ORIGINAL ARTICLE

Ischemic preconditioning improves stability of intestinalanastomoses in rats

Goran Marjanovic & Eva Jüttner & Axel zur Hausen &

Ulrich Theodor Hopt & Robert Obermaier

Accepted: 31 March 2009 /Published online: 18 April 2009# Springer-Verlag 2009

AbstractBackground The aim of our study was to establish whetherischemic preconditioning (IPC) directly before performinga small bowel anastomosis has an effect on anastomoticstability and healing.Material and methods Forty male Wistar rats wererandomized to five groups: control (CO, n=8) withpreparation of the superior mesenteric artery (SMA)but without IPC. IPC groups had different intervals ofischemia (occlusion of the SMA) and reperfusion: 10 minischemia and 20 min reperfusion (IPC10/20, n=7),10 min ischemia and 30 min reperfusion (IPC10/30, n=8),15 min ischemia and 20 min reperfusion (IPC15/20, n=8),and 15 min ischemia and 30 min reperfusion (IPC15/30, n=9). On the fourth postoperative day, the animals wererelaparotomized: bursting pressure, hydroxyproline concen-tration, and histological ischemia mucosal injury scale of theanastomosis were assessed.Results Four days after operation, the mean burstingpressure was 73±6 mmHg in the control group, whereasit was significantly higher in IPC10/20 (113±11 mmHg;p=0.018), IPC10/30 (110±13 mmHg; p=0.001), andIPC15/30 (124±9 mmHg; p=0.003). IPC15/20 did notshow a significant difference (63±2 mmHg; p=0.4). Wedid not find a significant effect regarding hydroxyprolineconcentration, but IPC diminished mucosal injury.

Conclusions IPC directly before performing a small bowelanastomosis has a time-dependent beneficial effect onanastomotic stability, thus indicating a new clinical ap-proach to improve the healing process of intestinalanastomosis.

Keywords Ischemic preconditioning . Anastomotichealing . Intestinal anastomoses . Stability of intestinalanastomoses . Ischemia reperfusion

Introduction

Despite correct operative technique and medical treatment,the integrity of intestinal anastomoses may be compro-mised, resulting in anastomotic dehiscence with signifi-cantly higher morbidity and mortality [1]. To date, it is notpossible to influence the general condition in most patientsbefore performing an intestinal anastomosis. Accordingly,research is focusing on the backgrounds and the directimprovement of wound healing in gastrointestinal surgery[2].

Ischemic preconditioning (IPC) has been found to be oneof the most promising approaches during the last decade,appearing to increase the tolerance of different tissues toischemia/reperfusion injury (IRI) [3–11]. Since Hotter et al.[12] first described IPC of the intestine in 1996, subsequentstudies have confirmed this phenomenon in rats [13–20].Intestinal IPC was shown to have several beneficial localand systemic effects like reducing bacterial translocation,mucosal injury [13], epithelial apoptosis [14], improvingmicrovascular perfusion, and oxygenation after IRI [19–21].

Historically, IPC was used to investigate the protectiveeffect on different variables related to a prolonged IRI.

Int J Colorectal Dis (2009) 24:975–981DOI 10.1007/s00384-009-0696-0

G. Marjanovic (*) :U. Theodor Hopt : R. ObermaierDepartment of General and Digestive Surgery,University of Freiburg,Hugstetter Strasse 55,79106 Freiburg im Breisgau, Germanye-mail: goran.marjanovic@uniklinik-freiburg.de

E. Jüttner :A. zur HausenInstitute of Pathology, University of Freiburg,Freiburg im Breisgau, Germany

Regarding the known effects of IPC, the aim of our study wasto investigate whether different brief intervals of vascularocclusion followed by different reperfusion intervals directlyprior to performing a small bowel anastomosis—without anIRI interval—have a beneficial effect on anastomotic stability.

