a population-based study.
      Moreover, we are planning a future project including a larger number of patients with cirrhosis to find the threshold value of the L3 SMI to treat as a dimensional variable, which best distinguishes patients with sarcopenia and those who did not have sarcopenia with respect to time to death.
      In summary, sarcopenia is a strong and independent predictor of mortality in cirrhosis. Sarcopenia does not correlate with degree of liver dysfunction evaluated with conventional scores (Child–Pugh and MELD). Further studies including sarcopenia with conventional scores might allow significantly better prediction of mortality among patients waiting for liver transplantation.

      Video Abstract

      Supplementary Table 1Features Associated With Mortality in Cirrhosis by Multivariate Cox Analysis
      Features associated with mortalityDeath (n = 51)Alive (n = 61)HR95% CIP value
      Sodium (mmol/L) (normal, 133–146)136 ± 0.7138 ± 0.40.960.89–1.05.4
      Sarcopenia28 (55)17 (30)2.111.13–3.94.02
      Bilirubin (μmol/L) (normal, <20)97.6 ± 17.744.9 ± 5.41.000.99–1.01.2
      Albumin (g/L) (normal, 35–50)30 ± 0.831 ± 0.71.020.97–1.08.4
      INR (normal, 0.8–1.2)1.5 ± 0.061.3 ± 0.033.671.50–8.99.004
      Creatinine (μmol/L) (normal, 50–115)93 ± 680 ± 32.640.99–1.02.1
      Ascites42 (82)41 (67)1.830.80–4.16.2
      Encephalopathy12 (24)7 (12)1.740.85–3.58.1
      NOTE. Numbers in parentheses are percentages. Variables included in the multivariate analysis were sodium, sarcopenia, ascites, encephalopathy, creatinine, INR, bilirubin, and albumin.
      Supplementary Table 2C-Statistics Analysis for Features Associated With Mortality in Cirrhosis
      Features associated with mortality3-Month mortalityP value6-Month mortalityP value
      MELD0.85 (0.75–0.95)<.0010.80 (0.69–0.91)<.001
      Child–Pugh (points)0.84 (0.74–0.94)<.0010.78 (0.68–0.88)<.001
      Bilirubin0.76 (0.65–0.88).0010.75 (0.62–0.87).001
      Albumin0.69 (0.56–0.82).020.65 (0.52–0.78).04
      INR0.81 (0.69–0.92)<.0010.78 (0.68–0.88)<.001
      Creatinine0.64 (0.48–0.80).070.60 (0.45–0.76).2
      Sodium0.72 (0.58–0.86).0070.70 (0.56–0.83).006
      Lumbar SMI
       All0.61 (0.47–0.75).20.67 (0.55–0.79).02
       Men0.71 (0.60–0.82)<.0010.76 (0.64–0.87).009
       Women0.64 (0.40–0.87).30.50 (0.26–0.75).9
      NOTE. Numbers in parentheses are 95% CI.

