Nutrition 27 (2011) 177181

Contents lists available at ScienceDirect

Nutrition
journal homepage: www.nutritionjrnl.com

Applied nutritional investigation

Probiotics for patients with compensated liver cirrhosis: A double-blind

placebo-controlled study
David Pereg M.D. a, b, *, Andy Kotliroff M.D. a, b, Natan Gadoth M.D. b, d, Ruth Hadary M.D. b, c,
Michael Lishner M.D. a, b, Yona Kitay-Cohen M.D. b, c
a
Department of Internal Medicine A, Meir Medical Center, Kfar-Saba, Israel
b
Sackler Faculty of Medicine, Tel-Aviv University, Tel-Aviv, Israel
c
Hepatology Unit, Meir Medical Center, Kfar-Saba, Israel
d
Department of Neurology, Mayanei Hayeshuah Medical Center, Bnei Barak, Israel

a r t i c l e i n f o a b s t r a c t

Article history: Background: Gut ora is related to the major complications of liver cirrhosis including hepatic
Received 19 October 2009 encephalopathy, spontaneous bacterial peritonitis, and variceal bleeding. Prior studies have
Accepted 10 January 2010 reported a benecial effect of gut ora modication with probiotic bacteria in patients with
minimal hepatic encephalopathy. We aimed to study the effect of probiotics on clinical and
Keywords: laboratory parameters of patients with compensated cirrhosis.
Gut ora
Methods: A double-blind placebo-controlled study that included patients with liver cirrhosis and at
Cirrhosis
least one major complication of cirrhosis in the past, clinical evidence of portal hypertension, or
Ammonia
Hepatic encephalopathy decreased hepatic synthetic function. Participants were randomly assigned to receive probiotic
capsules containing Lactobacillus acidophilus, Lactobacillus bulgaricus, Bidobacterium lactis, and
Streptococcus thermophiles or placebo for a period of 6 mo.
Results: A total of 36 patients were available for nal analysis (distributed equally between the
probiotic and placebo groups). The administration of probiotics was not associated with signicant
differences in either clinical or laboratory parameters between the two groups. Because the lack of
a benecial effect may be related to the compensated liver disease of patients, we conducted
a subanalysis of patients with baseline ammonia levels >50 mmol/L. In this subgroup, the
administration of probiotics appeared to signicantly reduce the ammonia levels starting after
1 mo of treatment. However, this effect diminished and lost its signicance following comparison
to the placebo group.
Conclusions: Our study did not show a signicant benecial effect of probiotic supplementation in
patients with compensated liver cirrhosis. Nevertheless, it points toward a possible positive effect
of probiotics in patients with above normal baseline ammonia levels. This issue requires further
investigation in larger cohorts.
2011 Elsevier Inc. All rights reserved.

