METHODS: In this open-label trial, 75 patients with NASH fed a low-fat/low-calorie diet were randomly divided into the control group and experimental group, with the latter receiving the probiotic cocktail once daily for 12 weeks.

RESULTS: All patients were diagnosed with fatty liver by ultrasound examination and had elevated levels of γ-glutamyl transferase (GGT) and alanine aminotransferase (ALT), and slightly increased body mass index (BMI) and cholesterol levels. BMI and serum cholesterol were reduced by the low-fat/lowcalorie diet but ALT was not. However, the short-term (12-week) treatment with the probiotic cocktail caused a significant (by >20%) reduction of serum ALT compared with controls, indicating mitigation of inflammation. Accordingly, liver stiffness was decreased in the probiotic-treated group compared with the control group (P < 0.05). Moreover, a more significant decrease in the BMI and serum cholesterol was observed in the probiotic-treated group compared with control (P < 0.05). However, the reduction of GGT as a steatosis marker was insignificant. The composition of stool microbiota in probiotic-treated patients demonstrated a shift towards a normal pattern for all bacterial species examined. No adverse events were observed in any patient during the trial.

CONCLUSION: Short-term treatment with the probiotic cocktail caused significant improvement of liver inflammation without adverse events and, thus, may represent a promising candidate therapeutic approach for NASH.

INTRODUCTION

Non-alcoholic steatohepatitis (NASH) is an inflammatory condition caused by the accumulation of fat in the liver, which presents a serious health concern throughout the world, reaching an estimated prevalence of 5% in industrialized countries. It has been assumed that NASH originates from the more frequently observed non-alcoholic fatty liver disease (NAFLD), as characterized by steato hepatitis and the elevation of γ-glutamyl transferase (GGT), with an estimated prevalence of 25% in the Western and developing countries.¹

Increased inflammation accompanied by the elevation of transaminases (more significantly elevated alanine aminotransferase [ALT] compared with aspartate aminotransferase [AST]) promotes the progression of NASH to liver fibrosis and, eventually (during a time period of over 20 years) to cirrhosis, together with a potential to develop into hepatocellular carcinoma (HCC), which can occasionally occur without cirrhosis.² The pathogenesis of NASH remains unknown, although the following predisposing factors are considered: obesity, high-calorie/high-fat diet, diabetes mellitus (DM), hypercholesterolemia and certain drugs, etc.^3,4 However, these risk factors may be completely absent, indicating a possible impact of genetic background; thus, mutations in the patatin-like phospholipase domain-containing-3 (PNPLA3) gene have been shown to influence the severity of NASH.⁵ As the pathogenesis of NASH has not been completely understood, effective therapeutic approaches are not yet available. The symptomatic normalization of increased body weight, a low-calorie/low-fat diet, the lowering of elevated cholesterol levels, optimized therapy for DM, and avoidance of harmful medications are often not completely successful. Nevertheless, these measures should be taken to reduce the risk of hepatic steatosis and inflammation.

The composition of the gastrointestinal (GI) microbial community has recently emerged as a new component in the pathogenesis of NASH.⁶ The conventional view is that bacteria are either destroyed by the gastric acid and phospholipase-containing pancreatic juice or released with stool. The absorption of bacteria is harmful to the organism and there is no evidence that this process occurs physiologically.⁷ Accordingly, only microbial breakdown products or metabolites generated by the intestinal microbiota have an impact on the host metabolism by entering the organism through absorption in the small intestine.⁷ Consequently, alterations in the microbial community structure of the GI tract may have potential pathogenic or therapeutic effects on the development of NASH.

