download Full Article (PDF file) - Journal of Gastrointestinal and

ORIGINAL PAPER
Available from: URL: http://www.jgld.ro/2014/3/9.html
DOI: http://dx.doi.org/10.1543/jgld.2014.1121.233.we2
Mucosal Expression of Basic Fibroblastic Growth Factor,
Syndecan 1 and Tumour Necrosis Factor-α in Crohn’s Disease in
Deep Remission under Treatment with Anti-TNFα Antibodies
Antonio Tursi1, Walter Elisei2, Mariabeatrice Principi3, Cosimo Damiano Inchingolo4, Rosanna Nenna4, Marcello Picchio5,
Floriana Giorgio3, Enzo Ierardi3, Giovanni Brandimarte6
1) Gastroenterology Service,
ASL BAT, Andria (BT)
2) Division of
Gastroenterology, ASL Roma
H, Albano Laziale (Roma)
3) Department of Medical
Sciences and Organ
Transplantation, Section of
Gastroenterology,
University of Bari, Bari
4) Division of Pathology,
Lorenzo Bonomo Hospital,
ASL BAT, Andria (BT)
5) Division of Surgery, Paolo
Colombo Hospital, ASL Roma
H, Velletri (Roma)
6) Division of
Gastroenterology,
Cristo Re Hospital, Roma
Italy
Address for correspondence:
Dr. Antonio Tursi, MD
Servizio di Gastroenterologia
Territoriale
DSS n°4, ASL BAT
Via Torino, 49
76123 Andria (BT) – Italy
[email protected]
Received: 10.06.2014
Accepted: 24.07.2014
ABSTRACT
Background & Aims: Both inflammation and fibrosis may be detected in Crohn’s disease (CD). The molecular
pattern of Basic Fibroblastic Growth Factor (bFGF) and Syndecan-1 (SD1) expression is altered in stenosing
CD, but we do not know what the behaviour of this teamwork factor is in CD in deep remission under
treatment with anti-TNFα antibodies. Our aim was to compare the expression of bFGF, SD1 and TNF-α in
patients with CD in deep remission under treatment with Infliximab (IFX) or Adalimumab (ADA) and a
control group of patients with active CD.
Methods: We assessed the expression of bFGF, SD1 and TNF-α in 10 patients with active CD and in 28 patients
with CD in sustained deep remission for at least 6 months. All patients underwent surveillance colonoscopy
with biopsies, while receiving maintenance therapy with IFX or ADA. Analysis was conducted by real-time
reverse transcriptase PCR (RT-PCR) in biopsy samples.
Results: We found that bFGF, SD1 and TNF-α were significantly reduced under treatment with anti-TNFα
versus controls (p=0.000). bFGF and SD1 expression were similar between IFX and ADA patients (p=0.335
and p=0.289, respectively), while TNF-α was significantly under-expressed in ADA patients (p=0.008).
Conclusions: bFGF, SD1 and TNF-α are significantly reduced in CD patients in deep remission under
treatment with anti-TNFα, likely as an expression of optimal control of inflammation. The significance of the
TNF-α under-expression in patients under treatment with ADA with respect to those under treatment with
IFX should be elucidated in further studies.
Key words: adalimumab ‒ basic fibroblastic growth factor ‒ Crohn’s disease ‒ infliximab ‒ syndecan 1 ‒ tumour
necrosis factor α.
INTRODUCTION
The goal of medical treatment
in patients with Crohn’s disease
(CD) has shifted from symptom
c ont ro l a l on e t o c l i n i c a l
remission in conjunction with
mucosal healing [1]. This latter
point seems to be particularly
important, because reaching
this goal would also produce a
domino effect, resulting in fewer
complications, hospitalization,
and related surgical procedures
[2, 3]. This is particularly true
when considering the therapeutic
use of biologic therapies. The
pro-inflammatory cytokine
Tumour Necrosis Factor-alpha (TNF-α) appears to play a
pivotal role in the pathogenesis of mucosal inflammation,
mediating the inflammatory cascade in CD [3]. TNF-α is
mainly produced by monocytes and macrophages, although
many other cells of the innate and adaptive immune system
produce significant amounts of this cytokine [4]. Anti-TNFα
therapy by using Infliximab (IFX) or Adalimumab (ADA) can
lead to endoscopically assessed mucosal healing in patients
with CD, avoiding surgery and maintaining a steroid-free
remission up to 12 months [5].
