Sarcopenia is the independent predictor of mortality in critically ill patients with cirrhosis

Background: Sarcopenia is strongly associated with poor outcome in cirrhosis. There are little prospective data that sarcopenia influences outcomes in critically ill cirrhotics (CICs). Computed tomography (CT) is the gold standard for sarcopenia assessment in the intensive care unit (ICU), as it is independent of hydration status. Aim: This study aims to assess the prevalence of sarcopenia and study its impact on clinical outcomes in CICs. Methods: In this prospective observational study, CICs admitted to the liver ICU were enrolled, if meeting inclusion (age 18–70 years, abdominal CT scan within three months before ICU admission) and exclusion criteria (survival likely to be <24 h, coexisting chronic diseases). Clinical, hemodynamic, biochemical, and nutritional parameters, including length of stay (LOS), duration of mechanical ventilation (MV), development of new-onset infections (NOI), incidence of new-onset acute kidney injury (AKI), and overall survival, were recorded. CT images at the L3 level were analyzed using Slice-O-Matic V4.3 software to assess the skeletal muscle index (SMI) expressed as skeletal muscle area (cm2)/height (m2). Sarcopenia was defined if SMI was <50 cm2/m2 – males and <39 cm2/m2 – females. Data were analyzed using SPSS version 22. Results: Altogether 111 patients (M-83.8%; age 48.4±11.3 years; etiology: Alcohol – 56 [50.5%], non-alcoholic steatohepatitis – 27 [24.3%], viral – 12 [10.8%], and others – 16 [14.4%]; Child-Turcotte-Pugh – 11.9±1.8; model for end-stage liver disease – 27.8±7.3; sequential organ failure assessment – 10.5±4.1; APACHE – 23±8; and MV – 54 [48.6%]) were enrolled. Of these, 76 (68.5%) were sarcopenic and 35 (31.5%) non-sarcopenic. Sarcopenic CICs had higher overall mortality (72.4%) compared to non-sarcopenics (40%) (P=0.001, OR [95% CI] – 3.93 [1.69–9.12]), and higher prevalence of sepsis at ICU admission (53.9% vs. 31.4%, P=0.027, OR [95% CI] – 1.7 [1.0–2.92]) than non-sarcopenics. LOS, duration of MV, incidence of NOI, and development of new-onset AKI were comparable between groups. Multivariate binary logistic regression showed that sarcopenia, sepsis, and APACHE II score were independently associated with mortality. Conclusion: Two-thirds of CICs have sarcopenia at ICU admission, making them 1.7 times more susceptible to sepsis and increasing the risk of mortality by almost 4-fold in the ICU. Relevance for Patients: Almost 70% of patients with chronic liver disease admitted to the ICU have low muscle mass (sarcopenia). The presence of sarcopenia per se makes them highly prone to infections and increases the chances of death by almost 4-fold; thus, highlighting the importance of nutrition optimization in this patient group.


Introduction
Chronic liver disease (CLD) is a progressive disorder that adds significantly to the global burden of disease. Nevertheless, over the past years, various supportive therapies and interventions have been introduced that have improved the clinical outcome of these patients, resulting in a significant increase in patient longevity. However, very often these patients face life-threatening complications such as variceal bleed, hepatic encephalopathy (HE), spontaneous bacterial peritonitis (SBP), hepatorenal syndrome, hepatopulmonary syndrome (HPS), and infections that necessitate intensive care unit (ICU) admission [1].
Malnutrition is yet another complication in CLD which has been described as the phenotype of skeletal muscle loss with or without loss of fat mass [2]. Likewise, it may also contribute significantly in determining the fate of critically ill cirrhotics (CICs). Loss of muscle mass termed as sarcopenia negatively impacts the overall outcome of the disease and is associated with increased mortality [3], impaired health-related quality of life [4], increased health-care cost [5], and poor liver transplant outcomes [6].
Even though nutritional assessment in the ICU is essential, it is very challenging. Moreover, the traditional markers of assessment often lose their specificity in CLD and critical illness [7]. Computed tomography (CT) has been considered the gold standard for assessing skeletal muscle mass [8], which is performed as a part of the routine investigation in these patients.
Sarcopenia has been associated with decreased ventilator-free days [9], difficulty in weaning [10], and increased mortality [11] in the general critically ill patients. In a retrospective study, rapid muscle decline was found to be associated with increased mortality in CICs also [12]. However, the impact of sarcopenia on the clinical outcomes, especially mortality in the CICs, has not been studied prospectively. Hence, we planned this study to assess the prevalence of sarcopenia in CICs at ICU admission and to gauge its impact on the clinical outcomes during the course of ICU stay.

