Results
Clinical and sociodemographic characteristics of the studied patient population are described in
Table 1. In total, data from 79 patients with cirrhosis were included and analyzed in the study according to the flow chart in
Figure 2. The cohort consisted of 62% men, the mean age was 54±11 years, and alcohol was the most common etiology of liver disease (34%). The mean MELD score of the group was 25±5. Nine percent of patients were classified as Child-Pugh A, 14% as Child-Pugh B, and 61% as Child-Pugh C. Moreover, the mean BMI was 26±5 kg/m
2. The mean SMI was 47.9±9.1 cm²/m², whereas means for creatinine, GFR and total proteins levels were, respectively, 89±41 mmol/L, 82±26 mL/min/1.73m
2 and 62±11 g/L. Prevalence of common complications was as follow: 58% fluid retention (edema and/or ascites), 52% HE, 47% renal dysfunction, 46% sarcopenia, 43% electrolyte imbalance, 35% upper gastrointestinal bleeding, 35% portal hypertension, 25% spontaneous bacterial peritonitis, 22% pleural effusion, 20% pneumonia, 19% others respiratory issues, 18% splenomegaly, 13% compromised bone health, and 13% dysglycemia.
The mean time for pre-LT and post-LT CT scans was 40 days and 25 days, respectively. The presence of sarcopenia increased from 46% pre-LT to 62 % post-LT. Among patients who were sarcopenic before LT, 14% had resolved sarcopenia after LT whereas 86% remained sarcopenic. Among patients who were not sarcopenic before LT, 58% did not develop sarcopenia; however, 42% had become sarcopenic after transplantation. Apart from patients where sarcopenia was resolved (pre-LT 39.4±3.7 versus post-LT 47.6±7.7 cm
2/m
2,
p < 0.02), all patients experienced a decrease of SMI after LT (persistent: pre-LT 42.2±5.8 versus post-LT 39.0±5.9 cm
2/m
2,
p = 0.0005; newly: pre-LT 50.0±7.7 versus post-LT 41.0±7.9 cm
2/m
2,
p <0.0001; never: pre-LT 55.2±8.5 versus post-LT 53.7±8.0 cm
2/m
2,
p = 0.12) (
Figure 3). The presence of sarcopenia post-LT was associated with worst clinical outcomes after LT. Sarcopenic patients post-LT had longer stays in the hospital after LT (54±37 versus 29±10,
p = 0.002) and experienced, at one year, greater number of complications (5±2 versus 3±2,
p < 0.001) and episodes of infection (3±1 versus 1±2,
p = 0.027) compared to non-sarcopenic patients (
Table 2). As well, SMI post-LT was negatively correlated with the number of complications (
r = –0.34,
p = 0.002), episodes of infections (
r = –0.26,
p = 0.019), and days of hospitalization (
r = –0.45,
p < 0.001) (
Figure 4).
Two predictive models for SMI post-LT were constructed using multiple linear regression adjusted for pre-LT factors namely age, BMI, MELD score, SMI, total protein levels, number of common complications, number of episodes of infections; and adding GFR for Model 1 and creatinine levels for model 2 (Model 1,
R² = 0.71,
p < 0.0001; Model 2,
R² = 0.69,
p < 0.0001). The analysis showed that the SMI post-LT was independently associated with pre-LT renal function markers, GFR and creatinine (Model 1, GFR:
β = 0.33; 95% CI 0.04–0.17;
p = 0.003; Model 2, Creatinine:
β = –0.29; 95% CI –0.10 to –0.02;
p = 0.009). In both models, the baseline SMI pre-LT was the greatest predictor of SMI post-LT (Model 1,
β = 0.52, 95% CI 0.22–0.84,
p = 0.001; Model 2,
β = 0.60, 95% CI 0.29–0.94,
p < 0.001) (
Table 3). In agreement with our predictive models, patients who had renal dysfunction before LT showed lower SMI post-LT (
p = 0.043) and were at higher risk of developing sarcopenia (RR = 3.11, 95% CI 1.19–8.16,
p = 0.021). In addition, post-LT sarcopenic patients had a lower baseline GFR (91±23 versus 75±28 mL/min/1.73 m
2,
p = 0.003) and higher baseline creatinine levels (75±25 versus 106±75 mmol/L,
p = 0.018) compared to non-sarcopenic patients (
Figure 5).
