. -» «W «3* _
‘ ‘ - , ifisafis-gpqrg-
. .‘ 5' ‘ég/ $0?” 3"}. $.Er..,
.tf'; . .3 ’ ‘ CF $‘ i. ‘ .7
qfil‘ww 15‘ fl” .
. I ~ “9". 5;! .

r};

.,-__
,3

ak'"
,

. ’ wit «5"; ' ‘
r: v .—.x‘ 1w. ‘:,~« §‘- n.
;.. 3an,,affi~. ‘ ‘- 361‘
pink “Wang
w «545*! ”9% x
a a?" 5' at As.

I
.r'

‘4'“)
A)» :u
; “at: Xvi fl 3“
I x» "‘ »' ~ I...
V ”MK-3 . u ‘ 3.,
n A‘. - ‘1 '. 51" , .1‘l‘ .~;’
H“ ”v Maw“ ”22:.
O ‘ ‘ ’ "a". 149,“;
“\E'} ' . .
(“«Z" wk
'” “."'u S“
4;“: xx ' > ‘1 . '
"rfl'* " ‘ . '
I "NJ! )9“ fl?!
v'. (0:1 3
- 3 m‘ yawn .rfi'
" . ‘ ‘ “MAL” "VI" ‘1 >' a" ’3‘ V ' a ‘V‘jylxai-m
VJ" - ',,.' . ‘M “I ‘ . v‘k‘
. n ._ .
- 2“"? ‘3’" 5:1:— ' .
3;??? ._ . - avggfihw .
:«r w .fi v

5' v \

' '1“ \
"3&4!" 12‘.“ .
yam-M
'5 ’ » r ,..

{y’i‘rgél‘;§~>.,
1 ‘:“«‘.-‘-'»‘,f."
wwuggmgg

,_
"an:
I: ‘r. A

n 5L
A .

I!
L!“
wigs;

'3')? '7: a
I??? ‘3‘: Va

4.‘
Rudy‘s-E"??- -

. , .2 . ‘
3"}: .

~rk. .r
“<3 \>
’J'lfir'

m: , ....‘_
{Sizwaa
"" 41.1.753,

I)!

I a
\-'
”no

u.
1 .1

l

A

l

..
. .n,
a".
-t.
-- ,
.

,i
'4‘;

a n
S'L'
“g?
a- .
'1!“ I

w

I . x
'3-1

r

“0535M

 

LIEIBRAR

IUIWHIHHW“WWWWWll‘ ILHIIWI

3 1293 0078

 

 

 

V \
LIBRARY

Nichlgan State

University
\ J

 

 

This is to certify that the

dissertation entitled

A Serial Evaluation of
Anastomotic Healing with Comparison of
Czerny-Lembert and Circular-Stapled Techniques

presented by

Anthony Senagore

has been accepted towards fulfillment
of the requirements for

Masters Science

degree in

 

 

W‘s’flw

 

 

 

Date §/"7/5’9 “Lj- flubey

MS U is an Affirmativc Action/Equal Opportunity Institution 0-12771

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return 'on or before date due.

|__J| |L__
:14:
AD;

I l:
| ll

MSU Is An Affirmative ActionEquel Opportunity Institution
omen!” '

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A SERIAL EVALUATION OF
ANASTOMOTIC HEALING WITH COMPARISON OF

CZERNY-LEMBERT AND CIRCULAR-STAPLED TECHNIQUES

BY

Anthony Senagore

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Physiology

1989

(004a o'SX

ABSTRACT

A SERIAL EVALUATION OF ANASTOMOTIC HEALING WITH COMPARISON
OF CZERNY-LEMBERT AND CIRCULAR-STAPLED TECHNIQUES

BY

Anthony Senagore

This project was performed in four sections.
Section I was a serial evaluation of the healing
process of Czerny-Lembert (2—layer handsewn, inverted)
and circular stapled (double staple row, inverted)
anastomoses, representing the first direct comparison.
Both anastomoses healed via second intention with no
significant differences in anastomotic blood flow,
bursting pressure, hydroxyproline content, gross or
inflammatory scoring, or the incidence of anastomotic
healing complications.

Section II assessed anastomotic healing in non—
diverted animals and those with proximal diverting
colostomy which has been associated with increased
anastomotic stenosis. No significant differences were
identified for any of the healing parameters. There
was no predilection for diverted anastomoses to stenose.

Section III allowed development of a reproducible

model of anastomotic ischemia and prospective

assessment of laser doppler velocimetry (LDV) and
tissue pH (pHi) measurements in predicting subsequent
tissue viability. It was apparent that pHi was
superior to LDV in identifying critical ischemia and
predicting anastomotic outcome. There was no
significant difference in anastomotic healing or
complications for CL or CS techniques.

Section IV assessed the serial effects of
preoperative radiation therapy on anastomotic healing.
We confirmed that preoperative radiation therapy
(4SOOcGy) induces an early and persistent ischemic
insult in the colorectum and increased the incidence of
healing complications. 60umol/kg ATP Magnesium
Chloride (60umol/kg) during each radiation treatment
session resulted in a marked improvement in early
anastomotic blood flow, increased Hydroxyproline
content and decreased gross inflammatory scoring.
Potential mechanisms of action of the ATP-MgCl2 in
this model include a decrease in the ischemia resulting
from the endarteritis or direct protection from

cytotoxic effect of radiation exposure.

Copyright by
Anthony Senagore

1989

This thesis is dedicated to my wife, Patricia,
and my children, Antonio and Christina, who
encouraged me during the difficult times of this
project, and understood the absences required
by this work. It is also dedicated to my parents,
Antonio and Kathleen, who instilled in me the
importance of the pursuit of knowledge.

ACKNOWLEDGMENTS

I am greatly indebted to Dr. Irshad Chaudry who
provided me with the opportunity to work in his
laboratory and learn from his fine example. I would
also like to thank Dr. Ching Chung Chou and Dr. John
Chimoskey for participating on my graduate committee.
In addition, I would like to thank Dr. Richard Walshaw
and Dr. Ulreh Mostoskey and Dr. Robert Dunston who
assisted this project greatly due to their expertise
with veterinary surgery, radiology, and pathology,
respectively.

I would like to express my sincere gratitude to
Jeffrey W. Milsom, M.D., who was an enthusiastic
advisor throughout this project, as well as becoming a
good friend.

A special thanks goes to Wendy Johnson for her
significant technical assistance throughout the project
and to Pam Brown and Linda Heiny for the significant
amount of work required to type and prepare this
manuscript.

This project was made possible by a generous
research grant provided by United States Surgical
Corporation.

vi

TABLE OF CONTENTS

 

 

 

 

 

Page No.
I. INTRODUCTION ................................ 1
A. GENERAL INTRODUCTION ................... l
B. SURVEY OF PUBLISHED WORK ............... 2
1. Review of intestinal wound
healing ........................... 2
2. Review of enteric suturing ........ 4
3. Review of anastomotic ischemia
and its assessment ................ 8
4. Review of effect of colostomy
on anastomotic healing ............ 12
5. Review of the effects of
therapeutic radiotherapy on
colorectal anastomotic healing....13
C. AIMS OF THIS PROJECT .................. 17
II. MATERIALS AND METHODS ....................... 19
A. ANIMALS/ANESTHESIA ..................... 19

 

vii

PAGE NO.

 

 

 

EXPERIMENTAL GROUPS ................... 20
1. Section I ........................ 21
2. Section II ....................... 24
3. Section III ...................... 25
4. Section IV ....................... 27
5. Section V ........................ 29
OPERATIVE TECHNIQUE ................... 31
1. First Surgery .................... 31
2. Second Surgery ................... 32
ANALYTICAL TECHNIQUES ................. 34
1. Laser Doppler Velocimetry ........ 34
2. Tissue pH ........................ 35
3. Bursting pressures ............... 36
4. Hydroxyproline assay ............. 36
5. Gross inflammatory grading ....... 37
6. Barium enema examination

for stenosis or leak ............. 38
7. Microscopic inflammatory

score ............................ 38
8. Statistical analysis ............. 39

viii

Page No.

 

 

 

 

III.RESULTS ...................................... 41
A. SECTION ONE: NORMAL COLORECTUM ......... 41
1. Laser Doppler Velocimetry results..41
2. Bursting pressures ................. 42
3. Gross and microscopic inflammatory.43
4. Hydroxyproline content ............. 44
5. Anastomotic diameter and
complications ...................... 45
B. SECTION TWO: COLOSTOMY (COL) VS ........ 46
CONTROL (CON)
1. General data ....................... 46
2. Laser doppler velocimetry results..46
3. Bursting pressure .................. 48
4. Inflammatory scores ................ 48
5. Hydroxyproline content ............. 49
6. Anastomotic complications
and diameter ....................... 49
C. SECTION THREE ........................... 51
1. Phase I ............................ 51
a. General data on critical
quantitative values ........... 51
2. Phase II ........................... 53
a. General data and
complication rates ............ 53

ix

SECTION FOUR ......................... 55

 

(Radiation Therapy vs Control)

 

1. General data ..................... 55
2. Laser doppler velocimetry data

(Control vs. RT) ................. 55
3 Bursting pressure ................ 57
4 Gross inflammatory scoring ....... 57
5. Hydroxyproline content ........... 58
6 Anastomotic complications ........ 58
SECTION FIVE .......................... 59
1. Laser doppler velocimetry data

(saline, ATP—30, ATP~60) ......... 59
2. Gross inflammatory scoring ....... 60
3. Incidence of cutaneous lesions

in radiotherapy portals

(saline, ATP-30, ATP—60) ......... 6O
4. Hydroxyproline data

(saline, ATP—30, ATP-60) ......... 61
5. Bursting pressure ................ 61
6. Incidence of leak or stenosis....62

Page No.
IV. DISCUSSION ................................. 63

 

A. GENERAL DISCUSSION OF COLORECTAL

ANASTOMOTIC HEALING ................... 63
B. HEALING IN THE NORMAL COLORECTUM ...... 66
C. ANASTOMOTIC ISCHEMIA .................. 67

D. EFFECTS OF PREOPERATIVE PELVIC

RADIOTHERAPY .......................... 71
E. EFFECTS OF ATP~MgCl2 ON ADVERSE
EFFECTS OF RADIOTHERAPY ............... 73

F. EFFECT OF PROXIMAL COLOSTOMY

 

 

 

STENOSIS ......................... 75
V. CONCLUSIONS ................................ 78
REFERENCES ..................................... 81
APPENDIX A
APPENDIX B

 

xi

LIST OF TABLES

Page No.
SECTION I:
Table 1: LDV Readings Initial Surgery (mV) ...... 41
Table 2: LDV Data Day 3-120(mV) ................. 42
Table 3: Gross Inflammatory Score Day
3-120 (max=24) ........................ 43
Table 4: Microscopic Inflammatory Score
Day 3—120 .............................. 44
Table 5: Hydroxyproline Day 3—120 (mg/ml) ....... 44
Table 6: Anastomotic Diameter and Ratios for....45
60 and 120 Day Groups
SECTION II:
Table 7: Initial Perianastomotic Blood
Flow (mV) .............................. 46
Table 8: Perianastomotic Blood Flow (mV)
5-120 Day .............................. 47
Table 9: Bursting Pressures (mmHg) Day 5 ........ 48

Table 10: Gross Inflammatory Score Day 5-120
(max=24) ............................... 48

SECTION III:

Table 11: Anastomotic Hydroxyproline Content
(Colostomy vs Control 1mg/ml) .......... 49

Table 12: Comparison of Anastomotic Outcome
with LDV and pHi Data in Phase I ....... 51

Table 13: Predictive Accuracy of LDV and pHi,
Phase II ............................... 54

xii

Page No.

