- Top of page
- Patients and Methods
In patients with end-stage liver disease requiring liver transplantation (LT), protein-energy malnutrition is common and is closely associated with risk for morbidity and mortality after LT [1-5]. In general, nutritional assessment is performed using subjective global assessment, anthropometry including body mass index and muscle arm circumference, and biological markers such as prealbumin (transthyretin) and albumin. In patients undergoing LT, however, these parameters are not available due to decompensated cirrhosis. Anthropometric parameters are overestimated due to edema and ascites, and biological markers are affected by underlying liver dysfunction. In 2008, we introduced body composition analysis into nutritional assessment using bioelectrical impedance analysis (BIA) in patients undergoing LT. BIA measures the body's resistance to flow (impedance) of alternating electrical current at a designated frequency between points of contact on the body. Water in body tissue is conductive; therefore, the measurement of body impedance can indirectly provide information on the body's tissue content including total body water, fat-free mass and skeletal muscle mass. The prevalence of indirect methods of estimating body composition using BIA is increasing because BIA is easy to perform, noninvasive and quick; exhibits high interobserver reproducibility; and has been highly correlated with hydrostatic weighing, dual energy X-ray absorptiometry, and deuterium isotope dilution techniques in specific populations [6-8]. As for reproducibility, Erceg et al.  examined the repeatability of measurement results by the BIA method in adults and children. The intraclass correlation coefficients for men, women and children were greater than 0.99 and the coefficients of variation for the BIA system for men, women and children were small, which suggested strong repeatability of measurement results by BIA.
Body mass can be grossly divided into two compartments: fat mass and fat-free or lean body mass. In a multicompartment body composition model, lean body mass may be partitioned into skeleton and integument, skeletal muscle and visceral organs and total body water, which is further partitioned into intracellular and extracellular water . The concept of body cell mass (BCM), proposed in 1963 by Moore et al. , reflects the cellular components of the body involved in biochemical processes and energy metabolism. BCM is defined as the sum of intracellular water and fat-free mass, including skeletal muscle and viscera, without bone mineral mass . Nutritional status, physical activity level and disease states alter BCM, which in turn serves as a biomarker of these processes . BCM comprises the metabolically active and protein-rich compartments in the body and is known to be depleted in patients with protein-energy malnutrition . BCM is thus regarded as a useful parameter for assessing malnutrition in patients with cirrhosis . We most recently reported that preoperative BCM was an independent risk factor for posttransplant sepsis and death due to infection in living donor LT (LDLT) . This strongly suggests the significance of nutritional status, including skeletal muscle mass, on posttransplant outcomes.
Sarcopenia, defined as a low level of muscle mass, is associated with an increased risk for age-related decline in muscular strength and functional ability [16, 17]. Recent evidence has shown that sarcopenia is an independent predictor of lower disease-free and overall survival in various kinds of diseases [18-21]. Sarcopenic obesity has been observed to reduce survival in patients with solid tumors of the respiratory and gastrointestinal tracts . Moreover, sarcopenia was found to be an adverse prognostic factor in patients with colorectal liver metastasis and pancreatic cancer [19-21]. In patients with liver cirrhosis, malnutrition is caused by the decreased protein synthesis and disturbed energy metabolism that result from liver dysfunction. Protein malnutrition can cause a decrease in skeletal muscle mass in cirrhotic patients. Hayashi et al. reported that the skeletal muscle index, measured as % arm muscle circumference and arm muscle area, in patients with liver cirrhosis was significantly lower than in healthy subjects . However, little is known about the impact of pretransplant sarcopenia on outcomes after LT. Englesbe et al.  recently reported that central sarcopenia strongly correlated with postliver transplant mortality, using the size of the psoas muscle as measured by computed tomography (CT) scan. However, the total psoas area is merely part of skeletal muscle mass and might not correctly reflect whole body skeletal muscle mass. In contrast, BIA can easily and automatically measure whole body skeletal muscle mass.
In the present study, we examined the skeletal muscle mass in patients undergoing LDLT using BIA, and investigated the impact of sarcopenia on outcomes after LDLT. Additionally, we examined the effect of our perioperative nutritional therapy on overall survival in patients with and without sarcopenia.
