Gut, 1991, 32, 1482-1487

Postoperative changes in collagen synthesis inintestinal anastomoses of the rat: differences betweensmall and large bowel

M FW C Martens, Th Hendriks

AbstractCollagen synthesis is an essential feature ofanastomotic healing in the intestine. Post-operative collagen synthesis, measured in vitroin intestinal anastomoses was studied fromthree hours to 28 days after operation. For thispurpose, an ileal and a colonic anastomosiswere constructed within the same animal andthe results in both intestinal segments werecompared. In the ileum, coliagen synthesiswas significantly increased, with respect tounoperated controls, three hours after opera-tion. It remained raised during the period ofstudy, with a maximal 10-fold stimulation fourdays after operation, and had nearly returnedto the preoperative level after four weeks. Thegeneral pattern was the same in the colon,although quantitatively different: the increasein synthetic activity was delayed in comparisonwith the ileum. Maximal stimulation wasapproximately six-fold. In addition, we calcu-lated the ratio for each rat between anasto-motic coliagen synthesis and the average valuefound in non-operated control animals. Post-operative stimulation in the ileum was higherthan in the colon in almost every animalexamined. The results show that the ileumresponds more quickly and strongly to wound-ing than the colon, at least as far as theproduction of new coliagen is concerned.Possibly, this phenomenon contributes to thelower failure rate apparent for anastomoses inthe smali bowel.

Department of GeneralSurgery, St RadboudUniversity Hospital,Niimegen, TheNetherlandsM FWC MartensTh HendriksCorrespondence to:Dr Th Hendriks, DeptGeneral Surgery, St RadboudUniversity Hospital, PO Box9101, 6500 HB Nijmegen, TheNetherlands.Accepted for publication11 February 1991

Dehiscence of intestinal anastomoses remainsa major complication after surgery. Despiteimproved surgical techniques leakage of colonicanastomoses frequently occurs, resulting in highmorbidity and mortality. In a much cited studyon surgery for large bowel cancer the overallincidence of leakage was 10.8% in anastomosesconstructed intraperitoneally. ' There is stillmuch to be learned about the molecular pro-cesses which contribute to normal healing of thebowel wall: such research is essential in order toseek measures which could benefit healing andultimately prevent anastomotic leakage.The strength of the intact and the anasto-

mosed bowel wall is derived from collagenfibrils, located predominantly in the sub-mucosa,2 3 which connective layer forms thebackbone of the intestine. During the firstpostoperative days anastomotic strength islow and anastomotic collagen levels changemassively.4 It is generally assumed that thetransiently lowered collagen concentrations,observed during normal healing, are the result of

early, although limited, collagen degradationfollowed by considerable collagen synthesis.While early anastomotic strength depends on thesuture holding capacity of existing collagenfibrils, newly formed collagen is needed to bridgethe gap and restore the original strength of thebowel wall. Thus, the progress of collagensynthesis plays a central role in the healingsequence: disturbance ofits regulation will affectanastomotic strength and might enhance chancesfor dehiscence.The occurrence of collagen synthesis around

experimental intestinal anastomoses has beenshown in vivo, in the ileum5 and in the colon6from two days after operation onwards. Themethodology used for such experiments requiresadministration of large amounts of radioactiveproline to animals and precludes the measure-ment of synthesis in the immediate postoperativeperiod as the precursor needs to be injected atleast 24 hours before death.6 A system to measurecollagen synthesis in intestinal explants hasrecently been established.7 This in vitro systemallows measurement immediately after operationand appears suitable to initiate research on theeffects ofpotentially regulatory factors. Here, wereport on the postoperative stimulation of thecollagen synthesis, as measured in anastomoticexplants collected from three hours after opera-tion onwards. As results from previous experi-ments have suggested that collagen synthesis isinitiated at an earlier stage in the small intestinethan in the large intestine,8 we have constructedan ileal and a colonic anastomosis in the sameanimal and compared the postoperative course ofcollagen synthesis in both intestinal segments.

Methods

ANIMALSMale Wistar rats with an approximate weight of225 g were used. They were fed a standard diet(Hope Farms, Woerden, The Netherlands) andallowed free access to water. The animals werekilled by an intraperitoneal overdose of sodiumpentobarbital and the intestinal biopsies werecollected immediately.

