Bursting Pressure in Anastomotic

Healing in Experimentally Induced

Colitis in Rats
Micha Rabau, M.D.,* Ami Eyal, M.D.,*
Yoram Kluger, M.D.,* Dan Dayan, D.M.D., M.Sc.t
From the *Department orb & C Surgery Proctology Unit, Tel Aviv Sourasky Medical Center, Saclaler Faculty
of Medicine, and the ~Department of Oral Pathology and Oral Medicine, The Maurice and GabHela
Goldschleger School of Dental Medicine, Tel Aviv University, Tel Aviv, Israel

BACKGROUND: Experimental. studies on healing of colonic tis in rats < 7 is an established experimental mode1 o f
anastomosis have been thoroughly investigated. However, chronic colitis in rats, w h i c h discloses several patho-
clinical parameters of the healing process of anastomosis in
the inflamed colon has not yet been reported. METHODS: logic features of h u m a n IBD, particularly Crohn's dis-
In the present study, healing of anastomosis in trinitroben- ease. Anastomotic bursting pressure a n d tension are
zene-sulfonic acid-induced colitis in rats was assessed by valid criteria for assessing anastomotic strength a n d
measuring the bursting pressure and bursting wall tension.
integrity during its healing. 4 In the present study, the
RESULTS: On postoperative day 4, bursting pressure and
bursting wall tension were significantly lower (P < 0.001) latter parameters w e r e used to investigate anasto-
in rats with cotitis with or without anastomosis and normal motic healing in the experimental m o d e l of TNB-
colon with anastomosis, compared with normal colon with-
i n d u c e d colitis in rats.
out anastomosis. On postoperative day 7, bursting pressure
and bursting wall tension of normal colon with anastomosis
approached that of normal colon without anastomosis.
However, bursting pressure and bursting wall tension of
rats with colitis with or without anastomosis remained MATERIALS AND METHODS
significantly lower (P < 0.001) than the latter. Furthermore, A total of 72 male Wistar rats, w e i g h i n g 280 to
unlike rats without colitis in which perforation occurred
mostly at the anastomotic line, the bursting site in colitic 320 g, w e r e maintained in a restricted access r o o m
rats was predominantly away from the anastomotic line. with a controlled temperature o f 23°C a n d a light/dark
CONCLUSIONS: These results suggest that in surgery for cycle of 14:10 hours. Rats w e r e h o u s e d in a metal b o x
inflammatory bowel disease, it is the adjoining inflamed
container, with a m a x i m u m of three rats per cage.
bowel wall that is vulnerable to be perforated in response to
increasing intraluminal pressure rather than the anastomo- Standard I a b o r a t o w pellet formula a n d tap water w e r e
sis that is braced by the sutures. [Key words: Anastomotic provided. All experimental procedures w e r e per-
healing; Colitis; Bursting pressure; Bursting wall tension] f o r m e d according to our institution's guide for care
Rabau M, Eyal A, KIuger Y, Dayan D. Bursting pressure in and use of laboratory animals.
anastomotic healing in experimentally induced colitis in
rats. Dis Colon Rectum 1998;41:468-472.