Adhesion molecules containing heparan sulphate (syndecan
family) are chemically proteo-glycans and play a significant role
in tissue repair [6]. At the intestinal level, syndecan 1 (SD1) is
located in the basolateral region of the columnar epithelium
and is a relevant factor for damage reversal in inflammatory
bowel disease [7]. In particular, it has been shown that SD1 is
significantly decreased in mucosa and increased in serum of
active CD, and seems to be able to differentiate between CD and
J Gastrointestin Liver Dis, September 2014 Vol. 23 No 3: 261-265
262
Tursi et al
other granulomatous intestinal diseases, such as tuberculosis
[8]. Moreover, basic fibroblast growth factor (bFGF) is a
peptide able to repair ulcerative lesions, thanks to its capacity
to bind epithelial and stromal cells [9]. It has been suggested
that SD1 promotes bFGF morphology changes and modulates
the structure of its receptors [10]. During the physiological
restoration of mucosa in inflammatory bowel disease, SD1
migrates from basolateral epithelium to both epithelial apical
surface and stroma and modulates bFGF activity. In detail, such
a phenomenon favours the binding between bFGF and those
cells dedicated to the repair process, which are located next
to an ulcerative lesion. When it is not activated by SD1, bFGF
may be destroyed by both luminal and circulating proteases,
thus failing its role in the tissue restoration [11].
The induction of collagen secretion from fibroblasts by
bFGF may be one of the mechanisms relevant to the stromal
process disease in inflammatory bowel disease, in particular in
CD, including transmural fibrosis and stricture formation, as
well as tissue repair and healing [12]. TNF-α down-regulation
presumably interacts with the adhesion molecules such as
SD1 [3] and growth factors such as bFGF [4] in the process of
mucosal repair. Indeed, SD1 is located in the basolateral region
of the columnar epithelium and plays a relevant role in the
course of inflammatory bowel disease damage reversal [5, 13].
It has been showed that CD, especially stenosing CD, shows
impaired relation among this cytokine/adhesion molecule/
growth teamwork factor. This “pro-fibrosis” behavior could
explain the tendency toward fibrosis in CD [14]. On the
contrary, we do not know what the behaviour of this teamwork
factor is in CD in deep remission under treatment with
anti-TNFα antibodies. Moreover, we do not know whether a
different relationship among the cytokine/adhesion molecule/
growth teamwork factor in patients taking IFX from those
taking ADA is detectable.
We aimed to assess the mucosal expression of this molecular
framework in CD patients in deep remission with IFX or ADA,
compared with a control group of patients with active CD.
MATERIAL AND METHODS
This was a retrospective, observational, multicentre
study, performed on patients with CD in clinical, laboratory
and endoscopic remission and undergoing surveillance
colonoscopy while receiving maintenance therapy with IFX
or ADA.
Patients under treatment with IFX or ADA with a confirmed
history of CD and under sustained deep remission were
enrolled. All patients were under scheduled treatment with IFX
5 mg/kg/i.v. every 8 weeks or ADA 40 mg subcutaneously every
other week in order to maintain remission. No patient received
mesalamine enemas, corticosteroids, or immunosuppressants
(tacrolimus, azathioprine, and 6-mercaptopurine) during
the study period. According to our standard procedure,
immunosuppressants were started at the beginning of the IFX
and ADA treatment, in order to speed the response and to
reduce the risk of anti-TNFα failure, and stopped six months
after starting anti-TNFα therapy.
Disease activity was assessed by the Harvey-Bradshaw
Index (HBI) [15], endoscopic severity was assessed by Simple
J Gastrointestin Liver Dis, September 2014 Vol. 23 No 3: 261-265
Endoscopic Score for CD (SES-CD) [16, 17]: to be considered
in sustained deep remission for enrolment in this study,
participants had to have a HBI score ≤5 points, a SES-CD score
0, and normal values of C-reactive protein (CRP) for at least
6 months. Moreover, patients should have had no changes in
their CD medications (IFX or ADA) or any steroid use during
the previous 6 months.
Ten patients with active CD (HBI score ≥6 points and
SES-CD score ≥2 points) and not taking anti-TNFα treatment
served as control group.