SECTION IV:
Table 14: Initial Blood Flow (mV), Control
vs. Radiation Therapy .................. 55
Table 15: Blood Flow (mV) 5-120 Day, Control vs.
Radiation Therapy ...................... 56
Table 16: Bursting Pressures (mmHg) Radiation
(RT) vs. Control Day 5 and 11 .......... 57
Table 17: Gross Inflammatory Score (Max=24) ...... 57
Radiation vs. Control
Table 18: Hydroxyproline Content ................. 58
Table 19: Initial Blood Flow (mV) Saline,
ATP—30, ATP-6O ......................... 59
Table 20: Blood Flow (mV) Day 5, Saline,
ATP-30, ATP—60 ......................... 59
Table 21: Blood Flow (mV) Day 11, Saline,
ATP-30, ATP-60 ......................... 60
Table 22: Cross Inflammatory Score (Max=24),
Saline, ATP-30, ATP-6O ................. 60
Table 23: Hydroxyproline (mg/ml) Day 5 and 11,
Saline, ATP-30, ATP—6O ................. 61
Table 24: Bursting Pressure (mmHg) Saline, ....... 61

ATP-30, ATP-60

xiii

FIGURE
FIGURE
FIGURE
FIGURE
FIGURE
FIGURE

FIGURE

LIST OF FIGURES

Page No.
Section I, Protocol Outline ........ 23
Section II, Protocol Outline ....... 24
Section III, Protocol Outline ...... 26
Section IV, Protocol Outline ....... 28
Section V, Protocol Outline ........ 30
Circular Stapling Technique ........ 33

Comparison of laser doppler
velocimetry and tissue pH readings,
Section III, phase I ............... 52

xiv

LIST OF ABBREVIATIONS

The following abbreviations have been employed:

CL : Czerny-Lembert two layer handsewn
anastomosis

CS : EEA® circular stapled anastomosis

LDV : Laser doppler velocimetry

pHi : Intramural pH reading

ATP-MgCIZ Adenosine Triphosphate-Magnesium Chloride

cGy : Centigrays (unit of radiation dose)

XV

IA. GENERAL INTRODUCTION

Benjamin Travers, an English surgeon trained by Sir
Astley Cooper, ushered in the modern age of intestinal
suturing, with his report on anastomotic healing in
1812, stressing the need for accurate, circumferential
serosal apposition for healing of enteric
anastomoses.1 Antoine Lembert, working in France in
1827, described a two layer inverting anastomotic
technique based on two tenets 1) sutures should be
placed in the submucosa because of its considerable
strength; 2) the serosal surface should be accurately
apposed circumferentially.2 It is this technique,
later modified by Czerny and Connell, which remains the
most widely used today.3

Healing of enteric hand-sewn anastomoses has been
widely studied, with comparisons made between single
and double layer techniques, inverting and everting
anastomoses, and between different suture materials.
Over the last two decades enteric stapling has become
more popular, especially when used to create end to and
low rectal anastomoses with the circular EEA®

instrument. However, little is known regarding the

differences in healing between two layer handsewn and

circular stapled anastomoses. In fact, there has been
only one direct comparison between those two techniques

in either clinical or basic research projects.

IB. SURVEY OF PUBLISHED WORK

 

1) Review of Intestinal Wound Healing

Wound healing in the gastrointestinal tract is
understood to follow the same natural history as wounds
elsewhere in the body. In general, wound healing
occurs in three stages: 1) lag phase (0—5 days); 2)
period of rapid collagen synthesis (5—17 days); 3)
remodeling phase (17 days-2 years).4 The lag phase
implies a time period of minimal activity when in
reality this is a very active phase dominated by the
inflammatory response. Immediately after wounding
there is platelet activation and activation of Hageman
factor which results in a cascade of humoral responses
including the complement system, coagulation cascade
kinin activation, and plasmin generation.5 The net
result of these responses is hemostasis and attraction
of a number of cell lines important in wound healing
and repair.5 Platelets release a number of growth
factors which attract and stimulate fibroblast ingrowth

as well as to stimulate neovascularization. In the

first 24 hours, polymorphonuclear phagocytes and
monocytes enter the wound area to begin the period of
inflammation with degradation and removal of necrotic
tissue.6 The amount of necrotic tissue present in
the wound can greatly affect the degree and rapidity
with which successful wound healing can proceed by
altering the intensity of this inflammation.

At approximately day 5, the rate of collagen
synthesis begins to increase dramatically and this
continues at a high rate to approximately day 17.7
During this period, tensile strength increases
correspondingly. However, the ability of the
fibroblast to produce collagen is dependent on delivery
of nutrients, especially oxygen. Initially, oxygen is
delivered by diffusion but soon neovascular in-growth
allows return to more normal levels of microvascular
blood flow and nutrient delivery.

The period of collagen remodeling begins at
approximately day 17 with collagen cross linking
accounting for any further increases in wound tensile
strength. During this time period, an equilibrium is
reached between collagen production and collagen
degradation so that no net increase in collagen content

occurs after this time.

Inflammation without infection is a significant
early determinant of the success of gastrointestinal
wound healing. However, it should be noted that the
mere act of performing an anastomosis results in
significant local inflammation at the anastomotic
line. First, tissue disruption by bowel division sets
into action the previously discussed humoral cascade
systems. Additionally, suture material represents a
foreign body in the wound. Shambaugh and Dumphy, in
1937, found that silk (less inflammatory material)
decreased the incidence of wound infection when
compared to catgut.8 Madsen studying 12 different
suture materials described a marked exudative tissue
reaction and delayed collagen formation in association
with absorbable suture material.9 Lord gt al,

identified marked injury to the submucosa, associated

with braided versus monofilament suture material.10
Similarly, Herrmann gt g1, identified histologic
changes of sloughing and necrosis in inverted handsewn
colonic anastomoses with marked inflammation.11

2) Review of Enteric Suturing

 

Suture material is a foreign body, and as such
produces a local tissue inflammatory reaction which

impedes anastomotic healing. In the 1940's, it was

found that silk produced less inflammatory response and
therefore shortened the lag phase of healing and
increased early bursting strength.12 In comparison,
plain cat gut and chromic cat gut incited a significant
inflammatory response which resulted in sloughing of
that portion of the anastomosis with healing by second

13 More recently, less inflammatory

intention.
absorbable suture materials such as polyglycolic acid
have been evaluated and found to be comparable to silk
in anastomotic strength and degree of inflammatory
response.14 Therefore, the majority of studies have
concentrated on chromic catgut, silk, or polyglycolic
acid suture material.

In comparing single versus double layer handsewn
anastomoses, Sako and Wangensteen in Minneapolis in
1951, found significant reduction in stomal size and
increased inflammation in association with a double
layer inverted suture technique.15 Hamilton, in
1967, confirmed the disadvantages associated with a
double layer handsewn anastomosis by identifying
increased edema of the inverted cuff and a
correspondingly higher risk of postoperative

16

obstruction at the anastomosis. Letwin and

Williams found that single layer anastomoses exhibited

greater bursting strength, improved anastomotic blood

flow (using alphazurine dye) and less tissue necrosis

at the anastomotic line. In addition, they identified
an earlier rise in collagen content anastomotically

17 Thus, studies have

with a single layer technique.
uniformly supported single layer suture technique over
double layer for optimal anastomotic healing. Despite
these experimental data, a two layer handsewn technique
is still widely employed in clinical surgery.

In comparing inverting and everting techniques the
differences have not been as clear cut. It has been
found, however, that everting anastomoses tend to be
more dependent on adhesions for viability and
anastomotic integrity, and may be somewhat weaker in
the early postoperative period.18-22

Mechanical instruments for performing intestinal
anastomoses originated in the work of Humor Hultel of
Budapest, Hungary. In 1908, he demonstrated a stapling
device to the Second Congress of the Hungarian Surgical
Society, designed for use in distal gastrectomies. He
had performed 21 of these procedures in the previous

23

year without complications using this stapler.

Alador Von Petz subsequently improved on Hultel's

instrument in 1924, producing a lighter, more easily
usable instrument.24

Significant advancement in surgical stapling
occurred in the 1950's at the Scientific Research
Institute of Experimental Surgical Apparatus and
Instruments in Moscow. They developed staplers for
vascular anastomoses, terminal staplers for the GI
tract, and a circular stapler for rectal and esophageal
anastomoses.25-27 These instruments provided
accurate reproducible staple lines which were readily
applied to virtually any clinical situation.

Stapling techniques for creating gastrointestinal
anastomoses were introduced in this country in the late
1960's by Ravitch. Since that time staplers, have
gained widespread popularity in gastrointestinal
surgery, and especially in colorectal procedures as
they allow an anastomosis to be performed lower in the
pelvis than previously technically feasible. As a
result, a surgeon is more likely to be able to preserve
anal sphincter function in patients with rectal tumors
or inflammatory disease, thus avoiding a permanent
colostomy. Clinical experience with stapling

techniques has generally been very good, however, some

concerns have been raised regarding anastomotic

integrity with leakage rates ranging from 2 to

68% 28-31

In addition, it is felt that low anterior
stapled colorectal anastomoses have a higher risk of
stenosis based on retrospective uncontrolled clinical
studies reporting incidences ranging from 13 to

28,32—34
35%.

Therefore, concern remains over the
integrity of these stapled anastomoses. Despite a
significant clinical experience with anastomotic
stapling there have been few experimental evaluations
of anastomotic healing, and in particular, there has
been no comparison between the standard Czerny-Lembert
(2 layer) sutured anastomosis and any of the stapling
techniques.
3. Review of Anastomotic Ischemia and its Assessment
Anastomotic ischemia is another factor which can
affect early wound healing. When the bowel is
transected, platelet activation and coagulation causes
occlusion of local blood vessels to secure hemostasis.
Thus, the anastomotic line is dependent on delivery of
nutrients by diffusion to survive until
neovascularization can occur. The decrease in local
blood flow is coupled with a local increase in
metabolic activity, thus worsening the impact of the

borderline delivery of nutrition at a time when healing

is dependent totally on oxygen delivery. Chung gt
gt, identified acute decreases in mucosal blood at
the anastomosis with either stapling or suturing using
laser doppler velocimetry (LDV).36’37
Although ischemia at the anastomotic line has been
proposed as the predominant pathophysiologic mechanism
for impaired healing, it has been difficult to
quantitate clinically. Currently, there is no readily
available, non—invasive, accurate method of quantifying
critical levels of ischemia in the GI tract which
result in irreversible tissue injury. Non-quantitative
data regarding gastrointestinal blood flow may be
obtained by gross inspection, fluorescein injection, or

38:39 Although, these teChniques

doppler ultrasound.
are readily available clinically, the qualitative data
supplied is inadequate to allow quantification of
critical levels of tissue ischemia and thus accurate
prediction of successful anastomotic healing.

There are a number of currently available methods
which allow quantitation of gastrointestinal blood
flow: radioactive microspheres, Hydrogen clearance,
laser doppler velocimetry, and tissue oximetry. Each

of these techniques has relative advantages and

disadvantages. Microsphere analysis allows accurate

10

determination of differential blood flow to the layers
of the bowel wall; however, this technique requires
tissue removal and therefore is not useful in the

clinical setting.38

Similarly, hydrogen clearance
results in accurate data in the experimental situation,
however, its accuracy can be altered by diffusional
shunting of the indicator or as a result of surgical
manipulation of the bowel.38"42

A quantitative technique for measuring bowel wall
perfusion which is readily available is laser doppler
velocimetry (LDV). The laser doppler emits a
monochromatic light with a wave length of 632.8nm. The
light penetrates to a depth of 1.5mm and is reflected
back to a photodiode resulting in a doppler shift in
the reflected light by moving red blood cells within
surface capillaries. The shift is expressed as
millivolts.

Laser doppler velocimetry offers the advantage of
continuous quantitative measurement of blood flow it
ytyg at any portion of the GI tract which can be

37’40-42 Major disadvantages in

exposed at surgery.
using this instrument on the gastrointestinal tract are
that peristalsis may limit accurate optical coupling

between the probe and bowel surface and intermittent

11

compression of the bowel wall by excess pressure
applied to the probe can drastically alter local blood
flow measurements in the bowel wall.43’44
Measurement of intramural pH (pHi) via the Khuri pH
probe offers many of the same advantages as LDV. The
hydrogen ion electrode and thermistor wire can be
inserted at any portion of the gastrointestinal tract
and allow continuous determination of pHi £2 ytyg.
This technique has a major advantage over LDV in that
it supplies quantitative data which is observer
independent, thereby increasing its overall accuracy.
This stability was not observed with use of LDV in our
study. Although blood flow is not measured with the pH
probe, the effects of ischemia on tissue metabolism are
directly measured, thus allowing prediction of
subsequent viability. [The Khuri tissue pH probe used
in this study is comprised of a silver chloride, leaded
glass coated, hydrogen ion electrode and thermistor
probe inserted adjacently into the seromuscular layer
of the colon. A remote potassium chloride salt bridge
is then placed subcutaneously as an electrical
reference point. Electrical potential difference is
then transmitted to a computer which calculates the

local tissue pH using the Nernst equation:

12

pH=E-EG(F)/RT

 

Where E = tissue electrical potential, E0:
observed potential, F=Faraday's constant, R = constant,
and T = absolute temperature.