MATERIALSL-[2,3-3H]Proline (300 mCi/mg) was purchasedfrom Amersham International, England.Dulbecco's Modified Eagles Medium (DMEM)was obtained from Gibco, Breda, The Nether-lands. Collagenase (type VII), deoxyribonucleicacid (calf thymus) and bovine serum albumin(BSA) were all obtained from Sigma, St-Louis,

1482

on August 22, 2020 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.32.12.1482 on 1 D

ecember 1991. D

ownloaded from

Postoperative changes in collagen synthesis in intestinal anastomoses ofthe rai: differences between small and large bowel

USA. The scintillation liquid used was picofluor-30 from Packard, Groningen, The Netherlands.Kanamycin was obtained from Gist-Brocades,Delft, The Netherlands. All other reagentswere of analytical grade (Merck, Darmstad,Germany). Suture material used was ethilon 8x0(Ethicon, Norderstedt, FRG).

OPERATIVE TECHNIQUESAll animals received an anastomosis in the ileumand the colon. Surgery was undertaken insemisterile conditions using a Zeiss operatingmicroscope. The rats were anaesthetised by anintraperitoneal injection with sodium pentobar-bital. The abdomen was opened through amidline incision of approximately 4 cm. Theileum was transsected at 15 cm proximal to theileal-caecal junction and an end-to-end anasto-mosis was constructed using eight single layerinverting interrupted 8x0 Ethilon sutures. Sub-sequently, the descending colon was transsected3 cm proximal to the peritoneal reflection andcontinuity was restored as described above. Theabdomen was closed in two layers using silk forthe fascia and staples for the skin.

After three and 12 hours and 1, 2, 3, 4, 7, 14,and 28 days the animals were killed (six at eachtime point) and the anastomotic segments wereresected, opened longitudinally and washedtwice with physiological salt solution.

Control segments were obtained from non-operated rats of similar age and weight:uninjured intestine was taken from the same siteswhere experimental animals received theiranastomoses.

0.2 M NaOH and neutralised by the addition of0.3 ml 1 M N-(2-hydroxyethyl)piperazine-N-(2-ethanesulphonic acid) (HEPES) and 0 3 ml 0 15M HCI. Aliquots from this solution (0- 1 ml) werecounted to determine the incorporation in totalprotein. In order to determine proline incorpora-tion into collagen 0-2 ml 20 mM Tris-HCl, pH7.6, containing 50 mM CaCl2 and 0.1 mlcollagenase (chromatographically purified on aG200 gelfiltration column) were added to a 0.5ml aliquot of the solubilised sample and themixture was incubated for five hours at 37°C.The digestion was terminated by the addition oftrichloroacetic acid and tannic acid up to finalconcentrations of 0-6 M and 3 mM, respectively.After centrifugation (10 min; 14-500 g) a 1-0 mlaliquot ofthe supernatant was counted in a liquidscintillation analyser. The same procedure wasfollowed without the addition of collagenase.Subtraction of the counts released in this blankincubation from those released in the presence ofcollagenase yielded the collagen specific incorpo-ration, which will be referred to as collagenasedigestible protein (CDP). Subtraction of theradioactivity in the collagenase digestible proteinfraction from that in total protein yields theincorporation into non-collagenous protein(NCP).The relative collagen synthesis was calculated

with the formula9 that takes into account theenrichment of proline in collagen compared withother proteins:

CDP% relative collagen synthesis=-

(NCPx 5-4)+CDP-x 100%

ASSAY OF COLLAGEN SYNTHESISCollagen synthesis was measured in tissueexplants, according to a procedure validatedbefore for rat intestinal tissue.7 The anastomosisproper (± 2 mm left and right of the transsectionline) was isolated and cut into pieces of approxi-mately 1-2 mm2. Two equal samples, 35-70 mgwet weight, were transferred to Petri dishes(diameter 35 mm). The pieces were washed oncewith physiological salt solution and once withincubation medium (DMEM containing 50 ,ug/ml ascorbate and 250 [ig/ml kanamycin). Subse-quently, 1.5 ml incubation medium was addedand the samples were incubated for 30 min at37°C (95% air; 5% C02). The medium was thenremoved by suction and replaced with 1.5 mlincubation medium containing 4.5 [tCi[2,3-3H]proline. Incubation proceeded for threehours. All subsequent steps were carried out at4°C. Both tissue and medium were transferred toa centrifugation tube and spun for five minutes at2500 g. The sediment was homogenised in 3 0 ml50 mM Tris-HCl, pH 7-6, containing 25 mMethylenediaminetetraacetic acid (EDTA), 10mM N-ethylmaleimide (NEM), 1 mM phenyl-methylsulphonylfluoride (PMSF) and 1 mMproline. Trichloroacetic acid (TCA) was added(final concentration 0.6 M) to the homogenatewhich was then centrifuged for five minutes at2500 g. The sediment was washed three timeswith 0 3 M TCA containing 1 mM proline.The final sediment was dissolved in 0 75 ml

