EXPERIMENTAL
AND
MOLECULAR
PATHOLOGY
52, 37-45 (1990)
The Effects of Lathyrogens on Intestinal Anastomoses in the Rat K. VAN DOORN, B. DE MAN,
AND TH. HENDRIKS
Department of General Surgery, St. Radboud University Hospital, Nijmegen, The Netherlands Received May 9, 1989, and in revised form August 29, 1989 Submucosal collagen provides strength to the intestinal wall. In order to assess the importance of collagen fibers for the developing strength of intestinal anastomoses we have sought to prevent postoperative collagen crosslinking by administration of lathyrogens. Rats, receiving both an ileal and a colonic anastomosis, were treated with either Dpenicillamine or f3-aminopropionitrile from 1 day before operation. Animals were sacrificed 7 days postoperatively and bursting pressures, bursting sites, and anastomotic collagen (hydroxyproline) content and solubility were determined. n-Penicillamine, in a dose of 500 mg/kg/day and administered orally, had no effect at all. g-Aminopropionitrile, in a dose of 625 mg/kg/day and given orally or intraperitoneally, signiticantly increased the acid solubility of anastomotic hydroxyproline in both ileum and colon without affecting total hydroxyproline content or concentration. Bursting pressures of the anastomotic segments were lowered, more significantly in colon than in ileum. Also, the bursting site was found more frequently in the anastomotic area in these animals. By inhibiting the formation of crosslinks in intestinal wounds with g-aminopropionitrile, the anastomotic strength was reduced. These results demonstrate the importance of collagen in maintaining anastomotic integrity and at the same time emphasize that not only the quantity but also, and perhaps even more so, the quality of the collagen should be taken as an index of healing. 0 1990 Academic press, Inc.
INTRODUCTION The importance of collagen for the development of wound strength is generally accepted. The essential role of this connective tissue component in wound healing has been extensively investigated, in particular in skin wounds where the increase of hydroxyproline parallels the increase in tensile strength (Madden and Peacock, 1968). Also, a direct relationship is observed between collagen fiber diameter and tensile strength (Doillon et al., 1985). However, concepts derived from experiments on skin wounds are not necessarily valid for other types of wounds (Forrest, 1983). The healing of colonic anastomoses is compromised relatively frequently (Fielding et al., 1980). Disturbed healing may result in anastomotic leakage, a most serious and potentially life-threatening complication (Turunen and Peltokallio, 1983). Thus, research into the pathophysiology of intestinal anastomotic leakage is certainly warranted. So far, it has been taken for granted that in this type of wound collagen also is of major importance. Postoperative changes in collagen concentrations are taken to represent a “collagenous equilibrium,” crucial to the healing sequence (Hunt et al., 1980): immediately after operation collagenolysis occurs, followed by de nova collagen synthesis. If collagenolysis is exaggerated or if synthesis is delayed or inadequate, wound strength will diminish and chances for dehiscence will increase. Studies on anastomotic healing thus often comprise the measurement of anastomotic collagen concentration next to the measurement of a mechanical parameter (cf. Ravo, 1988, and references therein). However, 37 0014-4800/!90 $3.00 Copyright 0 1990 by Academic Press, Inc. All iigbts of reproduction in any form reserved.