Study population
In this prospective observational study, 111 CICs, admitted to the liver ICU (LICU) of the Institute of Liver and Biliary Sciences, New Delhi, were enrolled from January 2020 to August 2020. The inclusion criteria were as follows: CICs of any etiology, age ≥ge years, and abdominal CT performed not more than 3 months before the LICU admission. Moribund patients and those with hepatocellular carcinoma or other malignancies were excluded from the study.

Study plan
The work plan was to screen all the patients for eligibility. Baseline and daily follow-up information including demographic, clinical, hemodynamic, and biochemical details were gathered on a structured pro forma till the patient's stay in the ICU. The clinical outcomes of sarcopenics and non-sarcopenics were compared after muscle mass was evaluated and data were recorded on an Excel sheet.

Diagnosis of the disease
Liver cirrhosis was diagnosed based on standard, clinical, biochemical, histological, and/or radiological criteria [13].
Acute-on-chronic liver failure (ACLF) was diagnosed using Asian Pacific Association for the Study of the Liver criteria [14].

Assessment of disease severity
The severity of liver disease was assessed according to Child-Turcotte-Pugh (CTP) and Model for End-stage Liver Disease (MELD) scores. Sequential Organ Failure Assessment (SOFA) and Acute Physiology and Chronic Health Evaluation (APACHE) scores were used to determine the severity of critical illness.

Assessment of sarcopenia
All CT scans were performed on a Discovery 750 HD 64-row spectral CT scanner (GE, USA). A transverse CT image from the third lumbar vertebrae (L3) was identified from each scan. The CT images were analyzed using Slice-O-Matic software V4.3 (Tomovision, Montreal, Canada). The skeletal muscle area (SMA, in cm 2 ) was defined as the sum of the paraspinal, psoas, transversus abdominis, interior/exterior oblique, and rectus abdominis muscles at the level of L3 region. The identification of skeletal muscle was based on a predefined CT density of −29 to +150 Hounsfield Unit (HU) (Figure 1) [15]. The SMA was then normalized for stature (height in m 2 ) and expressed as skeletal muscle index (SMI) in cm 2 /m 2 . Sarcopenia was defined if SMI was <50 cm/m 2 in males and <39 cm/m 2 in females [16].
Bilateral psoas muscle attenuation was expressed in HU and measured using Picture Archiving and Communication System imaging software. An oval region of interest of 1.5 cm 2 was placed in the most homogenous area of the psoas muscle ( Figure 2) and skeletal muscle density (SMD) was expressed as the mean of radiological muscle attenuation of both psoas muscles [17].

Medical management
All the CICs admitted to the LICU were managed as per the standard guidelines including nutritional support, endotracheal intubation, fluid resuscitation, and renal replacement therapy  (RRT). All the patients were screened for infection and treated empirically with broad-spectrum antibiotic combinations as per the treating ICU team.

Objective
The primary objective was to assess the prevalence of sarcopenia at ICU admission and study its impact on the overall ICU mortality in critically ill patients with cirrhosis. The secondary objective was to study the effect of sarcopenia on other clinical outcomes such as duration of mechanical ventilation (MV), length of ICU stay, prevalence of sepsis, incidence of new-onset infections (NOI), development of new-onset acute kidney injury (AKI), and feed intolerance.

Data collection
Apart from the routine baseline demographics, information was collected on duration in hospital prior to ICU admission; presence of comorbidities; decompensation status like ascites, jaundice, upper gastrointestinal bleed; presence of sepsis; reason of ICU admission including altered sensorium, respiratory distress, or metabolic acidosis; and disease severity scores including SOFA, APACHE II, CTP, and MELD scores. Follow-up data collected daily until the death or discharge of the patients included the hemodynamic parameters such as heart rate, mean arterial pressure, and requirement of vasopressors; biochemical parameters such as complete blood count, liver function test, kidney function test, coagulation factors, and random blood sugars; blood gas parameters such as pH, PaO 2, and FiO 2 ; and incidence of AKI, requirement of RRT, days of MV, development of NOI, and duration of ICU stay.

Definition of outcome variables 2.9.1. Length of ICU stay
It was considered from the day of admission to the ICU till discharge/shift-out from the ICU or death.

Duration of MV
It was counted from the day of intubation till the day of extubation or death.