Materials and methods

Animals

The local Ethics Committee at University of Freiburgapproved all experiments. Male Wistar rats (Charles River,Sulzfeld) weighing 220–320 g were used for the experi-ments. The animals were housed two per cage, fed standardchow, and given access to water ad libitum. Twelve hoursbefore anesthesia, animals were deprived of food but hadfree access to water. During the first 24 postoperative hours,animals had free access to water but were fed stepwise witha standard amount of chow. All animals had the sameconsumption of chow.

Experimental design

After a preoperative observation time of 5–7 days underlaboratory conditions, animals were randomly assigned toone of the five different groups (Table 1). Each groupconsisted of at least seven animals. After induction ofanesthesia with isoflurane (4% in an oxygen mixture 3 L/min) in a gas box, animals got continuous isofluraneanesthesia (1.5% in an oxygen mixture 3 L/min) througha mask. A 26-G silicon venous catheter was then placedinto a tail vein. The catheter was rinsed by a continuousinfusion (9 mL/h kg body weight) of an isoosmolarelectrolyte solution (contents: Na+ 137 mmol/L; K+

4 mmol/L; Ca++ 1.65 mmol/L; Mg++ 1.25 mmol/L; Cl−

110 mmol/L; CH3COO− 18 mmol/L; pH5.0–7.0; osmolar-

ity 291 mOsm/L; Jonosteril, Fresenius, Bad Homburg,Germany).

Operative procedure

The operative procedure was performed under sterile labora-tory conditions. All animals were operated on by the sameinvestigator (G.M.). The abdomenwas shaved and disinfectedwith polyvidone (Betaisodonna, Mundipharma, Limburg,Germany). A 5-cm midline incision was performed.

In animals randomly assigned to an IPC group, thesuperior mesenteric artery (SMA) was prepared at themesenteric root. It was clamped with an atraumaticmicrosurgical clamp (Medicon, Germany) for different timeintervals. As a consequence, the whole small bowel—except the first jejunal loops—the cecum, and the proximalpart of the ascending colon temporarily became ischemic.Different reperfusion intervals followed the ischemicinterval (Table 1). In control animals, the SMA wasdissected but not pinched. Finally, the anastomotic proce-dure followed.

A 1-cm control segment was resected approximately15 cm proximal to the cecum. The ileal continuity wasrestored by performing an end-to-end anastomosis usingeight inverting interrupted sutures (Prolene 8/0, Ethicon,Germany). In order to standardize the suture technique, asilicon catheter (diameter 5 mm; Heidelberger Verlänger-ung, Braun, Melsungen, Germany) was introduced into theproximal and distal lumen of the resected ileum. Afterperforming the front-wall sutures, the catheter was re-moved and the back-wall sutures followed. The distance ofthe single sutures to the resection margins and the distancebetween the single sutures was 1–2 mm. The abdominalwall was closed in two layers—musculoperitoneal layerwith Monocryl 4/0 SH+ (Ethicon, Germany) and thefasciocutaneous layer with Vicryl 4/0 SH+ (Ethicon,Germany).

Table 1 Experimental groups and results on the fourth POD

Group N Ischemia(min)

Reperfusion(min)

BP mean±SEM(mmHg)

HP concentration±SEM(µg/g dry tissue)

Averagehistologicalgrading±SEM

Min–maxhistologicalgrading

Complications/deaths

Control 8 No No 73.3±6 65.3±9 1.6±0.2 1–2 One abscess

IPC10/20 7 10 20 113.2±11* 68.4±9 1.3±0.3 0–2 –

IPC10/30 8 10 30 110.4±13* 72.3±7 0.9±0.2 0–2 One death due toaspiration

IPC15/20 8 15 20 63.2±2 53.2±5 1.4±0.2 1–2 One death due to partialmesenterial gangrene

IPC15/30 9 15 30 124.2±9* 52.4±4 1.3±0.3 0–2 Two abscesses

The histological grading was assessed by the mucosal injury scale [23]

BP bursting pressure, HP hydroxyproline, min minutes, Min minimal, Max maximal, SEM standard error of the mean

*p<0.05, significant difference to control

976 Int J Colorectal Dis (2009) 24:975–981

On the fourth postoperative day (POD; day of theoperation=day0), the animals were operated on again. Afterinduction of anesthesia in the box with isoflurane (4%),animals were killed by cardiac puncture and potassiuminjection in lethal dose. The abdomen was opened with acomplete midline incision and additionally with a horizontalincision, so that an optimal exposure of the operative situswas generated. After exploration for signs of inflammation,adhesions around the anastomosis were not dissected and a4-cm long segment containing the anastomosis was removedand carefully cleaned from fecal remnants.