Treatment options for NASH are currently limited, either because of they are ineffective, as is the case with ursodeoxycholic acid (UDCA),^8,9 or because of the side effects observed with glitazones or farnesoid X receptor (FXR) agonists, such as weight gain, arrhythmia, itching, DM or hypercholesterolemia.^10,11 The impact of adverse events related to the drugs may be significant, as the treatment can last for decades because of the natural course of NASH, and it may eventually be more damaging than the disease itself. Therefore, it is desirable to develop effective therapeutic approaches without side effects, and the use of natural products may provide a solution to the problem. Probiotics have long been investigated for their health benefits, including those for patients with NASH.^12,13 In this study we aimed to examine the effect of a cocktail containing Lactobacillus casei (L. casei), Lactobacillus rhamnosus (L. rhamnosus), Lactobacillus bulgaricus (L. bulgaricus), Bifidobacterium longum (B. longum), Streptococcus thermophilus (S. thermophilus) and fructooligosaccharides (LBSF) in patients suffering from NASH.

MATERIALS AND METHODS

This was a randomized, controlled, non-blinded, prospective clinical trial that enrolled patients who were diagnosed as NASH with an age range of 30–60 years. The diagnosis of NASH in all the patients was made based on an ultrasonographic detection of hepatic steatosis (presenting as enhanced echogenicity compared with kidney parenchyma), and elevated serum GGT (>45 U/L) and ALT levels (>40 U/L). Valid transient elastography (Fibroscan) test results were also required to participate in the trial. Exclusion criteria were patients with other chronic liver diseases such as alcoholic steatohepatitis (alcohol consumption of more than 40 g daily for men and more than 20 g daily for women¹⁴), chronic hepatitis B and C, autoimmune hepatitis, Wilson’s disease, hemochromatosis and cholestatic liver disease diagnosed based on the anamnesis and standard laboratory procedures. Furthermore, the protocol excluded very obese patients with the body mass index (BMI) >30 kg/m² , those with DM (blood glucose >5.6 mmol/L), hypertriglyceridemia (>1.7 mmol/L), with severe comorbities, and pregnant or lactating women. To avoid patient withdrawal from the trial due to fear of possible complications caused by liver biopsy as well as for safety issues, no biopsy was performed in this pilot trial. The enrolled patients were randomly assigned to the experimental and control groups by using closed envelop drawing. All patients with NASH were provided with basic treatment consisting of a low-fat/low-calorie diet (30–90 g fat/day and 1800 kcal/day) after appropriate instructions. In addition, the experimental group received LBSF (Lactiale; Farmak, Kiev, Ukraine), containing L. casei, L. rhamnosus, L. bulgaris, B. longum and S. thermophilus (10⁸ bacteria/capsule in total) as well as fructooligosaccharides, once daily for 12 weeks. Significant suppression of ALT and GGT levels were used as endpoints of the trial to indicate the therapeutic efficacy of LBSF.

Moreover, before and after the 12-week treatment period, all patients underwent liver ultrasound for the evaluation of hepatosteatosis and Fibroscan examination (Aixplorer, Supersonic Imagine, Aix-en-Provence, France) with shear wave elastography to obtain objective information about the stiffness (elastic properties) of the liver tissues. In case of a skin capsular distance >2.5 cm, the obese-specific XL probe was applied. Since in our study very obese patients were excluded, the XL sensor was rarely used. Stiffness was determined based on the propagation of elastic waves during 20–30 ultrasonic pulses followed by a calculation of the average valid deformation pressure (kPa) with an interquartile range/median value of <0.3.^15,16 Nonvalid measurements with a successful measurement rate <60% did not allow the exact determination of liver stiffness and these patients were excluded from the trial. Normal values for healthy livers were 4.1 +- 0.89 kPa (range 2.3–5.9 kPa), and in this trial increased stiffness besides fibrosis also indicated liver inflammation. In addition, glucose, triglycerides, cholesterol and high-density lipoprotein cholesterol levels were determined. Microbial composition in stool was quantified by standard techniques as recently described.^17,18 Another parameter in the study was adverse events, which included an increase in body weight or any subjective changes such as heart arrhythmia.