Introduction peritonitis [2,3]. There is also an accumulation of data regarding

Colonic bacteria clearly play a major role in the pathogenesis
of major complications of liver cirrhosis. Urease-producing gut
bacteria including Klebsiella and Proteus species are associated
with increased production of ammonia and endotoxins that can
cause hepatic encephalopathy [1], and their translocation from
the gut to the peritoneal cavity appears to be the major
mechanism for the development of spontaneous bacterial
* Corresponding author. Tel: 972-9-7472534; fax: 972-9-7460781.
E-mail address: david_pereg@post.tau.ac.il (D. Pereg).
0899-9007/$ see front matter 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.nut.2010.01.006
a possible relation between bacterial translocation and bleeding
from esophageal varices [4,5]. Finally, most of the therapeutic
measures for both the treatment and the prevention of these
complications, such as the administration of antibiotics and
lactulose, are at least partially directed against gut bacteria [5,6].
In recent years there has been a growing interest in the effect
of probiotic bacteria on gut ora and on the prevention of
infections. Although they were found to prevent pediatric
infectious diarrhea [79] and antibiotic-associated diarrhea
especially those caused by Clostridium difcile [10,11]their
effect on the prevention of diarrhea in adults not treated with
178 D. Pereg et al. / Nutrition 27 (2011) 177181
antibiotics was less striking [12,13]. Probiotic bacteria, the most
common among which are the lactose-fermenting Lactobacilli,
inhibit the growth of pathogenic bacteria by acidifying the gut
lumen, competing for nutrients, and producing antimicrobial
substances [14,15]. Furthermore, they adhere to the gastroin-
testinal mucosa and by that are thought to prevent bacterial
translocation from the gut [16]. It has been demonstrated that
the administration of probiotic preparations to cirrhotic patients
resulted in a modulation of the gut ora with a signicant
reduction in the quantity of several bacterial pathogens in fecal
bacteriological analysis [15]. The accumulation of data regarding
the effects of probiotics has raised the possibility that they may
be benecial in the treatment and prevention of the complica-
tions of cirrhosis. It has been reported that 30-d administration
of probiotic bacteria to cirrhotic patients with minimal chronic
hepatic encephalopathy resulted in 50% reversal of encepha-
lopathy [15]. Furthermore, a signicant decrease in blood levels
of ammonia, bilirubin, and ALT, and an improvement in the
Child-Turcotte-Pugh functional class, was noted. Another study
has assessed the clinical efcacy of a 3-mo administration of
Bidobacterium longum plus fructo-oligosaccharides in 60
cirrhotic patients with minimal hepatic encephalopathy [17].
Improvement in both biochemical and neuropsychological tests
was noted in the probiotics compared to the placebo group.
A third study has shown a signicant rate of minimal hepatic
encephalopathy reversal following 60-d consumption of a pro-
biotic yogurt [18]. These studies showed promising results
but were carried out on a very specic population of patients
with cirrhosis, and therefore, did not examine whether the
administration of probiotics is benecial in the general pop-
ulation of patients with compensated cirrhosis. Furthermore, it
is still unknown whether this benecial effect of probiotics
persists when patients are treated for a longer duration. There-
fore, our placebo-controlled study was aimed to investigate
these issues.
Material and methods
Patients
Our population consisted of patients with hepatic cirrhosis that were being
followed up at the liver clinic in our medical center. Cirrhosis was diagnosed
either histologically or based on clinical and radiologic grounds in the patients for
whom biopsy was contraindicated or refused. Patients were considered to have
alcohol-related cirrhosis if alcohol intake had been in excess of 80 g/d in men and
30 g/d in women for more than 5 y and testing for viral, metabolic, and immune
causes was negative.
Inclusion and exclusion criteria
Included in our study were patients with liver cirrhosis and at least one of the
following:
1) One or more major complication of cirrhosis in the past (including variceal
bleeding, encephalopathy, or spontaneous bacterial peritonitis).
2) Clinical evidence of portal hypertension (including the presence of ascites,
splenomegaly with hypersplenism, or esophageal varices).
3) Decreased hepatic synthetic function.
In this study we focused on ambulatory patients with compensated liver
