10. The Role of Changes in Subcutaneous and Visceral Adiposity, Sarcopenic Obesity, and Myosteatosis/Muscle Quality in Cirrhosis: How to Diagnose It and Its Contribution to Prognosis
KeywordsBody composition abnormalitiesVisceral obesityLow subcutaneous adipositySarcopenic obesityMyosteatosisPrognosis
Body mass index (BMI) has been widely used as a clinically accessible assessment of body composition. However, BMI does not differentiate between the two main body compartments, i.e., muscle and adipose tissue which have different functions. While muscle is mainly responsible for mechanical activity, myokines produced by muscle are involved in regulating metabolism in muscle and other tissues . Adipose tissue is metabolically active tissue involved in fat metabolism and energy homeostasis; it is also involved in regulating glucose metabolism, insulin sensitivity, angiogenesis, appetite, and inflammation by secreting proteins called adipokines .
Summary of studies on CT-determined body composition abnormalities in cirrhosis
Cutoff for body composition abnormality
Montano-Loza et al., 2016 
678 patients with cirrhosis evaluated for LT
L3 SMI ≤41 cm2/m2 for women and ≤53 cm2/m2 for men with BMI > =25 and ≤43 cm2/m2
(BMI > 25 kg/m2)
Sarcopenic obesity was determined in 20% of the patient population and was associated with mortality
MELD, and Child–Pugh scores were higher in patients with sarcopenic obesity
Kobayashi et al., 2017 
465 patients who went through primary hepatectomy for HCC
L3 SMI <40.31 cm2/m2 for men and 30.88 cm2/m2 for women
Visceral adipose tissue area ≥ 100 cm2
Sarcopenic obesity was observed in 7% of the population
Sarcopenic obesity was a significant predictor of mortality and HCC recurrence
Montano-Loza et al., 2016 
678 patients with cirrhosis evaluated for LT
L3 muscle attenuation <41 HU in patients with a BMI up to 24.9 and <33 in those with a BMI ≥25
Myosteatosis was determined in 52% of patients and was an independent predictor of mortality
Bhanji et al., 2018 
675 patients with cirrhosis evaluated for LT
L3 muscle attenuation <41 HU in patients with a BMI up to 24.9 and < 33 in those with a BMI ≥25
Myosteatosis was identified in 52% of the population and was an independent predictor of both hepatic encephalopathy and mortality
High visceral adiposity
Terjimanian et al., 2016 
348 liver transplant recipients
Continuous variable (cm2)
Visceral fat area from T12 through L4
Visceral fat area was significantly associated with post-transplant mortality
Montano-Loza et al., 2017 
289 male patients with HCC
L3 VATI ≥65 cm2/m2
High visceral adiposity is an independent risk factor for HCC incidence and recurrence after liver transplant in male patients with cirrhosis
Low subcutaneous adiposity
Ebadi et al., 2018 
221 female patients with cirrhosis
L3 SATI<60 cm2/m2
Low SATI was an independent predictor of higher mortality in female patients with cirrhosis
Alcohol-induced cirrhosis tended to be higher in patients with low SATI
Body composition can be assessed by an extensive range of indirect and direct modalities such as anthropometry, bioelectrical impedance (BIA), dual-energy X-ray absorptiometry (DEXA), ultrasound (US), magnetic resonance imaging (MRI), and computed tomography (CT). While most of these modalities may be applicable in the general population, some are not appropriate in cirrhosis. A fluid shift which is a common complication in patients with decompensated cirrhosis may influence the accuracy of some of these modalities such as DEXA and BIA.
Given the association between body composition abnormalities and worse outcomes in patients with cirrhosis, early identification is critical to facilitate proper interventions in order to reverse body composition abnormalities in these patients. Moreover, evaluating the prevalence and clinical impact of body composition abnormalities on morbidity and mortality in cirrhosis can be used to develop a predictive model to work as a clinical tool to estimate the individual mortality risk of patients while waiting for and after LT. Therefore, developing a predictive model to enhance patient selection may help to circumvent futile LT and thus lead to proper advances in the policy for the optimization of the organ allocation criteria in the LT setting.