A nastomotic dehiscence and leakage are still ma- Colitis I n d u c t i o n

jor complications in surgery o f the gastrointesti-
nal tract. ~ Several factors have b e e n r e c o g n i z e d to
significantly influence healing of intestinal anastomo-
ses. 1 Inflammatory b o w e l disease (IBD) is considered
to b e a predisposing condition to anastomotic leak-
age. >5 H o w e v e r , no experimental study has yet in-
vestigated the influence o f IBD o n anastomotic heal-
ing.
Trinitrobenzene-sulfonic acid (TNB)-induced coli-
Address reprint requests to Dr. Dayan: Department of Oral Pathol-
ogy and Orai Medicine, The Maurice and Gabriela Goldschleger,
School of Dental Medicine, Tel Aviv UniversitT, Tel Aviv, [srad.
468
Colitis w a s i n d u c e d as previously described, s
Briefly, u n d e r general anesthesia (intramuscular injec-
tion of 80 m g / k g ketamine with 4 m g / k g of 2 percent
xylazine), a 0.7-mm-diameter polyethylene catheter
with two distal side ports was inserted transanally into
the c o l o n in such a m a n n e r that the tip was 8 c m
proximal to the anus (approximately at the splenic
flexure). A mixture of 0.8 ml of 5 percent (40 mg) TNB
(Sigma Chemical Co, St. Louis, MO) and 0.4 ml of pure
ethanol was p r e p a r e d a n d administered via the cath-
eter, w h i c h was slowly w i t h d r a w n while the solution
was infused.
Vol. 41, No. 4 ANASTOMOTIC HEALING IN EXPERIMENTAL COLITIS
As previously described in this model, 9 the colitic
process was stabilized and attained its m a x i m u m
chronic inflammatory features 20 days after its induc-
tion. The inflammatory process subsided 40 days after
colitis induction, and the intestinal wall regained near
normal texture. Thus, experiments in colitic rats were
started 20 days after colitis induction and lasted for 21
days.
Surgical Model
All rats underwent surgery with general anesthesia
(intramuscular injection of 120 m g / k g ketamine and 6
m g / k g 2 percent xylazine). With the use of standard
operative techniques, a midline laparotomy was per-
formed in all animals. After surgery, 150 m g of cefa-
mezin was injected into the peritoneal cavity and the
a b d o m e n was closed with one layer of 4-0 silk su-
tures. Each studied group was divided into three sub-
groups of six rats each. :Rats were killed b y ether
inhalation on the 4th, 7th, or 21st day postsurgery.
Colon w i t h C o l i t i s and
Anastomosis ( 1 8 R a t s )
At laparotomy, the inflamed segment was identified
and carefully separated from the adjacent adherent
tissue. A 7-ram longitudinal incision at the distal edge
of the inflamed segment was performed. Retained
feces w e r e evacuated, and the enterotomy was closed
transversally b y continuous inverting 5-0 silk sutures.
Rats were killed at 4, 7, and 21 days, six rats at each
time point.
N o r m a l Colon and Anastomosis ( 1 8 R a t s )
At laparotomy, the normal descending colon was
mobilized, and similar enterotomy and anastomosis
w e r e performed as in the colon with colitis and the
anastomosis group. As in the previous group, six rats
were killed at each time point.
Colon with Colitis (18 Rats)
At laparotomy, the inflamed segment was mobi-
lized as in the colon with colitis and the anastomosis
group; however, the anastomoses were not per-
formed. Six rats were killed at each time point.
469
Normal Colon (18 Rats)
At laparotomy, the descending colon was mobi-
lized. No anastomoses were performed and, as in all
other groups, six rats were killed at each time point.
Measurement o f B u r s t i n g P r e s s u r e 0 3 P ) and
WaR Tension 03WI3
After animals w e r e killed, an immediate postmor-
tem examination was performed. The abdominal cav-
ity was widely opened; the diseased segment with or
without anastomoses in colitic rats and the respective
colonic segment with or without anastomoses in non-
colitic rats were identified, mobilized, and resected
with a margin of 2 cm on each side of disease-free
intestine.
The distal end of the examined colon segment was
tied. Through tile proximal end, a cannula was in-
serted into the intestinal lumen to the distance of 1 cm
and the intestine was tied over it. Proximally, the
cannula was connected to two channels: to an infu-
sion p u m p at a rate of 2 ml/min of normal saline and
to a pressure recorder (Mennen Medical) via a pres-
sure transducer. The pressure at which leakage/
rapture of the intestine occurred was recorded as the
BP. After bursting, the intestine was longitudinally
opened, the bursting sRe in relation to the anastomo-
ses was identified, and the diameter of the colon at
the level of perforation was measured. BWT was
calculated according to Laplace's law: BXXq7 (DYN/
cm) = K (1.33 × 103 DYN/cm2/IlmlHg) × BP
(mmHg) × R (cm). 1° Data were expressed as m e a n +
standard deviation.
Statistics
Statistical significance b e t w e e n m e a n values at each
time point in each group was determined b y analysis
of variance. A comparison was m a d e b e t w e e n all
investigated groups at each time point using Student's
t-test.
RESULTS
All animals survived the experimental procedures.
Mean Colonic BP (mmHg)
Four days postanastomosis, the m e a n BP of the
colitic colon with anastomosis was 100 m m H g (Fig.
470 RABAU ETAL Dis Colon Rectum, April 1998

300

250 -
,I-

g 200 -
o

150 -

m 100 -
._o
C --n-- Colon with Colitis + Anastomosis
_o
O --e-- Normal Colon + Anastomosis
O 50-
--&-- Colon with Colitis
--x-- Normal Colon

I I I
4 7 21
Days

Figure 1. Mean colonic burst pressure of anastomosis in induced colitis during the investigated period. *P < 0.001.

1.0
4"
0
>(

E 0.8
o
Z
>-

C
.2 0.6
C
I--

0.4
e-
;=
P
-I
m
.2
o
C

0
0.2
i
--~-- Normal Colon
0.0
1 I I
4 7 Days 21

Figure 2. Mean colonic bursting wall tension of anastomosis in induced colitis during the investigated period. *P <
0.001.