Ethical permission

All study participants were informed of the study design, nature of the treatment, and the properties of the drug, and all signed a written consent form. The trial was not registered in a publicly accessible database because this was not the policy in Ukraine. The approved study protocol was registered by the regional committee for research ethics in Kiev (no. 942 of 9 December 2015). This registration policy applies to prospective, randomized, controlled trials.

Statistical analysis

Statistical analyses were performed using SPSS 20 (IBM, Armonk, NY, USA). All data are expressed as the mean +- standard deviation or numbers and percentages, respectively. The Kolmogorov–Smirnov normality test was used for data distribution analysis; all values showed parametric distribution. Analysis of variance (ANOVA) was used for multiple comparisons followed by a post-hoc Tukey’s test for significant results. Comparison of the data before and after treatment was performed for each patient using the Student’s t-test for paired samples. The differences between groups were considered significant at P < 0.05.

RESULTS

For this trial, we chose 75 patients with NASH. The enrolled patients were randomly assigned to the experimental (n = 38) or the control group (n = 37). These patients did not show clear risk factors, such as overweight, DM or hypertriglyceridemia, for NASH development1 but were characterized by the NASH phenotype, defined as steatohepatitis with elevated levels of GGT (>45 U/L) and ALT (>40 U/L). None of the patients had alcoholic steatohepatitis, chronic viral hepatitis (B, C and D), autoimmune hepatitis, Wilson’s disease, hemochromatosis, cholestatic liver disease, or treatable or severe metabolic disorders such as obesity, DM and hypertriglyceridemia, which are known to cause NASH.

Both groups of patients with NASH (the control and the experimental groups) were instructed to maintain a low-calorie/low-fat diet (30–90 g fat/day and 1800 kcal/day) as basic therapy. The experimental group received the additional LBSF probiotic cocktail. The baseline characteristics of the patients did not differ between the two groups (Table 1). Patients in both groups had elevated ALT and GGT levels, increased BMI and serum cholesterol level, as well as increased liver stiffness compared with the normal range.

In all the participants, the low-fat/low-calorie diet resulted in the reduction of their BMI. In the control group, a significant decrease in serum cholesterol level was observed (P < 0.05), whereas serum ALT and GGT levels remained unchanged. In contrast, in the experimental group, 12-week therapy with LBSF significantly decreased the serum levels of ALT and AST (P < 0.05), returning both to within the normal range, and improved liver stiffness (P < 0.05; Table 2). However, the decrease in GGT level was not statistically significant between the two groups, and hepatosteatosis persisted as evidenced by ultrasound examination. Furthermore, the decreases in BMI and cholesterol level were more pronounced in the experimental group than in the control group (P < 0.05), while serum glucose and triglyceride remained constantly within the normal ranges in both groups (Table 2).

The microbial composition in the feces of our population was compared with that of standard distribution in healthy, age-matched control individuals on a regular diet (in-house historical data of 50 individuals). The borderline levels are indicated as colony forming unit (CFU) in Table 3. For our control study population an approximate reduction of 50% was noticed for Bifidobacteria, Lactobacilli, Escherichia coli (E. coli) with normal properties and Enterococcus faecalis. At the same time, higher proportions (21.6–45.9%) of enterohemorrhagic E. coli, pathogenic enterobacteria, non-pathogenic Staphylococcus, S. aureus, Klebsiella, Proteus, Citrobacter and Candida were detected (Table 3).

The application of the probiotic cocktail significantly changed the microbial structure in the feces of the experimental group. In all cases, an increase towards the normal range was detected. No significant change was registered for pathogenic enterobacteria. However, these showed low abundancy and remained largely within the normal range of distribution. The increase in Bifidobacteria and Lactobacilli in the experimental group was expected, because these microorganisms were constituents of the probiotic cocktail. For the other bacterial species the shift to normal levels could be attributed to the beneficial effects of LBSF because the low-fat/low-calorie diet was provided to both groups.