This chapter summarizes current knowledge regarding the various features of body composition in patients with cirrhosis, focusing on the prevalence and consequences of these abnormalities.
Association Between BMI and Outcomes in Patients with Cirrhosis
BMI equal or greater than 25 kg/m2 and 30 kg/m2 are, respectively, considered as overweight and obesity, according to the World Health Organization. Insulin resistance; alterations in the production of growth factors; hormones and adipokines including leptin, adiponectin, and inflammatory cytokines such as tumor necrosis factor-α (TNF- α); and interleukin-6 (IL-6) are main characteristics of obesity. The worsened physiologic state in obesity may associate with a rapid progression of chronic liver diseases. Seven to 10% weight loss has been suggested by European associations to improve liver histology and enzymes in nonalcoholic fatty liver .
In patients with end-stage liver disease, LT is the decisive management option. Despite higher peri- and postoperative complications in obese LT recipients, controversy remains regarding the prognostic significance of BMI extremes in predicting LT outcomes. Some studies suggested that obesity may be a paradoxical protective feature for mortality after LT in patients with cirrhosis. This contrary association between obesity and mortality, characterized as the “obesity paradox,” is well established in patients with various types of chronic diseases.
In one report, longer hospital stay and shorter survival were reported in patients at the extremes of BMI (BMI < 18.5 kg/m2 or a BMI > 40 kg/m2). Nevertheless, LT in severely obese patients with the lower MELD score (≤22) significantly improved the survival of patients suggesting allocation of additional MELD points to severely obese patients . Complications such as hepatic artery and portal vein thrombosis were the main reasons for elevated mortality rate in obese patients (BMI > 35 kg/m2) undergoing LT . Given confounding impacts of fluid accumulation on BMI, a modified BMI (conventional BMI ∗ serum albumin level) was suggested to predict outcomes after LT. Contrary to conventional BMI, the modified BMI was not a contraindication for LT .
When the impact of BMI on early post-LT outcome was investigated in the context of MELD score, elevated BMI was not associated with higher mortality risk or graft loss. Underweight patients (BMI < 18.5 kg/m2) had higher risk for mortality and graft loss, especially those with the lower MELD score (≤26), in comparison to normal weight recipients . Overall, these results suggest that in listed obese patients with an advanced liver disease, BMI in conjunction with MELD might be used to prioritize patients for organ allocation.
The present evidence concerning the impact of BMI on outcomes in patients with cirrhosis is not consistent, and part of the discrepancy may be explained by variation between studies in regard to the choice of a reference group and obesity definition (BMI upper-limit values). Besides, one of the major limitations of BMI in predicting outcomes in LT is that it is affected by fluid retention. Therefore, weight loss in malnourished patients may be masked by the accumulation of fluid. It also does not discriminate the composition and abnormalities of the body mass to properly predict the risk of adverse outcomes associated with these abnormalities. Recent application of body composition modalities has demonstrated that obesity masks the presence and progression of sarcopenia (i.e., sarcopenic obesity), and it is also associated with elevated fat infiltration into muscle (i.e., myosteatosis) . Both skeletal muscle abnormalities, i.e., sarcopenia and myosteatosis, are common in patients with cirrhosis. Therefore, the prognostic significance of body composition abnormalities in patients with cirrhosis should be evaluated using precise and unified diagnostic imaging rather than BMI across studies.
