- Email: [email protected]
Lower limb ischaemia and reperfusion alters gut permeability
Eur J VascSurg 6, 158-163 (1992)
Lower Limb Ischaemia and Reperfusion Alters Gut Permeability* R. J. C o r s o n 1, I. S. P a t e r s o n 1 , S. T. O ' D w y e r 1 , P. R o w l a n d 2, E. K i r k m a n 3, R. A. Little 3, C. N. M c C o l l u m 1 University Departments of lSurgery and 2pathology, University Hospital of South Manchester, U.K. and 3North Western Injury Research Centre, University of Manchester, U.K. The gut may be important in the aetiology of multiple organfailure (MOF) by amplifying of the inflammatory response to trauma. We investigated the effects of the reperfusion of ischaemic lower limbs on gut permeability. Male Wistar rats (n = 30) were randomised to (group 1) controls; (group 2) 3 h bilateral hind-limb ischaemia alone; or 3 h ischaemiafollowed by (group 3) 15 min reperfusion or (group 4) 2 h reperfusion. Gut permeability and plasma endotoxin were measured prior to tourniquet application, immediately before tourniquet release, and following reperfusion. To evaluate the effect of the hypotension that follows tourniquet release, (group 5) sodium nitroprusside was infused &further controls to maintain mean arterial pressure (MAP) at 75 mmHg for 2 h. Horseradish peroxidase was instilled into the isolated ileo-caecal loop 15 min before the animals were killed to measure permeability of horseradish peroxidase through mucosal intercellular tight junctions by electron microscopy. Mean arterial pressure increased from 105 + 5mmHg to 136 + 4mmHg on tourniquet application and fell to 79 ± 7mmHg following reperfysion (p < 0.05). In group i (controls), group 2 (ischaemia alone animals) and group 5 (ischaemia and nitroprusside) one animal out of six demonstrated permeability to horseradish peroxidase. Following reperfusion, horseradish peroxidase permeability had not developed by 15 min (group 3) but was present in all animals by_2 h (group 4) (p = 0.015 Fisher's exact test). Plasma endotoxin increased from 21.8 + 2.0pgmI -I to 30.7 ++_2.6pgm1-1 following 2h reperfusion (p
Introduction
Multiple organ failure (MOF) frequently occurs in the critically ill without evidence of sepsis. 1'2 Activated inflammatory mediators such as those liberated u p o n the reperfusion of ischaernic tissue m a y be important in the aetiology. 3-5 Loss of mucosal barrier function in the gut m a y result in bacterial translocation and/or endotoxin absorption in the critically ill6-8 and m a y amplify this inflammatory response particularly following trauma. 9-11 Gut damage has been demonstrated following aortic surgery 12 but the mechanism by which this promotes MOF remains unclear. We studied the effects of lower limb ischaemia and reperfusion on gut mucosal barrier function and examined the role of the hypotension that follows reperfusion.
Methods
Male Wistar rats weighing 300 + 30 g were housed in a standardised environment of 12 h light per day at a constant 20°C with free access to a standard rat diet (CRM, Special Diet Services, U.K.) and water until the night before each experiment w h e n food but not water was withheld. Animals were anaesthetised with 5% isoflurane (Abbot, U.K.) in oxygen and sterile polythene cannulae were inserted into (i) a tail vein for the intermittent intravenous administration of sodium pentobarbitone (RMB Animal Health U.K.) at a rate of 1 8 m g k g lh-1 to maintain general anaesthesia once isoflurane had ceased, (ii) the ventral tail artery to monitor m e a n arterial pressure and (iii) the right external jugular vein to allow hydration * Presented at the 5th Annual Meeting of the European Society for using an infusion of 0.9% saline at the rate of Vascular Surgery, Warsaw, September 1991. 3 . 3 m l k g - l h -1 using a Fresenius Injectomat 30 inPlease address all correspondence to: R. J. Corson, University fusion p u m p (Runcorn, U.K.) and to allow venous Hospital of South Manchester, M20 8LR, U.K. sampling.
0950-821X/92/020158+06 $03.00/0© 1992Grune & StrattonLtd.
Remote Ischaemia Reperfusion and Gut Damage
Maintenance of endotoxin-free conditions All infusion fluids were pyrogen-free; cannulae were inserted under aseptic conditions and all disposables were sterile and pyrogen-free. Non-disposable equipment was washed in laboratory detergent, rinsed with deionised water, flushed with methylated spirits and allowed to dry before re-use. Blood was obtained and cultured aerobically and anaerobically immediately prior to laparotomy. Soda-glass blood collection bottles were autoclaved twice and heparinised with pyrogen-free heparin (Kabivitrum U.K.).