Bursting pressure

One lumen of the anastomotic segment was attached to aninfusion pump by a 14-G silicon catheter filled withisoosmolar saline solution. The other lumen was attachedto a digital pressure transducer (Codman ICP Express,Ethicon, Norderstedt, Germany). The pressure was raisedwith an infusion rate of 60 mL/h and digitally monitored—both the bursting pressure (in millimeters of mercury[mmHg]), i.e., the maximum pressure recorded immediatelybefore sudden loss of pressure, and the site of rupture werenoted (mesenterial or the antimesenterial site). Then, thebowel wall of the anastomotic segment was completelycleaned from all adhering tissue and adhesions, opened atthe mesenteric site, and a 1-cm long segment containing thecomplete suture line in the middle was excised and washedgently with saline solution. This anastomotic segment wasthen divided into two longitudinal strips, each containing apart of the anastomotic ring. One strip was spread out in acassette for paraffin embedding and immediately fixed in4% phosphate-buffered formaldehyde (pH7.3). The secondstrip was stored in an Eppendorf tube at −4°C until themeasurement of hydroxyproline concentration.

Hydroxyproline concentration

One strip of the anastomosis was dried in an oven (HeraeusElectronic UT5042EK, Hanau, Germany) to constantweight, which was defined as dry weight. Here, the

hydroxyproline concentration was determined by using thechloramine-T spectrophotometric method which has beenpreviously described by Reddy et al. [22]. The procedure isbased on alkaline hydrolysis of the tissue homogenate andsubsequent determination of the free hydroxyproline inhydrolysate. Chloramine-T was used to oxidize the freehydroxyproline for the production of a pyrrole. Theaddition of Ehrlich's reagent resulted in the formation of achromophore that can be measured at 550 nm. Calculationswere made to express the results as micrograms ofhydroxyproline per gram of dry tissue.

Histological evaluation

Tissues were fixed in formalin and embedded in paraffin.Histological sections were stained with hematoxylin–eosin.Mucosal injury, inflammation, and hyperemia/hemorrhagewere assessed and graded in a blinded manner by twoindependent pathologists using the histologic injury scalepreviously described by Chiu et al. [23]. Briefly, mucosaldamage was graded from 0 to 5 according to the criteriashown in Table 2.

Statistical analysis

All data are expressed as the mean±standard error of themean (SEM). Subsequent comparisons between the groupswere made by Mann–Whitney U test; p<0.05 was assumedto be significant. SPSS 14.0.2 for Windows (Chicago, IL,USA) was used.

Results

General observations

There was no postoperative mortality in the control group.Two animals died in the IPC groups: one animal died onthird POD in the IPC10/30 group without an abdominalcomplication and the other died because of partial mesen-terial gangrene on the third POD in the IPC15/20 group. In

Table 2 Mucosal injury scale adapted from Chiu et al. [23] with definition of different grades describing the histological changes of intestinalmucosa after ischemic injury

Grade Definition

0 Normal mucosal villi

1 Development of subepithelial Gruenhagen's space at the apex of the villus, often with capillary congestion

2 Extension of the subepithelial space with moderate lifting of the epithelial layer from the lamina propria

3 Massive epithelial lifting down the sides of villi, possibly with few denuded tips

4 Denuded villi with lamina propria and dilated capillaries exposed, possibly with increased cellularity of lamina propria

5 Digestion and disintegration of the lamina propria, hemorrhage, and ulceration

Int J Colorectal Dis (2009) 24:975–981 977

the IPC15/20 group, one animal had diarrhea on the secondPOD. One animal of the CO group and two animals of theIPC15/30 group had an abscess at the anastomotic regionwithout macroscopically visible anastomotic insufficiencyat reoperation—consequently, bursting pressure measure-ment was not possible in these animals (Table 1). There wasno difference in perioperative weight gain.

