Patients with cirrhosis have high rates of progression to acute-on-chronic liver failure (ACLF), and predicting the development of ACLF is a key focus of clinical research. Two new studies approach the challenge from different angles: one identifies distinct clinical courses and pathophysiologies of patients admitted to hospital with acute decompensation of cirrhosis, whereas the other uses metabolomics to establish metabolite profiles to identify patients at high risk of developing ACLF.
In the first study, clinical and laboratory data for 1,071 patients hospitalized with acutely decompensated cirrhosis but without ACLF were collected and followed up for 3 months; 12-month outcomes were also recorded. The prospective, observational European study, called PREDICT, aimed to stratify these patients according to clinical course of acute decompensation and to predict ACLF development.
Credit: Laura Marshall/Springer Nature Limited
Three clinical courses were identified: pre-ACLF (n = 218; developed ACLF within 90 days; high systemic inflammation), unstable decompensated cirrhosis (n = 233; did not develop ACLF but had ≥1 readmission; low-grade systemic inflammation but exhibited complications related to severe portal hypertension), and stable decompensated cirrhosis (n = 620; did not develop ACLF and were not readmitted; no severe systemic inflammation or frequent portal hypertension complications). Mortality rates also differed between the three groups. “These data open a novel path for the stratification of care, development of treatment strategies and personalized medicine, for example, treating patients with pre-ACLF with therapies targeting systemic inflammation and patients with unstable decompensated cirrhosis with therapies targeting portal hypertension,” says Jonel Trebicka, corresponding author of the study.
In the second prospective study, researchers collected serum samples from 602 patients with cirrhosis at the time of admission to 12 North American hepatology centres. Metabolomic analyses using liquid chromatography, random forest analysis and area under the curve (AUC) analysis were used to identify metabolites associated with clinical outcomes. “We need to determine predictors of who develops ACLF rather than wait to see who has already developed ACLF to initiate a more proactive rather than reactive approach,” says Jasmohan Bajaj, corresponding author of the study.
Levels of microbial metabolites (such as bile acids and aromatic amino acid metabolites) at the time of admission were associated with ACLF development and mortality. For example, lower levels of conjugated, secondary and sulfated bile acids, lower levels of indoxyl-sulfate and indole propionate and higher levels of indoleacetate were associated with ACLF. Levels of microbial metabolites identified patients who developed ACLF with an AUC of 0.84 (95% CI, 0.78–0.88; P = 0.001), patients who died in the hospital with an AUC of 0.81 (95% CI, 0.74–0.85; P = 0.002), and patients who died within 30 days with an AUC of 0.77 (95% CI, 0.73–0.81; P = 0.02). In addition, lipid metabolism-associated metabolites (specifically, lower phospholipids and higher oestrogen metabolites) were associated with ACLF and mortality. The associations remained substantial after controlling for clinical variables, such as age and sex, serum albumin and sodium levels, and white blood cell counts. Stool samples from 133 of the patients were also collected, and levels of specific microbial metabolites associated with particular clinical outcomes correlated with microbiota composition.
“Levels of microbial metabolites… at the time of admission were associated with ACLF development and mortality”
Taken together, the findings of the studies raise hope that the development of ACLF in patients with cirrhosis could be predicted, by clinical course, pathophysiology and metabolite profiles. “The major task is to predict these phenotypes already at hospital admission using extensive biomarker research and to develop treatment strategies,” says Trebicka of the three clinical courses identified by the PREDICT study. “The implication for future research and potential application is to define a focused set of metabolites that can be used to guide clinicians and researchers to determine prognosis and pathophysiology of ACLF,” says Bajaj. “We plan to initiate pilot studies focused on modulation of the microbiota to potentially reduce the pathobionts that generate the metabolites associated with these poor outcomes.”
About this article
Cite this article
Hindson, J. Predicting development of ACLF.
Nat Rev Gastroenterol Hepatol (2020). https://doi.org/10.1038/s41575-020-0355-z