Biopsy samples were taken from the intestinal location in
which the disease was active before and in remission thereafter
(active group) or in which the disease was currently active
(control group) (Table I).
Molecular analysis
Molecular analysis was independently performed by two
expert gastrointestinal pathologists (E.I. and F.G.), blinded
about the type and the degree of the disease.
The method of detecting mucosal quantitative levels of
bFGF, SD1 and TNF-α was a real-time reverse transcriptase
PCR (RT-PCR) in biopsy samples from paraffin samples.
We chose this technique because it is simpler and faster
than classical immunohistochemistry, and it significantly
correlates with semiquantitative evaluations obtained
by immunohistochemistry [18]. RT-PCR has the ability
to reflect the altered pattern of the expression of genes
dedicated to the synthesis of a specific molecule and to
quantify its transcription levels [18]. Therefore, in this study
the technique allowed the assessment of the amount of the
mRNA codifying for the synthesis of these cytokines in
intestinal biopsy samples. The amount was expressed by a
numerical value (i.e. the fold change compared to controls,
represented by patients undergoing colonic biopsy in absence
of endoscopic and histological alteration, e.g., in the case of
suspected microscopic or collagenous colitis) [19]. The relative
expression of the studied genes levels was calculated using
the 2CT method.
Statistics
The collection and analysis of data were performed using
MedCalc® Release 11.2.0.0. Chi-square test for categorical
variables and Mann-Whitney U test for continuous variables
were used. All tests were two-tailed, and the level of significance
was p=0.05.
Ethics approval
This study was approved by the Institutional Review Board
and each subject gave written informed consent.
RESULTS
A total of 28 patients received maintenance therapy and
were in deep remission with IFX (11 patients) or ADA (17
patients). The characteristics of the treated patients and
controls are summarized in Table I. The characteristics of the
CD treated groups are summarized in Table II.
When evaluating the cytokine expression, we found that
bFGF, SD1 and TNF-α were significantly reduced under
TNF-α, SD1 and bFGF expression in Crohn’s disease
263
Table I. Characteristics of the study groups
Gender Male n (%)
Treated patients
(n=28)
Controls
(n=10)
P*
17 (60.7)
6 (60.0)
1.000
Age median (range), years
34 (20-63)
35 (21-64)
0.976
Age > 40 years n (%)
12 (42.9)
4 (40.0)
1.000
CD duration median (range), months
19.5 (9-60)
18 (10-56)
0.877
Previous surgery n (%)
3 (10.7)
-
0.551
Smoking n (%)
4 (14.3)
1 (10.0)
1.000
Family history n (%)
4 (14.3)
2 (20.0)
0.644
Terminal ileum
6 (21.4)
2 (20.0)
Ileo-colon
18 (64.3)
6 (60.0)
Colon
4 (14.3)
2 (20.0)
7 (25.0)
2 (20.0)
21 (75.0)
7 (70.0)
Location n (%)
Extraintestinal diseases n (%)
0.913
1.000
Crohn’s disease type n (%)
Inflammatory
Stenosing
3 (10.7)
2 (20.0)
Fistulizing
4 (14.3)
1 (10.0)
Mesalazine
28 (100.0)
10 (100.0)
Immunosuppressors
0 (0)
0 (0.0)
Steroids
0(0)
2 (10)
0.737
Medications n (%)
0.845
*Chi-square test for categorical variables and Mann-Whitney U test for continuous
variables
treatment with anti-TNFα versus controls (p=0.000; Fig.1).
Cytokine expression under therapy with IFX and ADA is
reported in Fig. 2; bFGF and SD1 expression were similar
between IFX and ADA patients (p=0.335 and p=0.289,
Table II. Characteristics of the CD treated population
IFX-treated patients
(n=13)
ADA-treated patients
(n=15)
P*
Gender Male n (%)
8 (61.5)
9 (60.0)
1.000
Age median (range), years
34 (20-63)
32 (19-66)
0.966
Age > 40 years n (%)
5 (38.5)
7 (46.6)
1.000
CD duration median (range), months
19.5 (9-44)
18 (9-60)
0.877
Previous surgery n (%)
1 (7.7)
1 (6.6)
0.981
CRP values (in mg, n.v: <5 mg)
3 (2-5)
3 (1-5)
1.000
SES-CD score (median)
0
0
1.000
HBI score (median)
3 (2-5)
3 (2-5)
1.000
Median TNFα treatment duration, months
20 (12-44)
24 (12-60)
0.946
Smoking n (%)
1 (7.7)
3 (20.0)
0.530
Family history n (%)
2 (14.3)
2 (20.0)
0.944
Terminal ileum
3 (23.)