Identification of sepsis
Microbiologically proven or presence of any one of the following -lung infiltrates, spontaneous bacterial empyema, SBP, cellulitis, urinary tract infection, or infection at the surgical site [18].

New-onset infections
Development of new infection manifesting as pneumonia on X-ray, positive blood or mini-BAL culture reports or line sepsis during the stay in ICU after 48 h of admission [19].

Acute kidney failure
The KDIGO has recommended a staging system for the severity of the AKI [20].

Requirement of RRT
The requirement of RRT for the patients was assessed if the patient was on sustained low-efficiency dialysis or on continuous RRT (CRRT) [21].

Incidence of feed intolerance
Feed intolerance was defined in terms of gastric residual volume >500 ml over 6 h [22].

Mortality
Death in the ICU is taken as ICU mortality. The day of death was also noted to assess the time to death in the ICU.

Statistical analysis
Data were analyzed in SPSS version 22 and presented as mean (±standard deviation), median (range), or number (%) as appropriate. All variables were checked for normal distribution by Kolmogorov-Smirnov test and variables which did not follow normal distribution were analyzed using non-parametric methods. Continuous variables were compared between sarcopenics and non-sarcopenics using Student's t-test or Mann-Whitney as appropriate, and categorical variables were tested by Pearson Chisquare test or Fisher's exact test. Factors associated with mortality were ascertained by univariate and multivariate binary logistic regression analysis. P<0.05 was considered statistically significant.

Baseline demographic and clinical characteristics of patients
Of the 111 patients enrolled, the majority were male (83.8%) with a mean age of 48.4±11.3 years. The cohort included 82.9% of patients with CLD and 17.1% with ACLF, in whom the most common etiology was alcohol-associated cirrhosis in 50.5% and comorbidities were present in 45.9%. The patients had a mean body mass index (BMI) of 24 kg/m 2 , with 92.3% having Child C status, with high MELD (27.87), SOFA (10.52), and APACHE II (23). The most common reason for ICU admission was HE (38.7%) ( Table 1). MV was required in 48.6% of patients, while 34.2% had shock and 46.8% had sepsis at ICU admission. AKI was present in 53.1% of the patients, although 22% required CRRT at ICU admission (Table 2). Baseline biochemical details of the CICs are summarized in Table 3.

Prevalence of sarcopenia
Of the 111 patients, 76 (68.5%) had sarcopenia with a mean SMA of 113.1 cm 2 , SMI of 40.67 cm 2 /m 2 , and SMD of 50.58 HU which was significantly lower than those without sarcopenia. Sarcopenics had a significantly lower dry BMI at ICU admission and a significantly higher prevalence of sepsis (OR=1.7 [1.0-2.92]; P=0.027). All the other biochemical and clinical parameters were comparable between sarcopenics and non-sarcopenics (Tables 1-3).

Effect of sarcopenia on other clinical variables during the ICU stay
As shown in Table 4, the length of ICU stay and duration of MV were comparable between the groups. Prevalence of sepsis at    baseline was significantly higher (P=0.027) in sarcopenics (53.9%) as compared to non-sarcopenics (31.4%), with lungs being the most common foci. Sarcopenics had a trend toward a higher incidence of feed intolerance during ICU although it was not significant (P=0.094). The incidence of new-onset sepsis and AKI, and requirement of RRT was comparable between sarcopenics and non-sarcopenics. In univariate analysis, the presence of shock (P=0.001), respiratory distress (P=0.02), sepsis (P≤0.001), MV (P≤0.001), and sarcopenia (P=0.001), total leukocyte count (P=0.027), hemoglobin (P=0.043), SOFA score (P=0.001), APACHE II score (P≤0.001), CTP score (P=0.015), and MELD score (P=0.038) was associated with mortality, as shown in Table 5. However, on multivariate analysis, sarcopenia (P=0.002), sepsis (P=0.009), and APACHE II (P=0.032) score were found to be independently associated with mortality (R 2 =0.394) ( Table 5).