Bursting pressure

Control versus IPC groups

During bursting pressure procedure, the location of burstingcould be observed within the anastomotic line equallydistributed either at the mesenteric or the antimesentericside in all animals. In the control group, bursting pressurewas 73±6 mmHg but it was significantly higher in three offour IPC groups (Fig. 1). The IPC10/20 group had abursting pressure of 113±11 mmHg (p=0.018), the IPC10/30 group had 110±13 mmHg (p=0.001), and the IPC15/30animals had 124±9 mmHg (p=0.003). We did not see asignificant difference in the mean bursting pressure be-tween control and IPC15/20 animals (63±2 mmHg; p=0.4).

Specific effect of ischemia and reperfusion on burstingpressure

Each of the IPC groups had significantly higher burstingpressure rates compared to IPC15/20. In order to investigatethe specific effects of ischemia and reperfusion on thebursting pressure values, we compared the IPC groups withthe same intervals for ischemia (IPC10/20+IPC10/30 versusIPC15/20+IPC15/30) and reperfusion (IPC10/20+IPC15/20versus IPC10/30+IPC15/30), respectively. There was nosignificant difference in mean bursting pressure comparingthe animals with 10 min ischemia (111±11 mmHg) to thosewith 15 min ischemia (94±6 mmHg) (p=0.09). However,animals with the longer reperfusion interval of 30 min had a

significantly higher bursting pressure (117±10 versus 88±7 mmHg; p=0,015) compared to those with 20 minreperfusion.

Hydroxyproline levels

Hydroxyproline was assayed as a measure for collagencontent in the anastomosis. We found a hydroxyprolineconcentration of 65±9 µg/g dry tissue in the control groupand no significant difference in the IPC groups. The IPC10/20 group had 68.4±9 µg/g dry tissue, IPC10/30 had 72.3±7 µg/g dry tissue, IPC15/30 group 52.4±4 µg/g dry tissue,and IPC15/20 animals 53.2±5 µg/g dry tissue (p>0.05).

Histology

A histological score for intestinal mucosa damage, accord-ing to Chiu et al. [23], was used and we could show that, inall groups, including the control group, slight damage of themucosal villi appeared at the anastomotic line. Maximumdamage was seen up to grade 2 with an extension of thesubepithelial space with moderate lifting of the epitheliallayer from the lamina propria. It was preferentially found inthe first few villi of both sides of the ileo-ileal anastomoticline. Villi more distant from the anastomotic line were notaffected. We could not show any significant difference inthe grading of mucosal damage between IPC and controlgroup (Fig. 2). Nevertheless, there were two animals in theIPC10/20 group, three animals in the IPC10/30 group, andalso two animals in the IPC15/30 group with no mucosaldamage (grade 0) but none in the CO and IPC10/20 groups.

Discussion

In gastrointestinal surgery, research is focusing on theinvestigation of mechanisms to improve the anastomotichealing. We are the first to present experimental data on thedirect effect of IPC on anastomotic stability, immediatelyprior to the creation of a small bowel anastomosis in rats.

Wound healing in the gastrointestinal tract includes afine balance between collagen synthesis and collagenolysisin a brief period from 3 to 5 days postoperatively which canbe influenced by local or systemic factors like hypoperfu-sion, tension, malnutrition, hypovolemia, or immunodefi-ciency [2, 24]. Different methods exist to determine thestability of intestinal anastomoses including bursting pres-sure which is a sensitive method to assess anastomoticstrength in the first five PODs, which we used to investigateearly anastomotic healing.

One of the most important factors affecting anastomotichealing is a good blood supply with sufficient microcircu-lation, which has been studied in several clinical trials.