OTHER PROCEDURESDNA was measured in the NaOH-solubilisedtrichloroacetic acid sediment, by the method ofBurtonl' using calf thymus DNA as a standard.Statistical methods are mentioned with theresults.

ResultsThe intestinal wall synthesises collagen. Wereported in an earlier communication thatcollagen synthesis can be measured in short termexplants of previously uninjured intestinaltissue.7 The same procedure was applied toexplants from intestinal anastomoses. In a firstseries of experiments, increasing amounts oftissue from four day old anastomoses, con-structed both in the ileum and the colon, wereincubated with [3H]proline for various timeperiods. Incorporation into the collagenasedigestible protein fraction increased linearlywith time and tissue weight (results not shown).The standard conditions chosen for subsequentexperiments, 50 mg tissue incubated for threehours, fall well within the linear part of therespective curves.

Collagen synthesis around intestinal anasto-moses is strongly enhanced as compared withthat in uninjured intestine. For practicalreasons, it was impossible to measure synthesisin control segments removed during operation,which would have enabled comparison betweenanastomotic and control segments from the same

1483

on August 22, 2020 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.32.12.1482 on 1 D

ecember 1991. D

ownloaded from

Martens, Hendniks

TABLE I Collagen synthesis in ileal anastomoses

Collagenase CollagenaseTime digestible digestible Relativeafter protein dpmlmg protein dpmlpg collagenoperation wet weight DNA synthesis (%)

3 h 177 (69)t 57 (20)* 0-36 (0.09)12 h 262 (68)t 71 (19)t 0 43 (0.08)*1 d 277(87)t 108 (52)* 0-52(0 10)t2d 310(88)t 101 (41)* 0-61 (0-17)t3 d 509 (82)t 167 (64)t 1-09 (0-21)t4 d 687(268)t 300(126)t 2-42(0.72)t7 d 284 (104)t 184 (120)t 1-69 (0.67)t

28 d 109 (32) 70 (32) 0.54 (0 25)Control 70 (18) 41 (7) 0.30 (0.04)

Explants from anastomotic tissue were incubated for three hourswith 4-5 ICi [3H]proline. Results are expressed as dpmcollagenase digestible protein and as percentage relative collagensynthesis. Data represent average values (SD) from six animals.Differences between anastomoses and control intestine fromunoperated rats was tested for significance using a one-sidedWilcoxon's test: *O.O1<p-0o05; tp-OOl.

10 -

6-

2-

C-n0COg 10-

0E 6-

0m 2-

10-

animal. The fact that anastomotic constructionnot only affects collagen synthesis in the anasto-motic area but also at locations further removed,particularly in the small intestine (Martens andHendriks, in preparation), precluded the use ofcontrol segments removed at time of death.Because the combination of anaesthesia andabdominal incision did not influence intestinalcollagen synthesis during the first week afterlaparotomy (results not shown) we used non-operated control rats to establish normal values.

First, it was established that anastomoticconstruction in the large bowel did not signifi-cantly increase collagen synthesis in the smallbowel and vice versa. For this purpose, rats wereoperated either in the ileum or the colon andcollagen synthesis was measured after fourdays in the unoperated intestine at the sitewhere normally the second anastomosis wouldhave been constructed. Collagen synthesis inuninjured ileum in animals with a colon anasto-mosis was 30 (18) (average (SD) n=6) dpm CDP/Rg DNA, compared with 41 (17) dpm CDP/RgDNA in ileum of unoperated control animals.Likewise, collagen synthesis was 52 (11) dpmCDP/[g DNA in intact colon in rats with anileum anastomosis and 77 (8) dpm CDP/[gDNA in colon of unoperated controls.