38
VAN
DOORN,
DE
MAN,
AND
HENDRIKS
direct evidence for such an essential contribution of collagen in the development of anastomotic strength is still scarce. One way to assess the importance of newly formed collagen fibers for the restoration of intestinal strength would be by preventing their development. Molecular stability of collagen fibers depends on the formation of crosslinks. The mechanism of crosslinking is based on the availability of lysyl-derived aldehydes (Eyre er al., 1984), generated by the enzyme lysyl oxidase. Lathyrogens interfere with this process: B-aminopropionitrile (BAPN) inhibits lysyl oxidase activity, while D-penicillamine (n-pen) acts by chelating aldehydic functional groups (Siegel, 1979). Here, we describe the effects of administration of BAPN and D-pen on the healing of intestinal anastomoses in the rat. MATERIALS
AND METHODS
Animals Forty male Wistar rats were divided into four groups of 10 rats each. One group served as controls, and the others received o-pen orally, BAPN orally, or BAPN intraperitoneally. The animals were fed a standard diet (Hope Farms, Woerden, The Netherlands) and allowed water ad fibitum. Average body weight at the start of the experiment was 256 + 40 (SD) g and body weights were measured daily. Lathyrogens The doses of lathyrogens used in this study were chosen from the literature as being effective in influencing collagen characteristics in granulation tissue in rats (Arem et al., 1979; Junker et al., 1981). D-pen (Serva, Heidelberg, Federal Republic of Germany) was given in a dose of 500 mg/kg/day, dissolved in 2 ml water via a gastric tube. BAPN fumarate (Sigma, St. Louis, MO) was administered in a dose of 625 mg/kg/day, dissolved in 2 ml saline, either via a gastric tube or by means of an intraperitoneal injection. Animals received these compounds daily from 1 day before operation until the day of sacrifice. A pilot study had shown this dose of n-pen not to cause a reduction in weight gain in unoperated rats, while this regimen of BAPN slightly retarded the normal increase in weight but did not result in loss of weight during an 8-day course (cf. Haney et al., 1973). Operation Operations were performed as described before (de Roy van Zuidewijn et al., 1986). Briefly, after induction of anesthesia with sodium pentobarbital (0.1 ml/l00 g body wt) and a median laparotomy, 1 cm of both ileum and colon was resected at 15 cm proximal to the ileocecal junction and 3 cm proximal to the rectal-peritoneal reflection, respectively. An inverting one-layer end-to-end anastomosis was constructed microsurgically with 8 x 0 monofilament suture material (Ethicon R). After 7 days the rats were sacrificed by an intracardial overdose of sodium pentobarbital. Both anastomoses were prepared free and cleaned from surrounding tissue and a segment of approximately 5 cm containing the anastomosis was excised. After determination of the bursting pressure, l-cm samples containing the anastomosis were collected.
LATHYROGENS
AND
ANASTOMOTIC
39
HEALING
Analytical Procedures Bursting pressures and bursting sites were determined as described in an earlier communication (Hesp et al., 1984). All intestinal samples were frozen immediately after collection and stored in liquid nitrogen. Subsequently, samples were pulverized, lyophilized, and kept at - 30°C. The dry weights were measured. In all samples hydroxyproline, as a measure for collagen, was measured essentially according to Prockop and Udenfriend (1960). Hydroxyproline concentration was expressed as micrograms per milligram dry weight and hydroxyproline content as micrograms per centimeter intestine. The percentage acid-soluble collagen was assayed after suspending 10 mg sample in 1.1 ml 0.5 M acetic acid, incubating overnight at 4°C under rotation, and centrifuging for 1 hr at 42,OOOg. The supematant was lyophilized and assayed for hydroxyproline. The pellet was resuspended in 1.1 ml 0.5 M acetic acid containing 1 mg pepsin (Boehringer, Mannheim, Federal Republic of Germany) and the suspension was treated as described above. The resulting supematant contained the pepsin-soluble collagen. Statistical methods employed are mentioned with the results. RESULTS All animals lost weight after operation. However, the weight loss was limited and although the lathyrogen groups appeared to lose a little more and to take a little longer to recover as compared to the control group (Table I) these differences remained nonsignificant (Kruskal-Wallis test). Bursting pressures were determined as a measure of strength of the healing intestine. In ileum, the average bursting pressure in the control group was slightly higher-although statistically nonsignificant-than those in the lathyrogen groups (Fig. 1). In colon (Fig. 2), comparison of the four groups showed significant (P = 0.0002, Kruskal-Wallis test) differences. In particular, the difference between the control group and both BAPN groups was pronounced, the average bursting pressure in the latter being 26% lower. Values in the o-pen group were close to those observed in the controls. An 8-day course of either D-pen or BAPN did not result in changed bursting pressures in intact intestine (results not shown). During the determination of bursting pressures the leakage site was also recorded (Table II). The majority (67%) of the intestinal segments containing a 7-day-old anastomosis bursted outside the anastomosis proper. Administration of D-pen to the animals had no effect. However, administration of BAPN, either oral TABLE I Weight Loss in Control and Experimental
Groups Weight 6 days after operation
Maximal weight loss (%) Control (10) D-pen (10) BAPN-o
(8)
BAPN-ip (10) Note.