The instrument has been shown to have a 95%
response time less than 15 seconds and drift less than
.01 pH units over a six hour period.47
4) Review of Effect of Colostomy on Anastomotic

Healing

Narrowing or stricture is a recognized complication

of circular stapled large bowel anastomoses.28

Several reports have associated a marked increase in
the incidence of anastomotic stenosis with the circular
staple technique following proximal diversion.28’33'35
It is generally felt that the absence of the fecal
stream for an extended period of time plays a prominent
pathophysiologic role in stenosis. This could result
from the loss of normal dilating effect of the fecal

35

bolus. Others have proposed an absence of short

chain fatty acids which are the primary fuel source for

. . . 48
colonic mucosa, as a potential mechanism.

Additionally, Buchman demonstrated an increased amount
of submucosal fibrosis in diverted stapled anastomoses

when compared to a two—layer, hand sewn technique.34

13

A major flaw in many of these studies is a small number
of subjects as well as a lack of uniformity for the
indications for proximal diversion. The lack of
uniformity with regards to indications for proximal
diversion may be very important, as the reason for
diversion in the clinical study (i.e. abscess, multiple
trauma, or shock) may be a more important factor in
inducing anastomotic stenosis than the surgical
technique utilized.

5) Review of the Effects of Therapeutic Radiotherapy

on Colorectal Anastomotic Healing

 

Preoperative adjuvant radiation therapy in the
range of 4-5000 rads, has demonstrated significant
beneficial effects in rectal cancer patients.48-50
These data have demonstrated tumor shrinkage or
ablation, a decrease in the incidence of metastatic
local/regional lymph nodes, and a decrease in local
recurrence rates. Additionally, it should be noted
that current experience has shown these dosages to be
well tolerated by patients. Widespread acceptance of
this adjuvant treatment modality has been slow over
concern of adverse affects of radiation therapy when

resection and primary anastomosis is considered. A

number of studies have attempted to define the untoward

14

effects of preoperative radiation therapy on
anastomotic healing. Bubrick and Shauer experienced a
20-30% leak rate for low anterior anastomoses performed
in dogs following preoperative doses of 2,000 and

51

4,000cGy's. They noted that the leak rate

increased dramatically to 70-80% however for animals

52 Thus, they

who received 6,000cGy preoperatively.
recommended that dosages less than 4,000cGy be employed
to minimize complications.

The pathophysiology of anastomotic healing
complications induced by radiotherapy has not been well
studied.

It is clear, however, that 5—10% of patients who
receive pelvic radiation will develop significant

56’57 One mechanism for impaired

sequelae.
anastomotic healing, following radiation therapy,
appears to be direct cellular injury and death with
healing occurring by fibrosis. Crowley demonstrated a
significant early inflammatory reaction with loss of
epithelial cells, reduction of villous height, and

59 He noted, however, that by 22

submucosal edema.
days these effects had subsided almost entirely, thus
the reason for the current recommendation of a three

week rest period prior to surgical intervention.

15

Although histologic studies have implicated endothelial
and connective tissue damage as the hallmark of late
radiation injury, the key factor may be a failure or
depletion of stromal stem cells required for adequate

healing of the involved tissue.65

This is modeled by
the subsequent bone marrow failure which occurs after
repeated exposures to either radiation therapy or cyto
toxic drugs in lymphoreticular disorders.

It was this hypothesis of repeated stromal stem

cell injury which led us to investigate ATP-MgCl2 as

a cytoprotective agent in radiotherapy. ATP-MgCl

2
has demonstrated ability to accelerate recovery in a
model of post ischemic renal failure.62 Similarly,

ATP-MgCl2 has demonstrated preservation of liver and

cardiac function following exposure to low flow

states.6l’63’64

Finally, it has been used in an in
vitro model of cyclosporine renal toxicity and demon—
strated a blunting of the renal dysfunction resulting
from exposure to high dosages of this drug.64 Black
evaluated histologic changes following rectal
irradiation in rats and identified a chronic,
progressive increase in mural fibrosis and vascular
sclerosis up to one year following radiation exposure,

57

either early or late. This progressive vascular

16

sclerosis and endarteritis has been invoked as the
cause of mural ischemia, resulting in the subsequent
increase in anastomotic complications. Although this
progressive mural ischemia has often been mentioned in
the surgical literature, there has never been an
attempt to quantify the magnitude of mural ischemia
induced by therapeutic radiation exposure, or

quantitate interventions to reverse the damage.

IC. AIMS OF THIS PROJECT

 

1) In particular, to assess healing in normal
colorectum and under several adverse situations
encountered in clinical medicine.

2) Serially compare colorectal anastomotic healing
with the standard Czerny—Lambert (CL; two layer;
suture; inverting) or EEA (CS; double row staples;
inverting) technique a porcine model.

3) Healing was assessed at 5, 11, 60, or 120 days
postoperatively to evaluate the three phases of wound
healing (lag, collagen synthesis, and collagen
remodeling). Parameters assessed include: a)
perianastomotic blood flow; b) bursting strength; c)
gross inflammatory score; d) microscopic inflammatory
score; and e) hydroxyproline content.

4) Compare colorectal anastomotic healing for CL
and CS techniques with or without proximal fecal
diversion (colostomy).

5) Compare colorectal anastomotic healing for CL
and CS techniques in the presence of ischemia.

6) Assess the accuracy of LDV and pHi to accurately
quantitate critical levels of ischemia at the
colorectal anastomosis and predict subsequent tissue

healing.

17

18

7) Compare colorectal anastomotic healing for CL
and CS techniques following 4500 cGy of preoperative
pelvis radiotherapy.

8) Assess the affects of ATP—MgCl2 as a

cytoprotective agent during preoperative pelvic

radiotherapy.

II. MATERIALS AND METHOD

 

A. Animals/Anesthesia

 

The model selected was female mixed breed pigs, age
three to five months old. This animal is readily
accessible, relatively inexpensive, and the distal
colon resembles the human rectosigmoid anatomically
with a circumferential circular and longitudinal muscle
layer. Our laboratory is well equipped and experienced
in porcine anesthesia and perioperative care.
Preoperative mechanical bowel preparation consisted of
8oz. of magnesium citrate in the drinking water the day
prior to surgery and cleansing tap water enemas
immediately prior to operation. Perioperatively they
received three intramuscular doses of Penicillin G
(lOOU/kg) and Gentamicin (1mg/kg) as prophylactic
antibiotic coverage administered immediately prior to
surgery, and at 6 and 12 hours postoperatively.

Preoperative sedation consisted of Ketamine
(2.5mg/kg) and sodium Pentothal (30mg/kg) administered
intramuscularly thirty to forty-five minutes prior to
surgery. Anesthesia was induced using Pentobarbital
(5mg/kg IV), and orotracheal intubation was performed
allowing inhalational anesthesia with halothane. The

animals were allowed to recover spontaneously from

19

20

anesthesia in their pens. They were allowed food and
water ad libitum postoperatively. Use of animals in
this project was in accordance with the National

Institute of Health Publications Principles For

 

Use gt Animals and Guide for the Care

 

and Use gt Laboratory Animals. {Office for

 

Protection from Research Risks, National Institutes of
Health, Bethesda, MD 20205}. The Animal Research
Review Committee of Michigan State University monitored

this project in routine fashion.

B. Experimental Groups

The study was performed in five sections. Section
I compared the Czerny—Lembert (CL) anastomotic
technique with EEA stapled technique (CS) in control
animals. Section II evaluated the effects of proximal
diverting colostomy (COL) to non—diversion (CON) using
the same two anastomotic techniques. Section III
evaluated the effects of ischemia on anastomotic
healing in CL and CS anastomoses and prospectively
compared the accuracy of laser doppler velocimetry
(LDV) and tissue pH (pHi) in quantitating the effects
of anastomotic ischemia. Section IV evaluated the

effects of preoperative pelvic radiation therapy (RT)

21

on anastomotic healing and CL or CS anastomoses. In

Section V, ATP-MgCl was infused during RT treatment

2
sessions in two subgroups (ATP-30: 30umol/kg; ATP-60:
60umol/kg) in an attempt to diminish RT induced side

effects.

1) Section I

 

In Section I (see Figure 1), the animals were
randomly allocated to either CL (N=21) technique or CS
(N=21) technique. The animals underwent a second
surgery on postoperative day 3 (N=10), day 5 (N=10),
day 11 (N=10), day 60 (N=6), or day 120 (N=6) to assess
the healing process during the well documented three
stages of wound healing. At the initial surgery, laser
doppler velocimetry (LDV) readings were obtained 1cm
proximal, at, and 1cm distal to the anastomosis.
Readings were obtained at the mesenteric, anti—
mesenteric and right and left aspects of the bowel at
these three locations. The four values were then
averaged for each level. At second surgery, assessment
of the anastomosis consisted of repeat laser doppler
velocimetry (same locations), bursting pressure, gross
inflammatory or microscopic inflammatory, and

hydroxyproline determination. In addition, the 60 and

 

22

120 day animals underwent preoperative barium enema to

assess the degree of anastomotic stenosis.

23

FIGURE 1: Section I, Protocol Outline
N=42
Handsewn (CL) Circular Stapled (CS)
(N=21) (N=21)

\ /

LDV readings

t

Second surgery (Day 60 and 120 with
preoperative barium enema)

2’ L1 L \4 \1
Day 3 Day 5 Day 11 Day 60 Day 120
(N=10) (N=10) (N=10) (N=6) (N=6)

\J \l/ J/ 4/ L/

1) Repeat LDV readings

2) Gross Inflammatory Scoring

3) Bursting Pressure

4) Specimens for Hydroxyproline and
histopathologic evaluation

24

2. Section II

Section II (Figure 2) consisted of animals randomly
assigned to diverting colostomy (COL, N=24) or
nondiversion (CONT N=28) groups. Diverting colostomy
was performed by dividing the bowel 20cm proximal to
the anastomosis oversewing the distal end (leaving it
intraabdominally) and bringing the proximal bowel to
surface as a colostomy. Animals were also randomized
to CL or CS technique with second surgery performed at
5, 11, 60 or 120 days. Similar data to Section I was

collected except for microscopic inflammatory scoring.

FIGURE 2: Section II, Protocol Outline
N=52
</ \1
Control Colostomy
N=28 N=24
if” Bi éf/ \El
CL CS CL CS
N=14 N=14 N=12 N=12
\1 \l/ (LDV) 3/ Z

Second Surgery
(5, 11, 60, 120 Days)
All with preop barium enema

1) Repeat LDV reading

2) Gross Inflammatory Scoring
3) Bursting Pressure

4) Specimens for Hydroxyproline

25

3. Section III

 

Section III (Figure 3) was performed in two
segments. First a pilot group Phase I, (N=8) in which
variable portions of the mesentery were divided
proximal and distal to the anastomosis to cause
devascularization sufficient to induce a color change
in the anastomotic segment. Subsequent anastomotic
outcome was retrospectively compared to the
quantitative values obtained by laser doppler
velocimetry (LDV) and tissue pH (pHi) determination to
allow identification of quantitative values reflecting
critical levels of ischemia. The ability of these
critical values to accurately predict anastomotic
healing outcome (good or bad) was tested in a
prospective manner (Phase II) using 24 animals
allocated evenly to either CL or CS technique. Again,
follow up in the prospective portion of section III was
performed at 5, 11, 60 or 120 days. The purpose of
this section of the study was to define the accuracy of
the quantitative data in identifying critical levels of
anastomotic ischemia, and to use this level of ischemia

to develop a reproducible model of anastomotic stenosis.

26

FIGURE 3: Section III, Protocol Outline
N=32
4/ 8
Phase I Phase II
=8 =24
M’ ‘S
CL cs CL / N cs
=4 N=4 N=12 N=12
Mesenteric Mesenteric
devascularization devascularization
W V

LDV Readings
pHi Readings

W

Second Surgery
Day 5
1) Preop barium enema
2) Repeat LDV
3) Bursting pressure
4) Gross Inflammatory
Scoring

(This portion allowed
identification of "critical"
quantitative data reflecting
levels of ischemia which
would impair healing.)

LDV Readings
pHi Readings

\l/

Second Surgery
Day 5,11,60, or 120
1) Preop barium enema
2) Repeat LDV
3) Bursting pressure
4) Gross Inflammatory
Scoring

(Tested accuracy of
"critical" levels to
predictive good or
poor anastomotic
healing.)