TABLE II Collagen synthesis in colonic anastomoses

Collagenase CollagenaseTime digestible digestible Relativeafter protein dpmlmg protein dpmltg collagenoperation wet weight DNA synthesis (%)

3 h 267 (77) 73 (20) 0-62 (0 14)12 h 263 (35)t 88(16) 0X64 (010)Id 370(21))t 123(40) 0-88(0.14)t2 d 399 (67)t 126 (48)* 0.93 (011)t3 d 1002 (462)t 253 (88)t 2.65 (1.23)t4 d 1086 (375)t 373 (188)t 2-87 (0.72)t7 d 421 (148)t 154 (55)t 1-48 (O. Sl)t

28 d 192(85) 108(36) 1.01 (0.24)tControl 188 (29) 77(8) 0.52 (0.03)

Explants from anastomotic tissue were incubated for three hourswith 4-5 [tCi [3H]proline. Results are expressed as dpmcollagenase digestible protein and as percentage relative collagensynthesis. Data represent average values (SD) from six animals.Differences between anastomoses and control intestine fromunoperated rats was tested for significance using a one-sidedWilcoxon's test: *O Ol<p- 0-05; tp-O Ol.

6-

2-

C1 ileumJ colon

0

3h 1 ih1d 2d 3d 4d 7d 28d

Time after operation

Figure 1: Relative increase ofcollagen synthesis inanastomotic explantsfrom both ileum and colon. In eachanastomosis collagen synthesis was measured as dpmcollagenase digestible proteinlmg wet weight (A), dpm CDPIpgDNA (B) or percentage relative collagen synthesis (C)and the ratio between this value and the average value in sixunoperated controls was calculated. Values depicted representthe average (SD)from six experimental animals.

Collagen synthesis was measured from threehours to 28 days after operation. In Table I theaverage values for ileal anastomoses are given.Three hours after operation incorporation oflabel into collagenase digestible protein, expres-sed on a wet weight and DNA basis, was alreadysignificantly increased with respect to theunoperated controls. The relative collagensynthesis was significantly enhanced after 12hours. From three hours to four days afteroperation the absolute and the relative collagensynthesis rose steadily. Thereafter, syntheticactivity started to fall: the average values in twofive day old anastomoses were 230 dpm collage-nase digestible protein/mg wet weight, 132 dpm/Fg DNA, and 1-73% relative collagen synthesis,respectively. After four weeks collagen synthesisin the anastomotic area had nearly returnedto the preoperative level; average values notbeing significantly different from thosemeasured in the controls.The general pattern was the same in the colon

(Table II), although the increase in syntheticactivity appeared delayed in comparison with theileum. The absolute collagen synthesis wassignificantly raised after 12 hours, if expressedon a wet weight basis, and only after two days, ifexpressed on a DNA basis. Likewise, the relativecollagen synthesis in the anastomotic area wassignificantly higher than in control intestine from24 hours after operation onwards. In the colon,synthesis also seemed maximally stimulated afterfour days. The average values measured in twofive day old colonic anastomoses were 288 dpm

1484

on August 22, 2020 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.32.12.1482 on 1 D

ecember 1991. D

ownloaded from

Postoperative changes in collagen synthesis in intestinal anastomoses ofthe rat: differences between smalland large bowel

relative increase ileumrelative increase colon4-

A

3-

2- * S ,S : :0

2-~~~~~S @ 0

1 -----__-_+____ __ --1 8

B5-

4-

3 0

2- 0

2-+-r---b---g 1

c~~~~

14-

3l

2 m

3h 12h 1d 2d 3d 4d 7d 28dtime after operation

Figure 2: Ratio between ileal and colonic stimulation ofanastomotic collagen synthesis. In each rat, the collagensynthesis in both ileal and colonic anastomotic explants wasmeasured as dpm CDP/mg wet weight (A), dpm CDPIgDNA (B) and percentage relative collagen synthesis (C) andthe ratio between these values and the average value incorresponding intestinal segmentsfrom six unoperated controlswere calculated. The data depicted (each point representingone animal) were obtained by dividing the relative increase inthe ileal anastomosis by the relative increase in the colonicanastomosis. The postoperative rise in collagen synthesis issignificantly (p.cO-5) higher in ileal than in colonicanastomoses ifall six values exceed 1 (one-sided signed ranktest).