Discussion
Notwithstanding the huge advancements in medical therapies and salvage interventions, patients with liver cirrhosis do end up developing life-threatening complications such as HE, variceal bleed, SBP, and HPS which require intensive medical care [1]. Loss of muscle mass or sarcopenia is also one of the major complications seen in patients with cirrhosis [2].We planned to study the effect of sarcopenia on the clinical outcomes of this select group of patients in the ICU [2]. To the best of our knowledge, this is the only study examining such an aspect in a very closed group of CICs.
Our study of 111 CICs including both decompensated cirrhotics and ACLF patients revealed a very high prevalence of sarcopenia at the time of ICU admission. We found that almost two-thirds of our patients had sarcopenia as defined by SMI studied with the help of CT scan images. We kept our study CT based as it is a widely studied method, which is free from hydration status. Moreover, CT scans are mostly available in the ICU setup. The muscle area assessed from a single-slice CT image has shown a good correlation with total body skeletal muscle quantified by multi-slice analysis [23].
A recent retrospective study found that 90.4% of CICs had sarcopenia at the time of ICU admission [12]. This figure is higher than any other study of critically ill patients, where prevalence ranges from 31% to 71% [9,10,24,25]. In patients with cirrhosis, particularly transplant waitlisted candidates, sarcopenia is present in almost 45% [26][27][28][29]. It is well known that cirrhotics have a higher prevalence of sarcopenia. We also found sarcopenia in 68.5% of CICs. This prevalence may be underestimated as we only included those patients in the study who had a CT scan within≤3 months of ICU admission. Despite that, we found that two in every three patients had sarcopenia. Hence, the true prevalence of sarcopenia may be much higher in this select group. Cirrhotics lose a significant amount of skeletal muscle mass with disease progression [30], and critical illness only worsens the already present sarcopenia in these patients. Interestingly, only 8% of the   Our results also demonstrated that patients with sarcopenia had a significantly lower SMD as compared to those without sarcopenia, suggesting that not only the quantity but the quality of the muscles is also compromised.
On assessing the effect of sarcopenia on survival, we found that 62.2% of the patients died in the ICU, with maximum mortality at 7 days. Further, we observed that sarcopenics had significantly higher mortality in the ICU. Nevertheless, we did not observe any significant difference in time to death between sarcopenics and non-sarcopenics. Sarcopenia per se increases the risk of mortality by almost 4 times. Sarcopenia was also identified as an independent predictor of in-hospital mortality in patients admitted to the LICU (OR=3.5 [1.32-9.33]; P=0.011). Likewise, in a retrospective study of general critically ill patients, sarcopenia, as assessed by psoas muscle area, was identified as the independent predictor of in-ICU mortality [11]. Moreover, a meta-analysis investigating 14 studies and 3000 patients in all, also identified CT-defined sarcopenia as a predictor of mortality among critically ill patients [31].
In our study, the presence of sarcopenia was associated with a higher prevalence of sepsis in CICs, with sarcopenics having more sepsis-related mortality in the ICU. Likewise, in yet another study from our center, the incidence of post-liver transplant sepsis was found to be higher in sarcopenics and sepsis was identified as a perfect predictor of 90-day mortality in post-liver transplant patients [32]. Another study in liver transplant candidates reported a higher sepsis-related death among sarcopenics as compared to non-sarcopenics [26]. Higher post-liver transplant infections have also been reported in sarcopenics. However, in our study, the incidence of NOI was comparable between sarcopenics and nonsarcopenics. Plausibly a reduced muscle mass impairs immunity and physiological functions, making sepsis as one of the leading causes of death in sarcopenic patients with cirrhosis [33].
We also found a trend toward a higher incidence of feed intolerance among patients with sarcopenia. The reasons for feed intolerance are not clear but could be due to a high prevalence of shock, abdominal Koch's, and SBP in this particular group of patients. Since ours is the first prospective follow-up study in CICs, any comparisons with other studies on this aspect are not easy.
One retrospective study in critically ill patients with cirrhosis found a strong correlation between rapid muscle loss (of >2% per year) before ICU admission and mortality [12]. For the 1 st time, in our study, we have demonstrated the impact of sarcopenia on the clinical outcomes prospectively among CICs. It is duly noted that sarcopenia influences the outcomes in chronic diseases [34] and it becomes irreversible once the patient reaches the ICU. Cirrhosis is a state of accelerated starvation, which leads to a reduction in the glycogen stores, thus enhancing gluconeogenesis from muscle protein breakdown, thereby precipitating sarcopenia [35,36]. Sarcopenia in cirrhosis is secondary to the disease per se apart from the primary age-related phenomenon. Hence, early identification of sarcopenia along with appropriate nutritional therapy is the need of the hour in these patients with cirrhosis to combat sarcopenia and improve the outcomes in the ICU.

Conclusion
The prevalence of sarcopenia is high among CICs at the time of ICU admission, which not only predisposes them to a high risk of sepsis but also affects their survival in the ICU. The presence of sarcopenia itself should alert clinicians to the same extent as any other complication in cirrhosis, thus warranting timely identification and management.