Experimental design Thirty animals were randomised to five groups: group 1 (controls) were maintained under general anaesthesia for 5 h. Bilateral hind-limb ischaemia was produced by the application of rubber band tourniquets as described by Rosenthal. ~3 Six animals underwent: (group 2) 3h ischaemia without reperfusion; (group 3) 3 h ischaemia followed by 15 rain of reperfusion; (group 4) 3 h ischaemia followed by 2 h reperfusion and (group 5) 3 h of ischaemia followed by an infusion of sodium nitroprusside (Sigma, U.K.; 6 txg ~1 ~ diluted in 5% glucose) infused at the rate of 28 txg min -1 kg -1 for 2 h with the tourniquets remaining in situ.
159
the ileal loop. After 15min the gut was removed, irrigated with 2.5% glutaraldehyde, 2% formaldehyde and 0.4% CaC12 in 0.1M sodium cacodylate buffer (pH 7.4) and fixed for 15rain. The loop was removed from fixative and cut into 0.4-cm transverse sections which were re-immersed in fresh fixative for 24 h, after which they were cut into 1-mm rings and bisected. The tissue was then rinsed in 0.1 M sodium cacodylate buffer (pH 7.4) for l h at room temperature, treated with 0.055M 3,3-diaminobenzidine tetrachloride in Tris pH 7.6 in the dark at 4°C for i h, then with added 0.1% H202 for I h, rinsed with 0.1 M sodium cacodylate buffer (pH 7.4) and finally postfixed in reduced 1% osmium tetroxide with 2% potassium ferrocyanide for l h at room temperature. Following a rinse with buffer and dehydration through a graded series of alcohols and propylene oxide the tissue was embedded in Agar 100 epoxy resin and 1 txm traverse sections were cut and stained with toluidine blue to check tissue orientation by light microscopy. If orientation was satisfactory ultra-thin 90-nm sections were cut, collected onto copper 200 mesh grids and stained with uranyl acetate and lead citrate. -Over 30 fields per animal were examined, each containing three to four intercellular junctions, by a pathologist who had no knowledge of the experimental group using transmission electron microscopy (Associated Electrical Industries U.K., El 801). Gross mucosal disruption and/or evidence of penetration of horseradish peroxidase beyond intercellular tight junctions indicated abnormal permeability.
Measurements Endotoxin measurements Physiological monitoring Arterial blood pressure was recorded continuously from the ventral tail artery using an Elcomatic blood pressure transducer, connected to a Lectromed chart recorder. Core temperature was monitored using a thermistor inserted into the rectum, attached to an ELLAB du3s monitor (Kings Lynn, U.K.) and maintained at 37.5 + 0.5°C by external heating as required.
Intestinal permeability Abnormal permeability was detected by the method originally described by Rhodes. 14 Fifteen minutes prior to killing, the distal 10 cm of ileum and caecum was isolated in continuity by silk ligatures. Using a 23 G needle, 5 mg of type VIA horseradish peroxidase (Sigma, U.K.) in I ml 0.9% saline was instilled into
Plasma endotoxin was measured before tourniquet application, immediately prior to tourniquet release and prior to laparotomy. Blood (0.5ml) was Withdrawn from the external jugular vein into a sterile soda-glass bottle containing preservative-free heparin and placed on ice. To maintain isovolaemia 0.5ml sterile 0.9% saline was administered. The blood was spun at 850 x g for 10min at 4°C and the resultant platelet-rich plasma was stored at -20°C in a Falcon test tube. Endotoxin was measured in pg m1-1 using the Coatest chromogenic Limulus amoebocyte lysate assay (Kabivitrum, U.K.).
Statistical analysis Data were expressed as means + s.E. Analysis of variance using the Scheffe F-test and Fisher's exact Eur J VascSurg Vol 6, March 1992
160
R.J. Corson et aL
150
test for small numbers were used as appropriate and values deemed statistically significant at a p-value of <0.05.
I10
Results CGI
Blood pressure M e a n a r t e r i a l p r e s s u r e r e m a i n e d u n a l t e r e d in c o n t r o l a n i m a l s ( g r o u p 1) t h r o u g h o u t t h e 5 h of a n a e s t h e s i a at 107 + 2 m m H g . O n a p p l i c a t i o n of t o u r n i q u e t s m e a n a r t e r i a l p r e s s u r e ( M A P ) i n c r e a s e d s i g n i f i c a n t l y to 133 + 3 m m H g b y I h of i s c h a e m i a a n d r e m a i n e d signific a n t l y e l e v a t e d at 118 + 3 m m H g b y 3 h (Fig. 1). 15o
7(
:50 0
I I
I 2
I 5
I 4
I 5
Time (h)
Fig. 2. Mean + s.F. arterial pressure throughout the experiment. Mean arterial pressure increased following tourniquet application in the nitroprusside animals (group 5). Mean arterial pressure fell after commencing the infusion of nitroprusside at 3 h. There was no significant difference when compared with the 2 h reperfusion animals. D, 3h ischaemia + 2h reperfusion (group 4); B, 3h ischaemia + 2 h nitroprusside (group 5).