8.3 13.2 10.9 10.0
-+ 1.3 -+ 5.9 f 3.0 f 2.5
Results are given as average values f SD.
Day of minimum weight 2.2 3.9 4.8 4.5
* rt + *
0.9 2.0 1.3 0.9
Weight 1 day before operation 0.95 0.89 0.90 0.92
r f + +
0.03 0.08 0.03 0.02
40
VAN DOORN, DE MAN, Bursting
____-
pressure
SC
. 3oc
AND HENDRIKS
”
1
. .
f -+
.
269
. .
.
200
. . .
. . .
100
I
Control
1
D-pen
I
BAPN-o
1
BAPN-ip
FIG. 1. Bursting pressures recorded in the ileal anastomotic segments. Results of the individual measurements are given together with the average values 2 SD.
or intraperitoneal, did result in more leaks within the anastomotic area. This effect was particularly evident in the ileum (P = 0.018, x2 test). Hydroxyproline concentrations in the l-cm segments containing the anastomosis showed no variations between groups, either in ileum or in colon (Table III). We also calculated the hydroxyproline content in micrograms per centimeter anastomotic segment. This parameter showed a greater variation within the various groups than the hydroxyproline concentration, due to the fact that the collection of reproducible l-cm segments from intestine is very difficult as a result of variations in degree of contraction. While hydroxyproline concentrations in coionic anastomoses were clearly higher than those in ileal anastomoses, there was no such difference for hydroxyproline content. Comparison (Kruskal-Wallis test) of hydroxyproline content in the four experimental groups revealed no difference in colon but a significant (P = 0.0389) difference in ileum. This was due to the elevated values in the D-pen and BAPN-ip groups as compared to the others. However, in these groups the average hydroxyproline content in uninjured intestine, as measured in the control segment removed at operation, was also slightly elevated: 120 + 29 (SD, n = 10) and 110 + 29 (n = 10) for the o-pen and BAPNip groups, respectively, and 105 f 41 (n = 10) and 100 + 43 (n = 8) for the control and BAPN-o groups, respectively. It should be emphasized that these latter values, representing the hydroxyproline content in intact intestine expressed in mi-
LATHYROGENS
Bursting
300
AND
pressure
41
HEALING
-
.
.
3
ANASTOMOTIC
242
T ;,
= Ti
200
222
’
3’9
.
.
. . II30
l
-2
++ f
.
27
”
176
23
:
.
100
I Control
D-pen
BAPN-o
BAPN-ip
FIG. 2. Bursting pressures recorded in the colonic anastomotic segments. Results of the individual measurements are given together with the average values ? SD.
crograms per centimeter, may not be used in comparison with those in anastomotic segments in order to draw conclusions about the degree of collagen synthesis. Construction of an anastomosis per se will, by the inversion of the intestinal wall necessary to sew the divided ends together, by definition increase the volume of intestinal tissue per centimeter. Therefore, anastomotic segments should be compared only to anastomotic segments. The fact that collagen synthesis occurred around intestinal anastomoses during the first week of healing is TABLE II Leakage Site in Anastomosed Intestine Leakage site Ileum In anastomosis Outside anastomosis Colon In anastomosis Outside anastomosis Total In anastomosis Outside anastomosis
Control
D-pen
BAPN-o
BAPN-ip
4 6
1 9
7 3
7 3
1 9
3 7
4 6
4 6
5 15
4 16
11 9
11 9
42
VAN
DOORN,
DE
MAN,
AND
TABLE
Hydroxyproline
HENDRIKS
III
Concentration and Hydroxyproline
Content in Anastomotic Segments
Ileum
Colon
whvz Control D-peal
BAPN-o BAPNip
10.7 11.0 9.9 9.5
k&m
2 1.6 (10) -+ 1.9 (10) f 2.0 (8) + 1.4 (10)
248 313 262 318
2 + 2 f
69 70 60 73
i-dmfs (10) (10) (8) (10)
15.4 15.3 13.4 14.8
2 * f 2
kdcm
1.7 (9) 1.1 (10) 3.8 (8) 2.1 (10)
252 268 244 286
If: 48 2 57 f 76 + 65
(9) (10) (8) (10)
Note. Results are expressed as average values + SD and (in parentheses) number of samples.