disease. Therefore, patients were excluded from the study if they presented with
any sign of decompensation from any precipitant including gastrointestinal
bleeding, infections, acute renal failure, electrolyte impairment, or hepatocellular
carcinoma. Patients were also excluded if they were chronically treated with
antibiotics or lactulose. Patients with alcoholic cirrhosis, for whom alcohol
abstinence for at least 2 months prior to enrollment could not be conrmed, were
not included either.
Outcome denitions
Hospital admissions due to decompensation of liver disease as well as
ammonia levels during the 6-mo study period were the two primary endpoints of
our study. Secondary endpoint included the other liver function parameters
(including AST, ALT, albumin, bilirubin, and prothrombin time) and the electro-
encephalogram (EEG) and psychometric tests.
Ethical considerations
The study was approved by the research ethics committee at Meir Medical
Center and written informed consent was obtained from each participant. The
study was registered in the ClinicalTrials system (ID: NCT00312910).
Supplementation with the probiotic product
Both probiotic and placebo capsules were delivered in identical pillboxes that
were consecutively numbered. All the boxes were pooled and one box was
randomly drawn for each participant at the time of enrollment. Thereafter,
during every follow-up visit, each patient received new boxes with the same
consecutive number. Whether each box contained probiotic bacteria or placebo
remained unknown to both the investigators and the participants until after all
patients had completed the study and codes were broken.
Patients drawn to the probiotic group received capsules containing four
freeze-dried bacteria, namely Lactobacillus acidophilus, Lactobacillus bulgaricus,
Bidobacterium bidum, and Streptococcus thermophiles (Bio-plus, Supherb,
Israel), each at a daily dose of 2  1010 colony forming units. Participants in the
placebo group received capsules containing wheat-based non-fermentable
bers. Both groups were treated for a 6-mo period. Patients compliance was
monitored by collecting empty boxes at each follow-up visit.
Patients assessment
Patients enrolled in the study were scheduled for three follow-up visits
after 1, 3, and 6 mo (the end of the study). At enrollment and during all follow-up
visits, blood tests were obtained (see below). A psychometric test and an EEG
were conducted only on enrollment and at the completion of the study.
Laboratory analysis
The analyses of blood samples, including the measurement of serum
albumin, bilurubin, both transaminases and prothrombin time, were performed
on fresh samples in our laboratory facility. The laboratory is authorized to
perform tests according to the international quality standard ISO-9002.
Prothrombin time was presented as international normalizing ratio (INR)
values. Ammonia levels were measured using the ammonia kit of Integra 700
analyzer (Roche, Indianapolis, IN, USA) according to the manufacturers
instructions.
Electroencephalogram
Patients underwent a routine scalp EEG using a Nihon Kohden EEG machine,
with recording of 14 channel montages, at enrollment and at the end of the study.
Minimal hepatic encephalopathy was considered when slowing of the occipital
basic rhythm or runs of teta activity were observed.
Psychometric test
This was performed using a battery of the digit span, trail making, and the
visual scanning and attention test. All three tests were previously used for
assessment of hepatic encephalopathy [19]. Patients whose performance on any
of the above tests was at least two standard deviations below that of age- and
educational-matched controls were considered to have minimal hepatic
encephalopathy.
Statistical analysis
Sample size was estimated with respect to the single randomized study that
was available at the time of the study design [15]. In this study the probiotic and
placebo groups included 20 and 15 patients, respectively. We assumed that the
much longer follow-up period of our study (6 mo compared to only 1 mo in the
previous report) would balance the better clinical status of patients enrolled in
our study.
The differences between the two groups were analyzed using two-tailed
t test and c2 test, each when appropriate. A difference was considered signi-
cant when the P value was  0.05. For comparing changes in linear variables
between the two groups including the effect of treatment period length, we used
a general linear model with repeated measures. All statistical analyses were
performed with the use of SAS statistical software, version 9.1.
 D. Pereg et al. / Nutrition 27 (2011) 177181 179