collagenase digestible protein/mg wet weight,151 dpm collagenase digestible protein/FIg DNA and a relative collagen synthesis of1.39%, respectively. Values after four weekswere not significantly different from controlvalues, except for the relative collagen synthesiswhich was still raised.We then calculated at each time point the

average ratio between the values measured inanastomoses and those observed in non-operatedrats. Figure 1 shows that the increase in collagensynthesis, expressed on basis ofeither wet weightor DNA was always higher in the ileum than inthe colon. For instance, maximal increase, withrespect to normal value, was 9-8 (3.8) in ileumand 5 8 (2.2) in colon (if synthetic activities wereexpressed on a wet weight basis, Fig IA). Incontrast, the stimulation of the relative collagensynthesis appeared comparable in both small andlarge bowel, certainly up to two days (Fig 1C).Three days after operation, the average ratio inthe colon was even higher than in the ileum,although at four and seven days the effect wasagain the other way around.

In order to further substantiate the idea that

anastomotic construction stimulates collagensynthesis more strongly in the ileum than in thecolon, the relative increase in both segments wascompared within each rat. For this purpose, theratio between collagen synthesis in the ilealanastomosis and the average value found in thenon-operated controls was calculated. The samewas done for the colonic anastomosis in the sameanimal. Subsequently, a final ratio was calcu-lated between the ratio's found for ileal andcolonic anastomoses. A ratio exceeding the valueof 1 indicates that the collagen synthesis in theanimal in question is enhanced more strongly inthe ileum than in the colon. The data for all ratsare represented in Figure 2. Because the numberof rats killed at each time point is relatively small(six), the ileum-colon difference is only signifi-cant (one-sided signed rank test) ifa ratio above 1is found in all six animals. For the syntheticactivities, expressed on a wet weight basis (Fig2A) this was the case at 12 hours and two daysafter operation. Still, at the other time points aratio above 1 was observed in five of six animals.Altogether, the reverse - a ratio lower than 1,indicating a stronger stimulation in colonicanastomoses - was the case in only six of 48animals. This effect was even more pronouncedif synthetic activity was expressed on the basis ofDNA (Fig 2B). Here, a significant effect wasnoted at 12 hours, one day, four days, and sevendays after operation. A different picture emergedfor the relative collagen synthesis (Fig 2C): noconsistent difference was noted between theileum and the colon. Stimulation of collagensynthesis in ileal anastomoses was consistentlyand significantly higher than in colonic anasto-moses at seven days after operation only.

DiscussionCollagen synthesis is an essential feature ofanastomotic healing: the strength of the suturedintestinal wall has to be restored by newlyformed collagen fibrils. Inhibition of post-operative fibrillogenesis impairs the build up ofanastomotic strength." Disturbance of collagensynthesis in the anastomotic area will lead to lossof strength and probably to increased chances foranastomotic failure. The frequency rate of dehi-scence of bowel anastomoses remains ratherhigh.' Apparently, anastomoses are compro-mised regularly, even under optimal conditions,and a multitude of circumstances may aggravatethe risk considerably,'2 even to such extent thatanastomotic construction has to be delayed.Because the underlying mechanisms are poorlyunderstood, examination of the course andregulation of the primary molecular processes,such as the synthesis of collagen, is of majorimportance. Dehiscence of bowel anastomosesappears mainly a problem in the large intestine. 11

Explanations for this difference between smalland large bowel are speculative and it isunknown ifmolecular events which occur duringhealing differ between both bowel segments. Ifso, they might indicate how to proceed in orderto improve colonic anastomotic repair. Indeed,our comparison of postoperative collagensynthesis in ileal and colonic anastomoses withinthe same animal shows that this process starts

1485

on August 22, 2020 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.32.12.1482 on 1 D

ecember 1991. D

ownloaded from

1486 Martens, Hendriks

earlier and is more strongly stimulated in thesmall bowel.We have used explants of intestinal anasto-

moses and thus measured collagen synthesis invitro. In an earlier communication we haveshown that such a system is useful to assesschanges which occur in vivo, such as induced byageing.7 The present results show that the con-siderable stimulation of anastomotic collagensynthesis, which had been shown to occur bymeans of in vivo measurements,5'6 is also foundthis way. The maximal stimulation observed,nine-fold in the ileum and six-fold in the colon ascompared with uninjured intestine, is similar inboth types of experiments.