clearly illustrated by comparison of the anastomotic hydroxyproline concentration and hydroxyproline content in 7-day-old anastomoses (Table III) with those in 3-day-old anastomoses (data of this separate control group of 10 animals have not been used further in the present paper). Here, average hydroxyproline concentration and content were, respectively, 6.9 4 1.1 (SD) pg/mg and 122 2 37 t&cm in ileum and 10.2 2 2.8 ug/mg and 184 + 42 l&cm in colon. Thus, anastomotic hydroxyproline content increased considerably between 3 and 7 days postoperatively. Acetic acid solubilized a significant amount of collagen from uninjured intestine, ileal hydroxyproline always being better solubilized than colonic hydroxyproline. For instance, in the control group an average 11.6 ? 2.5% (SD, n = 9) of total hydroxyproline in the ileal control segment could be extracted with acid as opposed to 6.4 + 1.4% (n = 10) in the colonic control segment (P = 0.0092, signed rank test). Lathyrogen treatment during 8 days did not affect these values (results not shown). Administration of BAPN had a very significant effect on the size of the acid-soluble hydroxyproline fraction in the anastomosis (Table IV). Clearly, BAPN enhanced the solubility of anastomotic hydroxyproline, both in ileum and in colon, while D-pen had no effect. On the contrary, neither lathyrogen had any effect on the amount of hydroxyproline extracted when the acetic acid residues were treated subsequently with pepsin. TABLE
Solubility of Hydroxyproline
IV
in Anastomotic Segments
Ileum HAc-soluble Control D-pen BAPN-o BAPN-ip P value
10.3 + 2.9 (10) 10.2 ” 2.5 (10) 17.5 f 5.8 (9) 15.1 2 3.5 (10) 0.0002
Colon Pepsin-soluble 6.8 f 2.1 (7) 8.9 f 2.5 (7) 7.2 f 3.2
(8) 9.2 f 3.8 (10) ns
HAc-soluble 5.9 f 1.1 (10) 6.6 f 2.2 WV 14.8 + 9.6 (9) 10.1 f 2.9 (10)
AND
MOLECULAR
PATHOLOGY
52, 37-45 (1990)
The Effects of Lathyrogens on Intestinal Anastomoses in the Rat K. VAN DOORN, B. DE MAN,
AND TH. HENDRIKS
Department of General Surgery, St. Radboud University Hospital, Nijmegen, The Netherlands Received May 9, 1989, and in revised form August 29, 1989 Submucosal collagen provides strength to the intestinal wall. In order to assess the importance of collagen fibers for the developing strength of intestinal anastomoses we have sought to prevent postoperative collagen crosslinking by administration of lathyrogens. Rats, receiving both an ileal and a colonic anastomosis, were treated with either Dpenicillamine or f3-aminopropionitrile from 1 day before operation. Animals were sacrificed 7 days postoperatively and bursting pressures, bursting sites, and anastomotic collagen (hydroxyproline) content and solubility were determined. n-Penicillamine, in a dose of 500 mg/kg/day and administered orally, had no effect at all. g-Aminopropionitrile, in a dose of 625 mg/kg/day and given orally or intraperitoneally, signiticantly increased the acid solubility of anastomotic hydroxyproline in both ileum and colon without affecting total hydroxyproline content or concentration. Bursting pressures of the anastomotic segments were lowered, more significantly in colon than in ileum. Also, the bursting site was found more frequently in the anastomotic area in these animals. By inhibiting the formation of crosslinks in intestinal wounds with g-aminopropionitrile, the anastomotic strength was reduced. These results demonstrate the importance of collagen in maintaining anastomotic integrity and at the same time emphasize that not only the quantity but also, and perhaps even more so, the quality of the collagen should be taken as an index of healing. 0 1990 Academic press, Inc.