3 (20.0)
Ileo-colon
8 (61.5)
10 (66.6)
Colon
2 (15.3)
2 (13.3)
4 (30.7)
3 (20.0)
Inflammatory
9 (69.2)
12 (80.0)
Stenosing
2 (14.3)
1 (6.6)
Fistulizing
2 (15.3)
2 (13.3)
Location n (%)
Extraintestinal diseases n (%)
0.813
0.780
Crohn’s disease type n (%)
0.737
*Chi-square test for categorical variables and Mann-Whitney U test for continuous variables
J Gastrointestin Liver Dis, September 2014 Vol. 23 No 3: 261-265
264
Tursi et al
Fig. 1. Expression of bFGF, SD1, and TNF-alpha, and SD1 in controls
and treated patients. Median values and minimum to maximum values
are denoted by bars and error bars.
Fig. 2. Expression of bFGF, SD1, and TNF-alpha, and SD1 in the
Infliximab and Adalimumab group. Median values and minimum to
maximum values are denoted by bars and error bars.
respectively), while TNF-α was significantly under-expressed
in ADA patients (p=0.008).
DISCUSSION
The achievement of mucosal healing is a critical endpoint
in the treatment of CD patients. Achievement of mucosal
healing in patients with inflammatory bowel disease carries the
prospect of influencing the natural history of this disease by
the prevention of complications, such as the need for surgery
or hospitalization. The understanding of basic mechanisms of
wound generation and healing is crucial for the improvement
of existing and the development of future therapies.
The mechanism by which it occurs has been recently
hypothesized. Ierardi et al demonstrated that a decrease
of TNF-α, induced by IFX treatment is accompanied by a
decrease in both SD1 and bFGF when mucosal healing occurs
[20]. A possible explanation is that IFX may down-regulate,
via a marked reduction of TNF-α mucosal levels, the bFGF/
SD1 link. This molecular profile might represent a pathway of
J Gastrointestin Liver Dis, September 2014 Vol. 23 No 3: 261-265
mucosal healing. However, the parallel trend of TNF-α, SD1
and bFGF might be just a simultaneous consequence of the
control of inflammation. To clarify this last point, a further
study on cultured biopsy samples, taken from patients with
both CD and ulcerative colitis and incubated in a medium
containing comparable amounts of IFX similar to those reached
in the serum of treated patients was performed. After 24 hours,
TNF-α, SD1 and bFGF were assayed in tissue homogenates.
TNF-α was decreased, while SD1 and bFGF levels were
still high when evaluated by both a molecular method and
immunohistochemistry [21]. This last finding supports the
hypothesis that a mucosal TNF-α reset, induced by biological
drugs, is followed by a mucosal restoration in which SD1
modulates the strong reparative bFGF aptitudes.
This mechanism of mucosal restoration has been recently
confirmed by De Bruyn et al [22]. The authors found that gene
expression of many matrix metalloproteinases (MMPs), tissue
inhibitors of metalloproteinases (TIMPs), and metalloprotease
with thrombospondin motifs [ADAM(TS)s] and growth factor
was upregulated in patients with active IBD. After controlling
inflammation with IFX, most gene dysregulations observed
at baseline were restored in responders. This confirms that
suppression of inflammation results in the arrest of epithelial
damage and subsequent mucosal healing [22].
However, two issues arise from these preliminary analyses:
first, what happens when patients are in deep remission
(namely under clinical, laboratory and endoscopic remission);
second, what happens when we use ADA instead of IFX to
treat CD?
With regard to the first question, we found that the
framework TNF-α, SD1 and bFGF is significantly decreased
in CD in deep remission when compared with active disease.
In particular, all molecules are simultaneously decreased.
