Radiological assessment of skeletal muscle index and myosteatosis and their impact postoperative outcomes after liver transplantation

Since 1963 when the first liver transplantation (LT) was performed by Starzl¹, it has become a standard treatment modality for patients with acute liver failure and chronic liver disease.² Ninety percent 5-year survival rate and better quality of life are the two most important outcomes of LT.² Most liver transplant centres use the Model for End-stage Liver Disease (MELD) score for organ allocation.³ Patients with MELD score 15 or more, patients with poor quality of life due to chronic liver disease symptoms (diuretic-intractable ascites, variceal bleeding, pruritus, cachexia), and patients with acute liver failure are those who benefit most from LT.³ MELD score can underestimate the severity of liver disease in specific groups of patients (acute on chronic liver disease, presence of sarcopenia, chronic kidney disease, etc.).³ Several modifications of the MELD score have been introduced, but none offers a more reliable and accurate scoring system. Albumin-bilirubin (ALBI) score is mainly used as an objective method to assess liver function and predict postoperative complications, particularly after hepatectomy in patients with hepatocellular carcinoma (HCC).⁴ Its role in determining post-LT outcomes is not yet determined. Two main objective parameters of nutritional status are sarcopenia and myosteatosis (MI). The European Working Group on Sarcopenia defines sarcopenia as the presence of low muscle mass (under the 5^th percentile) and low muscle function (strength or performance) in patients with advanced age, cancer, or other diseases.^5,6 Myosteatosis is defined as the abnormal fatty transformation of skeletal muscle. It negatively affects muscle strength and is common in advanced age^7,8, diabetes^7,8, obesity^9,10, chronic⁹, and malignant diseases.^11,12 Overview of the literature shows large number of different methods used for body composition assessment in patients with liver cirrhosis.¹³ There is still no consensus on the best tools for each body component in patients with liver cirrhosis. Most frequently used are computed tomography (CT), bioimpedance analysis (BIA), dual-energy X-ray absorptiometry (DXA) and anthropometry.¹³ Some of them BIA, DXA and body mass index (BMI)) are not applicable in patients with end-stage liver disease due to frequent water retention.^14,15 In the last decade, computed tomography with automated or semiautomated body composition analysis at the third lumbar vertebra has emerged as an objective method of defining the nutritional status of patients with chronic liver disease.^15,16 Nutritional assessment with CT is not affected by water retention or presence of ascites. Skeletal muscle volume and myosteatosis can be measured from CT images obtained as a part of routine pre-transplant evaluation.¹⁴ Skeletal muscle mass index (SMI) is calculated as muscle mass area divided by the square of the height. The Global Leadership Initiative on Malnutrition (GLIM)¹⁶ score was introduced as a potential nutritional assessment tool in recent years. It was shown to have good predictive value as a risk assessment tool for postoperative morbidity and mortality in patients after colorectal surgery.¹⁷ However, its role as a predictive factor in LT is not yet established.

This study aimed to investigate feasibility of radiological assessment of a nutritional status of a patient and the predictive value of SMI and MI on postoperative complications, length of hospitalization, liver graft rejection, and mortality. We further compared the predictive value of SMI and MI with MELD, ALBI, and GLIM scores.

We retrospectively analyzed 138 adult patients who had first orthotropic LT from brain dead donors between 1.1.2012 and 1.1.2020 in our institution. We excluded patients who had re-LT procedure and those whose abdominal CT scan could not be obtained from data base or received reduced size and liver graft from donation after cardiac death (DCD) donor. From the medical database, we collected recipient age, gender, body mass index, underlying liver disease, presence of ascites, hepatocellular carcinoma (HCC), and laboratory parameters (serum levels of sodium, creatinine, albumin, protein, bilirubin, and International normalized ratio (INR)). We calculated MELD and ALBI scores from laboratory parameters. Among GLIM criteria, we used chronic liver failure and end-stage liver disease as etiologic and SMI or MI as phenotypic criteria. Length of intensive care unit (ICU) stay, hospitalization, postoperative complications according to Clavien-Dindo classification¹⁸, infections, and 90-day mortality were collected from the database of liver recipients and analyzed. Diagnosis of liver rejections was confirmed with laboratory tests and histological examination of all liver graft specimens obtained by ultrasound-guided biopsy. Acute rejection was defined with 6 points or more according to the liver allograft fibrosis score.¹⁹

The Slovenian National Medical Ethics Committee approved our study design (approval number 0120–230/2018–10) and waived the need to obtain informed consent from participants.