INTRODUCTION The importance of collagen for the development of wound strength is generally accepted. The essential role of this connective tissue component in wound healing has been extensively investigated, in particular in skin wounds where the increase of hydroxyproline parallels the increase in tensile strength (Madden and Peacock, 1968). Also, a direct relationship is observed between collagen fiber diameter and tensile strength (Doillon et al., 1985). However, concepts derived from experiments on skin wounds are not necessarily valid for other types of wounds (Forrest, 1983). The healing of colonic anastomoses is compromised relatively frequently (Fielding et al., 1980). Disturbed healing may result in anastomotic leakage, a most serious and potentially life-threatening complication (Turunen and Peltokallio, 1983). Thus, research into the pathophysiology of intestinal anastomotic leakage is certainly warranted. So far, it has been taken for granted that in this type of wound collagen also is of major importance. Postoperative changes in collagen concentrations are taken to represent a “collagenous equilibrium,” crucial to the healing sequence (Hunt et al., 1980): immediately after operation collagenolysis occurs, followed by de nova collagen synthesis. If collagenolysis is exaggerated or if synthesis is delayed or inadequate, wound strength will diminish and chances for dehiscence will increase. Studies on anastomotic healing thus often comprise the measurement of anastomotic collagen concentration next to the measurement of a mechanical parameter (cf. Ravo, 1988, and references therein). However, 37 0014-4800/!90 $3.00 Copyright 0 1990 by Academic Press, Inc. All iigbts of reproduction in any form reserved.
38
VAN
DOORN,
DE
MAN,
AND
HENDRIKS
direct evidence for such an essential contribution of collagen in the development of anastomotic strength is still scarce. One way to assess the importance of newly formed collagen fibers for the restoration of intestinal strength would be by preventing their development. Molecular stability of collagen fibers depends on the formation of crosslinks. The mechanism of crosslinking is based on the availability of lysyl-derived aldehydes (Eyre er al., 1984), generated by the enzyme lysyl oxidase. Lathyrogens interfere with this process: B-aminopropionitrile (BAPN) inhibits lysyl oxidase activity, while D-penicillamine (n-pen) acts by chelating aldehydic functional groups (Siegel, 1979). Here, we describe the effects of administration of BAPN and D-pen on the healing of intestinal anastomoses in the rat. MATERIALS
AND METHODS
Animals Forty male Wistar rats were divided into four groups of 10 rats each. One group served as controls, and the others received o-pen orally, BAPN orally, or BAPN intraperitoneally. The animals were fed a standard diet (Hope Farms, Woerden, The Netherlands) and allowed water ad fibitum. Average body weight at the start of the experiment was 256 + 40 (SD) g and body weights were measured daily. Lathyrogens The doses of lathyrogens used in this study were chosen from the literature as being effective in influencing collagen characteristics in granulation tissue in rats (Arem et al., 1979; Junker et al., 1981). D-pen (Serva, Heidelberg, Federal Republic of Germany) was given in a dose of 500 mg/kg/day, dissolved in 2 ml water via a gastric tube. BAPN fumarate (Sigma, St. Louis, MO) was administered in a dose of 625 mg/kg/day, dissolved in 2 ml saline, either via a gastric tube or by means of an intraperitoneal injection. Animals received these compounds daily from 1 day before operation until the day of sacrifice. A pilot study had shown this dose of n-pen not to cause a reduction in weight gain in unoperated rats, while this regimen of BAPN slightly retarded the normal increase in weight but did not result in loss of weight during an 8-day course (cf. Haney et al., 1973). Operation Operations were performed as described before (de Roy van Zuidewijn et al., 1986). Briefly, after induction of anesthesia with sodium pentobarbital (0.1 ml/l00 g body wt) and a median laparotomy, 1 cm of both ileum and colon was resected at 15 cm proximal to the ileocecal junction and 3 cm proximal to the rectal-peritoneal reflection, respectively. An inverting one-layer end-to-end anastomosis was constructed microsurgically with 8 x 0 monofilament suture material (Ethicon R). After 7 days the rats were sacrificed by an intracardial overdose of sodium pentobarbital. Both anastomoses were prepared free and cleaned from surrounding tissue and a segment of approximately 5 cm containing the anastomosis was excised. After determination of the bursting pressure, l-cm samples containing the anastomosis were collected.