27

4. Section IV

 

Section IV (Figure 4) consisted of animals (N=24)
receiving 4500cGy of preoperative RT over five weeks
followed by a three week rest period. The pelvic
irradiation was administered in 425cGy fractions in a
right to left then left to right fashion. A 300KV,
20mA orthovoltage x-ray unit was used with a tube to
skin distance of 50cm. The radiation dose delivered
was measured using a Radicon 3 with 550-5 probe which
was inserted 20cm above the dentate line. The
radiation was administered through a 7cm by 15cm
portal. This technique assured that the total dose
administered to the animals over the course of
treatment was 4500cGy. This technique of radiation
administration is very analogous to the treatment
schedules used in clinical medicine and avoided the use
of the Normal Standard Dose (NSD) formula commonly used
in experimental protocols. This is especially
important because recent work has demonstrated the lack
of validity for the NSD formula when attempting to
study chronic radiation changes.53 The animals were
then randomly allocated to either CL or CS technique.
Similar data to Section I was collected except for

microscopic inflammatory scoring.

28

FIGURE 4: Section IV, Protocol Outline

4500 cGy preoperative
RT over 5 weeks

1/ \
Handsewn (CL) Circular
N=24 Stapled (CS)
N=24
\ /

LDV readings

\V
Second surgery (Day 60 and 120 with
preoperative barium enema)

Z J/ J, \1
Day 5 Day 11 Day 60 Day 120
(N=12) (N=12) (N=12) (N=12)
\1 ‘1’ 4r ‘5

1) Repeat LDV readings

2) Gross Inflammatory Scoring
3) Bursting Pressure

4) Specimens for Hydroxyproline

29

5. Section V

Animals (N=24) in this section (Figure 5) received
4,500cGy of preoperative RT as in Section IV. However,
12 animals (ATP-30) received intravenous infusions of

30umol/kg ATP-MgCl over two hours during each RT

2
session. The other group of 12 animals (ATP—60)

received 60umol/kg of ATP-MgCl over four hours

2
during each RT session. Infusion rates were selected
based on earlier pilot studies to identify a maximal
rate of infusion which avoided hemodynamic
alterations. Postoperative evaluation was limited to

Day 5 and Day 11 with similar data to Section IV being

collected.

30

FIGURE 5: Section V, Protocol Outline
4500cGy Preop RT 4500cGy Preop RT
A(30umoles/kg/RT treatment) B(60umoles/kg/RT treatment
N=12 N=12
\4 )1 t/ N
CL CS CL CS
=6 N=6 N=6 =6
53 V’ dz :4
Second Surgery at Day 5 and 11
\L

1) Repeat LDV readings

2) Gross Inflammatory Scoring

3) Bursting Pressure

4) Specimens for Hydroxyproline

5) Preoperative barium enema for
stenosis or leak

RT = Radiation Therapy
CL = Czerny-Lembert (Handsewn)
CS = Circular Stapled

C. OPERATIVE TECHNIQUE

1) First Surgery

The animals were anesthetized, endotracheally
intubated, placed in the supine position and surface
hair was clipped. An abdominal scrub with Betadine
soap followed by Betadine solution was performed. A
vertical midline incision allowed entry into the
abdominal cavity. The distal colon was mobilized at
the pelvic brim with resection of a 4-5cm segment of
colorectum. A primary anastomosis was then performed
using either the Czerny-Lembert (CL; two layer hand
sewn) technique or EEA (CS) stapled technique. For the
CL technique, an inner layer of running 3—0 chromic was
placed using full thickness bites of the bowel wall
circumferentially, followed by an interrupted outer
layer of 3-0 silk seromuscular Lambert sutures. The
EEA stapled anastomoses (CS) were performed using
either a 25 or 28mm stapler head depending on the
luminal diameter of the bowel. A 2—0 polypropylene
pursestring suture was placed at the opening in the
proximal and distal bowel. The stapler was then
inserted transanally and the anvil extended out of the
distal colon. The proximal bowel was then placed over

the anvil and the pursestring suture was tied securely

31

32

(Fig. 6a and b). Similarly the distal pursestring was
tied securely. The stapler head was approximated and
fired completing the anastomosis. The presence of a
complete ring or "doughnut" of colonic wall was
required from both proximal and distal colon to assure
accurate firing of the stapler. Leak testing was then
done following both techniques by filling the pelvis
with saline and injecting air transanally while
observing for bubbles as evidence of leak. If a leak
was identified the location was recorded and it was
repaired with interrupted 3-0 silk seromuscular
sutures. Laser doppler velocimetry readings were then
obtained 1cm proximal, at, and 1cm distal to the
anastomosis. The abdomen was then thoroughly irrigated
with saline and the fascia was approximated in a single
layer using running #2 polypropylene suture. The skin
was approximated using metal staples. The animals were
returned to their cages and allowed to recovery

spontaneously from anesthesia.

2) Second Surgery

 

The midline laparotomy incision was reopened
following general anesthesia. The anastomotic segment

was carefully exposed to allow repeat laser doppler

33

 

.t...‘.. L.

FIGURE 6A: Circular Stapling Technique

 

FIGURE GB: Circular Stapling Technique

 

34

velocimetry readings and gross inflammatory scoring.
The segment was then resected allowing at least 5cm of
bowel proximal and distal to the anastomosis. The
animal was euthanized using 0.5ml/kg IV of T61 solution

immediately prior to removal of the anastomotic segment.

D. ANALYTICAL TECHNIQUES

1) Laser Doppler Velocimetry

Laser doppler velocimetry (LDV) was utilized as a
means of quantifying seromuscular blood flow proximal,
at and distal to the anastomosis. Seromuscular blood
flow was measured because these are the layers which
provide the majority of the tensile strength of the
anastomosis and figure prominently in the healing
process. The instrument used was the MedPacific 5000
Laser Doppler Velocimeter. This is a helium-neon Laser
Doppler Velocimeter using a monochromatic light source
with a wave length of 632.8mm. The light penetrates up
to 1mm and is reflected back to a photo diode via a
second set of fiber optics. The Doppler-shifted light,
reflected from moving red blood cells within surface
capillaries of the bowel is processed and expressed in
millivolts. The doppler shift is proportional to the

velocity of blood flow in the vessels. Readings were

35

obtained at four points proximal (1cm proximal) at, and
distal (1cm distal) to the anastomosis. The four
readings at each level (proximal, at, and distal) were
then used to obtain an "average" blood flow for that

level. All readings are presented as millivolts (mV).

2) Tissue pH

The Khuri tissue pH probe was used to measure
perianastomotic seromuscular colonic pH in both normal
and ischemic colon. The instrument consists of a
silver chloride, leaded glass, microhydrogen ion
electrode which is inserted seromuscularly. A
thermistor wire is inserted seromuscularly in an
adjacent location. These data (hydrogen ion current
and temperature) are used to calculate the tissue pH
via the Nernst equation (pH=Eo-E(F/RT). A reference
salt (KCl) bridge is placed subcutaneously as an
electrical reference point. Overall the instrument has
a 95% response time of fifteen seconds with a drift of

45 A pilot

less than 0.01 pH units per six hours.
study was performed as part of Section III to determine
normal and critical pH values induced by ischemia.

These critical values were then tested prospectively in

36

the second portion of Section III of this project. All

readings are presented as pH units.

3)Bursting Pressures

 

Bursting pressure was determined by fixing an 18
French foley catheter in the proximal and distal end of
the anastomotic segment. The catheters were secured in
place by means of a pursestring suture at the ends of
the bowel. One of the catheters was attached to an air
inflow which allowed air to be insufflated at 1.0
liter/min. The second catheter was attached to a
pressure transducer and a physiograph machine recorded
pressure changes. The site and pressure (mmHg) of the

burst was recorded for each anastomotic segment.

4) Hydroxyproline Assay

The assay utilized was defined by Woessner and is a
colorimetric method of assessing hydroxyproline tissue
content.68 The technique is briefly described
below. The tissue is hydrolyzed in 6N HCl at 60°C (lml
HCl/lOmg tissue), following which the pH is raised to
6—7 using 2N NaOH. The sample is then brought to a 2ml

volume by adding the appropriate amount of buffer

solution as necessary. One milliliter of Chloramine T

37

is then added. Twenty minutes later lml of
aldehyde/perchloric acid is added to the solution which
is warmed to 60°C for fifteen minutes. Absorbance is
measured using a spectrophotometer set at 550nm.
Standard solutions were prepared at 0.1mg/ml, 0.3mg/ml
and 0.5mg/ml to allow construction of a standard curve
(linear relationship) on which to plot the results of
the tissue samples and allow calculation of the
quantity of hydroxyproline present. Data is presented

as mg/ml.

5) Gross Inflammatory Grading

A 24 point scale was used to grade the degree of
serosal and mucosal inflammation at the time of second
surgery. The degree of serosal inflammation,
hemorrhage, edema, and adhesion formation were each
graded on a three point scale. Similarly the amount of
mucosal inflammation, necrosis, hemorrhage, and edema
was recorded. Thus a maximum score of 24 points was
available with 12 points contributed by the serosal
reaction and 12 points contributed by the mucosal
reaction. Grading was performed by the surgeon at the

time of second laparotomy.

38

6) Barium enema examination for stenosis or leak

Barium enemas were performed immediately prior to
second surgery for animals at the 60 and 120 day
intervals in all groups. In the ischemic study
(Section III) all animals had studies performed prior
to second surgery to identify leak or stenosis. All
animals receiving preoperative RT (Section IV and V)
had barium studies. From the barium study, an
anastomotic index was calculated by dividing the inner
diameter at the anastomosis by the sum of the inner
diameter of the bowel 3cm proximal and distal to the
anastomotic line. A non—stenotic anastomosis would
have an index of 0.5. Values less than 0.3 were
considered stenotic. Leaks if present were also

recorded, as well as their location.

7) Microscopic Inflammatory Score

A histologic grading scale was utilized by Dr.
Robert Dunston of the Michigan State University,
College of Veterinary Medicine, Department of
Pathology, to assess the degree of inflammatory
reaction at the anastomosis. Factors graded on a three
point scale were: 1) erosion/ulceration; 2) necrosis;

3) congestion; 4) fibrosis; 5) inflammation; 6) PMN

39

infiltration; 7) lymphocyte infiltration; 8) macrophage
infiltration; 9) eosinophil infiltration. Thus a
maximum score available was 27 points. Examination was
performed separately on the mucosa, muscularis, and
serosa. In addition, overall reaction to the material
used in the anastomosis was graded (3 points each) by
assessing: l) fibrosis; 2) inflammation; 3) PMN
infiltration; 4) lymphocyte infiltration; 5) Macrophage
infiltration; 6) eosinophil infiltration. A maximum
score available for this overall grade was 18 points.
Specimens were fixed in formalin for at least 48
hours and then embedded in paraffin from which three
sections were cut and placed on slides. Tissue was

stained with routine hematoxylin and eosin.

8)Statistical Analysis

 

Blood flow data was analyzed using Analysis of
Variance and BonFeroni's Least Significant Difference,
a posteriori test. Student's t—test was also used to
compare CL and CS LDV data at each level. Bursting
pressure, and tissue hydroxyproline levels were
analyzed with Student's t-test. Inflammatory scores

were analyzed with the rank sum test. In all cases,

4O

significance was set at p<.05. Results are expressed

as mean 1 standard deviation.

III.RESULTS

A. SECTION ONE: NORMAL COLORECTUM

1. Laser Doppler Velocimetry Results

Forty-two (42) female mixed breed pigs with an
average weight of 63.7kg (58kg—69kg) were used in this
study. The animals were randomized to CL (N=21) or CS
(N=21) technique of colorectal anastomosis. The
average perianastomotic LDV data at the initial surgery

are shown in Table 1.

TABLE 1, SECTION I: LDV Readings (mV) at Initial
Surgery (mV)

Prox At Dist
CL (N=21) 4921154 264:121* 356:156
CS (N=21) 620i221# 329il74* 546i215#

*p<.001 ANOVA (comparing Prox, At, Dist)
#p<.05 t—test (comparing CL to CS)

For each technique there was a significant
decrease in seromuscular blood flow as measured by LDV
at the anastomotic line compared to flow 1cm proximal
and distal. Additionally, seromuscular blood flow was
significantly greater 1cm proximal and 1cm distal to
the anastomotic line for the CS anastomoses compared to

CL. There was no significant difference between CL or

41

42

CS techniques with respect to LDV readings at the

anastomosis.

readings demonstrates that the decrease in flow at the

The data for perianastomotic seromuscular LDV

anastomotic line resolved by day three (Table 2).

