http://het.sagepub.com/Human & Experimental Toxicology

http://het.sagepub.com/content/33/1/64The online version of this article can be found at:

 DOI: 10.1177/0960327113489049

2014 33: 64 originally published online 23 May 2013Hum Exp ToxicolP Ypsilantis, M Lambropoulou, I Kourkoutas, A Pechlivanis and C Simopoulos

Liver radiofrequency ablation compromises the biological gut barrier  

Published by:

http://www.sagepublications.com

can be found at:Human & Experimental ToxicologyAdditional services and information for    

  http://het.sagepub.com/cgi/alertsEmail Alerts:

 

http://het.sagepub.com/subscriptionsSubscriptions:  

http://www.sagepub.com/journalsReprints.navReprints:  

http://www.sagepub.com/journalsPermissions.navPermissions:  

What is This? 

- May 23, 2013OnlineFirst Version of Record  

- Dec 19, 2013Version of Record >>

at Imperial College London Library on June 26, 2014het.sagepub.comDownloaded from at Imperial College London Library on June 26, 2014het.sagepub.comDownloaded from

Article

Liver radiofrequency ablationcompromises the biological gutbarrier

P Ypsilantis1, M Lambropoulou2, I Kourkoutas3,A Pechlivanis4 and C Simopoulos1

AbstractAim: Liver radiofrequency ablation (RFA) has been shown to disrupt the mechanical component of the gut bar-rier. The aim of the present study was to investigate the consequences of liver RFA on the biological gut barrier interms of the effects of bile production rate and bowel inflammatory state on intestinal microflora balance.Method: A total of 25 New Zealand rabbits were assigned to five groups (n ¼ 5 per group): group CBD: sub-jected to common bile duct (CBD) extracorporeal bypass; group CBD-RFA: subjected to CBD bypass plus onesession of open liver RFA; group RFA: subjected to liver RFA; group sham: subjected to sham operation; andgroup TBD: subjected to total bile deviation (TBD). In groups CBD and CBD-RFA, bile production rate wasassessed for 48 h. In groups sham and RFA, measurement of biliary glycine conjugates of cholic and deoxycholicacid levels, histopathologic examination of the non-ablated liver tissue, morphometric analysis, and histopatho-logic examination of the terminal ileum and microbiological analysis of fecal and tissue samples collected fromthe jejunum and the cecum (and in group TBD) were performed at 48 h post-operation. Results: One sessionof liver RFA resulted in ablation of 18.7 + 2.7% of liver weight. Following liver RFA, bile production rate wasreduced, while the levels of biliary bile salts were not affected. There was mild injury of the non-ablated liverparenchyma, mild intestinal wall inflammation, intestinal mucosa atrophy, and intestinal microbial populationovergrowth. Conclusion: Reduced in bile production and mild bowel inflammation secondary to liver RFAimpaired the biological gut barrier as manifested by intestinal microflora imbalance.

KeywordsRadiofrequency ablation, liver, bile, intestinal microflora, gut barrier

Introduction

Radiofrequency ablation (RFA) is a contemporary

method applied for the local destruction of primary and

metastatic liver tumors. Although the method is consid-

ered safe, there is a small incidence (2.4–4.6%) of post-

RFA complications of potentially microbial origin that

may pose a life-threat to treated patients. These include

hepatic and perihepatic abscesses, liver failure, perito-

neal infection, unspecified sepsis, pleural effusion,

pneumonia, acute respiratory distress syndrome, renal

failure, fever, and the post-ablation syndrome.1–4 Infec-

tious complications in other pathologic entities, such

as hemorrhagic shock, trauma, acute pancreatitis, burn

injury, and obstructive jaundice, have been related

to migration of bacteria from the intestinal lumen

through the circulation and/or the lymphatic root to

1Laboratory of Experimental Surgery and Surgical Research,School of Medicine, Democritus University of Thrace, Alexan-droupolis, Greece2Laboratory of Histology and Embryology, School of Medicine,Democritus University of Thrace, Alexandroupolis, Greece3Applied Microbiology and Molecular Biotechnology ResearchGroup, Department of Molecular Biology and Genetics, Democri-tus University of Thrace, Alexandroupolis, Greece4Laboratory of Analytical Chemistry, Department of Chemistry,Aristotle University of Thessaloniki, Thessaloniki, Greece

Corresponding author:P Ypsilantis, Laboratory of Experimental Surgery and SurgicalResearch, School of Medicine, Democritus University of Thrace,University Hospital of Alexandroupolis, Alexandroupolis 68100,Greece.Email: pipsil@med.duth.gr

Human and Experimental Toxicology2014, Vol 33(1) 64–73ª The Author(s) 2014

Reprints and permission:sagepub.co.uk/journalsPermissions.nav

DOI: 10.1177/0960327113489049het.sagepub.com

at Imperial College London Library on June 26, 2014het.sagepub.comDownloaded from

extraintestinal tissues as a result of intestinal mucosa

barrier disruption.5–9 According to recent animal stud-

ies, liver RFA at approximately 30% of rat liver par-

enchyma disrupted the gut barrier leading to bacterial

translocation,10,11 which offers a plausible explanation

for the occurrence of septic complications.