- Top of page
- Patients and Methods
The median ratio of preoperative skeletal muscle mass was 92% (range, 67–130%) of the standard mass (Figure 1). A significant positive relationship was observed between skeletal muscle mass and BCM (r = 0.636, p < 0.001; Figure 2A) and a significant negative relationship between skeletal muscle mass and BTR (r = −0.254, p = 0.005; Figure 2B). No significant correlations were observed between the skeletal muscle mass and age (p = 0.079), sex (p = 0.401), total lymphocyte count (p = 0.273), zinc (p = 0.594), prealbumin (p = 0.224), ABO compatibility (p = 0.386), Child–Pugh classification (p = 0.278) or MELD score (p = 0.754). No significant correlations were observed between BCM and age (p = 0.161), sex (p = 0.121) or MELD score (p = 0.781).
Figure 2. Correlation between the skeletal muscle mass value and BCM (A) and BTR (B). BCM, body cell mass; BTR, branched-chain amino acids to tyrosine ratio.
Download figure to PowerPoint
The overall survival rate was significantly lower in patients with low skeletal muscle mass (n = 47) than in patients with normal/high skeletal muscle mass (n = 77; p < 0.001; Figure 3A). Similarly, the overall survival rate was significantly lower in patients with low BCM (n = 48) than in patients with normal/high skeletal muscle mass (n = 76; p = 0.005; Figure 3B).
The causes of death in 20 patients with low skeletal muscle mass were as follows: sepsis (n = 9); pulmonary complications including hepato-pulmonary syndrome and pneumocystis pneumonia (n = 5); graft failure including antibody-mediated rejection and chronic rejection (n = 3); disseminated intravascular coagulation (n = 2); and bleeding in the brain (n = 1). The causes of death in 11 patients with normal/high skeletal muscle mass were as follows: sepsis (n = 4); graft failure (n = 3); pulmonary complications (n = 3); and recurrence of hepatocellular carcinoma (n = 1). The causes of death were similar between the two subgroups.
Next, we investigated the effect of perioperative nutritional therapy on survival according to skeletal muscle mass. Twenty-one of 47 patients with low skeletal muscle mass and 42 of 77 patients with normal/high skeletal muscle mass received perioperative nutritional therapy with more than 1 week of preoperative nutritional therapy. Perioperative nutritional therapy significantly improved overall survival in patients with low skeletal muscle mass (p = 0.009; Figure 4A). Neither preoperative Child–Pugh classification nor MELD score was significantly different between patients with or without nutritional therapy in this subgroup (p = 0.338, p = 0.247, respectively). In contrast, perioperative nutritional therapy had only a slight impact on overall survival in patients with normal/high skeletal muscle mass (p = 0.550; Figure 4B).
Figure 4. Overall survival rates according to perioperative nutritional therapy in patients with low skeletal muscle mass (A) and normal/high skeletal muscle mass (B).
Download figure to PowerPoint
Univariate analysis revealed that pretransplant low skeletal muscle mass, pretransplant low BCM and lack of perioperative nutritional therapy were significant risk factors for death after LT (Table 1). Due to the fact that skeletal muscle mass and BCM were closely correlated, multivariate analysis was performed incorporating either parameter. As a result, pretransplant low skeletal muscle mass, pretransplant low BCM and lack of perioperative nutritional therapy were found to be risk factors for death after LT (Table 2).