Likewise there were no significant differences between

CL and CS techniques for Day 5-120 (see Table 2).

TABLE 2,

Post-Op Day

Day
Day
Day
Day
Day
Day
Day
Day
Day
Day

2.

The

3CL(N=5)
3CS(N=5)
5CL(N=5)
5CS(N=5)
11CL(N=5)
11CS(N=5)
60CL(N=3)
60CS(N=3)
120CL(N=3)
120CS(N=3)

CL=2 layer handsewn

SECTION I:

Prox

475i54

338i220
684i365
533i201
640i273
4621242
356i81

331i104
2901142
240:60

Bursting Pressures

 

At

353i62
230:111
707:436
315i102
5411345
407i218
213i40
228178
224i193
144i60

LDV (mV) Data Day 3-120

Dist

421i115
4081220
682i226
448i103
561i277
342i178
249i89

344i48

3201157
280181

CS=circular stapled

bursting pressures at day three were not

significantly different for the two techniques

(CL:

157i9mmHg;

CS:133i33mmHg).

The bursting pressures

43

at day five were also not significantly different

(CL:188154; CS:231157mmHg).

3. Gross and Microscopic Inflammatory

Analyzing the gross and microscopic inflammatory
scores did not reveal any significant differences
between the two techniques at any of the time periods

(Tables 3 and 4).

TABLE 3, SECTION I: Gross Inflammatory Score Day 3—120
(max=24 point)

Post—Op Day CL CS
Day 3 (N=10) 4.111.0 3.911.2
Day 5 (N=10) 5.411.3 5.21.9
Day 11 (N=6) 6.41.9 5.21.9
Day 60 (N=6) 4.21.9 4.41.5
Day 120 (N=6) 4.11.8 4.61.7

(Score is based on gross findings at the time of second
surgery with a maximum of 12 points contributed by both
the serosal and mucosal surfaces).

44

TABLE 4, SECTION I: Microscopic Inflammatory Score Day

3-120

Post—op Day Mucosa Muscularis Serosa
Day 3CL(N=5) 10.212.9 1511 1312.1
Day 3CS(N=5) 13.514.5 1411.6 14.2513.3
Day 5CL(N=5) 11.614.6 11.014.4 15.213.7
Day 5CS(N=5) 11.414.2 9.213.8 10.812.4
Day 11CL(N=5) 8.213.l 12.611.9 11.013.7
Day llCS(N=5) 11.2513.0 10.514.5 8.2513.2
Day 60CL(N=3) 7.711.5 10.012.6 7.011
Day 60CS(N=3) 10.712.5 7.712.9 4.712.l
Day 120CL(N=3) 6.711.2 6.311.2 3.712.1
Day 120CS(N=3) 9.013.5 4.012.6 5.012.0

CL=2 layer handsewn CS=circular stapled

4. Hydroxyproline Content

 

The anastomotic hydroxyproline content for each
technique at the time interval evaluated is shown below

(Table 5).

TABLE 5, SECTION I: Hydroxyproline Day 3-120 (mg/ml)

Post—op Day CL CS
Day 3 (N+10) 0.041.03 0.071.04
Day 5 (N=10) 0.141.13 0.051.04
Day 11 (N=6) 0.201.08 0.141.10
Day 60 (N=6) 0.151.09 0.201.l
Day 120 (N=6) 0.221.07 0.141.02

CL=2 layer handsewn CS=circular stapled

There was a trend for

45

more rapid gain in

hydroxyproline content for handsewn anastomoses but

this did not achieve statistical significance.

5. Anastomotic Diameter

 

Anastomotic diameters and anastomotic ratios were

similar for both CL and CS techniques (Table 6).

TABLE 6, SECTION I: Anastomotic Diameter and Ratios
for 60 and 120 Day Groups

Diameter
(mean 1 S.D.)
60 CL 4.611.4
60 CS 3.41.6
120 CL 3.611.4
120 CS 5.11.5

CL=2 layer handsewn

CS=circular stapled

Ratios
0.43
0.40
0.37
0.42

46

B. SECTION TWO: COLOSTOMY (COL) vs. CONTROL (CON:
NON-COLOSTOMY)

1. General Data

 

Fifty-two (52) mixed breed female pigs were
employed in this portion of the study. The average
weight was 65.3kg (57—69kg). The animals were randomly
allocated to colostomy (COL, N=24) or noncolostomy

control group (CON, N=28).

2. Laser Doppler Velocimetry Results

The perianastomotic blood flow as measured by LDV,
at the initial surgery was similar in both techniques
irrespective of the presence or absence of proximal

diversion (Table 7).

TABLE 7, SECTION II: Initial Perianastomotic Blood

Flow
(mV)
CL CL CL
prox at dist
Control 6271107 3301181* 5451230
Colostomy 4411180 2471122* 5221190
CS CS CS
prox at dist
Control 4921153 2691121* 3561157
Colostomy 5171218 274177* 5391170

*p<.001 ANOVA (comparing prox, at, and dist
flow; no significance between
flow at any level comparing CON
to COL).

47

The early decrease in blood flow at the anastomotic
line was identified in the colostomy and control
animals without significant difference between the two
groups. This decrease in flow resolved by day five in
these animals and there was no significant difference

subsequently with respect to anastomotic technique or

the presence or absence of proximal diversion (Table 8).

TABLE 8, SECTION II: Perianastomotic Blood Flow
(mV1SD) 5-120 day

Prox At
Post-op Day Con(N=6) Col(N=3) Con(N=6) Col(N=3)
Day 5 CL 6411344 474163 7071436 313169
Day 5 CS 5111206 4421143 3151102 4141171
Day 11 CL 6401273 5311204 5411345 395153
Day 11 CS 4541243 600162 4071218 343146
Day 60 CL 356181 333117 213140 223155
Day 60 CS 3311104 338170 228178 2641114
Day 120 CL 2901142 5461124 2241193 6371402
Day 120 CS 240160 4961127 144160 3281195
Distal

Con(N=6) Col(N=3)
Day 5 CL 6811226 3991111
Day 5 CS 4471103 485172
Day 11 CL 5611277 4661168
Day 11 CS 3421178 4731155
Day 60 CL 250189 257119
Day 60 CS 344148 281173
Day 120 CL 3201157 5801211
Day 120 CS 280181 4421252

CON=control
COL=colostomy

48

3. Bursting Pressure

 

There was no significant difference in bursting
pressures between diverted anastomoses and nondiverted
anastomoses with respect to anastomotic technique

(Table 9).

TABLE 9, SECTION II: Bursting Pressures (mmHg) Day 5

JW *4. .ur- *

CL CS
Control (N=6) 188154 231157
Colostomy(N=6) 262158 177138

4. Inflammatory Scores

 

With respect to gross inflammatory scoring, there
was no significant difference between control and

proximally diverted anastomoses (Table 10).

TABLE 10, SECTION II: Gross Inflammatory Score Day

5-120
(Max=24)

Post-Op Day Control Colostomy
Day 5 CL N=6 5.411.3 5.31l.2
Day 5 CS N=6 5.210.8 11.313.0
Day 11 CL N=6 6.410.9 410.5
Day 11 CS N=6 5.210.8 410.4
Day 60 CL N=6 4.21.08 5.813.9
Day 60 CS N=6 4.410.5 4.513.0
Day 120 CL N=6 4.110.8 6.011.7
Day 120 CS N=6 4.610.7 6.011.7

CL=2 layer handsewn CS=circular stapled

49

5. Hydroxyproline Content

 

There was no significant difference in
hydroxyproline content for anastomoses regardless of
colostomy or anastomotic technique (Table 11).

TABLE 11: Anastomotic Hydroxyproline Content
(Colostomy vs. control mg/ml).

Post—Op Day COL CON

Day 5 CS .111.04 ' .071.04
Day 5 CL .091.04 .041.03
Day 11 CS .211.05 .201.08
Day 11 CL .191.04 .141.10
Day 60 CS .121.05 .201.1

Day 60 CL .121.02 .151.09
Day 120 CS .181.9 .141.02
Day 120 CL .191.10 .221.07

6. Anastomotic Complications and Diameter

There was one anastomotic leak in the colostomy
(COL) animals compared to no leaks in control (CON)
animals. The anastomotic diameter was significantly
more narrow for colostomy stapled anastomoses than
control (non-diverted) stapled anastomoses (CON,
4.4411cm vs 2.851.62, p<.001). However, the
anastomotic ratios were similar (CON, 0.41 vs COL,
0.43). The diameter for handsewn anastomoses was also
more narrow for diverted animals (COL, 3.141.31 vs CON,

4.0911.46) which did not achieve statistical

50

significance. The ratios were again similar (0.48 Vs

0.41).

C. SECTION III

 

1. Phase I

a) General Data on Critical Quantitative Values

The criteria for impaired anastomotic healing were
leak, stenosis, or bursting pressure <100mmHg. Using
these criteria we retrospectively identified
quantitative values associated with critical levels Of
ischemia, (LDV 3200mV; pHi 57.0 pH units), (Table

12).

TABLE 12, SECTION III: Phase I, Comparison of
Anastomotic Outcome with LDV and pHi Data
in Phase I

LDV (mv) pHi
>200 <200 >7.0 <7.0
"Good" healing 5 0 5 0
Leakage or stenosis l 2 0 3

LDV=laser doppler velocimetry
pHi=intramural pH

These critical values (<200mV, <pH7.0) were tested
prospectively in phase II of the study. In phase I,
there was no significant difference identified for
either pHi or LDV readings when we compared HS to CS

techniques (Fig. 7).

51

Figure 7:

LDV 150
(millivolts)

100

50

 

52

A comparison of laser dopper velocimetry
(LDV in millivolts) and tissue pH (Khuri
pH probe) determinations in the Phase I
study. There were four animals in each
of the hand sewn (HS) and circular
stapled (CS) groups. No statistical

significance existed between groups.

UI

 

QQI-‘I-I
UI

 

 

cmuuamo‘qf‘f‘f'j'
'U
3:

 

 

 

 

 

 

 

 

 

 

HS CS

:1
on

CS

53

The average intramural pH (pHi) in these pilot
animals was 7.06 (6.6-7.40) significantly less than the
average arterial pH of 7.39 (7.23—7.48), (p<0.05,

t-test).

2. Phase II

a) General Data and Complication Rates

Thirteen animals (N=13) were randomized to the HS
technique and 14 to the CS technique. The standardized
ischemic insult utilized in this portion of the study
induced a 70% (19/27) incidence of anastomotic healing
complications. At 5 and 11 day intervals, contained
small leaks and decreased bursting pressures were
predominantly seen (ll/14). By 21 and 60 day
intervals, stenosis was the predominant healing
complication and occurred in 53% (7/13) animals. There
was no significant difference in the incidence for
healing complications between handsewn (9/13) versus CS
(8/14) anastomoses.

Overall, LDV correctly predicted subsequent
anastomotic healing outcome (Table 12) in 70% (19/27).
This is contrasted with the significantly greater

accuracy (p<0.05, Fishers exact t—test) of pHi at 93%

54

(25/27). The positive and negative predictive values

were also greater for pHi (Table 13).

TABLE 13: Predictive Accuracy of LDV and pHi, Phase II

Overall

Accuracy
LDV 19/27* (70%)
pHi 25/27 (93%)

*p<.05 Fisher's exact t—test

IV. SECTION IV

 

1. General Data

 

Twenty-four (N=24) mixed breed female pigs were
used in this portion of the study. All animals
received 4,500cGy of preoperative radiation (RT) over

four and a half weeks.

2. Laser Doppler Velocimetry Data (Controls vs
331
The initial blood flow data for the RT animals
compared to control animals demonstrated significant
decrease in seromuscular anastomotic blood flow at all

levels (Table 14).

TABLE 14: Initial Blood Flow (mV) Control vs.
Radiation Therapy (RT)

Prox At
CL CS CL CS
Control (N=24) 6441251 6271187 3451151 3291181
RT (N=24) 4091147* 3911103* 181168*# 170159*#
DIST
CL CS
Control (N=24) 5781131 5451230
RT (N=24) 382177* 375170*

*p<.05 T—test (comparing control to RT at each
level)

#p<.05 ANOVA (comparing prox, at, and distal for
each technique)

55

11,

56

The blood flow data for the second surgery at 5,

60 or 120 days revealed a persistent decrease in

blood flow at the anastomotic line in irradiated

animals when compared to controls (Table 15).