The pathogenetic mechanism of gut barrier dysfunc-

tion following liver RFA is still under investigation. The

intestinal mucosa barrier comprises of the immune, the

biological, and the mechanical components.12 The bio-

logical gut barrier is represented by a balanced intestinal

microflora, which prevents invasion of pathogenic or

opportunistic bacteria to extraintestinal tissues by

repressing their colonization and growth, enhancing dif-

ferentiation and proliferation of intestinal epithelial

cells and promoting the development of gut’s mucosal

immune system.13 Among the key factors recognized

to affect intestinal microflora balance are bile acids

present in the bile9 as well as the inflammatory state

of the bowel.14 Bile acid microbial modulatory func-

tion is achieved primarily by the inhibition of growth

of bacterial species, such as Bacteroides, Clostridia,

Lactobacillus, and Streptococci. As shown in obstruc-

tive jaundice rat models, absence of bile salts leads to

gram-negative bacteria overgrowth.9 Inflammation of

the intestinal wall provides a suitable tissue substrate

for certain bacterial population overgrowth, which leads

to commensal microbiota imbalance.14

The aim of the present project was to investigate

whether liver RFA affects the biological gut barrier

in terms of the effect of bile production rate and bowel

inflammatory state on intestinal microflora balance.

Materials and methods

Animals

A total of 25 New Zealand white rabbits, aged 4 months,

weighing 3.0–3.5 kg, which were provided by our

inbred rabbit colony, were used in the present study.

They were housed individually in stainless steel cages

under controlled environmental conditions (room tem-

perature 20–22�C, humidity 50–60%, and 12 h photo-

period). They were fed with 125 g of commercially

available pelleted diet per day per animal and tap water

ad libitum. The facilities were in accordance with Direc-

tive 86/609/EEC.

Experimental design

The rabbits were randomly assigned into five groups of

five animals in each group. In a first set of experiments,

the rabbits were subjected to common bile duct (CBD)

bypass by inserting a catheter toward the hepatic duct

(proximal catheter) and another one toward the sphinc-

ter of Oddi (distal catheter) to establish an extracorpor-

eal bile flow bypass, and then they were subjected to

one session of open liver RFA (group CBD-RFA) or

CBD bypass (group CBD). Bile was collected extra-

corporeally and its volume was measured for 48 h post-

operatively. An aliquot of the collected bile (2 mL/h for

5 h/day, between 9:00 a.m. and 2:00 p.m.) was rein-

fused to the duodenum via the distal CBD catheter to

simulate the enterohepatic circulation.

In a second set of experiments, the rabbits were sub-

jected to either one session of liver RFA (group RFA)

or sham operation (group sham) without prior cannula-

tion of their CBD. At 48 h postoperation, (a) bile was

collected by gallbladder puncture to determine biliary

bile salts concentration, (b) a tissue sample was excised

from the non-ablated liver portion for histopathologic

examination, (c) a tissue sample was excised from the

terminal ileum for morphometric analysis and histo-

pathologic evaluation, and (d) fecal and tissue samples

were collected from the jejunum and the cecum for

microbiological analysis.

An extra group of rabbits was subjected to CBD sin-

gle cannulation toward the hepatic duct to totally divert

bile extracorporeally without reinfusion of bile to the

intestine (group TBD) in order to investigate the effect

of total bile deprivation on intestinal microflora. The

distal part of the CBD toward the intestine was ligated.

After 48 h, fecal and tissue samples were collected

from the jejunum and the cecum for microbiological

analysis.

At the end of the experiment, all animals were eutha-

nized by exsanguination under general anesthesia. The

experimental protocol was approved by the Animal

Care and Use Committee of the local veterinary service

since it complied with Directive 86/609/EEC.

Animal preparation

After 24-h food and 12-h water deprivation, anesthe-

sia was induced by intramuscular injection of a xyla-

zine (5 mg/kg)—atropine (0.04 mg/kg)—ketamine

(50 mg/kg) mixture. Following endotracheal intuba-

tion, animals were connected to an anesthetic

machine to receive 30 breaths/min of 80 mL tidal

volume. Anesthesia was maintained by sevoflurane

(2% in oxygen) administration. A self-adhesive

gelled grounding pad was placed on a shaved surface

at the back of the animals. The animals were placed

Ypsilantis et al. 65

at Imperial College London Library on June 26, 2014het.sagepub.comDownloaded from

at dorsal recumbency and their abdominal wall was

clipped and prepared for aseptic surgery.

