Vol. 199, No.

3S, September 2004 Alimentary Tract III S21

hrs. (p⬍0.005). Tail vein administration of a Ski siRNA expression
vector into nude mice with flank tumors resulted in a nearly 2-fold
decrease in growth at 3 weeks compared to administration of control
vector.
CONCLUSIONS: Human pancreatic cancers overexpress Ski. Inhi-
bition of Ski through RNA interference restored TGF-B signaling
and growth inhibition in vitro and decreased tumor growth in vivo.
These findings support a role for Ski as a suppressor of TGF-B sig-
naling in pancreatic cancer.
Mast cell function is essential to intestinal
anastomosis healing
John B Conneely MD, Desmond Winter MD, Joseph Dowdall MD,
David Bouchier-Hayes MA, BCH, FACS
Royal College of Surgeons in Ireland, Beaumont Hospital
Dublin 9, Ireland
INTRODUCTION: Despite meticulous surgical technique, impaired
intestinal healing may lead to anastomotic leakage with attendant
morbidity and mortality risks. A key event in tissue healing is the
degranulation of mast cells with liberation of cytokinetically active
products. We hypothesised that Mast Cell degranulation is essential
to healing of intestinal anastomoses and reduces leak rates.
tial for anastomotic healing. Manipulation of mast cell function may
represent a beneficial therapy for surgical patients.
A substance P receptor antagonist (SPRA) that
reduces intraabdominal adhesion formation
decreases peritoneal matrix metalloproteinase
(MMP) activity
Philip A Cohen MD, Karen L Reed PhD, Adam C Gower MS,
Arthur F Stucchi PhD, Jill Lehrmann MD, A Brent Fruin MD,
Susan E Leeman PhD, James M Becker MD, FACS
Boston University Medical Center
Boston, MA
INTRODUCTION: Intraabdominal adhesions remain a significant
complication of abdominal surgery. Previous data from our labora-
tory demonstrate that administration of a SPRA (CJ-12,255, Pfizer)
reduces adhesion formation. There is a growing body of evidence
suggesting that dysregulation of extracellular matrix (ECM) degra-
dation and remodeling by MMPs is a key element of adhesion for-
mation. The aim of this study was to utilize the SPRA to determine if
the proinflammatory peptide substance P (SP) modulates peritoneal
MMP activity.

METHODS: 30 male Sprague-Dawley rats were randomised into METHODS: Following laparotomy, 6 ischemic buttons were cre-
three groups (n ⫽ 10 each). Group A were treated with mast cell ated on the peritoneum to induce adhesions in 46 rats. These rats
stabiliser Ketotifen (1mg/kg) and group B treated with mast cell were randomized to receive either intraperitoneal injections of a
degranulator Compound 48/80 (1mg/Kg) intravenously 24 hours SPRA (5 mg/kg) or sterile water. Peritoneal fluid (PF) and peri-
before surgery. Group C were controls, receiving isovolumetric ve- toneal adhesion tissue (PT) was collected at 24 hours and 7 days
hicular injection. Animals underwent right colonic anastomosis and and compared with 27 non-operated controls. Total MMP activ-
were sacrificed at 4 days. Anastomoses were excised, examined by ity was determined by a quenched fluorescent substrate assay
light and electron microscopy, for radiological contrast leakage, anas- while the activity of MMP-2 and MMP-9 was assessed by gelatin
tomotic bursting pressure and Hydroxyproline content. Statistical zymography.
analysis of continuous and discrete variables was performed using
ANOVA and Chi-squared tests. Results are expressed as mean, P RESULTS: SPRA administration resulted in significant reduc-
value ⬍ 0.05 considered significant. tions in total MMP activity at 24 hours and 7 days in PF and PT
compared with both non-operated controls and operated, un-
RESULTS: Mast cells were identified in abundance at the site of treated animals. In contrast, SPRA administration was ineffective
anastomoses by light and electron microscopy. Prevention of mast in reducing the observed increase in MMP-2 and MMP-9 activ-
cell degranulation (group A) impaired healing compared with con- ities in PT and PF following laparotomy.
trols (see table). Mast cell degranulation (group B) significantly en-
hanced healing compared with group A (P ⬍ 0.01).