Results Table 2
Comparison of laboratory values between the probiotic and placebo groups
according to the time from enrollment (1, 3, 6 mo following enrollment)
Baseline characteristics
Laboratory values Probiotics Placebo P value
Forty patients with cirrhosis were enrolled in our study and (mean  SD)

were randomly assigned to the probiotic (n 20) and the Albumin

Enrollment 3.6  0.5 3.7  0.6 0.57
placebo (n 20) groups. Four participants (two from each
1 mo 3.75  0.6 3.8  0.6 0.8
group) dropped out of the study within the rst month post 3 mo 3.7  0.6 3.8  0.6 0.74
enrollment (all of them due to low compliance), leaving 36 6 mo 3.7  0.6 3.6  0.5 0.66
participants eligible for analysis. There were no signicant AST
differences in the age, etiology of cirrhosis, baseline laboratory Enrollment 58.4  25.9 62.2  32.2 0.7
1 mo 56.4  25.4 67.2  36 0.31
characteristics, and the Child-Turcotte-Pugh score between the 3 mo 54.3  26.3 70.4  38.9 0.13
probiotic and placebo groups (Table 1). Seven patients (19.4%) 6 mo 54.4  31 66.4  29.4 0.28
were considered to have minimal hepatic encephalopathy ALT
according to the abovementioned EEG ndings or their perfor- Enrollment 50.2  32.6 55  34.5 0.67
1 mo 49.3  37.6 63.7  45 0.3
mance on the psychometric tests. Four of them were from the
3 mo 47.2  32 64  30.2 0.22
probiotic group (22.2%) and three belonged to the placebo group 6 mo 49.6  29.3 61.4  31 0.35
(16.6%). Total bilirubin
Enrollment 1.2  0.5 1.3  0.6 0.6
1 mo 1.2  0.6 1.4  0.8 0.35
3 mo 1.1  0.5 1.1  0.6 0.96
Comparison between the probiotic and placebo groups
6 mo 1.1  0.4 1.1  0.6 0.95
Creatinine
During the 6-mo follow-up period of our study, there were six Enrollment 1.1  0.4 1  0.3 0.37
cases of hospital admissions, three from each group. One of the 1 mo 1.1  0.4 1  0.4 0.17
cases (from the probiotics group) was admitted for tense ascites 3 mo 1.1  0.4 0.9  0.2 0.12
6 mo 1.1  0.3 0.9  0.2 0.14
that required paracentesis and another case (from the placebo
Ammonia
group) due to upper gastrointestinal bleeding secondary to Enrollment 49.6  30.2 46.4  23.7 0.72
gastric erosions. The other four patients were admitted due to 1 mo 42.3  23.2 47.3  25.3 0.54
acute illnesses not related to their liver disease. None of the 3 mo 42.1  26.9 45.2  23.6 0.7
participants in the study demonstrated any change in his Child- 6 mo 43.2  22.1 48.5  22 0.5
INR
Turcotte-Pugh score during the study period. Table 2 compares Enrollment 1.1  0.08 1.1  0.1 0.34
the laboratory values for the probiotic and placebo groups 1 mo 1.1  0.1 1.1  0.14 0.48
obtained 1, 3, and 6 mo from enrollment. There was no signi- 3 mo 1.1  0.09 1.1  0.13 0.25
cant difference in the plasma levels of albumin, ALT, AST, bili- 6 mo 1.1  0.08 1.1  0.1 0.37
rubin, creatinine, ammonia, and in the INR, between the two ALT, alanine aminotransferase; AST, aspartate aminotransferase; INR, international
groups. Of the seven patients that were diagnosed with minimal normalizing ratio
hepatic encephalopathy at enrollment, none demonstrated
reversal of encephalopathy either on the EEG or on the psycho-
Subanalysis of patients with high baseline levels of ammonia
metric test. There were no new cases of minimal hepatic
encephalopathy diagnosed at the end of the study. Throughout
We further compared the effect of probiotics on ammonia
the study, none of the participants developed clinical or
levels of patients with baseline ammonia levels above the
electroencephalographic ndings of overt encephalopathy.
normal range (which is considered >50 mmol/L in our labora-
tory). A total of 14 patients (7 patients from each group) were
suitable for this subanalysis. Among the seven patients from the
Table 1 probiotic group, the ammonia level decreased signicantly after
Comparison of baseline characteristics between the probiotic and placebo groups 1 mo of treatment from 82.6 to 54.3 mmol/L (P 0.018). This
Probiotics Placebo P value effect remained signicant but to a lesser extent after 6 mo of
(n 18) (n 18) treatment (82.6 to 59.6 mmol/L, P 0.046). However, following
Age (mean  SD) 63.2  10.5 65.9  8.4 0.4 a comparison with the placebo group, the apparent reduction in
Etiology of cirrhosis, n (%) ammonia levels in the probiotic group lost its statistical
HCV 9 (50%) 10 (55.6%) 0.82 signicance (P 0.537) (Fig. 1).
Other (HBV, alcoholic, cryptogenic) 9 (50%) 8 (44.4%) 0.74
Laboratory data (normal values)
Albumin (3.55.5 g/dL) 3.6  0.5 3.7  0.6 0.57 Discussion
