|
METHODS
Animals
Female Sprague-Dawley rats weighing between 275 and 325 grams
were used. All rats were kept at least 1 week before operation
to adjust to the laboratory environment. During this time they
received standard lab chow and water, ad lib. They were housed
in a temperature- and humidity-controlled environment with a
12-hour light-dark cycle. The animals were cared for by the University
of Minnesota Research Animal Resources Department in accordance
with the principles in the Guide For Care and Use of Laboratory
Animals, NIH publication, revised 1993.
Surgical technique
Rats were anesthetized with an IP injection of sodium pentobarbital
(30 to 40mg/k). The abdomen was shaved with an electric clipper
and prepared with povidine-iodine solution. Procedures were done
under aseptic conditions.
Abdominal wall resection and omentectomy: necropsy studies
Preliminary studies were done to assess the effects, with
respect to adhesions, of the tissue backing the porous mesh—subcutaneous
tissue versus peritoneum. It was our initial impression in this
rat model that omentum had a greater propensity to adhere to
polypropylene mesh than did intestine. This proposition was tested
by removing the omentum in some animals.
In one group the polypropylene mesh was sewn around its periphery
to the peritoneal surface on either side of a midline incision.
Peritoneum, muscle, and skin were then reapproximated, so the
prosthetic material had underlying peritoneum. In other animals
an abdominal wall defect was created, preserving the skin and
subcutaneous tissue but excising muscle and peritoneum. The outer
surface of the mesh was exposed to the subcutaneous tissue rather
than to peritoneum, mimicking the clinical situation. In another
group of rats omentectomy was performed to observe the consequences
on adhesion formation. So, four preparations were studied: with
or without abdominal wall resection and with or without omentectomy.
The groups were prepared sequentially with no attempt at randomization.
Adhesions to mesh were evaluated by sacrificing the animals 30
days or more after operation. Data were analyzed and groups compared
by two-way analysis of variance followed by inspection of differences
between pairs of means by Duncan's multiple range test.
Adhesion dynamics: laparoscopy
In subsequent studies we used the model found to have maximum
adhesion formation, namely, abdominal wall resection and preserved
omentum. These animals were studied by repeated laparoscopy starting
at day 1 and at chosen intervals, 3, 7, 14, and 28 days after
mesh installation. Additional animals were subjected to this
preparation and sacrificed at days 1, 3, 5, 7, 14, and 28 solely
to obtain samples for scanning electron microscopy. The rats
were arbitrarily designated as to sacrifice date at the time
of mesh placement.
Technical considerations
Pilot studies demonstrated that interrupted sutures encouraged
omentum or intestine to adhere to the cut edge of the prosthetic
material, occasionally even to permit visceral herniation between
sutures into the subcutaneous space. We subsequently used continuous
monofilament polypropylene suture, everting the edge of the prosthesis
and the cut edge of the abdominal wall away from the abdominal
cavity. Even with this method there was some propensity for adhesions
to form at the abdominal wall to prosthesis intersection.
A technical detail that proved important
was that of skin closure. Some investigators have reported problems
with skin dehiscence in the rat after laparotomy.8-10
In our hands, the association of an abdominal wall defect, polypropylene
mesh, and simple skin approximation with interrupted or running
nonabsorbable material frequently failed because the animal chewed
on the sutures with resulting skin separation. The use of running
intradermal absorbable suture eliminated this problem. An associated
observation was that when the skin edges did separate and expose
the mesh, extensive dense adhesions covered the intraabdominal
surface under the exposed area. So any animal that developed
skin dehiscence was discarded from further consideration. Of
the 28 animals prepared for these studies 3 were eliminated because
of skin breakdown (Table 1).
Table 1. Effect of Omentectomy and Abdominal Wall Resection,
or Both, on the Area of Polypropylene Mesh Involved with Adhesions
|
|
|
|
Mean |
|
|
|
|
|
|
|
Group |
No. of animals |
n |
% |
Range (%) |
|
|
O + AWR + |
7 |
8* |
48 |
18–75 |
|
O – AWR + |
7 |
3* |
93 |
5–100 |
|
O + AWR – |
5 |
10* |
16 |
0–51 |
|
O – AWR – |
6 |
13 |
67 |
25–100 |
|
* Values are significantly different from each
other, p < 0.05. AWR, abdominal wall resection; O, omentectomy.
Data are means with standard error of the mean in parentheses.
Laparoscopy
To test the tolerance of the rat to abdominal insufflation
for laparoscopy a bulb pump was used to inject room air. A 25-gauge
needle was inserted into the abdominal cavity and room air was
slowly instilled while ventilatory rate and pattern were observed.
Intraabdominal pressure up to 20 mmHg did not cause respiratory
alterations. Seven mmHg was found to provide enough space for
laparoscopic observations; subsequent exams were done without
pressure measurements, simply instilling sufficient air to allow
the needed visualization.