Malnutrition with skeletal muscle mass loss is a frequent complication in patients with chronic or end-stage liver disease.²⁵ The aetiology of malnutrition is multifactorial.²³ Insufficient calorie intake (early satiety, loss of appetite, alcohol consumption, diet restriction), metabolic abnormalities, catabolic state of metabolism, and malabsorption are the main contributors to muscle loss.^25,26 It affects from 30 to 70%^27,28 of patients with end-stage liver disease and up to 80%²⁷ of patients with alcohol-related liver disease. Myosteatosis affects more than half of patients with chronic liver disease.⁹ In our patient population, we used selected parameters for low SMI (men < 52.4 cm²/m², women < 38 cm²/m²)^9,21 and MI (< 33 HU in patients with a BMI ≥ 25 kg/m² and < 41 HU in those with a BMI < 25).^22,23,24 These cut-off values were determined specifically for patients on LT waiting list are the most widely used in literature.^22,23,24 Low SMI was present in 63% of males and 28.9% of females. More than one-third of the patients (32.6%) had fatty infiltration of the skeletal muscle. Patients with alcoholic aetiology had more fat-infiltrated muscles than other aetiologies for end-stage liver disease (P < 0.008). Alterations in the metabolism of fatty acids, eating habits, and alcohol consumption probably result in fatty infiltration of skeletal muscle fibres. However, due to the small numbers of patients (3 patients) with non-alcoholic steatohepatitis (NASH) or non-alcoholic fatty liver disease (NAFLD) in our study group, we cannot properly assess the relationship between the aetiology of underlying liver disease and the degree of MI.

Several studies have shown that sarcopenia and myosteatosis are associated with a higher incidence of postoperative complications, infection rate, duration of hospitalization, and mortality in a wide range of gastrointestinal cancers, including tumours of the hepatopancreatobiliary system.^29,30,31,32 Lack of a strictly defined patient population, different methods of measurement of muscle mass and function, and the absence of uniformly accepted cut-off points for patients with chronic or end-stage liver disease are the leading causes for wide ranges of sarcopenia incidence reported in the literature.^14,21,33 Therefore, the selected cut-off points for patients with cancer are used for evaluation, which may affect the quality of results.

Sarcopenia is associated with poorer outcomes in patients with chronic liver disease on the wait-list or after LT.^34,35 Patients with end-stage liver disease and sarcopenia have shorter survival than non-sarcopenic patients (22 ± 3 vs. 95 ± 22-month, P < 0.001).^9,36 The main cause of death in patients with end-stage liver disease is sepsis.³⁷ Incidence of sepsis as a cause of death is even higher in presence of sarcopenia (36% vs. 16%, P < 0.001) or myosteatosis (32% vs. 19%, P = 0.020).⁹ Lower muscle mass is also associated with a higher incidence of hepatic encephalopathy in patients with end-stage liver disease.^38,39 On the other hand, a group from USA⁴⁰ showed that sarcopenic patients had a higher tendency for pulmonary complications than the nonsarcopenic group (38% vs. 18%, P = 0.100); however, there was no significant difference in morbidity and mortality.