      References

        • Durand F.
        • Valla D.
        Assessment of prognosis of cirrhosis.
        Semin Liver Dis. 2008; 28: 110-122
        • Pugh R.N.
        • Murray-Lyon I.M.
        • Dawson J.L.
        • et al.
        Transection of the oesophagus for bleeding oesophageal varices.
        Br J Surg. 1973; 60: 646-649
        • Kamath P.S.
        • Kim W.R.
        • Advanced Liver Disease Study Group
        The model for end-stage liver disease (MELD).
        Hepatology. 2007; 45: 797-805
        • Child C.G.
        Surgery and portal hypertension.
        in: Child C.G. The liver and portal hypertension. W B Saunders, Philadelphia1964: 50-72
        • Planas R.
        • Ballesté B.
        • Alvarez M.A.
        • et al.
        Natural history of decompensated hepatitis C virus-related cirrhosis: a study of 200 patients.
        J Hepatol. 2004; 40: 823-830
        • Albers I.
        • Hartmann H.
        • Bircher J.
        • et al.
        Superiority of the Child-Pugh classification to quantitative liver function tests for assessing prognosis of liver cirrhosis.
        Scand J Gastroenterol. 1989; 24: 269-276
        • Malinchoc M.
        • Kamath P.S.
        • Gordon F.D.
        • et al.
        A model to predict poor survival in patients undergoing transjugular intrahepatic portosystemic shunts.
        Hepatology. 2000; 31: 864-871
        • Kamath P.S.
        • Wiesner R.H.
        • Malinchoc M.
        • et al.
        A model to predict survival in patients with end-stage liver disease.
        Hepatology. 2001; 33: 464-470
        • Salerno F.
        • Merli M.
        • Cazzaniga M.
        • et al.
        MELD score is better than Child-Pugh score in predicting 3-month survival of patients undergoing transjugular intrahepatic portosystemic shunt.
        J Hepatol. 2002; 36: 494-500
        • Angermayr B.
        • Cejna M.
        • Karnel F.
        • et al.
        Child-Pugh versus MELD score in predicting survival in patients undergoing transjugular intrahepatic portosystemic shunt.
        Gut. 2003; 52: 879-885
        • Said A.
        • Williams J.
        • Holden J.
        • et al.
        Model for end stage liver disease score predicts mortality across a broad spectrum of liver disease.
        J Hepatol. 2004; 40: 897-903
        • Freeman R.B.
        • Wiesner R.H.
        • Harper A.
        • et al.
        The new liver allocation system: moving toward evidence-based transplantation policy.
        Liver Transplant. 2002; 8: 851-858
        • Merion R.M.
        • Wolfe R.A.
        • Dykstra D.M.
        • et al.
        Longitudinal assessment of mortality risk among candidates for liver transplantation.
        Liver Transplant. 2003; 9: 12-18
        • Heuman D.M.
        • Abou-assi S.G.
        • Habib A.
        • et al.
        Persistent ascites and low serum sodium identify patients with cirrhosis and low MELD scores who are at high for early death.
        Hepatology. 2004; 40: 802-810
        • Merli M.
        • Riggio O.
        • Dally L.
        Does malnutrition affect survival in cirrhosis?.
        Hepatology. 1996; 23: 1041-1046
        • Gunsar F.
        • Raimondo M.L.
        • Jones S.
        • et al.
        Nutritional status and prognosis in cirrhotic patients.
        Aliment Pharmacol Ther. 2006; 24: 563-572
        • Abad-Lacruz A.
        • Cabré E.
        • González-Huix F.
        • et al.
        Routine tests of renal function, alcoholism, and nutrition improve the prognostic accuracy of Child-Pugh score in nonbleeding advanced cirrhotics.
        Am J Gastroenterol. 1993; 88: 382-387
        • Alvares-da-Silva M.R.
        • Reverbel da Silveira T.
        Comparison between handgrip strength, subjective global assessment, and prognostic nutritional index in assessing malnutrition and predicting clinical outcome in cirrhotic outpatients.
        Nutrition. 2005; 21: 113-117
        • Carey E.J.
        • Steidley D.E.
        • Aqel B.A.
        • et al.
        Six-minute walk distance predicts mortality in liver transplant candidates.
        Liver Transplant. 2010; 16: 1373-1378
        • Baumgartner R.N.
        • Koehler K.M.
        • Gallagher D.
        • et al.
        Epidemiology of sarcopenia among the elderly in New Mexico.
        Am J Epidemiol. 1998; 147: 755-763
        • Lang T.
        • Streeper T.
        • Cawthon P.