Common bile duct cannulation

A midline laparotomy was performed and the CBD was

exposed and blindly dissected approximately 2–3 cm

from the duodenal wall. After longitudinal dissection of

the CBD wall, a 5-gauge umbilical catheter (Vygon/

Ecouen, France) was inserted approximately 1 cm

toward the liver (proximal catheter) and secured using

3-0 Vicryl ligatures (groups CBD, CBD-RFA, and

TBD). A 3.5-gauge umbilical catheter was inserted

toward the sphincter of Oddi (distal catheter) and

secured (groups CBD and CBD-RFA). The free ends

of the catheters were tunneled subcutaneously and

exteriorized between the scapulae; the free end of the

proximal catheter was connected to a 250-mL sterile

plastic bag attached to the back of the animals to col-

lect the bile, while that of the distal catheter was

tapped and facilitated for bile reinfusion. A tailor-

made jacket was placed over the rabbits to protect the

catheters from being torn apart.

Radiofrequency ablation

A Radionics Cool-tip RFA System (Valleylab/Tyco

Healthcare, Gosport, UK) consisting of a radiofre-

quency generator, a peristaltic perfusion pump, a

grounding pad, and a single-shaft, 15 cm long, needle

electrode with a 2-cm exposure tip was used. After

midline laparotomy, the left lateral hepatic lobe was

exposed. The tip of the electrode was inserted into the

hepatic parenchyma from the caudal surface of the

lobe at a 90� angle. The power delivered was 60 W for

a 3-min period per session. The final tissue tempera-

ture reached between 60 and 70�C. During the RFA

session, the tip of the electrode was cooled by contin-

uous perfusion of ice-cold distilled water delivered by

the peristaltic perfusion pump. Sterile gauzes soaked

in cold normal saline were placed around the liver

lobes to prevent transmission of heat to the surround-

ing tissues. Finally, the abdominal wall was closed in

layers using 2-0 Vicryl sutures. During autopsy, both

total liver and the ablated liver portion were weighed

after being dissected from the rest of the liver. The

percentile portion of the ablated liver weight was cal-

culated with the following equation

ablated liver weight

total liver weight� 100

Determination of bile salts concentration

Bile samples were kept at�80�C until analyzed. Cholic

(CA) and deoxycholic acids (DCA) were determined in

the form of glycine conjugates (glycine cholic acid

(GCA) and glycine deoxycholic acid (GDCA), respec-

tively) on a ultrahigh-performance liquid chromato-

graphy coupled with quadrupole time-of-flight–mass

spectrometry (UPLC-qTOF-MS) system (Waters,Mil-

ford, Massachusetts, USA). The system, comprising

of an acquity UPLC and a qTOF Ultima MS system

(Waters, Milford, Massachusetts, USA), operated on a

MassLynx platform in negative electrospray ionization

with the following parameters: capillary: 2.20 kV; cone:

25 V; source temperature: 120�C; desolvation tempe-

rature: 350�C; desolvation gas flow: 800 L/h; TOF:

9.10 kV; Micro-channel plate (MCP): 1950. Chromato-

graphic separations were performed in an acquity

UPLC high strength silica T3 column using a binary

gradient solvent system consisting of 0.1% (volume per

volume; v/v) formic acid in high-performance liquid

chromatography grade water (solvent A) and 0.1% (v/

v) formic acid in acetonitrile (solvent B) with the fol-

lowing program: 30% solvent A constant for 1 min and

then linear increase to 100% solvent B within 15 min;

isocratic 100% solvent B for 1 min and back to initial

condition of 30% solvent B, where the system was held

isocratic for 4 min for column equilibration prior to the

subsequent injection. The flow rate was 0.4 mL/min

and the injection volume was 5 mL. GCA and GDCA

gave characteristic ions at 464.27 m/z and 448.27 m/z,

respectively. To enable the quantification, GCA and

GDCA injections of reference standards were used to

construct linear calibration curves in the range of

100 ng/mL–100 mg/mL. Unknown samples were ana-

lyzed thrice and their average was used for quantifica-

tion. Samples were diluted at the appropriate ratio

(500- to 4000-fold) in order to bring their concentration

within the linear dynamic range.