TABLE 15, SECTION IV: Blood Flow (mV) 5—120 day,

Day
Day
Day
Day
Day
Day
Day
Day

Day
Day
Day
Day
Day
Day
Day
Day

Control vs.

Post—Op Day

5CL(N=3)
5CS(N=3)

11CL(N=3)
11CS(N=3)
60CL(N=3)
60CS(N=3)
120CL(N=3)
120CS(N=3)

Post-Op Day

5CL(N=3)
5CS(N=3)

11CL(N=3)
11CS(N=3)
60CL(N=3)
60CS(N=3)
120CL(N=3)
120CS(N=3)

Prox
5791346
4681201
5691114
5791186
356180
3311103
5281145
654150

Prox
3041127
155176*
360188
3531165
2211120
225192
2901142*
240116*

Control

RT

Radiation Therapy

At
404127
341191
4421126
5251182
213140
228178
4831199
643122

At

1651100*

77127*
197139*
3261232*
110179*
171169*
2241193*
144160*

Dist
485174
3921135
4981105
4241154
250189
344133
5411176
661139

Dist
2891150
1571129*
3781124
3621260
1881130
2211127
3201157*
280181*

*p<.05 t—test (comparing control to RT at each

level in paired fashion)

CL=2 layer handsewn

CS=circular stapled

57

3. Bursting Pressure
There was no significant difference between
bursting pressures regardless of time or technique

(Table 16).

TABLE 16: Bursting Pressure (mmHg) Radiation (RT) vs.
Control, Day 5 and Day 11

Post—Op Day Radiation Control
CS 5 217156 213157
HS 5 195148 188154
CS 11 196151 188153
HS 11 201148 191143

4. Gross Inflammatory Scoring

Gross inflammatory scores were considerably higher
in irradiated animals versus controls through Day 60
but by Day 120 there was no significant difference

(Table 17).

TABLE 17: Cross Inflammatory Score (Max=24): Radiation
vs. Control

Post—Op Day Radiation Control

Day 5 CS 10.014.3* 4.211.?
Day 5 HS 10.611.2* 2.610.9
Day 11 CS ll.013.6* 2.010

Day 11 HS 10.312.1* 2.610.5
Day 60 CS 4.611.2 3.610.6
Day 60 HS 7.012.0* 2.610.6
Day 120 CS 6.011.0 6.61l.2
Day 120 HS 4.310.6 4.010

*p<.05 Rank Sum Test (comparing RT to Control)

58

5. Hydroxyproline Content

TABLE 18: Hydroxyproline Content

Post—Op Day RT Control
Day 5 CS .131.02 .071.04
Day 5 CL .111.06 .041.03
Day 11 CS .111.05 .201.08
Day 11 CL .131.03 .l41.10
Day 60 CS .101.01 .201.1

Day 60 CL .111.04 .151.09
Day 120 CS .181.06 .141.02
Day 120 CL .211.09 .221.07

6. Anastomotic Complications

There was a significant increase in anastomotic
leaks or stenosis in the RT group compared to normal
controls (RT:7/24 vs Control:2/24, p<.05 Fishers exact

test).

SECTION V

 

1. Laser Doppler Velocimetry Data (Saline, ATP—30,
ATP-60)
Comparing blood flow data at the time Of initial
surgery for saline, ATP-30, and ATP-60 animals revealed
a significant improvement in perianastomotic blood flow

in the laboratory (Table 19).

TABLE 19, SECTION V: Initial Blood Flow (mV) Saline,
ATP-30, ATP-60

Prox At Dist
Saline(N=12) 4101147 181168 382177
ATP—30(N=12) 3271126 1911129 3831191
ATP-60(N=12) 5861243* 3791152* 6011175*

ATP—30=30umol/kg/dose ATP-60=60umol/kg/dose
p<.001 ANOVA (comparing saline, ATP-30, and ATP-60)

This difference, however, did not persist at days 5 or

11 (Table 20 and 21).

TABLE 20, SECTION V: Blood Flow (mV) Day 5, Saline,
ATP—30, ATP-60

Prox At Dist
Saline(N=12) 2371114 121182 2231145
ATP—30(N=12) 186130 111146 226146
ATP-60(N=12) 226165 184146 265197

ATP-30:30umol/kg/dose ATP—60:60umol/kg/dose

59

60

TABLE 21, SECTION V: Blood Flow (mV) Day 11 Saline,
ATP—30, ATP-60

Prox At Dist
Saline(N=12) 4031183 2621164 3671186
ATP-30(N=12) 3111118 2231177 2971155
ATP—60(N=12) 215182 146179 238166

ATP-30=30umol/kg/dose ATP-60:60umol/kg/dose

2. Gross Inflammatory Scoring
Gross inflammatory scoring was also significantly
less at second surgery in the ATP-60 group of animals

(Table 22).

TABLE 22, SECTION IV: Gross Inflammatory Score (Max=24)
Saline, ATP—30, ATP-60

Day 5 Day 11
Saline 10.312.9 10.712.7
ATP-30 7.212.3 6.813
ATP-60 6.811.9* 6.011.9*

ATP—30:30umol/kg/dose ATP—60=60umol/kg/dose
*p<.001 Rank Sum Test (comparing ATP—60 to saline)

3. Incidence of Cutaneous Lesions in Radiotherapy
Portals (Saline, ATP—30, ATP-60)

Additionally, the incidence of cutaneous side sores
in the radiation portals was much less in the ATP-30
*
and ATP—60 groups (saline: 9/12; ATP-30:0/12 ;

*
ATP—60:2/12 ; p<.001 Fisher's exact test).

61

4. Hydroxyproline Data Saline, ATP-30, ATP-60
Comparing hydroxyproline data at days 5 and 11
revealed a significant increase in hydroxyproline
content at the anastomotic line at day 5 and 11 for the
ATP—2 group compared to the other two groups (Table 23).

TABLE 23, SECTION V: Hydroxyproline (mg/ml) Day 5 and
11 (Saline, ATP—30, ATP—60)

Day 5 Day 11
Saline 0.121.05 .121.04
ATP-30 .061.02 .111.05
ATP—60 .241.05* .201.03*

ATP-30=30umol/kg/dose ATP-60=60umol/kg/dose
*p<.05 ANOVA (comparing saline, ATP—30, and ATP-60)

5. Bursting Pressure
There was no significant difference in bursting
pressures between the three groups of animals (Table

24).

TABLE 24: Bursting Pressure (mmHg) Saline, ATP-30,

ATP-60
Post-Op
Day Saline ATP—30 ATP-60
CS 5 217156.2 198146.3 186147.3
HS 5 195147.5 187149.l 189151.4
CS 11 196151.3 210150.3 200149.1
HS 11 201147.8 221153.2 202153.7

62

6. Incidence of Leak or Stenosis
There was no significant difference in the
incidence of leak or stenosis between the three groups

of animals (Saline:3/12; ATP—30:4/12; ATP-60:5/12).

IV. DISCUSSION

 

A. General Discussion of Colorectal Anastomotic

 

Healing

The incidence of colorectal cancer continues to
increase with over 43,000 cases predicted in 1988.58
The majority of these patients will undergo colorectal
resection of the tumor and primary anastomosis. In
addition, approximately 112,000 patients underwent
colorectal resection and anastomosis for inflammatory
bowel disease or complications of diverticular disease
in 1983 alone.63 These facts, coupled with the high
incidence (2-65%) of anastomotic complications,
emphasize the need for greater understanding of wound

28-30,32—35
In

healing in the colorectal anastomosis.
particular, a careful study of the patterns of healing
in a mechanically stapled anastomosis (CS) versus the
standard two—layer hand sutured technique (CL) may
allow identification of attributes of each technique
for use in specific situations.

The colorectum was chosen as the site to evaluate
the CL and CS techniques because of the relatively high
incidence of leak and stenosis associated with this
location compared to other areas of the GI

tract.32—35 The colorectum is an especially hostile

63

64

environment for anastomotic wound healing because it is
a watershed area for blood flow, as there is a
transition from visceral arterial supply to systemic
arterial supply. Another risk factor is performance of
anastomoses at or below the peritoneal reflection, as
the absence of the serosa may affect anastomotic
healing.29 In addition, unique clinical problems in
the region greatly increase the hazards of anastomotic
healing such as diverticular abscess, inflammatory
bowel disease and neoplasms. Therefore, the study was
designed to allow comparison between the two
anastomotic techniques in normal colorectum, and in
situations which are potential causes of adverse
anastomotic healing.

Wound healing in the gastrointestinal tract follows
the same natural history as wounds elsewhere in the
body, with three stages: 1) lag phase (0-5 days); 2)
period of rapid collagen synthesis (5-17 days); 3)
remodeling phase (17 days-2 years).5 The lag phase
is dominated by an inflammatory response during which
necrotic tissue is removed.

At approximately day 5 the rate of collagen
synthesis begins to increase dramatically and this

continues at a high rate for approximately 17 days.5

65

During this period, tensile strength increases
correspondingly. However, the ability of the
fibroblast to produce collagen is dependent on delivery
of nutrients, especially oxygen. Initially, oxygen is
delivered by diffusion but over a 3—5 day period,
neovascular in-growth allows return to more normal
levels of blood flow and nutrient delivery.

The period of collagen remodeling begins at
approximately day 17 with collagen cross linking
accounting for further increases in wound tensile
strength.5 During this time period, an equilibrium
is reached between collagen production and collagen
degradation so that no net increase in collagen content
occurs after this time.

Because surgical technique impacts heavily on the
degree of inflammation and subsequent pattern of
anastomotic healing, it is important to elucidate any
difference in healing between a two-layer handsewn and
circular stapled anastomosis. A potential advantage
for the stapled anastomosis is that it is formed using
relatively inert metal staples which theoretically
cause minimal inflammatory reaction compared to the

suture material.4

66

B. Healing in the Normal Colorectum

 

The present study (Section I) is the first to
directly compare an inverted two-layer handsewn
anastomosis (CL) with an inverted circular stapled (CS)
technique. We were unable to identify any significant
differences in the pattern of healing for these two
techniques. Unexpectedly, gross and microscopic
grading of inflammation at the anastomosis was similar
at the intervals assessed (days 3, 5, 11, 60, or 120,
Table 3, page 40 and Table 4, page 41). When we
specifically examined the degree of mucosal ulceration
and subsequent submucosal fibrosis, there was no
significant difference between these two techniques.
Additionally, bursting strength (page 41) and
hydroxyproline content (Table 5, page 41) were similar
for the two techniques. Therefore, it appears that
both anastomotic techniques cause similar degrees of
anastomotic inflammation without significant effect on
the pattern or quality of healing. In fact,based on
these data the pattern of both CL and CS techniques are
identical. Initially there is an inflammatory reaction
and sloughing of the inverted anastomotic line, leading

to ulceration, which then heals by epithelial ingrowth

67

(second intention healing) irrespective of the
technique.

These results are contrary to those reported by
Ravitch gt _t who described primary healing
(initial mucosal apposition without ulceration) along
the sutured anastomotic line.14 A significant
difference exists between the stapling techniques used
in Ravitch's work compared to the present study.
Ravitch used a TA stapler to create an everted staple
line thus providing mucosal apposition at the time the
anastomoses is created and allowing healing by first
intention. The inversion created by both the EEA
stapler and 2 layer HS technique create identical
pathophysiologic changes at the anastomotic line

resulting in ulceration of the mucosa at the

anastomosis requiring healing by second intention.

C. Anastomotic Ischemia

Another factor which can affect early wound healing
is the relative ischemia at the anastomosis. The
anastomotic line is dependent on diffusional delivery
of nutrients, and in particular, oxygen, to survive
until neovascularization can occur. The LDV data

(Table 1, page 38) confirms the fact that microvascular

68

blood flow is diminished at the anastomotic line and
that this decrease is similar with either CL or CS
techniques. This is in accordance with the findings of
Chung gt a; who identified acute decreases in

mucosal blood at the anastomosis with either stapling

41’42 An interesting finding was the

or suturing.
significant decrease in the blood flow 1cm proximal and
1cm distal to the anastomosis for the CL technique when
compared to CS (Table 1, page 38). This may reflect a
greater degree of adjacent tissue trauma due to
handling during the performance of a handsewn
anastomosis. In a precarious situation, stapling may
provide an edge with respect to preserving blood flow
locally at the anastomosis by minimizing tissue trauma
due to handling during surgery. We did note, however,
that by day 3 (Table 2, page 39) perianastomotic blood
flow had improved in both types of anastomoses without
significant differences between techniques.