So far, no data have been published concern-ing anastomotic collagen synthesis during thefirst two days of the healing period. Clearly,synthesis is enhanced very quickly after wound-ing, certainly in the ileum where incorporation of[3H]proline into CDP has risen significantlyalready after three hours. As the influx offibroblasts, which are generally believed to bethe source of collagen produced during woundhealing, is only apparent two days after anasto-motic construction'4 this means that residentcells respond to injury by increasing theircapacity for collagen production. The fact thatthe relative collagen synthesis is also significantlyenhanced in the early period indicates that this isnot a general stimulation of protein synthesis butindeed a more specific response. It is impossibleto decide from the present results if indeedfibroblasts are responsible for the collagen pro-duction. Recent experiments have shown thatintestinal smooth muscle cells, isolated fromhuman jejunum, are capable of synthesising typeI, III, and V collagen.5 16 The authors suggestthat these cells could play an important role inrepair processes after damage and inflammationof the intestinal wall.

Apart from the description of the early courseof postoperative collagen synthesis in intestinalanastomoses, the comparison of this process insmall and large bowel was the major goal of thisstudy. So far, one report exists in the literaturewhich attempts such a comparison: in vivocollagen synthesis, expressed as specific activity(dpm [3H]hydroxyproline/tmol hydroxypro-line) and measured four days after operationonly, was not significantly different in iieal andcolonic anastomoses. '7 The same is probably trueif we compare the average collagen synthesis,expressed as dpm/,ug DNA, in ileal and colonicanastomoses at this time point (Tables I and II).Such an approach, however, does not take intoconsideration the fact7 that basal synthesis incolon is significantly higher than in the ileum. Inorder to allow comparison within each individualrat, we constructed both anastomoses in thesame animal. Control experiments confirmedthat anastomotic construction in the ileum doesnot affect collagen synthesis in unoperated colonand vice versa. If collagen synthesis is calculatedas dpm collagenase digestible protein, expressedon the basis of wet weight or DNA, postopera-tive stimulation in the ileum is higher than in thecolon in almost every animal examined (Fig 2).Moreover, ileal synthesis is clearly stimulated atan earlier stage than synthesis in colon. For

instance, incorporation of [3H]proline intocollagenase digestible protein and expressed perFtg DNA, is significantly higher than inunoperated controls at three hours after opera-tion in the ileum and at 48 hours after operationin colon. Taken together, we believe that theseresults indicate that the ileum responds morequickly to wounding than the colon, at least asfar as production of new collagen is concerned.This conclusion is in agreement with earlierstudies48 which showed that the transient post-operative lowering of anastomotic hydroxypro-line concentrations was less pronounced andmore rapidly undone in the ileum.The reasons why the small bowel reacts more

rapidly in this respect than the large bowelremain as yet to be determined. The cellsresponsible for collagen synthesis, resident fibro-blasts and/or smooth muscle cells,'5 16 arestimulated by wounding, possibly by mediatorsreleased from platelets.'8 Somehow, coloniccollagen producing cells are less responsive tomediators released - for example, plateletderived growth factors - in the immediate post-operative period. Because the preoperative levelof collagen synthesis is higher in the colon, itcould be that the cells in question reside at ahigher level of activity than those in the ileumand therefore need a greater stimulus in order toincrease their collagen production.

It is generally accepted that anastomoticcollagen determines anastomotic strength, andnext to synthesis, degradation of this matrixprotein will affect wound strength. Productionof collagenase has been shown immunohisto-chemically around colonic anastomoses'9 and wehave found transiently increased extractablecollagenolytic activity in the anastomotic areabut no differences between ileal and colonicanastomoses.20 While degradation of anasto-motic collagen is probably limited under con-ditions optimal for healing,2' it should beemphasised that this process could be enhancedby local or systemic factors and thus contributeto loss of strength.

In conclusion, we suggest that the differencesin postoperative collagen synthesis, observedbetween ileal and colonic anastomoses con-structed under normal conditions, may help toexplain the lower failure rate of anastomoses inthe small intestine. Investigation of the regula-tion of collagen synthesis, either in tissueexplants or on the cellular level, is needed inorder to determine optimal conditions and allowimprovement of this important aspect of repair,particularly in anastomoses which have to beconstructed in high risk conditions

The authors are grateful to Prof Dr J J H H M de Pont for helpfuldiscussions, to BM de Man for expert technical assistance, and toTh M de Boo for statistical analysis of the data.

1 Fielding FP, Stewart-Brown S, Blesovsky L, Kearney G.Anastomotic integrity after operations for large bowelcancer: a multicentre study. BM3r 1980; 281: 411-4.