Fig. 1 Mean bursting pressures. *p<0.05, significant difference tocontrol and to IPC15/20

978 Int J Colorectal Dis (2009) 24:975–981

Because of the high risk of anastomotic insufficiency,implementation of IPC prior to gastric tube formation andgastric pull-up at esophageal resection is considered to beimportant [25, 26]. Different techniques of IPC as vascularaugmentation of graft (i.e., gastric conduit) to recipient (i.e.,internal thoracic vein/artery) vessels [supercharging] [27] orpreoperative partial disruption of conduit blood supply[“embolic delay”] [28, 29] were shown to lead to reductionof the risk of anastomotic leakage and partial necrosis oftransferred gut conduit at esophageal resection.

In our setup, we focused on a novel (IPC in anastomotichealing) but “old” (IPC before IRI) method and coulddemonstrate a beneficial effect of IPC on the anastomoticstability of small bowel anastomoses. These functionalchanges (higher bursting pressure) were not associated withstructural changes (collagen content). A positive correlationbetween bursting pressure and hydroxyproline concentra-tion was found in several experimental studies for anasto-motic healing [30, 31] as well as a negative correlation [32,33]. A possible explanation for the different data isdiscussed to be an increase of noncollagenous substancesat the anastomotic region, thus the collagen concentrationremaining relatively unchanged [34]. Furthermore, since arapid increase of collagen content 3 days after surgery with

maximal values on the seventh POD was shown in the ratcolon [35, 36], significant differences in hydroxyprolineconcentration might be measured in further healing of thesmall bowel anastomoses but are not seen on fourth POD,even if the healing of small bowel anastomoses might bedifferent [34]. However, in a study on ileal anastomoses inrats, Posma et al. could show that profound mesentericischemia of 30 min resulted in a significantly decreasedanastomotic breaking strength on the third, fifth, andseventh PODs but differences in collagen content werenot seen when compared to the control. They discussed apossible difference in collagen quality which could havebeen damaged by IRI [32]. These data support our findingsthat IPC, i.e., a type of a “controlled IRI,” does not seem toaffect collagen quantity. Nevertheless, a mismatch betweenfunctional and structural changes in early anastomotichealing is not necessarily contradictory, since anastomoticstability does not only depend on the quantity and qualityof submucosal collagen but especially on local adhesions tofurther intestinal loops or fatty tissue. In our study, localadhesions were seen in all animals, but such macroscopicsmall differences are hard to quantify in a reliable manner.As we did not dissect the adhesions around the anastomoticsegment, differences in bursting pressure might reflect anindirect measure of varying adhesion development.

As data of the direct effect of IPC on anastomotichealing are lacking, we have to discuss the known impacton IRI. Certainly, the two procedures, IPC and IRI, areidentical as far as the technical performance goes—ischemia is induced followed by a consecutive reperfusionphase. The basic difference lies especially in the length ofthe ischemia time, since a short ischemic phase of 5–20 minis included in the preconditioning and has variouslypositive effects in different organ systems which have beenexposed to a longer ischemia interval (i.e., 45 min) andconsequent reperfusion injury [3–9]. Prolonged ischemiaresults in deleterious local and remote organ complicationswhich are significantly aggravated by the reperfusioninduced inflammatory response [37].

Hotter et al. [12] were the first to show the beneficialeffects of IPC prior to IRI in the intestine. He postulatedthat preconditioning triggers an increase in nitric oxidesynthesis. Sola et al. described the protective effect of IPCdue to nitric oxide in a small bowel transplantation modelin rats, where he performed a cold 90-min ischemiafollowed by a warm 60-min reperfusion period. IPC(10 min ischemia/5 min reperfusion) protected the intestinalgrafts from cold preservation and reperfusion injury [15]. Ina small bowel transplantation model in dogs, Ferencz et al.demonstrated that even four cycles (10 min ischemia/5 minreperfusion) of preconditioning prior to cold preservationcould moderate the severity of oxidative stress decreasingoxygen free radicals and increasing glutathione and