LATHYROGENS
AND
ANASTOMOTIC
39
HEALING
Analytical Procedures Bursting pressures and bursting sites were determined as described in an earlier communication (Hesp et al., 1984). All intestinal samples were frozen immediately after collection and stored in liquid nitrogen. Subsequently, samples were pulverized, lyophilized, and kept at - 30°C. The dry weights were measured. In all samples hydroxyproline, as a measure for collagen, was measured essentially according to Prockop and Udenfriend (1960). Hydroxyproline concentration was expressed as micrograms per milligram dry weight and hydroxyproline content as micrograms per centimeter intestine. The percentage acid-soluble collagen was assayed after suspending 10 mg sample in 1.1 ml 0.5 M acetic acid, incubating overnight at 4°C under rotation, and centrifuging for 1 hr at 42,OOOg. The supematant was lyophilized and assayed for hydroxyproline. The pellet was resuspended in 1.1 ml 0.5 M acetic acid containing 1 mg pepsin (Boehringer, Mannheim, Federal Republic of Germany) and the suspension was treated as described above. The resulting supematant contained the pepsin-soluble collagen. Statistical methods employed are mentioned with the results. RESULTS All animals lost weight after operation. However, the weight loss was limited and although the lathyrogen groups appeared to lose a little more and to take a little longer to recover as compared to the control group (Table I) these differences remained nonsignificant (Kruskal-Wallis test). Bursting pressures were determined as a measure of strength of the healing intestine. In ileum, the average bursting pressure in the control group was slightly higher-although statistically nonsignificant-than those in the lathyrogen groups (Fig. 1). In colon (Fig. 2), comparison of the four groups showed significant (P = 0.0002, Kruskal-Wallis test) differences. In particular, the difference between the control group and both BAPN groups was pronounced, the average bursting pressure in the latter being 26% lower. Values in the o-pen group were close to those observed in the controls. An 8-day course of either D-pen or BAPN did not result in changed bursting pressures in intact intestine (results not shown). During the determination of bursting pressures the leakage site was also recorded (Table II). The majority (67%) of the intestinal segments containing a 7-day-old anastomosis bursted outside the anastomosis proper. Administration of D-pen to the animals had no effect. However, administration of BAPN, either oral TABLE I Weight Loss in Control and Experimental
Groups Weight 6 days after operation
Maximal weight loss (%) Control (10) D-pen (10) BAPN-o
(8)
BAPN-ip (10) Note.
8.3 13.2 10.9 10.0
-+ 1.3 -+ 5.9 f 3.0 f 2.5
Results are given as average values f SD.
Day of minimum weight 2.2 3.9 4.8 4.5
* rt + *
0.9 2.0 1.3 0.9
Weight 1 day before operation 0.95 0.89 0.90 0.92
r f + +
0.03 0.08 0.03 0.02
40
VAN DOORN, DE MAN, Bursting
____-
pressure
SC
. 3oc
AND HENDRIKS
”
1
. .
f -+
.
269
. .
.
200
. . .
. . .
100
I
Control
1
D-pen
I
BAPN-o
1
BAPN-ip
FIG. 1. Bursting pressures recorded in the ileal anastomotic segments. Results of the individual measurements are given together with the average values 2 SD.
or intraperitoneal, did result in more leaks within the anastomotic area. This effect was particularly evident in the ileum (P = 0.018, x2 test). Hydroxyproline concentrations in the l-cm segments containing the anastomosis showed no variations between groups, either in ileum or in colon (Table III). We also calculated the hydroxyproline content in micrograms per centimeter anastomotic segment. This parameter showed a greater variation within the various groups than the hydroxyproline concentration, due to the fact that the collection of reproducible l-cm segments from intestine is very difficult as a result of variations in degree of contraction. While hydroxyproline concentrations in coionic anastomoses were clearly higher than those in ileal anastomoses, there was no such difference for hydroxyproline content. Comparison (Kruskal-Wallis test) of hydroxyproline content in the four experimental groups revealed no difference in colon but a significant (P = 0.0389) difference in ileum. This was due to the elevated values in the D-pen and BAPN-ip groups as compared to the others. However, in these groups the average hydroxyproline content in uninjured intestine, as measured in the control segment removed at operation, was also slightly elevated: 120 + 29 (SD, n = 10) and 110 + 29 (n = 10) for the o-pen and BAPNip groups, respectively, and 105 f 41 (n = 10) and 100 + 43 (n = 8) for the control and BAPN-o groups, respectively. It should be emphasized that these latter values, representing the hydroxyproline content in intact intestine expressed in mi-
LATHYROGENS
Bursting
300
AND
pressure
41
HEALING
-
.