The Importance of Myosteatosis in Cirrhosis
Excess accumulation of lipids within skeletal muscle is a pathological phenomenon called myosteatosis. It is reflected by low skeletal muscle radiodensity on CT images and might be a manifestation of muscle loss. Pre-defined HU ranges for demarcating skeletal muscle cross-sectional area and attenuation on CT is −29 HU to 150 HU . Although the cutoff values to define normal and low attenuation muscle have not been standardized, muscle radiodensity of <33 HU in patients with a BMI ≥25 and < 41 in those with a BMI <25 was associated with shorter survival in cancer . Even though myosteatosis may be an extension of obesity, fluid retention in majority of patients with cirrhosis questions the applicability of these BMI-dependent cutoffs in cirrhotic population.
Myosteatosis is relatively poorly characterized in cirrhosis; however, it is emerging as a poor prognostic factor in these patients. Using cutoffs established in oncologic population , myosteatosis was identified in 52% of 678 patients with cirrhosis. Presence of myosteatosis was independently associated with worse survival and appeared to denote deterioration in physical condition rather than the severity of the liver disease. Despite longer intensive care unit stay in patients with myosteatosis, no significant difference in post-LT complications was reported between patients with and without myosteatosis . In our recent exploration of the association between myosteatosis and overt hepatic encephalopathy (HE) in patients with cirrhosis, frequency of myosteatosis was higher in patients with HE compared to the patients without (70% vs. 45%, p < 0.001); both sarcopenia and myosteatosis were independent predictors of HE .
Although myosteatosis has been defined as an excess accumulation of lipids within skeletal muscle, the composition of lipids seems to play more important role in pathology of muscle rather than the total amount of lipids per se . Comprehensive evaluation of muscle transcriptome in cancer recognized diabetogenic-like alterations in carbohydrate and lipid metabolism such as impaired lipid metabolism and diminished lipid oxidation as potential pathological mechanisms of myosteatosis . Impaired mitochondrial function and age-related differentiation of muscle stem cells into adipocytes  are further probable contributors to myosteatosis. Association between myosteatosis and deficits of physical function has been exhaustively characterized in aging; however, the prognosis of muscle quality in cirrhosis has not been well identified. In summary, poor muscle quality (low attenuation on CTs) was associated with adverse outcomes including mortality and complications such as HE in patients with cirrhosis. While muscle abnormalities such as sarcopenia and myosteatosis constitute important prognostic factors, they are not incorporated in conventional scores for prognosis in cirrhosis, such as the MELD or Child–Pugh scores. This requires further investigation.
Sarcopenia is associated with poor prognosis in various chronic diseases that is ignored in patients with higher BMI. Concordance of these two features, sarcopenia and obesity, called sarcopenic obesity has been gaining attraction in oncology. However, clinical predictors and prognostic significance of sarcopenic obesity in cirrhosis have not been widely investigated. A recent review reported the frequency of 20–35% for sarcopenic obesity in cirrhosis which was significantly associated with mortality . Montano-Loza et al. have profiled this body composition phenotype in patients with cirrhosis as the concurrent presence of sarcopenia and overweight or obesity (BMI > 25 kg/m2). Sarcopenic obesity was presented in 20% of 678 patients with cirrhosis and was more frequent in male, older patients with higher MELD and lower muscle attenuation when compared to patients without muscle abnormalities .
Possible mechanisms for the pathophysiology of sarcopenic obesity in cirrhosis have not been clearly determined. Age-related deterioration in body composition and decreased physical activity, chronic inflammation, and insulin resistance associated with both sarcopenia and elevated visceral obesity might play an important role .
Obesity is commonly defined as BMI of ≥25 or ≥30 kg/m2; however, its definition varies as defined by BMI, visceral obesity, and ratio of visceral to subcutaneous adipose tissue or whole body fat mass. Moreover, inconsistency in literature regarding the modalities and cutoffs used to define sarcopenia and obesity (visceral) exists and may lead to lack of standardized definition for sarcopenic obesity. Lastly, the accumulation of visceral adipose tissue but not subcutaneous adipose tissue appears to deliberate the majority of obesity complications. Thus, given limitations of using BMI to assess body compartments, prognostic significance of obesity should be evaluated based on adiposity index rather than BMI. Prospective studies are required to improve these definitions in patients with cirrhosis, considering their capability of predicting relevant clinical outcomes.