Adverse events were not recorded in any patient during the trial. During the study period, subjective sensations remained unchanged in regard to abdominal discomfort. There were no indications of more frequent urinary infections or heart arrhythmias.

DISCUSSION

In this short and small-sized trial, the efficacy of LBSF for treating NASH was evaluated. The diagnosis of NASH was simplified and based on the imaging evidence of hepatosteatosis and the elevation of serum ALT and GGT levels. Other chronic liver diseases with obvious etiologies, as well as metabolic disorders including obesity, diabetes and hypertriglyceridemia were ruled out. The diagnostic criteria applied here describe a subpopulation of patients with NASH who have no other obvious risk factors for NASH development and for whom a potential pathogenic role of gut microbiota could be postulated.¹³ Therefore, the rationale for this study was that changes in the bacterial community structure of the gastrointestinal tract with a probiotic cocktail may have a therapeutic effect on patients with NASH. If proved successful, this strategy may be beneficial as adjunctive therapy also for metabolic disorder-caused NASH.

Our results indicated that after a short treatment period of 12 weeks, LBSF decreased ALT levels and liver stiffness without any adverse events, indicating that the patients’ inflammatory status had been improved. The change in the microbial composition of the feces was also observed after the application of the LBSF probiotic cocktail, which partially passed through the small intestine to the colonic lumen. In the control group the microbial spectrum showed significant deviations from the normal condition, suggesting that an abnormal bacterial profile in NASH may present a risk for its pathogenesis. However, the application of the LBSF cocktail reversed these changes and stabilized the microbial composition in patients with NASH, which was correlated with the improvement of their inflammatory conditions. These data indicate that the intake of LBSF is clearly beneficial for patients with NASH, although the underlying mechanism for doing so is unknown. It remains to be determined whether the anti-inflammatory effect, such as ALT downregulation, can be attributed to metabolic products generated by the restructured microbial community or the absorption of distinct bacterial fragments, which exert positive effects on the metabolism of patients with NASH. The restoration of the microbial structure to the normal state may also be a result of the suppression of pathogenic microbes by the probiotics ingested in the cocktail therapy.

The strength of the present investigation is the discovery of a potential strategy to improve the inflammatory status in patients with NASH by applying a harmless probiotic preparation. It should be noted that in our patients, inflammation, but not steatosis, was decreased. The reduction of liver stiffness may reflect improved tissue elasticity through decreased inflammation rather than the resolution of fibrosis, which cannot be expected over such a short trial period. Thus, the anti-inflammatory activity of this probiotic cocktail may not only be applicable to the treatment of NASH but also to other inflammation-associated liver conditions, including alcoholic steatohepatitis or drug-induced liver injury.

Dietary fat and calorie restriction caused significant weight loss in both patient groups, although the decrease in BMI was more significant in the experimental group than in the control group. The control group also had a reduction in total cholesterol levels compared with baseline, which confirms the wellknown benefits of a low-fat/low-calorie diet on body weight and cholesterol metabolism. In fact, elevated serum cholesterol was common for our patients with NASH, and it remains to be determined whether hypercholesterolemia is a cause of NASH or just an epiphenomenon.

Our pilot study had some limitations. First, it was a non-blinded trial based on a small number of patients treated for a short time. Second, patients were diagnosed with NASH based only on GGT and ALT elevation together with the detection of steatosis by ultrasonography and enhanced liver stiffness by fibroscan. However, we did not perform liver biopsy, and, consequently, did not analyze proinflammatory tissue markers. In a confirmatory trial, biopsies at the beginning and end of the study period should be conducted. Third, the trial covered only a small fraction of patients with NASH, because those with overt metabolic disorders were excluded. Finally, the mechanism behind the changes of the microbial pattern in patients with NASH, and their influence on patients’ inflammatory state remain to be elucidated.

In conclusion, this study reveals the strength of introducing a new probiotic therapeutic strategy with no adverse events for the treatment of NASH, which should be further evaluated in larger double-blinded randomized controlled trials.