Histopathology—histomorphometric analysis

Tissue specimens excised from the non-ablated liver

portion (right median lobe) and the terminal ileum were

fixed in formalin and embedded in paraffin according

to standard procedures. Histopathologic examination

was performed at 4-mm hematoxylin–eosin stained sec-

tions. The endpoints evaluated for (a) liver tissue were

hyperemia/distension of sinusoidal space, hepatocellu-

lar degeneration/steatosis, distention of bile ducts, por-

tal infiltration, and necrosis and (b) ileal tissue were

neutrophil granulocyte, lymphocyte, and plasma cell

66 Human and Experimental Toxicology 33(1)

at Imperial College London Library on June 26, 2014het.sagepub.comDownloaded from

infiltration, edema, hyperemia/vascular dilatation, and

hyperplasia. The severity of lesions was quantified

according to the following scoring system: 0: none;

1: mild; 2: moderate; 3: severe. Lesion severity scores

were added to obtain the histopathologic score.

Intestinal mucosal morphometric characteristics

were assessed by the measurement of villous height and

density. Villous height was measured in 20 well-

preserved villi per sample using the Nikon Digital

Sight DC-L1 software (Nikon Eclipse 50i microscope,

Kawasaki, Japan) and expressed as average villous

height. Villous density was measured in 10 low-power

optical fields (10�) per sample, and their average was

expressed as number of villi per optical field. All

examinations were performed in a blinded fashion.

Microbiological analysis

Fecal samples collected aseptically from the jejunum

and the cecum were homogenized in sterile buffered

peptone water. Tissue samples, 3 cm long, were also

aseptically excised from the jejunum and the cecum.

Tissue samples were washed twice with sterile buf-

fered peptone water and vortex-mixed to break down

bacterial clumps and remove loosely attached bacteria

and then homogenized in 5 mL sterilized buffered pep-

tone water using a tissue grinder. Fecal and tissue sam-

ples were subjected to serial dilutions. The following

tests for microbiological analysis were performed: (i)

total aerobic counts in plate count agar (Fluka, Buchs,

Switzerland) at 30�C for 48 h, (ii) staphylococci in

Baird Parker egg yolk tellurite medium (Fluka) at

37�C for 48 h and confirmed by a positive coagulase

test, (iii) coliforms in violet–red bile agar (Fluka) after

incubation at 30�C for 24 h, (iv) enterobacteria in vio-

let–red bile glucose agar (Fluka) at 37�C for 24 h, (v)

streptococci in bile esculin azide agar after incubation

at 37�C for 24 h (Fluka), (vi) lactobacilli (gram (þ),

catalase (–)) in acidified MRS agar (MRS: de Man,

Rogosa, and Sharpe, named after the inventors; Fluka)

at 37�C for 48 h anaerobically (Anaerobic Jar, Anero-

cult C, Merck, Germany) and confirmed by Gram

staining and catalase test, (vii) lactococci in M17 agar

(Fluka) at 30�C for 24 h, and (viii) yeasts and molds in

malt agar (Fluka; pH was adjusted to 4.5 by sterile

solution of 10% lactic acid) at 30 C for 48 h. All incu-

bations were further extended up to 120 h, but no extra

colonies were observed. Results were presented as

logarithm of mean colony-forming units on solid media

culture plates containing between 30 and 300 colonies/

g of fecal sample or tissue.

Statistical analysis

Data were expressed as mean + SD. After normality

of data was tested with the Kolmogorov–Smirnov

test, these were subjected to analysis of variance. The

Bonferoni test was used for multiple comparisons

among groups and the Student’s t test for comparisons

between pairs of groups. A p < 0.05 was considered

statistically significant.

Results

One session of liver RFA resulted in the ablation of

18.7 + 2.7% (range 15.4–22.7%) of total liver weight.

Bile production rate

Bile volume produced per day was corrected for body

weight by adjusting the daily volume of bile to a standard

body weight (4 kg). Animals subjected to liver RFA

produced significantly less bile per day (p < 0.01) at

24 and 48 h compared with CBD bypassed animals

(Figure 1).

Figure 1. Daily rate (mean, n ¼ 5) of bile production fol-lowing CBD bypass (group CBD) or CBD bypass plus liverRFA (group CBD-RFA). Error bars represent SD. *p < 0.05versus CBD, **p < 0.01 versus CBD. CBD: common bileduct; RFA: radiofrequency ablation.

Table 1. Bile salts (GCA and GDCA) concentration (means+ SD, n ¼ 5) following sham operation (group sham) orliver RFA (group RFA).

Group GCA (mg/mL) GDCA (mg/mL)

Sham 2122.1 + 349.7 11282.7 + 6556.3RFA 2011.3 + 955.5 13130.5 + 4316.1

GCA: glycine cholic acid; GDCA: glycine deoxycholic acid; RFA:radiofrequency ablation.

Ypsilantis et al. 67

at Imperial College London Library on June 26, 2014het.sagepub.comDownloaded from

Bile salt concentration

No differences in the concentration of biliary GCA

and GDCA were found between groups RFA and

sham (Table 1).