Table 1. Univariate analysis of factors affecting posttransplant patient survival
|Variable||1-year OS (%)||p-Value|
|Recipient age (year)|
|<60 (n = 92)||74||0.500|
|≥60 (n = 32)||84|| |
|Donor age (year)|
|<50 (n = 78)||77||0.733|
|≥50 (n = 46)||76|| |
|Male (n = 60)||77||0.771|
|Female (n = 64)||77|| |
|HCV (n = 49)||84||0.155|
|NonHCV (n = 75)||72|| |
|ABO blood type|
|Compatible (n = 88)||78||0.525|
|Incompatible (n = 36)||72|| |
|A, B (n = 50)||78||0.763|
|C (n = 74)||76|| |
|<20 (n = 85)||78||0.649|
|≥20 (n = 39)||74|| |
|<0.8% (n = 37)||86||0.101|
|≥0.8% (n = 87)||72|| |
|Right (n = 69)||81||0.131|
|Left (n = 55)||71|| |
|Operative time (h)|
|<12 (n = 30)||83||0.733|
|≥12 (n = 94)||74|| |
|Operative blood loss (L)|
|<10 (n = 84)||77||0.762|
|≥10 (n = 40)||75|| |
|Pretransplant body cell mass|
|Low (n = 48)||65||0.005|
|Normal/high (n = 76)||84|| |
|Pretransplant skeletal muscle mass|
|Low (n = 47)||59||<0.001|
|Normal/high (n = 77)||87|| |
|With (n = 63)||86||0.012|
|Without (n = 61)||67|| |
Table 2. Multivariate analysis of factors affecting posttransplant patient survival
|Variable||Odds ratio||95% CI||p-Value|
|Pretransplant low skeletal muscle mass||4.846||2.092–11.790||<0.001|
|Without nutritional therapy||2.894||1.200–7.386||0.018|
|Variable||Odds ratio||95% CI||p-Value|
|Pretransplant low body cell mass||3.175||1.396–7.435||0.007|
|Without nutritional therapy||2.798||1.188–6.953||0.018|
- Top of page
- Patients and Methods
As far as we know, this retrospective study of 124 patients is the first to investigate the impact of sarcopenia on posttransplant outcomes in patients undergoing LDLT. The present study showed that sarcopenia was an independent risk factor for poor survival after LDLT. We most recently reported that BCM was closely involved with posttransplant sepsis and death due to infection . However, BCM is a sum of intracellular fluid and total body protein, including skeletal muscle mass. In the present study, therefore, we focused on skeletal muscle mass and examined the impact on posttransplant outcomes. The present study supports our previous findings and specifically clarifies the significance of skeletal muscle mass. In deceased donor liver transplantation (DDLT), other investigators have reported that sarcopenia evaluated by psoas area as determined by CT or magnetic resonance imaging (MRI) was associated with increased waiting list mortality and posttransplant mortality [23, 31]. Therefore, pretransplant sarcopenia is closely involved with posttransplant survival in LDLT as well as DDLT.
The exact mechanisms by which sarcopenia affects survival in patients undergoing LT are unclear. In such patients, sarcopenia reflects protein-energy malnutrition, which is a characteristic of decompensated liver cirrhosis. Our previous study showed that the most frequent cause of in-hospital death after LT was infection, including sepsis . Montano-Loza et al.  reported that the higher mortality risk in cirrhotic patients with sarcopenia was related to a higher frequency of sepsis-related death and not to liver failure. In the present study, however, no apparent differences were found in the causes of death between patients with or without sarcopenia. We are now investigating the relationship between nutritional condition and immunological status, which plays an important role not only in infection, but in posttransplant rejection as well.
Interestingly, a significant negative relationship was seen between skeletal muscle mass and the BTR (r = −0.254, p = 0.005). Before analysis, we expected that there must be a positive relationship between skeletal muscle mass and the BTR, since patients with decompensated cirrhosis sometimes develop hepatic encephalopathy. However, no significant relationship was found between skeletal muscle mass and prealbumin level (p = 0.224), Child–Pugh classification (p = 0.278) or MELD score (p = 0.754), in line with the report by Montano-Loza et al. , who examined patients with cirrhosis. Regarding the reason for the negative correlation between skeletal muscle mass and the BTR, we speculate that BCAAs in patients with cirrhosis are mainly metabolized in skeletal muscle. Therefore, the more skeletal muscle mass, the more BCAA consumption, which leads to a decrease in the BTR. Sarcopenia would be an independent parameter that reflects mainly liver dysfunction in conventional scores, including the Child–Pugh and MELD scores.