Prognostic Significance of Adipose Tissue Depots in Cirrhosis
Sex-dependent body composition disparity exists in patients with cirrhosis with females having higher adiposity and males having more skeletal muscle. Body composition discrepancy by sex might be related to differences in fat metabolism, hormonal characteristics, and anatomic location of adipose tissue deposition between male and females .
Visceral (VAT) and subcutaneous adipose tissue (SAT) are two main types of adipose tissue with differences in anatomic location, size of adipocytes, lipolytic capacity, insulin response, and adipokine secretion. VAT not only contains insulin-resistant adipocytes that are more responsive to catecholamine-induced lipolysis but also is an active producer of cytokines such as IL-6, TNF-α, and monocyte chemotactic protein 1 (MCP-1) . Elevated lipolysis of VAT, in response to catabolic stimuli, enables direct delivery of free fatty acids to the liver and consequently can cause elevated hepatic triglyceride deposition . SAT is involved in uptake and storage of circulating free fatty acids and triglycerides and also regulates insulin sensitivity, glucose and lipid metabolism, as well as immune response by producing adipokines mainly leptin [18, 20].
Although metabolic differences exist between SAT and VAT, potential differences in their contribution to the outcomes in cirrhosis have not been consistently demonstrated, partially due to the use of modalities with limited applicability and the variability among studies with regard to the time point in cirrhosis trajectory (pre- vs. post-LT) that patients are studied.
Cross-sectional imaging provides the opportunity to precisely quantify two main adipose tissue depots. CT-measured VAT and SAT, calculated as an area at the third lumbar vertebrate (L3) normalized to height and reported as indexes in cm2/m2, i.e., visceral adipose tissue index (VATI) and subcutaneous adipose tissue index (SATI), are the most commonly used indicator of body adiposity. Using this techniques, differences in adipose tissue distribution by sex have been characterized by males having more VATI, while SATI was predominant in female patients with cirrhosis . Variability in regional adipose tissue distribution by sex  and within each BMI category, as well as different behavior of adipose depots [18, 20], demonstrates the need to understand the prognostic significance of adiposity in cirrhosis by sex and depot.
Controversy remains regarding the association between visceral adiposity and outcomes in patients with cirrhosis as high visceral adiposity is reported to be related to poor survival in patients with hepatocellular carcinoma  as well as liver transplant recipients , whereas no association between VATI and mortality was observed in neither female nor male patients with cirrhosis evaluated for LT [21, 24]. Visceral adipose tissue has been linked to the higher probability of HCC incidence, recurrence, and outcome. CT-determined high visceral adiposity (VATI > 65 cm2/m2) was associated with higher risk of HCC before LT as well as recurrence after LT. However, this association between high VATI and HCC was only observed in male patients. The lack of association between VATI and HCC in female patients might be related to estrogen impact on prioritizing subcutaneous over visceral adipose tissue accumulation in females. Higher deposition of adipose tissue in the visceral region might be associated with chronic inflammatory state which favors neoplastic expansion and clarifies the higher risk for HCC before and after LT. Furthermore, high VATI is associated with insulin resistance, which might affect the risk of HCC .
In another large study of 1257 patients with various stages of HCC, among five measured body composition parameters including skeletal muscle index, muscle attenuation, VATI, SATI, and visceral to subcutaneous adipose tissue ratio, the main predictors of survival were sarcopenia, myosteatosis, and visceral adiposity. Interestingly, the prevalence of patients presenting with two or three poor prognostic body composition features was higher in underweight patients compared to normal weight (42% vs. 19%, P < 0.001) . However, the use of visceral to subcutaneous adipose ratio as an indicator of visceral obesity was questioned in previous studies as it may lead to the misclassification of patients. This ratio would be similar in people with high adiposity, with large amounts of both VATI and SATI, and in people with low adiposity who have small amounts of both VATI and SATI .