Liver histopathology

Liver RFA induced tissue injury to the non-ablated

liver parenchyma. Histopathologic lesions were mild

and included hyperemia/distension of the sinusoidal

space, distention of bile ducts, portal infiltration, hepa-

tocellular degeneration/steatosis and necrosis (Table 2;

Figure 2). The histopathologic score in group RFA was

significantly increased (p < 0.01) compared with that

of the sham-operated animals (Table 2).

Intestinal histopathology

Histopathologic lesions of mild severity characteriz-

ing the inflammatory state were noted in all layers

of the intestinal wall of rabbits subjected to liver RFA

(Figure 3). Overall, the histopathologic score in group

RFA was significantly higher (p < 0.001) than that in

group sham (Table 3).

Intestinal mucosa morphometry

Both intestinal villous height and density were signif-

icantly decreased in the rabbits subjected to liver RFA

(Table 4).

Intestinal microflora

In the jejunal content, the populations of Enterobac-

teria, Lactobacilli, and Lactococci were increased in

group RFA, of yeasts and molds in group TBD, while

of Coliforms in both groups TBD and RFA. In the cecal

tissue and content, total aerobic counts, Enterobac-

teria, Coliforms, Lactobacilli, Lactococci, yeasts, and

molds were increased in both groups TBD and RFA

(Figure 4). No change was noted in the populations

of Streptococci and Staphylococci (data not shown).

Table 2. Severity of histopathologic lesions (means + SD, n¼ 5) in the non-ablated liver tissue following sham operation(group sham) or liver RFA (group RFA).

Type of lesion

Group HSS HD/S DBD PI N Histopathologic score

Sham 0.5 + 0.0 0.3 + 0.4 0.0 + 0.0 0.0 + 0.0 0.0 + 0.0 0.8 + 0.4RFA 1.2 + 0.4 0.5 + 0.5 1.0 + 0.0 1.0 + 0.0 0.6 + 0.4 4.3 + 1.0a

HSS: hyperemia/distention of sinusoidal space; HD/S: hepatocellular degeneration/steatosis; DBD: distention of biliary ducts; PI: portalinfiltration; N: necrosis; RFA: radiofrequency ablation.ap < 0.01 versus group sham.

Figure 2. Non-ablated liver tissue section. (a) Sham operation, normal liver parenchyma; (b) Liver RFA, focal steatosis(arrow 1), distension of bile ducts and portal infiltration (arrow 2), and distension of blood vessels and parenchymal infil-tration (arrow 3; hematoxylin and eosin, �100.) RFA: radiofrequency ablation.

68 Human and Experimental Toxicology 33(1)

at Imperial College London Library on June 26, 2014het.sagepub.comDownloaded from

Discussion

The present experimental study contributes to the

understanding of the pathogenetic mechanism of gut

barrier dysfunction following liver RFA. It provides

evidence supporting the view that reduced bile pro-

duction and mild bowel inflammation secondary to

liver RFA alter intestinal microflora and therefore dis-

turb the biological gut barrier.

Figure 3. Ileal tissue section. (a) Sham operation, normal ileal tissue; (b) Liver RFA, stroma hyperemia (arrow 1), intestinalcrypt infiltration (arrow 2), mucosa infiltration (arrow 3; hematoxylin and eosin, �40.) RFA: radiofrequency ablation.

Table 3. Severity of histopathologic lesions (means + SD, n¼ 6) of ileal tissue following sham operation (group sham) orliver RFA (group RFA).

Histopathologic grade

Intestinal layer Lesions Sham RFA

Mucosa Neutrophil granulocyte infiltration 0.2 + 0.4 0.9 + 0.2a

Lymphocyte/plasma cell infiltration 0.3 + 0.5 1.0 + 0.0b

Intestinal crypts Hyperplasia 0.0 + 0.0 0.0 + 0.0Neutrophil granulocyte infiltration 0.0 + 0.0 0.8 + 0.4b

Lymphocyte/plasma cell infiltration 0.5 + 0.5 1.0 + 0.0Stroma Edema 0.0 + 0.0 0.0 + 0.0

Hyperemia 0.7 + 0.4 1.0 + 0.0Neutrophil granulocyte infiltration 0.0 + 0.0 0.9 + 0.2a

Lymphocyte/plasma cell infiltration 0.3 + 0.5 1.0 + 0.0b

Submucosa Neutrophil granulocyte infiltration 0.0 + 0.0 1.0 + 0.0a

Lymphocyte/plasma cell infiltration 0.0 + 0.0 1.0 + 0.0a

Hyperemia/vascular dilatation 0.7 + 0.3 1.0 + 0.0b

Serosa Hyperemia/vascular dilatation 0.7 + 0.4 1.0 + 0.0Histopathologic score 3.3 + 1.4 10.6 + 0.7c

RFA: radiofrequency ablation; SD: standard deviation.ap < 0.01 versus sham group.bp < 0.05 versus sham group.cp < 0.001 versus sham group.