AST (737 U/L) 58.4  25.9 62.2  32.2 0.7
ALT (034 U/L) 50.2  32.6 55  34.5 0.67
Total bilirubun (0.21.5 mg/dL) 1.2  0.5 1.3  0.6 0.6 Main ndings
Creatinine (0.51.2 mg/dL) 1.1  0.4 1  0.3 0.37
Ammonia (1150 mmol/L) 49.6  30.2 46.4  23.7 0.72 Our study did not show any signicant clinical or laboratory
INR 1.1  0.08 1.1  0.1 0.34
effect of probiotic administration in patients with compensated
Patients with MHE (%) 4 (22.2) 3 (16.6) 0.67
CTP score (mean  SD) 61 5.8  1.1 0.56 liver cirrhosis. Although among the subgroup of patients with
CTP score distribution (A/B/C) 13/5/0 14/4/0 0.7 high baseline levels of ammonia the administration of probiotics
ALT, alanine aminotransferase; AST, aspartate aminotransferase; CTP,
appeared to signicantly reduce the ammonia levels, especially
Child-Turcotte-Pugh; HBV, chronic hepatitis B; HCV, chronic hepatitis C; MHE, after 1 mo of treatment, this effect diminished and lost its
minimal hepatic encephalopathy signicance following comparison to the placebo group.
180 D. Pereg et al. / Nutrition 27 (2011) 177181
90
85
80
75
70
65
Ammonia
60
(mmol/L)
55
50
45
40
35
30
0 1 3 6
Time (months)
Placebo Probiotics
Fig. 1. A comparison of ammonia levels between the placebo and probiotic groups
in the subpopulation of patients with baseline ammonia levels above the normal
range (n 7 patients in each group). The x axis represents time from enrollment (in
months) and the y axis represents ammonia level (in mmol/L). P 0.537 for the
comparison between the 2 groups.
Comparison to prior studies
Only three randomized double-blind placebo-controlled
studies [15,17,18] have investigated the administration of
probiotics to patients with cirrhosis. These studies have reported
a dramatic clinical and laboratory improvement in cirrhotic
patients following 1 to 3 mo of treatment with probiotics.
However, there are several differences between these studies
and ours. First, while they included only patients with minimal
hepatic encephalopathy, only a minority of patients in our study
demonstrated electroencephalographic or psychometric ndings
suitable with this diagnosis. Furthermore, the percentage of
patients with Child-Turcotte-Pugh class > A (score > 6) at
enrollment to our study was relatively low (25%). Therefore, it
may be speculated that different patient selection and less
advanced level of cirrhosis in our population led to the different
results. Other reasons may be attributed to various demographic,
environmental, and nutritional differences between the two
populations because the prior studies were conducted in China
and the USA, while ours included both Jewish and Arab Israeli
patients.
Another controversial point common to all studies on
probiotics is the duration of treatment required to achieve the
optimal effect. While several reports have shown a positive
effect after a few days of treatment [7,8], others speculated that
a benecial prophylactic effect can only be expected with
regular consumption of the probiotic agent [20,21]. Regarding
patients with liver cirrhosis, the existing evidence demon-
strated a benecial effect after 1 mo of treatment [15]. In our
study patients were treated for 6 mo without any apparent
effect.
Study limitations
Our study was designed to study the role of probiotics in
patients with compensated cirrhosis. These patients comprise
the largest portion of patients with cirrhosis that are being
followed up in the ambulatory setting. Indeed the mean Child-
Turcotte-Pugh score of our population was relatively low and
only few had features of minimal hepatic encephalopathy. To
partially address this limitation, we conducted a subanalysis of
patients with high baseline levels of ammonia. We speculated
that by adding this criterion we could identify patients with
more advanced liver disease who may benet more from
probiotics that have been shown previously to reduce ammonia
levels. Indeed, in this small subgroup, the administration of
probiotics appeared to signicantly reduce the ammonia levels
especially after 1 mo of treatment. However, this effect dimin-
ished and lost its signicance following comparison to the
placebo group. It seems that probiotics administration appears
to have a higher benecial impact as the liver disease is more
decompensated.
It is possible that the negative results of the study may be
due to the relatively small population. Nevertheless, the study
is, to our knowledge, the rst ever to examine the role of pro-