Fig. 2 a Overview of an anastomosis (IPC15/30). HE staining, ×10.Suture hole in the submucosa (black arrow). b The anastomotic region(CO), broken during histological preparation at the anastomotic line.HE staining, ×25. Ischemic injury of two villous tips, grade 2 (ellipse,see c). c Extension of the subepithelial space with moderate lifting ofthe epithelial layer from the lamina propria (star). HE staining, ×100

Int J Colorectal Dis (2009) 24:975–981 979

superoxide dismutase and activate the endogenous cellularadaptation in bowel tissue [18]. Unfortunately, both authorsdid not evaluate anastomotic healing after transplantation,but histologic samples of the small bowel mucosa showed asignificant mucosal injury after cold preservation whichdiminished significantly in preconditioned groups [15].Histomorphological changes during low flow states (i.e.,ischemia) occur in the intestinal mucosa. In a histologicalanalysis of small bowel samples, IPC (10 min ischemia/10 min reperfusion) was shown to attenuate morphologicalchanges (villus height, mucosal architecture, inflammatorychanges of lamina propria) by prolonged ischemiareperfusion [38]. In our study, a mucosal injury was seenin all groups, but it was focused to the direct surrounding ofthe anastomotic line, which, to some extent, might be dueto the sutures which hereby caused a reduced microcircu-lation. Although our histologic results did not showsignificant difference of mucosal injury grading at theanastomotic site in IPC groups, it should be mentioned thatwe found several animals with a normal villus structure(grade 0) at the anastomotic site in the three IPC groupswith significant difference in bursting pressure values butthere was none in the control and IPC15/20 groups. Thesemorphological changes could be related to a possibly bettervillus microcirculation due to antioxidative, antiapoptotic,and anti-inflammatory effects as shown by Mallick et al.[20] and Vlasov et al. [16] and underline our results offunctional anastomotic stability (bursting pressure).

As far as our data is comparable to the studies mentionedabove, we actually performed “only one cycle of acontrolled IRI,” i.e., IPC prior to anastomotic constructionwith similar short ischemia times of 10 and 15 minfollowed by reperfusion intervals of 20 and 30 min.Critically considering the preparation of the bowel walland the placing of the sutures while performing a smallbowel anastomosis, one should note that a locally pro-longed ischemia at the anastomotic region is alwayscreated. In a later phase, reperfusion develops duringanastomotic healing with a possible systemic reperfusioninjury. Regarding the different IPC protocols in our groups,we found a time dependency of the beneficial effect.Different IPC protocols were used in aforementionedstudies: after an ischemic interval of 5–20 min, the vascularclamp was removed and the small intestine reperfused for5–15 min—in some studies, even more cycles of IPC wereused [12, 13, 15, 16, 39]. A direct comparison of the studiesis not possible due to different experimental setups.However, to date, there is no ideal IPC protocol, so wedefined different times of brief vascular occlusions anddifferent intervals of reperfusion, which were even up todouble of those in previously published studies. On thebasis of our data, it is not possible to give a convenientexplanation for the discrepant data of the IPC effect, but

when comparing the mixed groups with either the sameischemia interval or the same reperfusion interval, the mixedgroup with the longer reperfusion interval of 30 min hadsignificantly higher bursting pressure values. In fact, thebeneficial effect seems to depend not only on ischemia butespecially on the reperfusion interval until performing theanastomosis. So, it could be speculated whether a prolongedreperfusion time in relation to the ischemia would result in amore beneficial effect on anastomotic stability.

When comparing our data to the previously mentionedclinical trials in surgery of the esophagus [25–27, 29] anddiscussing a potential clinical impact, it is important toaddress that one animal of an IPC group died due to anintestinal gangrene. This could be related to the occlusionof the SMA. Additionally, anastomotic insufficiency mustbe assumed in one animal of the CO group (incidence 14%)and two animals of the preconditioned groups (incidence6%) as perianastomotic abscesses were found. And even ifa coincidental event is possible, IPC might have influencedthese serious complications.