2 Thomson HJ, Busuttil A, Eastwood AN, Elton RA. Sub-mucosal collagen changes in the normal colon and indiverticular disease. IntJr Colorectal Dis 1987; 2: 208-13.

3 Graham MF, Diegelmann RF, Elson CO, et al. Collagencontent and types in the intestinal strictures of Crohn'sdisease. Gastroenterology 1988; 94: 257-65.

4 Hesp WLEM, Hendriks Tb, Lubbers EJC, de Boer HHM.Wound healing in the intestinal wall. A comparison betweenexperimental ileal and colonic anastomoses. Dis ColonRectum 1984; 27: 99-104.

on August 22, 2020 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.32.12.1482 on 1 D

ecember 1991. D

ownloaded from

Postoperative changes in collagen synthesis in intestinal anastomoses ofthe rat: differences between smalland large bowel 1487

5 Jonssen K, Jiborn H, Zederfeldt B. Collagen metabolism insmall intestinal anastomoses. AmJrSurg 1987; 154: 288-91.

6 Jiborn H, Ahonen J, Zederfeldt B. Healing of experimentalcolonic anastomoses. IV. Effect of suture technique oncollagen metabolism in the colonic wall. Am J Surg 1980;139: 398-405.

7 Martens MFWC, Hendriks Th. Collagen synthesis in explantsfrom rat intestine. Biochim Biophys Acta 1989; 993: 252-8.

8 Hendriks Th, Vereecken THLB, Hesp WLEM, SchillingsPHM, de Boer HHM. Loss of collagen from experimentalintestinal anastomoses: early events. Exp Mol Pathol 1985;42:411-8.

9 Peterkofsky B, Chojkier M, Bateman J. Determination ofcollagen synthesis in tissue and cell culture systems. In:Furthmayer H, ed. Immunochemistry of the extracellularmatrix, vol 2. Boca Raton: CRC Press, 1981: 19-47.

10 Burton K. A study of the conditions and mechanisms of thediphenylamine reaction for the colorimetric estimation ofdeoxy-ribonucleic acid. BiochemJI 1956; 62: 15-23.

11 van Doom K, de Man B, Hendriks Th. The effects oflathyrogens on intestinal anastomoses in the rat. Exp MolPathol 1990; 52: 37-45.

12 Khoury GA, Waxman BP. Large bowel anastomoses. 1.The healing process and sutured anastomoses - a review.BrJ7 Surg 1983; 70: 61-3.

13 Hesp WLEM, Lubbers EJC, de Boer HHM, Hendriks Th.Anastomotic insufficiency in small bowel surgery: incidenceand treatment. LangenbecksArch Chir 1986; 368: 105-11.

14 Hesp WLEM, Hendriks Th, Schillings PHM, Lubbers EJC,de Boer HHM. Histological features of wound repair:a comparison between experimental ileal and colonicanastomoses. BrJr Exp Pathol 1985; 66: 511-8.

15 Graham MF, Drucker D, Diegelmann RF, Elson C. Collagensynthesis by human intestinal smooth muscle cells inculture. Gastroenterology 1987; 92: 400-5.

16 Perr HH, Graham MF, Diegelmann RF, Downs PN.Cyclic nucleotides regulate collagen production by humanintestinal smooth muscle cells. Gastroenterolog 1989; 96:1521-8.

17 Jonssen K, Jiborn H, Zederfeldt B. Comparison of healing inthe left colon and ileum. Changes in collagen content andcollagen synthesis in the intestinal wall after ileal and colonicanastomoses in the rat. Acta ChirScand 1985; 151: 537-41.

18 Ross R, Bowen-Pope DF, Raines EW. Platelet-derived growthfactor and its role in health and disease. Philos Trans R SocLond[Biol] 1990; 327: 155-69.

19 Chowcat NL, Savage FJ, Hembry RM, Boulos PB. Role ofcollagenase in colonic anastomoses: a reappraisal. BrJ Surg1988; 75: 330-4.

20 van der Stappen JWJ, Hendriks Th, de Boer HHM.Collagenolytic activity extracted from intestinal anastomosesof the rat. Matrix 1989; 9: 238-43.

21 Hendriks Th, Mastboom WJB. Healing of experimentalintestinal anastomoses. Parameters for repair. Dis ColonRectum 1990; 33: 891-901.

on August 22, 2020 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.32.12.1482 on 1 D

ecember 1991. D

ownloaded from