.
3
ANASTOMOTIC
242
T ;,
= Ti
200
222
’
3’9
.
.
. . II30
l
-2
++ f
.
27
”
176
23
:
.
100
I Control
D-pen
BAPN-o
BAPN-ip
FIG. 2. Bursting pressures recorded in the colonic anastomotic segments. Results of the individual measurements are given together with the average values ? SD.
crograms per centimeter, may not be used in comparison with those in anastomotic segments in order to draw conclusions about the degree of collagen synthesis. Construction of an anastomosis per se will, by the inversion of the intestinal wall necessary to sew the divided ends together, by definition increase the volume of intestinal tissue per centimeter. Therefore, anastomotic segments should be compared only to anastomotic segments. The fact that collagen synthesis occurred around intestinal anastomoses during the first week of healing is TABLE II Leakage Site in Anastomosed Intestine Leakage site Ileum In anastomosis Outside anastomosis Colon In anastomosis Outside anastomosis Total In anastomosis Outside anastomosis
Control
D-pen
BAPN-o
BAPN-ip
4 6
1 9
7 3
7 3
1 9
3 7
4 6
4 6
5 15
4 16
11 9
11 9
42
VAN
DOORN,
DE
MAN,
AND
TABLE
Hydroxyproline
HENDRIKS
III
Concentration and Hydroxyproline
Content in Anastomotic Segments
Ileum
Colon
whvz Control D-peal
BAPN-o BAPNip
10.7 11.0 9.9 9.5
k&m
2 1.6 (10) -+ 1.9 (10) f 2.0 (8) + 1.4 (10)
248 313 262 318
2 + 2 f
69 70 60 73
i-dmfs (10) (10) (8) (10)
15.4 15.3 13.4 14.8
2 * f 2
kdcm
1.7 (9) 1.1 (10) 3.8 (8) 2.1 (10)
252 268 244 286
If: 48 2 57 f 76 + 65
(9) (10) (8) (10)
Note. Results are expressed as average values + SD and (in parentheses) number of samples.
clearly illustrated by comparison of the anastomotic hydroxyproline concentration and hydroxyproline content in 7-day-old anastomoses (Table III) with those in 3-day-old anastomoses (data of this separate control group of 10 animals have not been used further in the present paper). Here, average hydroxyproline concentration and content were, respectively, 6.9 4 1.1 (SD) pg/mg and 122 2 37 t&cm in ileum and 10.2 2 2.8 ug/mg and 184 + 42 l&cm in colon. Thus, anastomotic hydroxyproline content increased considerably between 3 and 7 days postoperatively. Acetic acid solubilized a significant amount of collagen from uninjured intestine, ileal hydroxyproline always being better solubilized than colonic hydroxyproline. For instance, in the control group an average 11.6 ? 2.5% (SD, n = 9) of total hydroxyproline in the ileal control segment could be extracted with acid as opposed to 6.4 + 1.4% (n = 10) in the colonic control segment (P = 0.0092, signed rank test). Lathyrogen treatment during 8 days did not affect these values (results not shown). Administration of BAPN had a very significant effect on the size of the acid-soluble hydroxyproline fraction in the anastomosis (Table IV). Clearly, BAPN enhanced the solubility of anastomotic hydroxyproline, both in ileum and in colon, while D-pen had no effect. On the contrary, neither lathyrogen had any effect on the amount of hydroxyproline extracted when the acetic acid residues were treated subsequently with pepsin. TABLE
Solubility of Hydroxyproline
IV
in Anastomotic Segments
Ileum HAc-soluble Control D-pen BAPN-o BAPN-ip P value
10.3 + 2.9 (10) 10.2 ” 2.5 (10) 17.5 f 5.8 (9) 15.1 2 3.5 (10) 0.0002
Colon Pepsin-soluble 6.8 f 2.1 (7) 8.9 f 2.5 (7) 7.2 f 3.2
(8) 9.2 f 3.8 (10) ns
HAc-soluble 5.9 f 1.1 (10) 6.6 f 2.2 WV 14.8 + 9.6 (9) 10.1 f 2.9 (10)