Regarding the modality for assessing sarcopenia, we evaluated whole body skeletal muscle mass using BIA. In contrast, other investigators measured the psoas muscle area or the muscle cross sectional area at the L3 vertebral level using available CT and MRI images [23, 31] or calculated skeletal muscle mass using dual energy X-ray absorptiometry . We prefer BIA to other methods for the following reasons. First, BIA can measure whole body skeletal muscle mass. It might be more appropriate to evaluate not only the psoas muscle mass, but the skeletal muscle mass of the whole body as well, including the four extremities, to correctly examine the impact of sarcopenia. Second, BIA obtained by the InBody 720 can easily measure skeletal muscle mass within 2 min without calculating muscle mass volume using available CT and MRI images. Third, dual energy X-ray absorptiometry measurement exposes patients to radiation in evaluating muscle mass, whereas BIA works through the measurement of body resistance and reactance to alternating electrical current.
In the present study, we first reported that perioperative nutritional therapy significantly improved overall survival in patients with sarcopenia. This finding is clinically important, since perioperative nutritional intervention with our regimen could overcome the disadvantages associated with preoperative malnutrition. Importantly, a planned preoperative nutritional intervention can be performed in most cases of LDLT, since the date of transplant is known in LDLT, unlike in DDLT. We try to ask physicians to start nutritional therapy, as well as rehabilitation, at the time of referral of a potential recipient, approximately a few months before LT to more effectively increase skeletal muscle mass and BCM.
Several limitations must be borne in mind when considering this study. First, there was significant selection bias for patient inclusion in the study group. We included liver transplant candidates who could undergo BIA preoperatively. As a result, patients who could not stand independently for more than 2 min due to acute liver failure and severely ill general condition or who could not undergo BIA due to the unavailability of BIA for subemergent liver transplantation or the hospital's circumstances were excluded from this study. The median MELD score and the distribution of the Child–Pugh classification scores in this population were almost similar to those of our large study . Therefore, the patients enrolled in this study were not a special population in our institute. Second, since this was a retrospective study, there is a possibility of confounding bias for perioperative nutritional therapy. We compared the preoperative Child–Pugh classification scores and MELD scores in patients with and without nutritional therapy in the low skeletal muscle mass subgroup. As a result, neither the preoperative Child–Pugh classification scores nor the MELD scores were significantly different between the patients with and without nutritional therapy in this subgroup (p = 0.338, p = 0.247, respectively). Therefore, preoperative general condition and selection bias might not have affected the patients' prognosis in this study. Furthermore, the finding that neither Child–Pugh classification nor MELD score was not an independent risk factor for survival would support our idea that selection bias might have been at a minimum. Third, we could not collect data on other frailty parameters, including grip strength and levels of exhaustion, due to the study's retrospective design. We are now conducting a prospective study to evaluate the degree of perioperative sarcopenia by measuring body composition and grip strength. Finally, the optimal cut-offs of skeletal muscle mass that define sarcopenia require further investigation. In the present study, we defined sarcopenia as below 90% of the standard value analyzed by the InBody 720. There have been a few reports discussing the diagnosis of sarcopenia [35, 36], while indicating no consensus on the definition of sarcopenia. Chien et al.  advocated cut-off points for the skeletal muscle mass index based on two SD below the mean of 200 young adult patients divided by sex: men less than 8.87 kg/m2 and women less than 6.42 kg/m2. Janssen et al.  also used SD to define sarcopenia, measured in terms of the skeletal muscle index, which is the skeletal muscle mass/body mass × 100. In their study, Janssen et al. defined class I and class II sarcopenia to be present in subjects whose skeletal muscle index was within −1 to −2 SD and below −2 SD of young adult values, respectively . The present study comprised cirrhotic patients as opposed to healthy subjects. Therefore, the definitions of sarcopenia used in these reports are not necessarily applicable to cirrhotic patients.
In conclusion, sarcopenia has been observed to be closely involved with posttransplant mortality in patients undergoing LDLT. Additionally, our protocol of perioperative nutritional therapy was demonstrated to significantly improve overall survival in patients with sarcopenia. Until now, many investigators, including us, had reported the significance of perioperative nutritional therapy in patients undergoing LT [37-40]; however, there had been no report focusing on patients with sarcopenia. As far as we know, this is the first study to show the impact of perioperative nutritional therapy on outcomes after LT, even in patients with sarcopenia. Further prospective studies are needed to confirm our findings.