To accomplish the examination, an 8-mm skin incision was made
in the flank 4 cm lateral to the umbilicus. A 5-mm Auto Suture
Surgiport (Auto Suture Co, Norwalk, CT) beveled cannula was inserted,
through which a 4-mm Comeg arthroscope (American Medical Endoscopy,
Miami Beach, FL) was introduced. A Karl Storz Xenon Light Source
610 (Karl Storz Endoscopy-America, Culver City, CA) was attached.
Observations were video recorded.
Magnification obtained with this set-up varied depending on
the distance from the end of the scope to the tissue. With the
tip close to the observed surface, magnification on the monitor
was approximately 80 times. This allowed remarkably clear visualization
of the microvasculature.
To quantitate adhesions, the prosthetic patch was visually
divided into four equal quadrants (Fig. 1). Adhesions covering
the mesh were estimated according to area of attachment, ranging
from 0% to 100% (no adhesions to completely covered). The intraabdominal
structure involved, omentum or intestine, location of adhesions,
and surface appearance of uninvolved mesh were noted and recorded.
After the examination was completed the scope and cannula were
removed and the skin closed.
Figure 1. Laparoscopic images obtained on days 7, 14,
and 28 after initial operation. The mesh is visually divided
into four equal quadrants and adhesion area estimated. The edematous
omentum on day 7 becomes translucent on days 14 and 28, exposing
to the view the fibers of the underlying mesh. Individual fibers
of the polypropylene mesh are easily seen, as is the vasculature,
an advantage of laparoscopy.
Adhesiolysis
One group of six rats was studied for reformation of adhesions
after laparoscopic adhesiolysis 1 week after abdominal wall resection
and initial mesh placement and at intervals thereafter to assess
possible reformation of adhesions. In this study, the firmness
of the adhesions was scored on a scale of 1 to 3. A score of
1 meant they were loosely adhered and came free when the abdomen
was inflated or by gentle probing with the tip of the scope.
A score of 2 indicated moderate adherence such that the tissues
required separation with a grasper and more vigorous blunt dissection.
A score of 3 indicated firmly attached strong adhesions that
had to be divided sharply. Hemostasis was obtained by compressing
an instrument against the site of bleeding. Cautery was not used
to avoid new adhesion formation secondary to thermal injury.
Histology
Animals were sacrificed with 100mg/kg IP injection of sodium
pentobarbital. The abdomen was opened approximately 2cm from
the prosthesis on the caudal and both lateral aspects of the
prosthesis. The abdominal wall was carefully lifted in order
to observe and score adhesions to the mesh. The entire prosthesis,
adjacent abdominal wall, and adherent structures were then completely
removed by incising 2cm from the cephalad edge of the polypropylene.
The specimens were rinsed in saline, pinned on corkboards, and
fixed in 10% neutral buffered formalin for 48 hours. They were
cut into quarters to produce four samples, each of which included
normal abdominal wall, prosthesis, and the transition area between.
Tissues were routinely processed in an Autotechnicon (The Technion
Co, Chauncey, NY) and embedded in paraffin. Samples were sectioned
at 5 micrometers and stained with hematoxylin and eosin. To focus on collagen, a second set of slides
was stained with sirius red plus hematoxylin according to the
method of Junqueira and associates.11
Sirius red is a strong anionic dye with sulphonic acid groups
that stains collagen by reacting with basic groups in the collagen
molecule. Collagen birefringency was enhanced by polarization.
Histologic specimens were evaluated by a microscopist who was
always unaware of their origin.
Scanning electron microscopy
One rat was sacrificed on postplacement days 1, 3, 5, and
14, two on day 28, and three on day 7. Tissues were prepared
for scanning electron microscopy as follows. Immediately after
opening the abdomen the surgical mesh and surrounding area were
flooded with 2.5% glutaraldehyde in 0.1M cacodylate buffer. A
1×1-cm sample of mesh, including an area with and another
without adhesions, was removed and put in fresh glutaraldehyde
buffer fixative overnight. The sample was then rinsed in 0.1M
cacodylate buffer 3 times for 20 minutes each, and postfixed
on ice with 1% osmium tetroxide in 0.1M buffer for 30 minutes.
Samples were then rinsed in buffer, dehydrated in increasing
alcohol series (50%, 70%, 80%, 95%, and 100%), and critical point
dried in a Tousimis Model 780 A (Tousimis Research Corporation,
Rockville, MD). Sputter coating with gold and palladium particles
of 10nm was done. Observations and photographs were obtained
using a Hitachi S-450 S and a Hitachi 4700 computerized scanning
electron microscope (Hitachi Instrument Inc, Nissei Sangyo America,
Ltd, Mountain View, CA).
Introduction
| Methods | Results
| Discussion
| References
JACS |