Ypsilantis et al. 69

at Imperial College London Library on June 26, 2014het.sagepub.comDownloaded from

As previously demonstrated, RFA at approximately

30% of liver mass in the rat resulted in the disruption of

the mechanical component of the gut barrier leading to

translocation of intraluminal bacteria and endotoxins to

proximal and distal to the intestine organs. Atrophy of

the intestinal mucosa was attributed to increased crypt

cell apoptosis related to induction of oxidative stress.10

In the present liver RFA rabbit model, we investigated

the potential relation of altered bile production with

dysfunction of the biological gut barrier. The daily rate

of bile production was decreased in rabbits subjected to

RFA at approximately 15% of their liver mass. This

effect was anticipated since (a) a functional liver vol-

ume, equivalent to that necrotized by ablation, was

deprived and (b) the non-ablated liver parenchyma,

which undertook bile production, was injured as evi-

denced histologically in the current study confirming

previous experimental work.15–17

DCA is the main bile acid (90%) in the rabbit, fol-

lowed by the CA (8%).18 It has been estimated that the

bile acid pool in the rabbit depletes in 5 days when

bile flow to the duodenum is deviated.19 In light of the

impaired enterohepatic circulation of bile acids due to

extracorporeal diversion of bile flow and only partial

reinfusion of bile to the duodenum, bile salt levels

were determined in bile samples collected directly

from the gallbladder of animals with intact CBD. The

concentrations of biliary deoxycholic and cholic salts

were not affected. However, given the reduction in

bile production rate, the quantity of bile acids ulti-

mately reaching the duodenum was decreased.

A growing body of evidence highlights the impor-

tant role of bile acids in preserving a balance among

intestinal bacterial populations. While bile acids inhibit

the growth of gram-positive bacteria, they have little

effect on gram-negative ones.20 Therefore, an over-

growth of gram-negative bacteria, in the absence of

bile acids, is regarded as a consequence of intestinal

microbiota imbalance. Gram-negative bacteria have

variable endotoxic activity due to lipopolysaccharide

contained in their outer cell membrane. Bile salts also

bind directly to intraluminal endotoxins and bacteria,

to form poorly absorbed detergent-like complexes, pre-

venting them from being translocated through the gut

barrier.9

In vivo studies on the effect of bile acid deprivation

on intestinal microflora typically use obstructive jaun-

dice rat models induced by CBD ligation21,22 with

inevitable regression of bile to the liver and concurrent

pancreatitis due to pancreatic ducts obstruction. We

preferred bile flow extracorporeal deviation to avoid

complications derived from CBD ligation. In addition,

by determining microbial growth not only to fecal sam-

ples but also to intestinal tissue, the true number of

microbia adhered to the intestinal mucosa was evalu-

ated. Bile deprivation altered intestinal microflora as

manifested by overgrowth of total aerobic counts,

Enterobacteria, Coliforms, Lactobacilli, Lactococci,

yeasts and molds, especially in the cecum. The same

microbial populations were increased in those animals

subjected to liver RFA, recognizing thus the reduction

in bile production as a determining factor for the

changes noted in the intestinal microflora.

Small intestine bacterial overgrowth (SIBO) was

noted in the liver RFA group as evidenced by an

increase in the population of Enterobacteria, Coli-

forms, Lactobacilli, and Lactococci in the jejunal con-

tent. Bacterial overgrowth in the small bowel may be

a consequence of impaired intestinal motility. Attenua-

tion or interruption of sweeping contractions that

propagate from the stomach or the duodenum to the

terminal ileum allow proliferation of distal gut bacter-

ial populations into the small intestine.23 As shown in

animal studies, cutaneous burn injury leads to impaired

small bowel transit.24 This could offer a plausible

explanation for the liver RFA-induced SIBO, since

ablation of the liver results in severe burn trauma of the

hepatic parenchyma.

Injury of the gut wall characterized by mild inflam-

mation of all intestinal layers was noted in the rabbits at

48 h post liver RFA. Previous experimental work in

rats also reported the presence of inflammatory cells

confined mainly in the intestinal mucosa up to 48 h

after liver RFA, further expanding to the submucosa

at 72 h.10 It has been suggested that the composition

of gut microbial populations is related to the physiolo-

gical state of the intestine, although little is known

about the effect of inflammatory mediators of the

intestinal epithelium, such as pro-inflammatory cyto-

kines and chemokines, on intraluminal bacterial bal-

ance.23 Inflammation of the intestinal mucosa has

Table 4. Intestinal mucosa morphometric characteristics(means + SD, n ¼ 5) following sham operation (groupsham) or liver RFA (group RFA).