In conclusion, brief periods of ischemia followed bybrief periods of reperfusion directly before performing asmall bowel anastomosis have a time-dependent beneficialeffect on the anastomotic stability in the early postoperativeperiod. Beneficial histomorphological changes of themucosa have been shown in preconditioned animals. Ourdata point out a possible new clinical approach to improveanastomotic healing and thus to reduce the risk ofanastomotic insufficiency.

References

1. Fielding LP, Stewart-Brown S, Blesovsky L, Kearney G (1980)Anastomotic integrity after operations for large-bowel cancer: amulticentre study. Br Med J 281:411–414

2. Thompson SK, Chang EY, Jobe BA (2006) Clinical review:healing in gastrointestinal anastomoses, part I. Microsurgery26:131–136

3. Murry CE, Jennings RB, Reimer KA (1986) Preconditioning withischemia: a delay of lethal cell injury in ischemic myocardium.Circulation 74:1124–1136

4. Koti RS, Yang W, Dashwood MR, Davidson BR, Seifalian AM(2002) Effect of ischemic preconditioning on hepatic microcircu-lation and function in a rat model of ischemia reperfusion injury.Liver Transpl 8:1182–1191

5. Pang CY, Neligan P, Zhong A, He W, Xu H, Forrest CR (1997)Effector mechanism of adenosine in acute ischemic precondition-ing of skeletal muscle against infarction. Am J Physiol 273:R887–R895

6. Obermaier R, von Dobschuetz E, Drognitz O, Hopt UT, Benz S(2004) Ischemic preconditioning attenuates capillary no-reflowand leukocyte adherence in postischemic pancreatitis. Langen-becks Arch Surg 389:511–516

7. Glazier SS, O'Rourke DM, Graham DI, Welsh FA (1994)Induction of ischemic tolerance following brief focal ischemia inrat brain. J Cereb Blood Flow Metab 14:545–553

980 Int J Colorectal Dis (2009) 24:975–981

8. Sakurai M, Hayashi T, Abe K, Aoki M, Sadahiro M, Tabayashi K(1998) Enhancement of heat shock protein expression aftertransient ischemia in the preconditioned spinal cord of rabbits. JVasc Surg 27:720–725

9. Turman MA, Bates CM (1997) Susceptibility of human proximaltubular cells to hypoxia: effect of hypoxic preconditioning andcomparison to glomerular cells. Ren Fail 19:47–60

10. Du ZY, Hicks M, Winlaw D, Spratt P, Macdonald P (1996)Ischemic preconditioning enhances donor lung preservation in therat. J Heart Lung Transplant 15:1258–1267

11. Li Y, Roth S, Laser M, Ma JX, Crosson CE (2003) Retinalpreconditioning and the induction of heat-shock protein 27. InvestOphthalmol Vis Sci 44:1299–1304

12. Hotter G, Closa D, Prados M et al (1996) Intestinal precondition-ing is mediated by a transient increase in nitric oxide. BiochemBiophys Res Commun 222:27–32

13. Aksoyek S, Cinel I, Avlan D et al (2002) Intestinal ischemicpreconditioning protects the intestine and reduces bacterialtranslocation. Shock 18:476–480

14. Cinel I, Avlan D, Cinel L et al (2003) Ischemic preconditioningreduces intestinal epithelial apoptosis in rats. Shock 19:588–592

15. Sola A, De Oca J, Gonzalez R et al (2001) Protective effect ofischemic preconditioning on cold preservation and reperfusioninjury associated with rat intestinal transplantation. Ann Surg234:98–106

16. Vlasov TD, Smirnov DA, Nutfullina GM (2002) Preconditioningof the small intestine to ischemia in rats. Neurosci Behav Physiol32:449–453

17. Wu B, Ootani A, Iwakiri R et al (2004) Ischemic preconditioningattenuates ischemia–reperfusion-induced mucosal apoptosis byinhibiting the mitochondria-dependent pathway in rat smallintestine. Am J Physiol Gastrointest Liver Physiol 286:G580–G587

18. Ferencz A, Szanto Z, Borsiczky B et al (2002) The effects ofpreconditioning on the oxidative stress in small-bowel autotrans-plantation. Surgery 132:877–884