These results differ from what was previously described in
ulcerative colitis, in which TNF-α rapidly decreases under IFX,
followed by SD1 and bFGF [21]. The first and simple hypothesis
explaining this result is that the mechanisms involved in the
reparative process may be different in the two diseases. In fact,
we cannot forget that ulcerative colitis is a “mucosal” disease,
while CD is a “wall” disease, often causing fibrosis. Therefore,
in CD, but not in ulcerative colitis, the framework TNF-α,
SD1 and bFGF may decrease simultaneously under effect
of anti-TNFα drugs. However, a more attractive hypothesis
is that the parallel trend of TNF-α, SD1 and bFGF in CD in
deep remission might be just a consequence of the control of
inflammation. A mucosal TNF-α reset, induced by biological
drugs, is followed by a mucosal restoration in which SD1
modulates the strong reparative bFGF aptitudes. Finally, in
healed mucosa as in our population, cytokines, adhesion
molecules and growth factors resume their normal pattern.
Thus, the parallel trend of TNF-α, SD1 and bFGF in CD in
deep remission might be just a consequence of the optimal
control of inflammation.
Regarding the second question, we did not find any
difference in SD1 and bFGF expression between patients
under treatment with IFX or ADA, but TNF-α was significantly
under-expressed in ADA than in IFX patients. It is not easy
to explain these surprising results. We know that IFX heals
ulcers by the down-regulation of metallo-proteinase [23] and
TNF-α, SD1 and bFGF expression in Crohn’s disease
enhancement of tissue inhibitor of metalloproteinase [24], and
that it has distribution and elimination half-lives of 4.3 and 18.5
days, respectively [25]. On the contrary, mean concentrations
for patients receiving ADA quickly reach the steady-state, and
remain relatively constant during therapy, ranging from 15 to
19 days in relation to the dose administered [26, 27]. Indeed,
it is hypothesized that ADA inhibits TNF-α in a deeper way
thanks to its stable half-life, due to a 2-week administration.
However, there are many variables which could have biased
the results, i.e. site of biopsy sampling, time point of biopsy
sampling, TNFα antibody trough levels at the time point
of biopsy sampling, etc. Hence, this hypothesis needs to be
supported by further studies assessing the possible link between
ADA levels in blood/colonic mucosa and TNF-α expression
in intestinal mucosa.
CONCLUSION
Our results are limited by the small sample number of
CD patients enrolled to draw reliable conclusions. However,
the present study shows for the first time that bFGF, SD1
and TNF-α are significantly reduced in CD patients in deep
remission under treatment with anti-TNFα, likely to be an
expression of optimal control of inflammation. The significance
of the TNF-α under-expression in patients under treatment
with ADA with respect to those under treatment with IFX
should be elucidated in further studies.
Conflicts of interest: None to declare.
REFERENCES
1. Travis S, Van Assche G, Dignass A, Cabré E, Gassull MA. On the second
ECCO Consensus on Crohn‘s disease. J Crohns Colitis 2010;4:1-6.
2. Peyrin-Biroulet L, Ferrante M, Magro F, et al; Scientific Committee
of the European Crohn‘s and Colitis Organization. Results from the
2nd scientific workshop of the ECCO. I: Impact of mucosal healing
on the course of inflammatory bowel disease. J Crohns Colitis
2011;5:477–483.
3. Sandborn WJ, Rutgeerts P, Feagan BG, et al. Colectomy rate comparison
after treatment of ulcerative colitis with placebo or infliximab.
Gastroenterology 2009;137:1250-1260.
4. Hecht GA. Inflammatory bowel disease – live transmission. N Engl J
Med 2008;358:528-530.
5. Rutgeerts P, Van Assche G, Sandborn WJ, et al. Adalimumab induces
and maintains mucosal healing in patients with Crohn‘s disease: data
from the EXTEND trial. Gastroenterology 2012;142:1102-1111.e2.
6. Saunders S, Jalkanen M, O’Farrell S, Bernfield M. Molecular cloning
of syndecan, an integral membrane proteoglycan. J Cell Biol
1989;108:1547-1556.
7. Day R, Ilyas M, Daszak P, Talbot I, Forbes A. Expression of syndecan-1
in inflammatory bowel disease and a possible mechanism of heparin
therapy. Dig Dis Sci 1999;44:2508-2515.
8. Zhang S, Qing Q, Wang Q, et al. Syndecan-1 and heparanase: potential
markers for activity evaluation and differential diagnosis of Crohn‘s
disease. Inflamm Bowel Dis 2013;19:1025-1033.
9. Fernig DG, Gallagher JT. Fibroblast growth factors and their receptors:
an information network controlling tissue growth, morphogenesis and
repair. Prog Growth Factor Res 1994;5:353-377.