CT imaging is considered as the gold standard for body composition assessment. SMI calculated on the level of third lumbar vertebra is superior since it correlates best with the actual quantity of the muscles in the body.⁴¹ It is not affected by ascites and is part of preoperative evaluation in patients with hepatocellular carcinoma or the complications of portal hypertension. The value of SMI as a predicting factor for postoperative morbidity and mortality is yet to be determined. However, it has been shown in several reports that it has a negative impact on postoperative outcomes. Lower values of SMI are associated with more significant postoperative mortality, higher infection risk, and postoperative complications, more extended intensive care unit stay, ventilator dependency, and higher waitlist mortality rates in patients with end-stage liver disease.^33,42,43 SMI as a predicting factor for higher mortality rate in patients with end-stage liver disease was superior to other nutritional assessment indicators, such as BMI, upper arm muscle circumference, and triceps skinfold thickness.⁴⁴ However, some other reports showed contrasting findings. Using SMI for prespecified definitions of sarcopenia had no impact on mortality or delisting from the transplant waitlist between patients with and without sarcopenia.⁴⁵ Length of hospitalization following LT, days of hospitalization during the first year post-LT, survival at one year, or overall survival was not different between sarcopenic and nonsarcopenic patients.⁴⁶ Low SMI alone was not associated with graft and patient survival (P = 0.273 and P = 0.278) after LT.⁴⁷ These conflicting results are probably due to heterogeneity of used specific cut-off points (sex, age, race) and lack of strictly defined parameters representative uniquely for patients with chronic liver disease. We found a negative correlation between BMI and SMI, probably due to high-volume ascites in the patient population This confirms findings that BMI is not suitable for body assessment in patients with end-stage liver disease. In our study, the presence of ascites was associated with significantly lower SMI (P < 0.05) in patients with end-stage liver disease but did not influence MI. There is a weak, positive correlation between SMI and ICU length of stay in male patients, which was statistically significant (rs = 0.226, P < 0.025). We speculate that the main reason is a more pronounced systemic inflammatory reaction to surgical stress in patients with preserved muscle mass. It is known that muscle mass is mandatory for the normal function of the immune system.⁴⁸ Experience from Japan⁴⁹ shows a connection between low SMI and decreased incidence of graft rejection in living-donor liver transplantation. Analysis of our results shows no statistically significant difference in postoperative complications, rate of postoperative infection, and liver graft rejection rate between males and females with low or normal SMI (P > 0.05). BIA is commonly used technique in body composition analysis in every day clinical practice.⁵⁰ Its key parameters are resistance, reactance and phase angle (PA). PA is found to be associated with outcomes in different diseases and has been found to be useful for monitoring fluid changes and response to interventions.⁵⁰ The main limitation is the complexity of the determinants that requires its adjustment to the individual phenotypic diagnosis of each patient. Results can be affected by altered water and electrolyte balance, fluid retention and diuretic therapy.⁵⁰ DXA allows for the quantification of three body compartments (bone mass, fat mass, and bone fat-free mass (or lean mass)) based upon the differential tissue attenuation of X-ray photons.⁵¹ However, it can be affected by presence of ascites.⁵² Even though DXA can be modified to exclude influence of ascites or tissue oedema the correlation between the lean mass index and SMI was weaker (γ = 0.29, p = 0.035) and falsely high in patients with ascites before liver transplantation.⁵³ Main disadvantages of DXA compared to CT is inability to assess muscle mass quality (myosteatosis).⁵¹

Myosteatosis is a more clearly defined factor, and its influence is more uniformly established in literature.⁵⁴ However, there are many different cut off points (HU < 25 to HU < 39) reported in literature in patients with malignant disease.⁵⁵ Majority of reports uses a Martin cut-off point (< 33 HU in patients with a BMI ≥ 25 kg/m² and < 41 HU in those with a BMI < 25)⁵⁶ for determining presence of myosteatosis. Martin cut-off point is also recently defined threshold parameters for MI in a patient with a chronic liver disease.^22,23,24 It has been shown that severely ill patients with myosteatosis have a lower survival rate than those without fatty infiltration in muscles.⁵⁷ Myosteatosis negatively impacts the survival of patients with end-stage liver disease (28 ± 5 vs. 95 ± 22-month, P < 0.001)^9,36 and is associated with longer hospitalization and higher morbidity.⁴³ Patients with myosteatosis showed a higher mortality rate, most commonly due to respiratory and septic complication.⁴⁷ Myosteatosis had no influence on ICU length of stay (P = 0.161), hospitalisation (P = 0.771), postoperative complications (P = 0.839), infection rate (P = 0.703) and graft rejection (P= 0.449) in our patient population.