1). This was not significantly different from the m e a n
BP recorded in the normal colon with anastomosis
and in the colon with colitis without anastomosis,
111.6 and 85.8 mmHg, respectively. However, m e a n
BP of all groups was significantly lower ( P < 0.001)
than m e a n BP of the normal colon, which was 250
mmHg. On the seventh postanastomotic day, mean
BP in the colon with colitis and anastomosis and in
the colon with colitis without anastomosis rose insig-
nificantly to 120 and 113,3 mmttg. These recordings
were significantly lower (P < 0.001) than the BP of
the normal colon and the normal colon with anasto-
mosis, which attained levels of 264.1 and 201.6
mmHg, respectively. On the 21st postanastomotic
day, m e a n BP of the colon with colitis and anastomo-
sis was very similar to that of all control groups,
ranging between 203 and 253 mmHg.
Mean BWT (DYN/cm)
Mean BWT of the colon with colitis and anastomo-
sis on the fourth postanastomotic day was 0.3 )< 10 5
DYN/cm and not significantly different from 0.2 × 10 5
DYN/cm in the colon with colitis without anastomosis
and from that of the normal colon with anastomosis,
Vol. 41, No. 4 ANASTOMOTIC HEALING IN EXPERIMENTALCOLITIS
0.4 × 105 DYN/cm (Fig. 2). All data were significantly
lower ( P < 0.001) than the BWT of the normal colon,
which was 0.75 × 105 DYN/cm. On the seventh post-
operative day, BWT of the colon with colitis and
anastomosis rose to 0.45 × 105 DYN/cm and was not
different from the BWT recorded in the colon with
colitis without anastomosis, which was 0.4 × 105
DS,i~/cm. These values remained significantly lower
( P < 0.001) than those of the normal colon, 0.75 x
105 DYN/cm, and the normal colon with anastomosis,
0.7 × 105 DYN/cm (Fig. 2). On the 21st postoperative
day, there were no significant differences between
mean BWT of all study groups.
Localization of the bursting site revealed that in the
group in which anastomosis was performed in a nor-
mal colon, the burst occurred at the anastomosis site
in all animals on the fourth postoperative day and in
four of six rats on the seventh postoperative day
(Table 1). In the group in which anastomosis was
performed in the colon with colitis, the bursting site
was found in the diseased intestine remote from the
anastomosis line in all rats on the fourth postoperative
day and in five of six rats on the seventh postopera-
tive day. However, on the 21st postoperative day, the
bursting site was found to be remote from the anas-
tomotic line in both the colon with colitis and the
normal colon in all rats.
DISCUSSION
Early" septic abdominal complications related to
bowel resection and anastomosis including leakage,
abscess, and fistula formation are not rare occur-
rences aider surgical treatment of IBD. Anastomosis in
a macroscopically inflamed segment is done when
performing strictureplasty in Crohn's disease. The
present experiment was designed to investigate heal-
ing anastomosis simulating strictureplasty. TNB-
induced colitis in rats is a reproducible animal model
of a transmural chronic inflannnation and ulceration
Location of Perforation
At anastomotic line
R e m o t e f r o m anastomotic line
N = n o r m a l c o l o n and a n a s t o m o s i s ; C = c o l o n with colitis and anastomosis.
471
simulating but not similar to human Crotm's disease.
Strictureplasty was simulated by a longitudinal enter-
otomy in the inflamed segment, which was then
closed transversally.
In normal intestine, the natural course of anasto-
motic healing is characterized by a significant reduc-
tion of its bursting strength during the first four posta-
nastomotic days. 3 Thereafter, a progressive increase
of its bursting strength occurs, regaining almost pre-
anastomotic strength on day 21.
In the present study, in normal colons with anasto-
mosis, results were in accordance with the above-
mentioned course of healing anastomosis. 3 The re-
cording in an inflamed bowel with anastomosis
showed that on the fourth postanastomotic day, BP
was significantly reduced but not different from that
of normal colon with anastomosis. However, whereas
in normal colons with anastomosis BP gradually rose,
it remained significantly lower for at least three addi-
tional days in the inflamed bowel with anastomosis.
Furthermore, during the first postoperative week, the
bursting site in inflamed colons with anastomosis was
mainly remote from the anastomotic line compared
with normal colons with anastomosis in which the