biotics in patients with compensated cirrhosis and it presents
the experience of a single medical center. To completely assess
the effect of probiotics in compensated cirrhotic patients,
further investigation in larger multicenter studies is required.
Following our results it may be suggested to focus on patients
with compensated cirrhotic and elevated baseline ammonia
levels.
Conclusions
This study was, to our knowledge, the rst to study the effect
of probiotics on patients with well-compensated liver cirrhosis.
Although it failed to demonstrate any signicant clinical or
laboratory effect to probiotics administration to this population,
it points to a possible benecial effect in patients with above
normal baseline ammonia levels. This issue requires further
investigation in larger cohorts.
Acknowledgments
We thank Supherb Ltd, Israel, for the donation of the probiotic
and placebo capsules used in the study.
References
[1] Basile AS, Jones EA. Ammonia and GABAergic neurotransmission: Interre-
lated factors in the pathogenesis of hepatic encephalopathy. Hepatology
1997;25:13035.
[2] Guarner C, Soriano G. Spontaneous bacterial peritonitis. Semin Liver Dis
1997;17:20317.
[3] Campillo B, Pernet P, Bories PN, Richardet JP, Devanlay M, Aussel C. Intestinal
permeability in liver cirrhosis: relationship with severe septic complications.
Eur J Gatroenterol Hepatol 1999;11:7559.
[4] Thalheimer U, Triantos CK, Samonakis DN, Patch D, Burroughs AK.
Infection, coagulation, and variceal bleeding in cirrhosis. Gut 2005;54:
55663.
[5] Husova L, Lata J, Husa P, Senkyrk M, Jurankova J, Dte P. Bacterial infection
and acute bleeding from upper gastrointestinal tract in patients with liver
cirrhosis. Hepatogastroenterology 2005;52:148890.
[6] Riordan SM, Williams R. Treatment of hepatic encephalopathy. New Engl
J Med 1997;337:4739.
[7] Szajewska H, Kotowska M, Mikolajczyk W, Armanska M, Miko1ajczyk W.
Efcacy of Lactobacillus GG in prevention of nosocomial diarreha in infants.
J Pediatr 2001;138:3615.
[8] Isolauri E, Juntunen M, Rautanen T, Sillanaukee P, Koivula T. A human
Lactobacillus strain (Lactobacillus casei sp strain GG) promotes recovery
from acute diarrhea in children. Pediatrics 1991;88:907.
[9] Saavadra JM, Bauman NA, Oung I, Perman JA, Yolken RH. Feeding of Bi-
dobacterium bidum and Streptococcus thermophilus to infants in
hospital for prevention of diarrhoea and shedding of rotavirus. Lancet
1994;344:10469.
[10] Young RJ, Whitney DB, Hanner TL, Antonson DL, Lupo JV, Vanderhoof JA.
Prevention of antibiotic-associated diarrhea utilizing Lactobacillus GG.
Gastroenterology 1998;114:A435.
 D. Pereg et al. / Nutrition 27 (2011) 177181
[11] Pochapin M. The effect of probiotics on Clostridium difcile diarrhea. Am
J Gastroenterol 2000;95:S113.
[12] Pereg D, Kimhi O, Tirosh A, Orr N, Kayouf R, Lishner M. The effect of fer-
mented yogurt on the prevention of diarrhea in a healthy adult population.
Am J Infect Control 2005;33:1225.
[13] Hilton E, Kolakowski P, Singer C, Smith M. Efcacy of Lactobacillus GG as
a diarrhoeal preventive in travelers. J Travel-Med 1997;1:413.
[14] Gorbach SL. Probiotics and gastrointestinal health. Am J Gastroenterol
2000;95:S24.
[15] Liu Q, Duan ZP, Ha DK, Bengmark S, Kurtovic J, Riordan SM. Synbiotic
modulation of gut ora: effect on minimal hepatic encephalopathy in
patients with cirrhosis. Hepatology 2004;39:14419.
[16] Chiva M, Soriano G. Effect of Lactobacilli jhonsonii La1 and antioxidants on
intestinal ora and bacterial translocation in a rat model of experimental
cirrhosis. J Hepatol 2000;36:5016.
181
[17] Malaguarnera M, Greco F, Barone G, Gargante MP, Malaguarnera M,
Toscano MA. Bidobacterium longum with fructo-oligosaccharide (FOS)
treatment in minimal hepatic encephalopathy: a randomized, double-
blind, placebo-controlled study. Dig Dis Sci 2007;52:325965.
[18] Bajaj JS, Saeian K, Christensen KM, Hafeezullah M, Varma RR, Franco J.
Probiotic yogurt for the treatment of minimal hepatic encephalopathy.
Am J Gastroenterol 2008;103:170715.
[19] Weissenborn K, Ennen JC, Schomerus H, Ruckert N, Hecker H. Neuro-
psychological characterization of hepatic encephalopathy. J Hepatol
2001;34:76873.
[20] Vanderhoof JA, Whitney DB, Antonson DL, Hanner TL, Lupo JV, Young RJ.
Lactobacillus GG in the prevention of antibiotic-associated diarrhea in
children. J Pediatr 1999;135:5648.
[21] Ishibashi N, Yamazaki S. Probiotics and safety. Am J Clin Nutr
2001;73(suppl 2):465S70S.