        • et al.
        Sarcopenia: etiology, clinical consequences, intervention, and assessment.
        Osteoporos Int. 2010; 21: 543-559
        • Janssen I.
        Evolution of sarcopenia research.
        Appl Physiol Nutr Metab. 2010; 35: 707-712
        • Mitsiopoulos N.
        • Baumgartner R.N.
        • Heymsfield S.B.
        • et al.
        Cadaver validation of skeletal muscle measurement by magnetic resonance imaging and computerized tomography.
        J Appl Physiol. 1998; 85: 115-122
        • Kvist H.
        • Sjöström L.
        • Tylén U.
        Adipose tissue volume determinations in women by computed tomography: technical considerations.
        Int J Obes. 1986; 10: 53-67
        • Vehmas T.
        • Kairemo K.J.
        • Taavitsainen M.J.
        Measuring visceral adipose tissue content from contrast enhanced computed tomography.
        Int J Obes Relat Metab Disord. 1996; 20: 570-573
        • Mourtzakis M.
        • Prado C.M.
        • Lieffers J.R.
        • et al.
        A practical and precise approach to quantification of body composition in cancer patients using computed tomography images acquired during routine care.
        Appl Physiol Nutr Metab. 2008; 33: 997-1006
        • Prado C.M.
        • Lieffers J.R.
        • McCargar L.J.
        • et al.
        Prevalence and clinical implications of sarcopenic obesity in patients with solid tumours of the respiratory and gastrointestinal tracts: a population-based study.
        Lancet Oncol. 2008; 9: 629-635
        • Heinze G.
        • Gnant M.
        • Schemper M.
        Exact log-rank tests for unequal follow-up.
        Biometrics. 2003; 59: 1151-1157
        • Cox D.R.
        Regression models and life tables (with discussion).
        J Stat Soc B. 1972; 34: 187-220
        • Hanley J.A.
        • McNeil B.J.
        The meaning and use of the area under a receiver operating characteristic (ROC) curve.
        Radiology. 1982; 143: 29-36
        • Merli M.
        • Lucidi C.
        • Giannelli V.
        • et al.
        Cirrhotic patients are at risk for health care-associated bacterial infections.
        Clin Gastroenterol Hepatol. 2010; 8: 979-985
        • Cosquéric G.
        • Sebag A.
        • Ducolombier C.
        • et al.
        Sarcopenia is predictive of nosocomial infection in care of the elderly.
        Br J Nutr. 2006; 96: 895-901
        • Prado C.M.
        • Baracos V.E.
        • McCargar L.J.
        • et al.
        Sarcopenia as a determinant of chemotherapy toxicity and time to tumor progression in metastatic breast cancer patients receiving capecitabine treatment.
        Clin Cancer Res. 2009; 15: 2920-2926
        • Morley J.E.
        • Baumgartner R.N.
        • Roubenoff R.
        • et al.
        Sarcopenia.
        J Lab Clin Med. 2001; 137: 231-243
        • Cruz-Jentoft A.J.
        • Baeyens J.P.
        • Bauer J.M.
        • et al.
        Sarcopenia: European consensus on definition and diagnosis—Report of the European Working Group on Sarcopenia in Older People.
        Age Ageing. 2010; 39: 412-423
        • Heymsfield S.B.
        Development of imaging methods to assess adiposity and metabolism.
        Int J Obes (Lond). 2008; 32: S76-S82
        • Kim W.R.
        • Biggins S.W.
        • Kremers W.K.
        • et al.
        Hyponatremia and mortality among patients on the liver-transplant waiting list.
        N Engl J Med. 2008; 359: 1018-1026
        • Huo T.I.
        • Lee S.D.
        • Lin H.C.
        Selecting an optimal prognostic system for liver cirrhosis: the model for end-stage liver disease and beyond.
        Liver Int. 2008; 28: 606-613

      Linked Article

      • Prognostic Value of Muscle Wasting in Cirrhotic Patients
        Clinical Gastroenterology and HepatologyVol. 10Issue 9
        • Preview
          It was with great interest that we read the study by Montano–Loza et al1 reporting an independent association between sarcopenia and mortality among cirrhotic patients. This measurement helps to codify a previously subjective component of the gastroenterologist's clinical evaluation of the cirrhotic patient. If validated, these findings might help to better identify patients with the highest risk of mortality, with or without liver transplantation. We noted 2 important issues that were not addressed in the text of the article.
        • Full-Text
        • PDF