Comparison of the effects of probiotics, rifaximin, and lactulose in the treatment of minimal hepatic encephalopathy and gut microbiota

Abstract

Background: Minimal hepatic encephalopathy (MHE) is an early stage in the pathogenesis of hepatic encephalopathy. Intestinal microbiota is involved in the pathogenesis of hepatic encephalopathy and has become an important therapeutic target. Since there is no unified treatment principle for MHE, this study was conducted to determine the safety and efficacy of different intestinal microecological modulators in the treatment of MHE, and to explore the potential mechanism through intestinal microbiota analysis.

Methods: Patients with liver cirrhosis were screened for MHE using psychometric hepatic encephalopathy score test. Patients diagnosed with MHE were enrolled and received probiotics, rifaximin, or lactulose for 4 weeks. Adverse events were recorded. The psychometric hepatic encephalopathy score test was performed after treatment. Samples of blood and stool were collected at entry and 4 weeks. Blood samples were analyzed to assess blood ammonia, liver, kidney, and hemostatic functions. Stool microbiota were sequenced to confirm changes in microbial composition.

Results: Of 323 patients with liver cirrhosis, 74 patients were diagnosed with MHE. In all, 54 patients were enrolled and 52 who agree to follow-up were included in analysis. The recovery rates of MHE patients received probiotics, rifaximin, and lactulose were 58.8% (20/34), 45.5% (5/11), and 57.1% (4/7), respectively. Probiotics and rifaximin improved liver function in MHE patients to a certain extent. Taxonomic compositions of gut microbiota in MHE patients were distinct from healthy people before treatment; the differences were significantly reduced after treatment, and the gut microbiota gradually resembled the structure of healthy individuals. We found that the relative abundance of specific taxa associated with anti-inflammatory and good cognitive functions was increased in MHE patients after treatment. Accordingly, metabolic pathways in MHE patients were altered before and after treatment. Downregulated pathways after probiotics treatment included glycometabolism and degradation of aromatic compounds. After lactulose treatment, degradation pathways of arginine and ornithine showed a downward trend.

Conclusion: Probiotics, rifaximin, and lactulose are safe and effective in the treatment of MHE, and improve the composition of gut microbiota to some extent.

Keywords: gut microbiota; lactulose; minimal hepatic encephalopathy; probiotics; rifaximin.

FIGURE 1

Flow of patients of treatment for MHE.

FIGURE 2

Hematological indexes at baseline and 4 weeks showed significant improvement. TBIL (A), INR (B), and ALB (C) performance at baseline and 4 weeks in probiotics group and ALB (D) performance at baseline and 4 weeks in rifaximin group. TBIL, total bilirubin; INR, international normalized ratio; ALB, serum albumin.

FIGURE 3

The gut microbiota in MHE patients and controls. (A) Compositions of intestinal microbiota in different groups. (B) PCoA based on unweighted UniFrac matrix of bacterial taxonomy in all patients and control individuals.

FIGURE 4

Differential taxonomic features based on DESeq2 analysis. Changes in relative abundance of the gut microbiota between MHE patients and controls (A), MHE patients before and after probiotics treatment (B), MHE patients before and after rifaximin treatment (C), and MHE patients before and after lactulose treatment (D). Log2(baseMean): the log base 2 of base mean, the base mean is the mean of normalized counts of all samples, normalizing for sequencing depth. –loge(P): the negative of log base e of p value.

FIGURE 5

Differential metabolic features in patients before and after treatment. Comparison of differential metabolic features between MHE patients and healthy controls (A), MHE patients before and after probiotics treatment (B), MHE patients before and after rifaximin treatment (C), and MHE patients before and after lactulose treatment (D). q-value, adjusted p-value by Benjamini–Hochberg method. If q-value is not significant, p-value is used.

FIGURE 6

Key taxa related to MHE. Receiver operating curve (ROC)-analysis showed the performance of random forest model in distinguishing MHE and control group (A). The top 9 important taxa selected by random forest model (B).

FIGURE 7

Analysis of correlation. (A) Correlation coefficients between biomarkers and gut microbiota at genus level. (B) Key genus significantly correlated with different biomarkers. (C) Correlation coefficients between biomarkers and metabolic features. The Spearman correlation coefficients were calculated and labeled in the figure. AST, aspartate aminotransferase; ALT, alanine aminotransferase; ALB, serum albumin; TBIL, total bilirubin; PT, prothrombin time; INR, international normalized ratio; CRE, creatinine; CTP, Child–Turcotte–Pugh; MELD, model for end-stage liver disease. The Spearman correlation coefficients were calculated and labeled in the figure. Statistically significances are indicated: *p < 0.05, **p < 0.01.