24 hours 7 days
Peritoneal % of Non-operated control
Bursting Hydroxyproline
Contrast pressure content Assay sample ⴚSPRA ⴙSPRA ⴚSPRA ⴙSPRA
Groups leak (%) (mmHg) (micrograms/mg) Total
MMP PF 101 ⫾ 12 66 ⫾ 5*† 122 ⫾ 7 34 ⫾ 11*†
Group A (ketotifen) 75 2.5 4.1
PT 171 ⫾ 54 149 ⫾ 11 409 ⫾ 39* 273 ⫾ 50*†
Group B (compound
48/80) 25 23.5 5.3 MMP-2 PF 176 ⫾ 11* 221 ⫾ 17* 198 ⫾ 94 268 ⫾ 78*
PT Not detected Not detected 194 ⫾ 8* 205 ⫾ 15*
Group C (controls) 25 25.7 5.1
MMP-9 PF 305 ⫾ 43* 308 ⫾ 29* 116 ⫾ 4* 132 ⫾ 20*
P Value ⬍0.01 ⬍0.01 ⬍0.01
PT 234 ⫾ 26* 266 ⫾ 36* 162 ⫾ 10 143 ⫾ 20
Data expressed as mean ⫾ SEM.
CONCLUSIONS: Mast cells regulate tissue repair by releasing cyto- *p ⬍ 0.05 compared to non-operated controls.
kinetic products. Prevention of degranulation significantly impairs †
p ⬍ 0.05 compared to ⫺SPRA animals.
healing compared with controls. Normal mast cell function is essen- PF ⫽ peritoneal fluid; PT ⫽ peritoneal tissue.
S22 Alimentary Tract III J Am Coll Surg

CONCLUSIONS: By reducing the activation of MMPs, the SPRA
may prevent early peritoneal ECM remodeling, thus reducing adhe-
sion formation. These results suggest a prophylactic role for SPRA in
adhesion prevention.
Effects of long term exposure to lipopolysaccharide
on gastric injury and iNOS production
Emily K Robinson MD, James Suliburk MD,
David Mercer MD, FACS
The University of Texas Health Science Center at Houston
Houston, TX
INTRODUCTION: Short-term (⬍ 5 hrs) administration of lipo-
polysaccharide (LPS) does not cause macroscopic gastric injury,
but does render the stomach more susceptible to luminal insults,
an effect attenuated by inducible nitric oxide synthase (iNOS)
inhibition. However, the effect of prolonged exposure of LPS on
gastric mucosal integrity is unknown as is the role of iNOS. Thus,
this study was done to determine whether LPS alters gastric mu-
cosal integrity or iNOS expression when given for a longer
duration.
METHODS: Rats were given saline or the highly selective iNOS
inhibitor 1400W(1mg/kg IP) followed 15 minutes later by saline or
LPS (20mg/kg IP). After 24 hrs, rats were sacrificed, gastric fluid
aspirated, macroscopic injury (computerized planimetry) measured
and mucosa harvested for iNOS assessment (Western blot) (N ⬎4
group; ANOVA).
RESULTS: As shown in Table 1, LPS caused macroscopic gastric
injury, increased iNOS, and increased accumulation of alkaline fluid
within the gastric lumen compared to saline controls. 1400W inhib-
ited LPS induced macroscopic gastric injury and prevented gastric
luminal fluid alkalinization but had no effect on gastric luminal fluid
accumulation. LPS induced upregulation of gastric iNOS produc-
tion was not inhibited by 1400W.
Table 1. Effects of LPS with and without selective iNOS
inhibition on the stomach.