Discussion
To date, LT is the only curative treatment for patients with cirrhosis. Receiving a new liver is a rare privilege, so it is important to optimize recovery of patients following LT by identifying factors pre-LT that could promote negative outcomes. Using the cut-offs specific to the cirrhosis population on the transplant waiting list, we demonstrated that the prevalence of sarcopenia post-LT (62%) was higher than that of sarcopenia pre-LT (46%). Moreover, we observed newly onset sarcopenia in 42% patients within a three-month post-operative period. After adjusting for confounding variables, SMI and renal failure markers (GFR and creatinine) were found to be independent predictors for SMI post-LT. In addition, we showed that patients with renal dysfunction pre-LT were at higher risk of developing sarcopenia post-LT. Together, we demonstrated that renal dysfunction was associated with the development as well as the persistence of sarcopenia in patients following LT.
Imaging by CT scan is still the most well-validated method to assess sarcopenia in patients with cirrhosis (
16,
17). In the present study, L3 SMI cut-offs were used as follow: male SMI <50 cm
2/m
2; female SMI <39 cm
2/m
2 as defined by Carey et al and recommended by the European Association for the Study of the Liver (EASL) (
13,
18). We included patients who received a CT scan (to evaluate SMI) within the six-month peri-operative period (three months pre- and post-LT). Three months is a key period of time in nutrition assessment which distinguishes acute from chronic phase of malnutrition. Acute malnutrition appears and installs in less than three months and is commonly associated with illness or temporary situation affecting food intake (
19,
20). Malnutrition is known to be related to muscle weakness and loss (
21). However, malnutrition is a reversible risk factor, so early detection is important. The peri-operative period, referred to as pre-, intra-, and post-operative phases, is a critical and an important period in order to reduce complications and improve outcomes post-surgery (highlighted by the Enhanced Recovery After Surgery [ERAS] program) (
22,
23). In the context of LT, the pre-operative phase consists in prioritization and identification of patients at high risk and involves multiple biochemical analyses which require fasting (
24,
25). The frequency of these tests increases as the disease progresses and LT approaches. Moreover, invasive procedures such as varices ligation also require fasting and contribute to inadequate dietary intake, which may promote acute malnutrition and increase the prevalence of sarcopenia at the time of LT (
26). The post-operative phase is also important in peri-operative medicine as it determines potential future complications. Unfortunately, liver transplant recipients may also experience acute malnutrition in the early post-operative phase due to surgical stress and complications, immunosuppression therapy, surgery related hypermetabolism, or more fasting clinical tests, exposing them to early muscle wasting and other complications (
24). Of note, in our study, 86% of CT scans after LT were performed less than one month after surgery allowing us to have accurately describe the muscle status post-LT. In our cohort, 46% of patients were sarcopenic before LT, similar to findings observed in other studies which reported 41% to 48% of pre-LT sarcopenia (
4,
6,
15).
Interestingly, the majority of patients in our cohort developed a decline in muscle mass (ie, SMI) resulting in higher number of patients with sarcopenia post-LT (62%). Furthermore, 42% of patients who were non-sarcopenic before LT, became sarcopenic within three months post-LT. This finding supports our hypothesis of possible acute malnutrition related muscle wasting. Other studies have reported newly developed sarcopenia in patients but were assessed one year after LT (
12,
27). Using the psoas muscle area to define sarcopenia, Tsien et al reported an increase in the onset of sarcopenia from 62.3% to 86.8% (
27), while Jeon et al observed an increase from 36% to 46% (
12). However, it is unclear whether sarcopenia developed as early as three months or was due to factors occurring greater than three months post-LT. Few previous studies have addressed the onset of early sarcopenia (within three months post-LT). Some studies evaluated muscle function and showed early improvements in muscle function within six months after LT (
28–
30). In a prospective study measuring both muscle mass and function, it was demonstrated that improvement in hand grip values occurred independently to improvement in muscle mass, three months after LT (
31). Our study is the first to describe acute muscle wasting post LT which is critical to identify in order to prevent sarcopenia and manage related negative outcomes.
Pathogenesis of sarcopenia is a complex and multifactorial system. In patients with cirrhosis, in addition to age and malnutrition, increased catabolic state is one of the main causes of sarcopenia. Hypermetabolism increases consumption of amino acids as energy sources and decreases protein synthesis, resulting in muscle wasting and sarcopenia (
2,
32,
33). In the present study, a multivariate analysis after adjusting for age, BMI, MELD score, complications, and infections revealed that SMI, and GFR and creatinine levels pre-LT were independent predictors for SMI post-LT. In this regard, lower SMI pre-LT predicted lower SMI post-LT consistent with a high rate of persistent sarcopenia in our cohort (86%). This finding is in agreement with previous studies in literature that reported persistent sarcopenia in patients after LT (
4,
27).