CT image-based analysis of body composition in patients with cirrhosis evaluated for LT revealed that subcutaneous adipose tissue but not muscle is an independent predictor of mortality in female patients with cirrhosis. In a recent study of 221 female patients evaluated for liver transplantation, SATI was independently associated with mortality after adjusting for age, alcohol-related cirrhosis, albumin, and MELD . Interestingly, modification of MELD to include SATI showed an excellent discriminative performance for predicting mortality in female patients. This essentially highlights the poor prognostic significance of malnutrition in cirrhosis which has been neglected in the current organ allocation criteria in LT setting. Female patients with low SATI (<60 cm2/m2), evaluated or listed for LT, had higher risk of mortality. The clarification for the relationship between SATI and female mortality in cirrhosis has not been clinically recognized. However, severe energy exhaustion triggered by cirrhosis and low serum levels of leptin as an indicator of malnutrition are possible explanations. In addition, estrogen replacement therapy which favors fat accumulation in subcutaneous over visceral depots in females  was associated with prolonged survival in patients with HCC . Yet, these results need to be validated in larger prospective, multicentric studies in patients with cirrhosis.
Longitudinal Adipose Tissue Changes in Patients with Cirrhosis
Data on body composition changes in cirrhosis by sex has revealed that depletion of adipose tissue and skeletal muscle is more frequent in female and male patients, respectively . Liver cirrhosis is characterized by a significant reduction in body fat mass ; however, the timeline and pattern of loss of adipose tissue depots need to be established using validated body composition assessment tools in cirrhosis.
Assessment of whole body fat mass by triceps skinfold thickness (TSF) in patients with cirrhosis showed that fat loss, described as TSF below the fifth percentile of age- and sex-matched normal population, was more common in females than males [28, 29]. Sever depletion of body fat in female, malnourished hospitalized patients with cirrhosis was independently associated with lower survival rate . Body fat mass assessment by DEXA demonstrated a marked fat loss even in initial stages of liver disease  which was defined as total body fat <80% of the 50th percentile of the sex-matched control group. Fat loss augmented with advancing the severity of the liver disease [28, 29]. The metabolic pattern that drives fat loss in female patients with cirrhosis is similar to chronic diseases or starvation, whereas the metabolic pattern of muscle loss in male patients mimics the critical diseases .
Cross-sectional body composition assessment at the time of LT evaluation in patients with cirrhosis revealed association between body composition abnormalities and adverse prognosis. However, our understanding of the prognostic value of longitudinal changes in body composition parameters is limited. Result of studies analyzing sequential CT images should be interpreted cautiously as a change between −2% and 2% represents tissue maintenance.
In a retrospective cohort of 136 patients with cirrhosis, skeletal muscle loss occurred in 50% of patients, whereas 33% of patients were experiencing SAT loss with no changes in visceral adipose tissue. The baseline CT in this study was taken as part of the LT evaluation, and the second one was conducted at nearly 1 year later. While no association between mortality and skeletal muscle loss was observed in this study, loss of SAT increased mortality risk by twofold when compared to patients who gained/maintained SAT after adjusting for MELD score and serum albumin . SAT plays an imperative role as a key energy reservoir in the body; consequently, poor prognosis in patients with SAT loss might be related to the exhaustion of body energy reservoirs. Additional understanding of the prognostic significance of sex-specific body composition components, their interaction, as well as longitudinal changes is necessary for a better management of nutritional status of patients with cirrhosis.
Differences in body composition by sex that happen in patients with cirrhosis appear to influence prognosis. This suggests that cirrhotic patients with body composition abnormalities might be underlooked in liver transplant settings, and therefore prioritization of liver transplantation candidates merely based on the MELD score may lead to misclassification of patients. Inclusion of body composition abnormalities within the MELD score demonstrates promise to optimize the organ allocation criteria in the LT setting.