Although ischemia at the anastomotic line has been
proposed as the predominant pathophysiologic mechanism
for impaired healing, it has been difficult to
quantitate clinically. Currently, it is difficult

Clinically to quantify critical levels of ischemia in

the GI tract which result in irreversible tissue

69

injury. Non-quantitative data regarding
gastrointestinal blood flow may be Obtained by gross
inspection, fluorescein injection, or doppler

41’42 Although, these techniques are

ultrasound.
readily available, the qualitative nature of the data
is inadequate to allow identification of critical
levels of tissue ischemia and thus accurate prediction
of successful anastomotic healing.

Laser doppler velocimetry offers the advantage of
continuous quantitative measurement of blood flow in
ytyg at any portion of the GI tract which can be

43'46'48 Use of this instrument

exposed at surgery.
on the gastrointestinal tract is hampered by
peristalsis which impairs accurate optical coupling
between the probe and bowel surface. Intermittent
compression of the bowel wall by excess pressure
applied to the probe can drastically alter local blood
flow measurements in the bowel wall by inducing local
ischemia followed by a reactive hyperemia.

We compared LDV with the Khuri tissue pH probe
which also provides quantitative data regarding
adequacy of tissue perfusion as the degree of tissue

acidosis correlates well with impairment of oxygen

delivery.

70

Using the Khuri pH probe we were able to predict
subsequent impairment of anastomotic healing based on
initial intraoperative tissue reading in the
perianastomotic region. The predictive accuracy (Table
13, page 50) obtained with this device (93%) was
significantly greater than was obtained with LDV
(70%). Additionally, we did not identify any
significant differences in healing of the ischemic
anastomosis between handsewn or circular stapled
techniques. Tissue pH measurements allowed us to
reproduce a consistent ischemic insult which resulted
in impaired anastomotic healing in 63% of animals
without frank gangrene of the bowel segment.

Therefore, this tool can be used to reproduce a model
of anastomotic stenosis.

Because of its ease of use, tissue pH determination
may be very helpful in a number of clinical situations
when prediction of subsequent tissue viability is
difficult and highly desirable. It could be used as an
indicator of critical levels of ischemia at a worrisome
anastomosis as may be encountered following low
anterior resection or in resections for mesenteric
arterial or venous occlusion. It could also provide

useful information following reduction of a

71

strangulated Obstruction or following revascularization
Of acute mesenteric ischemia, possibly limiting the
amount of bowel resected or avoiding an unnecessary
high risk second look laparotomy. A worrisome
colostomy could easily be monitored at the bedside
using this technique. This instrument also provides
the ability to longitudinally follow tissue pH, as the
probe could be implanted seromuscularly at an impaired
segment of intestine with the probe wire brought out
through the abdominal wall much like a pacemaker wire
used in cardiac surgery, allowing one to identify
progression or regression of intestinal ischemia at the

bedside over hours or days.

D. Effects of Preoperative Pelvic Radiotherapy
Preoperative adjuvant radiation therapy in the
range of 4-5,000 rads, as utilized in this study, has
demonstrated significant beneficial effects in rectal
cancer patients.48~52 These data have demonstrated

tumor shrinkage or ablation, a decrease in the
incidence of metastatic disease local/regional lymph
nodes, and a decrease in local recurrence rates.

Additionally, it should be noted that current

experience has shown these dosages to be well tolerated

72

by patients. Widespread acceptance of this adjuvant
treatment modality has been slowed over concerns of
increasing anastomotic complications.

The pathophysiology of anastomotic healing
complications induced by radiotherapy is not well
studied. One mechanism for impaired anastomotic
healing, following radiation therapy, appears to be
direct cellular injury and death with healing occurring
by fibrosis. Crowley demonstrated a significant early
inflammatory reaction with loss of epithelial cells,
reduction of villous height, and submucosal edema.59
He noted, however, that by 22 days these effects had
subsided almost entirely, thus the reason for the
current recommendation of a three week rest period
prior to surgical intervention. The results of this
study indicated a significant increase in anastomotic
inflammatory reaction at the 5 and 11 day intervals for
irradiated animals (Table 17). Therefore, it is
demonstrated that inflammatory reaction is increased at
a time when one would assume that the majority of the
effects of radiation exposure had subsided.

Although progressive mural ischemia has often been
mentioned in the surgical literature, there has never

been an attempt to quantify the magnitude of mural

73

ischemia induced by therapeutic radiation exposure.
Using laser doppler velocimetry we were able to confirm
and quantify a significant decrease in seromuscular
blood flow which occurred as early as four weeks
following completion of the radiation treatment regimen
(Table 14, page 51). Again, it should be recalled that
this was identified at a time that the majority of the
inflammatory reaction induced by radiation exposure
should have subsided. We also noted that the level of
anastomotic ischemia persisted as long as four months
following exposure to the radiation protocol (Table 15,
page 52). There did not appear to be any significant
difference in perianastomotic blood flow based on the
anastomotic technique used. We were also able to
identify a significant increase in anastomotic
complications in the irradiated group of animals (RT:

7/64; Control: 2/24 p<.05 Fishers Exact Test).

E. Effects of ATP-MgCl2 on Adverse Effects of
Radiotherapy
Although a significant decrease in anastomotic
blood flow associated with preoperative RT was
identified, the mechanism remained unclear. Assuming

that the majority Of injury was due to direct cellular

74

injury to mesenchymal stem cells followed by healing in
the form of perivascular fibrosis and an endarteritis
attempts to reduce this injury with ATP-MgCl2 were
made. ATP-MgCl2 has demonstrated beneficial
cytoprotective effects during ischemic insults to the
liver, kidneys and myocardium during and following low
flow conditionael"63 Additionally, in vitro data
demonstrates its ability to diminish renal toxicity
from cis-platinum, a chemotherapeutic agent.64 We
found that administering 60umOl/kg of ATP-MgCl2
during each treatment session of radiation therapy
markedly reduced many of the measurable side effects.
Anastomotic blood flow at the initial surgery was
significantly improved (Table 19, page 55). The degree
of gross perianastomotic inflammation was diminished
(Table 22, page 56), the incidence of cutaneous lesions
was reduced, and hydroxyproline content (Table 23, page
57) was significantly greater in the 60umol/kg
ATP-MgCl2 group of animals. Therefore, these data
indicate that ATP-MgCl2 at a minimum dose of

60umol/kg diminishes many of the adverse effects of
preoperative radiation therapy in colorectal

anastomoses. This potential benefit coupled with its

exciting antineoplastic abilities to increase

75

permeability in transformed cells and to halt their
growth in the S—phase of the cell cycle may result in a
new, less dangerous combination adjuvant treatment

protocol.66-68

F. Effect of Proximal Colostomy

 

Stenosis
Stenosis or anastomotic narrowing is a well
recognized complication of circular stapled colonic
anastomoses. Several investigators have proposed that
a proximal diverting colostomy increases the risk of

28-30,32-35,45 Waxman and

anastomotic narrowing.
Ramsey using an ewe model found significant narrowing
in diverted stapled anastomoses independent of stapler
diameter.35 Similarly, Buchmann found that diverted
stapled anastomoses had greater submucosal fibrosis
than sutured anastomoses.34
In the current project the incidence of stenosis
was similar for both diverted (1 CL; O/CS) and
colostomy animals (O/CL; 1 CS). Although the stapled
colostomy animals had significantly more narrow
anastomoses (2.86cm versus 4.44cm, p<.001) there was no

significant difference between the anastomotic ratios

between diverted and nondiverted animals (0.43 versus

'. .mu"fl
I

76

0.48). Similarly, the handsewn anastomoses were more
narrow in diverted animals (3.14cm versus 4.09cm)
although this difference was not significant and again
the anastomotic rations (0.48 versus 0.40) were
similar. Most important, however, is the fact that
there was no significant difference between handsewn or
stapled anastomoses with or without fecal diversion.
Proximal diversion simply narrowed the entire length of
distal colon rather than any specific effect on the
anastomosis. Similarly, there was no significant
difference with respect to anastomotic blood flow
(Table 7 & 8, page 43—44), hydroxyproline content
(Table 11, page 46), or bursting pressures (Table 9,
page 45) with or without diversion.

Therefore, the results indicate that proximal
diverting colostomies do not inherently change the
pattern of colorectal anastomotic healing. It is more
likely, however, that the initial indication for
diversion (ie. lack of prepared bowel, local infection,
or shock state) is the etiologic factor in adverse
anastomotic healing. Specifically, proximal diversion
does not increase the risk Of poor anastomotic outcome

when the circular stapler is used to reconstitute bowel

77

continuity, as has been purported in the

literature.34'35

CONCLUSIONS

 

Circular stapled (CS) and two-layer handsewn (CL)
anastomotic techniques produced similar levels of
anastomotic ischemia as measured by laser doppler

velocimetry.

There is no difference in the incidence of
anastomotic healing complications (leak or

stenosis) between CS and CL techniques.

Healing of the mucosal aspect of CS and CL

anastomoses occurs by second intention. This work
is contrary to data Obtained by Ravitch (1967) that
stapled anastomoses healed by primary intention and

handsewn anastomoses healed by second intention.

Proximal fecal diversion did not increase the risk
of anastomotic stenosis or leak regardless of
anastomotic technique. In addition, bursting
pressures and hydroxyproline content at the
anastomosis were similar with or without fecal

diversion.

78

79

Tissue pH determination (pHi) is superior to laser
doppler velocimetry (LDV) in predicting tissue
viability following an ischemic insult based on

quantitative data.

Controlled anastomotic ischemia using pHi allows
reproduction of consistent ischemic insult to
produce a high incidence of anastomotic stenosis

without frank gangrene of the segment.

There is no significant difference in the incidence
of healing complications in the face of ischemia

between CL and CS techniques.

Preoperative external beam pelvic irradiation
(4500cGy) produced an early and persistent decrease
in seromuscular blood flow compared to

non—irradiated colon.

Preoperative radiotherapy significantly increases
the incidence of anastomotic healing complications

compared to control animals.

10.

11.

12.

13.

80

ATP—MgCl2 in doses of 30umol/kg or 60umol/kg
administered intravenously during each radiotherapy
treatment session significantly decreased the
incidence of cutaneous lesions in the radiation

portals compared to saline treated animals.

ATP-MgCl2 (60umol/kg) if used during each RT
treatment session significantly increased early

Hydroxyproline production at the anastomosis at day

5 and 11.

ATP—MgCl2 (60umol/kg) infused during each RT
treatment session increased anastomotic
seromuscular blood flow at the time of initial
surgery.

ATP-MgCl (60umol/kg) infused during each RT

2

session significantly decreased postoperative gross

inflammatory scores at day 5 and day 11.

LIST OF REFERENCES

10.

LIST OF REFERENCES

 

Travers B. An inquiry into the process of nature
in repairing injuries of the intestines. London,
Longman, 1812, pp128-75, 180-89.

Lembert A. Memoire sur l'enterographic. Rep
Gen d'Anat gt gg Physiol Pathol 1;

 

 

1826; II:101.

Connell ME. An experimental contribution looking
to an improved technique in enterorrhaphy, whereby
the number of knots is reduced to two or even one.
Medical Record 1982; 42:335—7.

 

Peacock EE Jr., VanWinkle W. "Wound Repair", 2nd
ed. WB Saunders Company, Philadelphia, 1976.

Weiss P. The Biological Foundations Of Wound
Repair. Harvey Lecture, (55)(1959-60):13, 1961.

Diegelman RF, Cohen IK, Kaplan AM. The role of
macrophages in wound repair: A review. Plast
Reconstr Surg, 1981; 68:107.

 

Madden JW, Peacock EE. Studies in the biology of
collagen during wound healing: III Dynamic
Metabolism of Scar Collagen and Remodeling of
Dermal Wounds, ADE Surg, 1971; 174:511.

Shambaugh P, Dunphy JE. Postoperative wound
infections and the use of silk: An experimental
study. Surgery, 1937; 1:379-85.

Madsen ET. An experimental and clinical evaluation
of surgical suture materials. S g Q, 1953;
97:73—80.

Lord MG, Broughton AC, Williams HTG. A
morphometric study on the effect of suturing the
submucosa of the large intestine. S g Q,

1978; 146:211-6.

81

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

82

Herrmann JB, Kelly RJ, Higgins GA. Polyglycolic
acid sutures. Laboratory and Clinical evaluation
of a new absorbable suture material. Arch Surg
1970; 100:485-90.