265
10. Day R, Forbes A. Heparin, cell adhesion and pathogenesis of
inflammatory bowel disease. Lancet 1999;354:62-65.
11. Beck PL, Podolsky DK. Growth factors in inflammatory bowel disease.
Inflamm Bowel Dis 1999;5:44-60.
12. Lawrance IC, Maxwell L, Doe W. Altered response of intestinal mucosal
fibroblasts to profibrogenic cytokines in inflammatory bowel disease.
Inflamm Bowel Dis 2001;7:226-236.
13. D’Haens G, Sandborn WJ, Feagan BG, et al. A review of activity indices
and efficacy end points for clinical trials of medical therapy in adults
with ulcerative colitis. Gastroenterology 2007;132:763–786.
14. Principi M, Giorgio F, Losurdo G, et al. Fibrogenesis and fibrosis in
inflammatory bowel diseases: good and bad side of same coin? World
J Gastrointest Pathopysiol 2013;4 100-107.
15. Best WR. Predicting the Crohn’s disease activity index from the HarveyBradshaw Index. Inflamm Bowel Dis 2006;12:304-310.
16. Daperno M, D’Haens G, Van Assche G, et al. Development and
validation of a new, simplified endoscopic activity score for Crohn’s
disease: the SES-CD. Gastrointest Endosc 2004;60:505-512.
17. Moskovitz DN, Daperno M, Van Assche G. Defining and validating
cut-offs for the Simple Endocopic Score for Crohn’s Disease.
Gastroenterology 2007;132:S1097.
18. Ierardi E, Giorgio F, Rosania R, et al. Mucosal assessment of tumor
necrosis factor alpha levels on paraffined samples: a comparison between
immunohistochemistry and real time polymerase chain reaction. Scand
J Gastroenterol 2010;45:1007-1008.
19. Bustin SA. Absolute quantification of mRNA using real-time reverse
transcription polymerase chain reaction assays. J Mol Endocrinol
2000;25:169–193.
20. Ierardi E, Giorgio F, Zotti M, et al. Infliximab therapy downregulation
of basic fibroblast growth factor/syndecan 1 link: a possible molecular
pathway of mucosal healing in ulcerative colitis. J Clin Pathol
2011;64:968-972.
21. Della Valle N, Giorgio F, Cantatore S, Zotti M, Ierardi E, Panella C.
Tumor Necrosis Factor α, syndecan 1 and basic fibroblast growth
factor levels and site in cultured biopsy specimens of patients with
inflammatory bowel diseases after incubation with infliximab
(Abstract). Dig Liver Dis 2013;45:S107.
22. de Bruyn M, Machiels K, Vandooren J, et al. Infliximab restores the
dysfunctional matrix remodeling protein and growth factor gene
expression in patients with inflammatory bowel disease. Inflamm Bowel
Dis 2014;20:339-352.
23. Di Sabatino A, Saarialho-Kere U, Buckley MG, et al. Stromelysin-1
and macrophage metalloelastase expression in the intestinal mucosa
of Crohn‘s disease patients treated with infliximab. Eur J Gastroenterol
Hepatol 2009;21:1049-1055.
24. Di Sabatino A, Pender SL, Jackson CL, et al. Functional modulation of
Crohn‘s disease myofibroblasts by anti-tumor necrosis factor antibodies.
Gastroenterology 2007;133:137-149.
25. Ternant D, Aubourg A, Magdelaine-Beuzelin C, et al. Infliximab
pharmacokinetics in inflammatory bowel disease patients. Ther Drug
Monit 2008;30:523-529.
26. Weisman MH, Moreland LW, Furst DE, et al. Efficacy, pharmacokinetic,
and safety assessment of adalimumab, a fully human anti-tumor
necrosis factor-alpha monoclonal antibody, in adults with rheumatoid
arthritis receiving concomitant methotrexate: a pilot study. Clin Ther
2003;25:1700-1721.
27. Kobayashi S, Harigai M, Mozaffarian N, et al. A multicenter, open-label,
efficacy, pharmacokinetic, and safety study of adalimumab in Japanese
patients with ankylosing spondylitis. Mod Rheumatol 2012;22:589-597.
J Gastrointestin Liver Dis, September 2014 Vol. 23 No 3: 261-265

Download Report