The nutritional status on the waitlist for LT as a possible risk factor is overlooked with the MELD score.⁵⁸ Another important limiting factor of the MELD score is using serum creatinine levels for score calculation.³ Serum creatinine level may vary significantly and is influenced by chronic kidney disease, ascites, paracentesis, and the influence of gender and liver disease on skeletal muscle mass.⁵⁹ Variation of serum creatinine levels may affect the MELD score and underestimate the severity of liver disease.⁵⁸ Thus, several modifications of the MELD score were developed to incorporate the nutritional parameters. Body composition MELD (BC-MELD)⁶⁰ has a better predictive value for waiting list mortality than MELD score. MELD-sarcopenia score showed a positive predictive value in patients with a lower score (< 15) on the postoperative course; however, it was not useful in patients with MELD above 15.^61,62 Combining SMI in MELD in multivariate analysis (AUROC= 0.812) is significantly better than MELD alone (AUROC = 0.787) for predicting 5-year mortality (P < 0.001).⁴⁴ We found a weak, positive correlation between the MELD score and ICU length of stay (rs = 0.261, P < 0.002). ALBI score showed promising results in more accurate prediction of liver disease severity and mortality on waitlist compared to the Child-Pugh score; however, its predictive value was inferior to the MELD score.⁶³ Patients with pre-transplant ALBI grade 3 liver disease had increased mortality after LT.⁶⁴ We found a weak, positive correlation between the ALBI score and ICU length of stay and hospitalization time, which was statistically significant (r_s = 0.279, P < 0.001; r_s = 0.197; P = 0.022). We found no correlation between the ALBI score and postoperative complications. The GLIM score¹⁶ was established as a potential assessment tool for evaluating patient nutritional status in recent years. Our results showed no difference between the patients who met GLIM criteria and those who did not, regarding ICU stay, length of hospitalization, postoperative complications, infection, and mortality.

We believe that nutritional status in a patient with end-stage liver disease is an essential aspect of pre-transplant workup. However, its precise role is not yet determined. LT is a complex surgical procedure influenced by numerous factors. In the early phase of LT, donor characteristics and comorbidities, quality of liver graft retrieval, liver graft quality, cold preservation duration, ischemic damage to the liver graft during static cold storage, recipient medical conditions and comorbidities, surgical procedure, and early postoperative therapy are by our opinion the most essential factors. However, muscle mass status, muscle function and myosteatosis are crucial factors in a period of rehabilitation.⁶⁵

Patient with cachexia or with high-risk (< 18.5 kg/m², Child-Pough C)⁶⁶ to lose muscle mass should be screened and involved in intensive but personalized nutritional support therapy. Nutritional requirements in end stage liver disease are 35 kcal/kg per day in non-obese patients (BMI 30 kg/m²) and 1,2g/kg per day intake of proteins.⁶⁶ It is safe to calculate requirements based on the dry weight of a patients.⁶⁶ In presence of hepatic encephalopathy (HE), animal protein should be substituted with vegetable protein origin. Randomized control study⁶⁷ showed that substitution of animal protein with vegetable protein for a period of six months (30–35 kcal/kg/d,1.0–1.5 g/kg/d protein) improved neuropsychiatric performance in patients with minimal HE and decrease their risk of developing overt HE compared to no intervention. It is unreal to expect to gain muscle mass in cirrhotic patients, but muscle mass preservation should be focus of such nutritional interventions. There are several strategies to prevent muscle mass loss in patients with end stage liver disease. First strategy is nutritional supplementations. Patients should have frequent small meals to avoid prolong fasting period (> 6 h).^66,68,69 Enteral supplements with side branched amino acids should be administered.^66,68,69 Snacks rich in carbohydrates should be taken as a late-night snack.^66,68,69 Second, physical activity in a form of resistance and endurance exercise, is probably appropriate and beneficial.⁷⁰ Micronutrition, especially administration of fat-soluble vitamins⁷¹ and ammonia lowering therapy⁷² is also important. Role of immunonutrition in patient with end stage liver disease is not yet established.^66,73 Post-transplant screening is advised in all patients after liver transplantation.^66,73

There are several drawbacks to our study. The two most important are the retrospective nature of data collection and the lack of functional assessment of recipients’ muscles. Another factor that may influence results is the relatively small number of cases that may affect some results or may not produce statistical significance due to the complexity of the treatment and numerous factors that determine its outcome.

In conclusion radiological assessment of a patient's nutritional status at the third lumbar vertebra represents an objective and reproducible method. It should become a standard screening tool in patients with acute or chronic end-stage liver disease. Due to complexity of liver transplant procedure, liver graft and liver recipients’ factors, it is difficult to established impact of a skeletal muscle index and myosteatosis on postoperative outcomes. However, nutritional interventions and physical activity should be part of the clinical pathway in patients with end-stage liver disease waiting for liver transplantation. Prospective randomized controlled studies are not possible due to ethical considerations; hence, standardization and uniformity in definitions, methods, and cutoff points are crucial to producing reliable data in the future.