A novel observation from our results is that high creatinine levels predicted low post-LT SMI, while low GFR predicts low post-LT SMI. Moreover, patients with renal dysfunction were at higher risk of having sarcopenia within three months after LT. Renal impairment is commonly observed in patients with cirrhosis (
34,
35) (47% of patients exhibited renal dysfunction in our study) and creatinine levels are carefully considered in the calculation of MELD score to predict survival after LT (
34,
36). The relationship between renal function and sarcopenia in patients with cirrhosis has not been thoroughly studied; however, a recent French study recently reported high rate of 72% sarcopenic in patients undergoing simultaneous liver and kidney transplantation using the same Carey’s criteria (
37). Sarcopenia is commonly observed in patients with chronic renal disease, affects 11% to 29% in early stages of renal disease, and can reach over 65% in patients requiring dialysis (
38–
40). Furthermore, several studies reported associations between sarcopenia and low GFR (
41), high proteinuria (
39,
42), and albuminuria (
43) in renal patients. Mechanisms underlying these observations are not clear. De Souza et al proposed a role of inflammation in the development of sarcopenia in patients suffering from renal dysfunction and reported high levels of hs-CRP and low levels of IL-4 in patients with renal dysfunction and sarcopenia (
39). This is in agreement with a meta-analysis which included 3,072 sarcopenic patients and stated significant high levels of inflammatory marker CRP in patients with sarcopenia than in controls. In addition, other studies highlighted the role of mitochondria in the pathogenesis of sarcopenia in the context of renal dysfunction (
40,
44). Impaired mitochondrial function, which plays an important energy metabolism and amino acid metabolism, results in muscle function impairment and consequently muscle mass loss (
45). Indeed, amino acids are necessary for the activation of the mammalian target of rapamycin (mTOR) signalling pathway which is required for the stimulation of human skeletal muscle protein synthesis (
46). Besides, Plank et al reported a loss of 10% of total body protein from skeletal muscle during the first 10 days after LT (
31). Muscle constituting a major reservoir of total body protein (
47), a reduction in total body protein would be associated with reduction in muscle mass. Overall, renal dysfunction may lead to changes in mitochondria metabolism leading to an increase in oxidative stress and inflammation. Conversely, inflammation and oxidative stress may induce mitochondria damage in patients with renal dysfunction resulting in muscle wasting (
45,
48). Alcohol consumption could explain the relationship between sarcopenia and renal impairment in the context of cirrhosis. Alcohol abuse induces liver inflammation and has been associated with myopathy (
49–
51). A recent study by de Silva et al suggested a role of the inducible nitric oxide synthase in ethanol-induced oxidative stress and pro-inflammatory cytokines production in the kidney (
51). These findings could explain our results since alcohol was the main cause of liver disease. However, alcohol-induced oxidative and inflammatory state on kidney function merits to be properly investigated. Taken together, acute malnutrition with impaired metabolism linked to renal dysfunction may synergistically contribute to muscle mass loss in the early stage of post-transplant recovery.
The prognostic value of sarcopenia in cirrhotic patients undergoing LT has attracted increasing interest in recent years. Previous studies have reported the role of sarcopenia on clinical outcomes after LT (
52). However, in comparison to our study, those studies assessed sarcopenia at the listing time for LT (
4,
15,
53). As our results showed that muscle mass was subject to changes during a six-month peri-operative period of time, it would be important to reconsider the timing for sarcopenia assessment and schedule CT scan for all patients within three months post-LT. Indeed, many patients become sarcopenic early following LT and might be at higher risk of long-term negative outcomes. Accordingly, we demonstrated that early (<3 months) post-transplant sarcopenia was associated with an increased number of complications, infections and longer hospital stays after LT. The association between sarcopenia and clinical outcomes post-LT is difficult to explain. Though, some explanations can be proposed. Sarcopenia is correlated to frailty and physical limitations (
54). Yet, sedentary lifestyle is common in liver transplant recipients especially in the early post-operative period (
52,
55). Both sarcopenia and physical inactivity may progress together after surgery leading to negative clinical outcomes such as metabolic disorders, including diabetes and cardiovascular diseases.
Study limitations
We acknowledge limitations of our study due to retrospective and cross-sectional design. In line with this, we recognize the following limitations: (1) small size of our cohort because of missing data and reliance on available CT scans acquired according to clinical indications, (2) the retrospective nature did not allow us to evaluate physical performance (For a thorough evaluation of muscularity, the four following criteria should be considered: quantity [SMI], quality [myosteatosis], strength, and functional capacity), and (3) statistical analyses were also limited due to retrospectively missing data. Thus, some confounding factors were not considered in multivariate analysis. Though, our predictive models were able to predict more than 68% variation of SMI.