 

Elkins DC. Wound infection: A comparison of silk
and catgut sutures. Ann Surg 1940; 112:280-3.

Adamsons RJ, Musco F, Enquist IF. The comparative
effects of silk and catgut on collagen lysis during
the lag phase of primary healing. Surg Gynecol
Obstet 1965; 106:216-24.

Herrmann JB, Kelly RJ, Higgins GA. Polyglycolic
acid sutures. Laboratory and clinical evaluation
of a new absorbable suture material. Arch Surg
1970; 100:485-90.

 

Sako Y, Wangensteen OH. Experimental studies on
gastrointestinal anastomoses. Surg Forum 1951;
2:117-23.

Hamilton JE. Reappraisal of open intestinal
anastomoses. Ann Surg 1967; 165:917-24.

Letwin E, Williams HTG. Healing of intestinal
anastomoses. Can g Surg 1967; 10:109-16.

Bell J. Excision of portions of the intestines.
Can Med Surg J 1883; 12:1257.

 

Mellish RW, Ty TC, Keller DJ. A study of
intestinal healing. g Ped Surg 1968;
3:286-91.

Gill W, Fraser J, Carter DC, Hill R. Everted
intestinal anastomosis. Surg Gynecol Obstet
1969; 128:1297-1303.

 

Hargreaves AW, Keddin NC. Colonic anastomosis: A
clinical and experimental study. B; Q Surg
1968; 55:774-7.

Ravitch MM, Canalis F, Weinshelbaum A, McCormick
J. Studies on intestinal healing: III.
Observations in everting intestinal anastomoses.
gag Surg 1967; 166:670—80.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

83

Hultl H. II Kongress der Ungarischen Gesellschaft
fur Chirurgis, Budapest, May 1908. Pester
Med-Chir. Presse 45:108-10, 121-22, 1909

Von Petz A. Zur Technik der Magenrasaktim. Ein
neuer Magen—Darmnahapparat. Zentralbl Chir
1924; 51:179.

 

Kukushin LE. Apparatus for suture of blood
vessels. Czesalevitch gm: Experience with

clinical use of new surgical apparatus and

instruments. Musen, Meditsina, 1964.

 

Kalinim TV. Mechanical sutures for intestinal and
esophageointestinal anastomoses. Preceedings II
Congress of Kazakstan Surgeons Alma Ata, 1960.

Kalinim TV. The use of mechanical suturing for
creation of anastomoses between the rectum and the
small or large intestine. Klin Khir (Kiev)

1966; 10:56.

 

Graffner H, Fredlund P, Olsse S, Osearson J,
Petersson B. Protective colostomy in low anterior
resection of the rectum using the EEA stapling
instrument. Dis C01 and Rectum 1983;

26:87-90.

 

Reilling RD, Reilling WA, Bernie WA, Hutler AB,
Perkins NC, Elliot DW. Prospective controlled
study of gastrointestinal stapled anastomoses.
Am g Surg 1980; 139:147-152.

Scher KS, Conner CS, Jones CW, Leech RW. A
comparison of stapled and sutured anastomoses in
colonic operations. Surg Gynecol Obstet

1982; 135:489-93.

 

Sharefkin J, Jaffe N, Silen W, Fromm D.
Anastomotic adhesions after low anterior resection
of the rectum. Am g Sgtg 1981; 135:519-23.

Eliot TE, Albertazzi VJ, Danto LA. Stenosis after
stapler anastomosis. Ag g Surg 1977;
133:750-1.

33.

34.

35.

36.

37.

38.

39.

40.

41.

84

Polglase AL, Hughes ESR, McDermott FT, Pihl E,
Burke FR. A comparison of end-to-end staple and
suture colorectal anastomosis in the dog. S g
Q 1981; 152:792-6.

Buchmann P, Schneider K, Gebbers J. Fibrosis of
experimental colonic anastomosis in dogs after EEA
stapling or suturing. Dis C01 and Rectum

1983; 26:217-20.

 

Waxman BP, Ramsay AH. The effect of stapler
diameter and proximal colostomy on narrowing at
experimental circular stapled large bowel
anastomoses. Aust Ng g Surg 1986;

56:797—801.

Chung RS. Blood Flow in Colonic Anastomoses:
Effects of Stapling and Suturing. Ann Surg
1988; 206:335-339.

Chung RS, Hitch RC, Armstrong DN. The role of
tissue ischemia in the pathogenesis of anastomotic
stricture. Surgery 1988; 104:824—9.

Silverman DG, Harford WE, Cooper HS, et al.
Quantification of fluorescein distribution of
strangulated rat ileum. g Surg Egg 1983;
3:179—186.

Kvietys PR, Shepherd AP and Granger DN.
Laser-Doppler, H 0 clearance and microsphere
estimates of mucgsal blood flow. Amer g
Physiol 1985; 249:6221.

Gana TJ, Huhlewych R, Koo J. Focal gastric mucosal
blood flow by Laser-Doppler and Hydrogen gas
clearance: A comparative study. g Surg Egg

1987; 43:337—343.

Murakami M, Moriga M, Miyake T and Uchino H.
Contact electrode method in hydrogen gas clearance
technique: A new method for determination of
regional gastric mucosal blood flow in animals and
humans. Gastroenterology 1982; 82:457.

 

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

85

Semb BKH. Gastric flow measured with hydrogen
clearance technique. Scand g Gastroenterol
1979; 14:641.

 

Granger DN, Kuietys PR. Recent advances in
measurement of gastrointestinal blood flow.
Gastro 1985; 88:1073-76.

Shepherd AP, Riedel GL. Continuous measurement of
intestinal mucosal blood flow by laser doppler
velocimetry. Am g Physiol 1982; 242:6668-72.

Waxman PB. Large bowel anastomoses II: The Circular
Stapler. Brit g Surg 1983; 70:64—7.

Feld AD, Fondacaro JD, Holloway G, Jacobson ED.
Measurement of mucosal blood flow in the canine
intestine with laser doppler velocimetry. Life
Sgt 1982; 31:1509-17.

Shakri Khuri, M.D., Personal Communication.

Higgins GA Jr., Conn JH, Fordan PH Jr., Humphrey
EW, Roswitz B, Keehn RJ. Preoperative radiotherapy
for colorectal cancer. Amm Surg 1975;

181:624-31.

Roswitz B, Higgins GA Jr., Keehn RJ. Preoperative
irradiation for carcinoma of the rectum and
rectosigmoid colon: Report of a national Veterans
Administration randomized study. Cancer 1975;
35:1597—1602.

Leaming RH, Stearns MW Jr., Reddish MR.
Preoperative irradiation in rectal cancer.
Radiology 1961; 77:257—63.

Schauer RM, Bubrick MP, Ferney DA, Johnston GR,
Rolfsmeyer ES, Strom RL, Hitchcock CR. Effects of
low-dose preoperative irradiation on low anterior
anastomosis in dogs. Dis C01 and Rectum

1982; 25:401—405.

 

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

86

Blake DP, Bubrick MP, Kochsiek GG, Feeney DD,
Johnston GR, Strom RL, Hitchcock CR. Low anterior
anastomotic dehiscence following preoperative
irradiation with 6000 Rads. Dis Col and

Rectum 1984; 27:176-181.

Overgaard M, Bentzen SM, Christensen JJ, Madsen

EH. The value of the NSD formula: An equation of
acute and late radiation complications in normal
tissue following 2 and 5 fractions per week in
breast cancer patients treated with post-mastectomy
irradiation. Radiother Oncol 1987; 9:1—12.

 

Woesnner JF. The Determination of Hydroxyproline
in tissue and protein samples containing small
proportions of this imino acid. Biochem and
Bioghy 1961; 93:440-7.

 

Goligher JC, Graham NG, DeDornbul FT. Anastomotic
dehiscence after anterior resection of rectum and
sigmoid. B; Q Surg 1970; 57:109-18.

Morgenstern L, Yamakawa T, Ben-Shoshan M, Lipporan
H. Anastomotic leakage after low colonic
anastomosis: Clinical and experimental aspects.

Am g Surg 1972; 123:104—9.

Black WC, Gomez CS, Yuhas JM, Klugerman.
Quantitation of the late effects of X-radiation on
the large intestine. Cancer 1980; 45:444—51.

Silverberg E and Lubera J. Cancer statistics
1987. CA 37:2-19, 1987.

Crowley CG, Anders CJ, Nelson T, Bagshaw M. Effect
of radiation on canine intestinal anastomoses.
Arch Surg 1968; 96:423—8.

 

Black WC, Gomez LS, Yuhas JM, Kligerman MM.
Quantitation of the late effects of X—radiation on
the large intestine. Cance; 1980; 45:444-451.

Chaudry IH, Clemens HG, Ohkawa M, et a1.
Restoration of hepatocellular function and blood
flow following hepatic ischemia. Am g

Physiol 1983; 244:R695-R702.

62.

63.

64.

65.

66.

67.

68..

87

Siegel NJ, Glazium WB, Chaudry IH, et al. Enhanced
recovery from acute renal failure by the post
ischemic infusion of adenine nucleotides and
magnesium chloride in rats. Kidney Int 1980;
17:338-349.

 

McDonogh PF, Laks H, Chaudry IH, Boere AE.
Improved Myocardial Recovery from ischemia:
Treatment with low dose Adenosine
Triphosphate-Magnesium Chloride. Arch Surg
1984; 119:1379—1384.

 

Sumpio BE, Chaudry IH, Boere AE. Reduction Of the
drug induced nephrotoxicity by ATP-MgCl . 1.
Effects on the cis-Diamminodichloroplatinum treated
isolated perfused kidneys. g Surg Egg 1985;
38:429-437.

Vital and Health Statistics, 1986. National
Institutes of Health, Department of Health and
Human Services.

Rozengurt E and Heppel LA. A specific effect Of
ATP on the permeability Of transformed 3T3 cells.
Biochem and Biophys Res Communications

1975; 67(4):1581-88.

Rozengurt E, Heppel LA, Friedberg I. Effect of
exogenous ATP on the permeability properties of
transformed cultures of mouse cell lines. g
Biol Chem 1977; 252(13):4584—90.

 

Rapaport E, Fishman RF, Bercel C. Growth
inhibition of human tumor cells in soft—agar
cultures by treatment with low levels of adenosine
5-triphosphate. Cancer Res 1983; 43:4402-06.

 

APPENDICES

APPENDIX A
A. PRESENTATIONS

1. Colonic Mural Blood Flow: The Effects of
Radiation and Differing Anastomotic Techniques.

Resident Competition, 1988
Society of University Surgeons
San Antonio, Texas

Annual Meeting FASEB, 1988
Las Vegas, Nevada

2. Intramural pH: A Quantitative Measurement for
Predicting Colorectal Anastomotic Healing.

Annual Research Forum, April 1988
Michigan State University
East Lansing, Michigan

GRAMEC Annual Research Day, May 1988
Grand Rapids, Michigan

American Society Of Colon and Rectal
Surgeons, Annual Meeting, June 1988
Anaheim, California

3. Reduction of Preoperative Intestinal Radiation
Injury by ATP—Magnesium Chloride. '

Residents Program, February 1988
Society of University Surgeons

4. Does Proximal Colostomy Affect Colorectal
Anastomotic Healing?

Poster Presentation
American Society of Colon and Rectal
Surgeons, Annual Meeting, June 1989
Birmingham, England

88

89

A Direct Comparison Between Czerny-Lembert and
Circular Stapled Anastomotic Techniques in
Colorectal Anastomoses.

American Society Of Colon and Rectal
Surgeons, Annual Meeting, June 1989
Toronto, Canada

B.

APPENDIX B

PUBLICATIONS

1.

Senagore A, Milsom JW, Mazier WP, Walshaw R,
Chaudry IH. Colonic Mural Blood Flow: The
Effects of Radiation and Differing Anastomotic
Techniques. FASEB 1988; 2(4):A742.

Senagore A, Milsom JW, Chaudry IH, Walshaw RK,
Mazier WP. Intramural pH: A Quantitative
Measurement for Predicting Colorectal
Anastomotic Healing. Accepted for publication
to Diseases gt the Colon and

Rectum.

Reduction of Preoperative Intestinal Radiation
Injury by ATP-MgCl Administration.

Senagore A, Milsom JW, Walshaw RK, Mostosky
UV, Chaudry IH. FASEB 1989;3(3).

9O

"7'11?1111111111111“