1400W/ 1400W/
Saline LPS saline LPS
Gastric injury
(mm2) ⫾ SE 0.125 ⫾ 0.7 7.85 ⫾ 2.1* 0.25 ⫾ 0.7 2.8 ⫾ 0.7**
iNOS
(densitometric
units) ⫾ SE 0.24 ⫾ 0.01 0.68 ⫾ 0.09* 0.22 ⫾ 0.02 0.67 ⫾ 0.07*
Gastric pH ⫾ SE 2 ⫾ 0.1 7.3 ⫾ 0.3* 2 ⫾ 0.1 2.9 ⫾ 0.5**
Gastric volume
(ml) ⫾ SE 0.25 ⫾ 0.04 1.5 ⫾ 0.3* 0.37 ⫾ 0.04 0.9 ⫾ 0.2
*p ⱕ 0.05 vs saline (ANOVA).
**p ⱕ 0.05 vs LPS (ANOVA).
CONCLUSIONS: This is the first study to report macroscopic gas-
tric injury occurs after LPS administration alone. Moreover, the ef-
fects of LPS on gastric mucosal integrity and iNOS production per-
sist even after 24 hrs. These data further suggest that septic patients
may be at risk for gastritis due to upregulation of iNOS.
Bile-pancreatic juice (BPJ) exclusion exacerbates
Akt/NF-kB pathway activation and increases
chemokine production in ligation-induced acute
pancreatitis
Isaac Samuel MBBS, FACS, Smita Zaheer PhD, Mark Yorek PhD,
Asgar Zaheer PhD
University of Iowa Carver College of Medicine & VAMC
Iowa City, IA
INTRODUCTION: Using a unique surgical model (The Donor Rat
Model) we showed that duodenal replacement of BPJ, obtained fresh
from a Donor Rat, ameliorates pancreatic morphologic changes, hy-
peramylasemia and hypercholecystokininemia in ligation-induced
acute pancreatitis. We hypothesize that BPJ exclusion from gut exac-
erbates Akt/NF-kB pathway activation and induces proinflamma-
tory chemokine production in ligation-induced acute pancreatitis.
The cytosolic IkB/NF-kB complex dissociates when IkB is phos-
phorylated following Akt activation, allowing NF-kB to translocate
to the nucleus and induce transcription of various inflammatory
mediators including chemokines.
METHODS: We studied rats as follows: Diseased-controls had bile-
pancreatic duct ligation. Diseased-treated rats had duodenal BPJ
replacement fresh from a Donor Rat beginning immediately before
duct ligation. Sham controls had ducts dissected only. Rats were
killed after 1 or 3 hours (n ⫽ 7/group). Pancreatic homogenates were
used to determine Akt activation (immunoblotting, immunecom-
plex kinase assay and ELISA), IkB activation (immunoblotting), and
production of chemokines MCP-1 and RANTES (ELISA). NF-kB
was quantitated in pancreatic nuclear fractions using EMSA.
RESULTS: Compared to sham, duct ligation was associated with
marked increases in pancreatic Akt and IkB activation, NF-kB nu-
clear translocation, and MCP-1 and RANTES production. Akt/
NF-kB pathway activation and increased MCP-1 and RANTES pro-
duction after duct ligation were substantially ameliorated by BPJ
replacement (ANOVA, p⬍0.05).
CONCLUSIONS: BPJ exclusion from gut exacerbates Akt/NF-kB
pathway activation and increases chemokine production in
ligation-induced acute pancreatitis. These findings support our
central hypothesis that BPJ exclusion-induced acinar cell hyper-
stimulation exacerbates acute pancreatic inflammation in this ex-
perimental corollary of gallstone-induced acute pancreatitis.
(Support: ACS Faculty Research Fellowship, NIH-K08
Award#DK062805)
Depletion and inactivation of intestinal muscularis
macrophages prevent postoperative ileus
Florian F Behrendt, Sven Wehner PhD, Mariola Lysson,
Anthony J Bauer PhD, Andreas Hirner MD, Joerg C Kalff MD
University of Bonn
Bonn, Germany
INTRODUCTION: Abdominal surgery results in a massive inflam-
mation within the intestinal wall leading to postoperative ileus. Pre-
vious results suggested that resident macrophages within the intesti-