Group Villous height Villous density

Sham 681.9 + 32.7 6.5 + 0.5RFA 588.9 + 93.0a 5.6 + 0.3b

RFA: radiofrequency ablation.ap < 0.05 versus sham group.bp < 0.01 versus sham group.

70 Human and Experimental Toxicology 33(1)

at Imperial College London Library on June 26, 2014het.sagepub.comDownloaded from

been shown to promote overgrowth of certain bacteria

of the Enterobacteria family whose fitness is enhanced

in the inflamed gastrointestinal tract.14 Conversely, an

imbalance of commensal microbiota has been proven

to activate the immune response of the host and dete-

riorate the intestinal inflammatory state through a

decrease in beneficial bacterial metabolic end products

and a concurrent increase in detrimental bacterial toxic

products. This host–intestinal microbiota interaction

has been well illustrated in chronic inflammatory dis-

eases, such as the inflammatory bowel disease and the

irritable bowel syndrome.23

Intestinal mucosa atrophy, suggesting dysfunction

of the mechanical gut barrier, was noted in the liver

RFA group. Although the pathogenetic mechanism of

this phenomenon is multiparametric, reduced bile

production and disturbed intestinal microflora, docu-

mented by the present study, have been shown to

Figure 4. Intestinal microflora (means, n ¼ 5) of animals subjected to sham operation (group sham), total bile deviation(group TBD), or liver RFA (group RFA). Error bars represent SD. Jejunal-T, jejunal tissue; Jejunal-C, jejunal content;Cecum-T, cecal tissue; Cecum-C, cecal content. a,b,cFor each intestinal site, means with different letters differ significantly(p < 0.05). RFA: radiofrequency ablation.

Ypsilantis et al. 71

at Imperial College London Library on June 26, 2014het.sagepub.comDownloaded from

contribute toward this direction. Bile exerts a trophic

effect on the intestinal mucosa by promoting intestinal

epithelial cell proliferation and is also important for the

maintenance of the integrity of enterocyte tight junc-

tions.9 Short-chain fatty acids produced by intraluminal

bacteria stimulate differentiation and proliferation of

epithelial cells in the small and large intestine.13

In conclusion, the findings of the present study shed

some light to the pathogenetic mechanism responsible

for the dysfunction of the gut barrier following liver

RFA. A decreased bile production rate and mild in-

flammation of the bowel impaired the biological gut

barrier as manifested by alteration of the intestinal

microflora. Malfunction of the gut barrier could justify

the occurrence of infectious complications related to

this tumor ablating method that narrow the safety limits

of the procedure. Based on these results, future studies

may focus on designing effective preventive or thera-

peutic strategies involving bile acid supplementation,

anti-inflammatory drug administration, or intestinal

microbiota balance modulators.

Acknowledgments

The authors wish to thank Drs Eirini Papageorgiou and Con-

stantinos Garoufas and Mr Ioannis Maragos for their help in

conducting the experiments; Mrs Marianthi Sidira for help-

ing with the microbiological analysis; and Associate Profes-

sor Georgios Theodoridis for helping with bile salt analysis.

Conflicting Interests

The authors declared no conflicts of interest.

Funding

This research received no specific grant from any funding

agency in the public, commercial, or not-for-profit sectors.

References

1. Mulier S, Mulier P, Ni Y, Dupas B, Marchal G, De

Wever I, et al. Complications of radiofrequency coagu-

lation of liver tumours. Br J Surg 2002; 89: 1206–1222.

2. Curley SA, Marra P, Beaty K, Ellis LM, Vauthey JN,

Abdalla EK, et al. Early and late complications after

radiofrequency ablation of malignant liver tumors in

608 patients. Ann Surg 2004; 239: 450–458.

3. Dodd GD III, Napier D, Schoolfield JD, and Hubbard

L. Percutaneous radiofrequency ablation of hepatic

tumors: postablation syndrome. Am J Roentgenol

2005; 185: 51–57.

4. Jansen MC, van Duijnhoven FH, van Hillegersberg R,

Rijken A, van Coevorden F, van der Sijp J, et al.

Adverse effects of radiofrequency ablation of liver

tumours in the Netherlands. Br J Surg 2005; 92:

1248–1254.

5. Deitch EA, Bridges W, Berg R, Specian RD and

Granger DN. Hemorrhagic shock-induced bacterial

translocation: the role of neutrophils and hydroxyl

radicals. J Trauma 1990; 30: 942–951.

6. Peitzman AB, Udekwu AO, Ochoa J and Smith S. Bac-

terial translocation in trauma patients. J Trauma 1991;

31: 1083–1086.