burst occurred mainly at the anastomosis itself. These
results suggest that when an intraluminal pressure is
exerted to the inflamed intestinal wall, anastomosis
that is reinforced by sutures is less susceptible to
perforation than other inflamed sites. The similarity in
the recorded BP profiles of inflamed colons with or
without anastomosis also suggests that it was not the
anastomosis that was prone to leakage complications
but rather the untouched ulcerous friable sites in the
intestinal wall that were the locus of minor resistance.
A possible explanation for reduced BP would be a
wider radius at the bursting site. When calculating BWT,
the longer the radius, the less intraluminal pressure re-
quired to obtain a given B~T. 1° On the fourth postop-
erative day, when healing anastomosis is at its greatest
vulnerability, no difference between BWT of both
Table 1.
Site of Intestinal Perforation
Postoperative Day
4 7 21
N C N C N C
6/6 -- 4/6 1/6 -- --
-- 6/6 2/6 5/6 6/6 6/6
472 RABAU E T A L
groups of rats with colitis and that of normal colon with
anastomosis was noted. However, on the seventh post-
operative day, the BWT of these groups was signifi-
cantly lower than that of rats with normally healing
anastomosis. The bursting site in rats with colitis oc-
curred in an il-fflamed dilated segment remote from the
anastomosis w h e n performed. The radius at the perfo-
ration site was longer than that at the anastomosis.
Therefore, a lower intraluminal pressure was required to
obtain BTW, This observation suggests that an inflamed
diluted segment is more prone to perforate than the
adjacent healing anastomosis.
An increased incidence of postoperative leaks after
Finney or J-pouch strictureplasty c o m p a r e d with
Heineke-Mikulicz strictureplasty for intestinal Crohn's
disease has b e e n reported, 11 We suggest that it is not
the mere fact of longer anastomotic suture line but
rather the vulnerability of the inflamed dilated
"pouch" that probably makes it more prone to perfo-
rate.
The 21st postoperative day corresponded to 40
days after colitis induction. At this time, healing of the
colitic segment was in an advanced stage character-
ized by replacement of the inflammatory elements
with granulation and fibrotic components and regen-
eration of mucosal epithelium and muscle fibers. 9
Resolution of the inflammatory process that was re-
placed by scaITing tissue signified that the intestinal
wall was regaining its integrity and strength. These
findings explain results that on the 21st postoperative
day, BP and BWT in rats with colitis were almost level
with those rats without colitis,
CONCLUSION
We suggest that the intestinal anastomosis braced
with sutures in an inflamed colon is less prone to
Dis Colon Rectum, April 1998
perforate in response to an increasing intraluminal
pressure than other vulnerable inflamed sites in the
bowel wall.
REFERENCES
1. Schrock TR, Deveney CW, Dunphy JE. Factors contrib-
uting to leakage of colonic anastomoses. Ann Surg
1973;177:513-18.
2. Blomquist P, Ahonen H, Jibron H, Zedeffeldt B. The
effect of relative bowel rest on healing of colonic anas-
tomoses. Acta Chir Scand 1984;150:677-81.
3. Irving TT, Hunt TK. Reappraisal of the healing process
of anastomosis of the colon. Surg Gynecol Obstet 1974;
138:741-6.
4. Hendriks T, Mastboom WJ. Healing of experimental
intestinal anastomoses: parameters for repair. Dis Colon
Rectum 1990;33:891-901.
5. Kirshner JB, Shorter RG. Inflammatory bowel disease.
2nd ed. Philadelphia: Lea & Febiger, 1980:215-510.
6. Morris GP, Rebeiro L, Herridge MM, Szewczuk MR,
Depew WT. An animal model for chronic granutoma-
tous inflammation of the stomach and colon [abstract].
Gastroenterology 1984;86:1188.
7. Morris GP, Beck PL, Herridge MM, Depew WT,
Szewczuk MR, Wallace JL. Hapten-induced model of
chronic inflammation and ulceration in the rat colon.
Gastroenterology 1989;96:795-803.
8. Ondrula D, Nelson RL, Andrianopoulos G, et al. Quan-
titative determination of pentane in exhaled air corm-
lares with colonic inflammation in the rat colitis model.
Dis Colon Rectum 1993;36:457452.
9. Rabau M, Eyal A, Dayan D. Histomorphometric evaluation
of experimentally induced colitis with trinitrobenzene-
sulfonic acid in rats. IntJ Exp Pathol 1996;77:175-9.
10. Koruda MJ, Rolandelli RH. Experimental studies on the
healing of colonic anastomoses. J Surg Res 1990;48:
504-15.
11. Keighley MR, William NS. Surgery of the anus, rectum
and colon. London: WB Saunders 1993:1740.