19. Mallick IH, Yang W, Winslet MC, Seifalian AM (2005) Ischaemicpreconditioning improves microvascular perfusion and oxygena-tion following reperfusion injury of the intestine. Br J Surg92:1169–1176

20. Mallick IH, Yang W, Winslet MC, Seifalian AM (2005) Protectiveeffects of ischemic preconditioning on the intestinal mucosalmicrocirculation following ischemia–reperfusion of the intestine.Microcirculation 12:615–625

21. Mallick IH, Yang W, Winslet MC, Seifalian AM (2004) Ischemia–reperfusion injury of the intestine and protective strategies againstinjury. Dig Dis Sci 49:1359–1377

22. Reddy GK, Enwemeka CS (1996) A simplified method for theanalysis of hydroxyproline in biological tissues. Clin Biochem29:225–229

23. Chiu CJ, McArdle AH, Brown R, Scott HJ, Gurd FN (1970)Intestinal mucosal lesion in low-flow states. I. A morphological,hemodynamic, and metabolic reappraisal. Arch Surg 101:478–483

24. Thornton FJ, Barbul A (1997) Healing in the gastrointestinal tract.Surg Clin North Am 77:549–573

25. Schroder W, Stippel D, Gutschow C, Leers J, Holscher AH (2004)Postoperative recovery of microcirculation after gastric tubeformation. Langenbecks Arch Surg 389:267–271

26. Lorentz T, Fok M, Wong J (1989) Anastomotic leakage afterresection and bypass for esophageal cancer: lessons learned fromthe past. World J Surg 13:472–477

27. Sekido M, Yamamoto Y, Minakawa H et al (2003) Use of the“supercharge” technique in esophageal and pharyngeal recon-struction to augment microvascular blood flow. Surgery 134:420–424

28. Akiyama S, Kodera Y, Sekiguchi H et al (1998) Preoperativeembolization therapy for esophageal operation. J Surg Oncol69:219–223

29. Akiyama S, Ito S, Sekiguchi H et al (1996) Preoperativeembolization of gastric arteries for esophageal cancer. Surgery120:542–546

30. Kologlu M, Yorganci K, Renda N, Sayek I (2000) Effect of localand remote ischemia–reperfusion injury on healing of colonicanastomoses. Surgery 128:99–104

31. Tireli GA, Salman T, Ozbey H, Abbasoglu L, Toker G, Celik A(2003) The effect of pentoxifylline on intestinal anastomotichealing after ischemia. Pediatr Surg Int 19:88–90

32. Posma LA, Bleichrodt RP, van Goor H, Hendriks T (2007)Transient profound mesenteric ischemia strongly affects thestrength of intestinal anastomoses in the rat. Dis Colon Rectum50:1070–1079

33. Siemonsma MA, de Hingh IH, de Man BM, Lomme RM,Verhofstad AA, Hendriks T (2003) Doxycycline improves woundstrength after intestinal anastomosis in the rat. Surgery 133:268–276

34. Jonsson K, Jiborn H, Zederfeldt B (1987) Collagen metabolism insmall intestinal anastomosis. Am J Surg 154:288–291

35. Brasken P (1991) Healing of experimental colon anastomosis. EurJ Surg Suppl 566:1–51

36. Brasken P, Renvall S, Sandberg M (1991) Fibronectin andcollagen gene expression in healing experimental colonic anasto-moses. Br J Surg 78:1048–1052

37. Pasupathy S, Homer-Vanniasinkam S (2005) Ischaemic precondi-tioning protects against ischaemia/reperfusion injury: emergingconcepts. Eur J Vasc Endovasc Surg 29:106–115

38. Sileri P, Sica G, Gentileschi P et al (2004) Ischemic precondition-ing protects intestine from prolonged ischemia. Transplant Proc36:283–285

39. Ishida T, Yarimizu K, Gute DC, Korthuis RJ (1997) Mechanismsof ischemic preconditioning. Shock 8:86–94

Int J Colorectal Dis (2009) 24:975–981 981