7. Tokyay R, Zeigler ST, Traber DL, Stothert JC Jr, Loick

HM and Heggers JP. Postburn gastrointestinal vaso-

constriction increases bacterial and endotoxin translo-

cation. J Appl Physiol 1993; 74: 1521–1527.

8. Andersson R and Wang XD. Gut barrier dysfunction in

experimental acute pancreatitis. Ann Acad Med Singa-

pore 1999; 28: 141–146.

9. Assimakopoulos S, Scopa C and Vagianos C. Patho-

physiology of increased intestinal permeability in

obstructive jaundice. World J Gastroenterol 2007; 13:

6458–6464.

10. Ypsilantis P, Panopoulou M, Lambropoulou M, Tsigalou

C, Pitiakoudis M, Tentes I, et al. Bacterial translocation

in a rat model of large volume hepatic radiofrequency

ablation. J Surg Res 2010; 161: 250–258.

11. Ypsilantis P, Lambropoulou M, Grapsa A, Tentes I,

Tsigalou C, Panopoulou M, et al. Pringle maneuver

deteriorates gut barrier dysfunction induced by

extended-liver radiofrequency ablation. Digest Dis Sci

2011; 56: 1548–1556.

12. Balzan S, de Almeida Quadros C, de Cleva R, Zilberstein

B and Cecconello I. Bacterial translocation: overview of

mechanisms and clinical impact. Gastroenterol Hepatol

2007; 22: 464–471.

13. Guarner F and Malagelada JR. Gut flora in health and

disease. Lancet 2003; 361: 512–519.

14. Sekirov I, Russell SL, Antunes LC and Finlay BB. Gut

microbiota in health and disease. Physiol Rev 2010; 90:

859–904.

15. Ypsilantis P, Pitiakoudis M, Souftas VD, Lambropou-

lou M, Tsalikidis C, Foutzitzi S, et al. Liver regenera-

tion following radiofrequency ablation. J Surg Res

2008; 150: 60–65.

16. Ypsilantis P, Lambropoulou M, Anagnostopoulos C,

Tentes I, Tsigalou C, Pitiakoudis M, et al. Mesna pre-

serves hepatocyte regenerating capacity following liver

radiofrequency ablation under Pringle maneuver. J Surg

Res 2011; 169: 44–50.

17. Ypsilantis P, Lambropoulou M, Anagnostopoulos C,

Tsigalou C, Vasiliadis C, Kortsaris A, et al. Pringle

maneuver exacerbates systemic inflammatory response

and multiple organ injury induced by extended liver

72 Human and Experimental Toxicology 33(1)

at Imperial College London Library on June 26, 2014het.sagepub.comDownloaded from

radiofrequency ablation. Hum Exp Toxicol 2011; 30:

1855–1864.

18. Kasbo J, Saleem M, Perwaiz S, Mignault D, Lamireau

T, Tuchweber B, et al. Biliary, fecal and plasma deoxy-

cholic acid in rabbit, hamster, guinea pig, and rat: com-

parative study and implication in colon cancer. Biol

Pharm Bull 2002; 25: 1381–1384.

19. Nguyen LB, Xu G, Shefer S, Tint GS, Batta A, Salen

G, et al. Comparative regulation of hepatic sterol

27-hydroxylase and cholesterol 7alpha-hydroxylase

activities in the rat, guinea pig, and rabbit: effects of

cholesterol and bile acids. Metabolism 1999; 48:

1542–1548.

20. Floch MH. Bile salts, intestinal microflora and enterohe-

patic circulation. Digest Liver Dis 2002; 34(Suppl 2):

S54–S57.

21. Parks RW, Clements WD, Pope C, Halliday MI, Row-

lands BJ and Diamond T. Bacterial translocation and gut

microflora in obstructive jaundice. J Anat 1996; 189:

561–565.

22. Ogata Y, Nishi M, Nakayama H, Kuwahara T, Ohnishi

Y and Tashiro S. Role of bile in intestinal barrier func-

tion and its inhibitory effect on bacterial translocation

in obstructive jaundice in rats. J Surg Res 2003; 115:

18–23.

23. Othman M, Aguero R and Lin HC. Alterations in

intestinal microbial flora and human disease. Curr

Opin Gastroenterol 2008; 24: 11–16.

24. Oliveira HM, Sallam HS, Espana-Tenorio J, Chinkes D,

Chung DH, Chen JD, et al. Gastric and small bowel ileus

after severe burn in rats: the effect of cyclooxygenase-2

inhibitors. Burns 2009; 35: 1180–1184.

Ypsilantis et al. 73

at Imperial College London Library on June 26, 2014het.sagepub.comDownloaded from