& 7~ c~
70
o
I
I
I
2
1
I
I
3 4 5 Time (h) Fig. 1. Mean arterial pressure throughout the experiment. Following tourniquet application MAP increased significantly when compared with controls (p < 0.05) and remained elevated until tourniquet release when it fell and remained low during the 2h of reperfusion (p<0.05 compared with controls). Q - - - O , control (group 1); G G, 3h ischaemia (group 2); O O, 3h ischaemia + 15 min reperfusion (group 3); [2[B, 3 h ischaemia + 2 h reperfusion (group 4). R e l e a s e of t h e t o u r n i q u e t s a n d l o w e r l i m b r e p e r f u s i o n r e s u l t e d in a d r o p in b l o o d p r e s s u r e to 81 + 2 m m H g b y 15 m i n a n d 76 + 4 m m H g b y 2 h o f r e p e r f u s i o n . F o l l o w i n g 3 h of i s c h a e m i a n i t r o p r u s s i d e w a s t i t r a t e d to a c h i e v e a s i m i l a r r e d u c t i o n in M A P to 75 m m H g for 2 h (Fig. 2).
Permeability change Electron microscopic examination demonstrated horseradish peroxidase penetration from the intestinal l u m e n i n t o t h e i n t e r c e l l u l a r s p a c e (Figs 3 a n d 4). Eur J Vasc Surg Vol 6, March 1992
Fig. 3. Transmission electron micrograph (x30000) of the intercellular junction between two normal enterocytes showing normal slender microvilli (A), an intact tight junction and desmosome (B) and normal intercellular space (C).
Remote Ischaemia Reperfusion and Gut Damage
161
Table 1. H o r s e r a d i s h peroxidase permeability according to group
Group
Negative
Positive
Controls
5
1
2 h nitroprusside
5
1
3 h ischaemia
5
1
15 min reperfusion
6
0
2 h reperfusion
0
6*
*p = 0.015, Fisher's exact test, relative to controls (n = 6 all groups).
1.3pgml 1 increased minimally to 15.7 + 1.4pgm1-1 and 17.5 + 1.8pgml i after 3 and 5h respectively but this increase was not significant. Ischaemia alone (group 2) did not influence endotoxin at 16.3 + 1.7 pg ml- 1 initially and 18.0 + 1.6 pg ml - 1 by 3 h. The initial endotoxin level of 24.0 + 3.2pgm1-1 in the animals subjected to 15min reperfusion (group 3) was increased by ischaemia to 26.8 + 4.2pgm1-1 and fell following 15 min reperfusion 23.5 + 2.1 pg m1-1 but neither of these changes were significant. Following 2 h reperfusion (group 4) there was a significant increase in endotoxin to 30.7 + 2.6 pg m1-1 compared with the baseline value of 21.8 + 2.0 pg m1-1 and 25.5 + 1.m1-1 before tourniquet release (p < 0.05; Fig. 5). Fig. 4. Transmission electron micrograph (x30000) of the intercellular space between two damaged enterocytes showing shortened microvilli (A), damaged tight junction and desmosome (B) and horseradish peroxidase penetration into the intercellular space signifying disruption of the mucosal barrier (C).
Horseradish peroxidase penetration was evident in one animal out of six in group 1 (controls), group 2 (ischaemia alone) and group 5 (ischaemia plus nitroprusside), respectively. Despite the marked fall in arterial pressure on tourniquet removal there was no evidence of abnormal permeability by 15rain of reperfusion. However by 2h following tourniquet release there was clear evidence of gut damage with mucosal disruption, stunting of microvilli and horseradish peroxidase penetration beyond intercellular tight junctions in all six animals (p = 0.015; Table 1).
10[
E
£
8 6 4
E o
O
Control
3 h ischaemia
(group I)
(group 2)
3 h ischaemia + 5 h ischaemia + 15 min reperfusion 2 h reperfusion (group 3) (group 4)
Fig. 5. Changes in endotoxin levels w h e n compared to baseline. Increase in endotoxin during ischaemia (N.S.). No significant increase following 15min reperfusion; however, significant increase in endotoxin following 2 h reperfusion w h e n compared to baseline (* p < 0.05). II, ischaemia; D, reperfusion.
Endotoxin concentration
These levels were compared with the baseline level of each group and the normal limit for systemic endotoxin in this laboratory of 22.5 pgmin -1. In control animals (group 1) initial endotoxin levels of 15.3 +
Discussion
These studies demonstrate that reperfusion of ischaemic lower limbs is associated with a subsequent Eur J Vasc Surg Vol 6, March 1992
162
R.J. Corson et aL
increase in gut mucosal permeability and systemic endotoxaemia. This mucosal disruption was almost certainly the consequence of reperfusion as there were no important changes with ischaemia that was not reversed. We were concerned that the persistent hypotension following tourniquet release might have directly caused this injury. 15"16 For this reason MAP was monitored throughout the experiment to quantify the predicted changes in blood pressure following tourniquet application and release. 17This allowed us to mimic these changes with an infusion of sodium nitroprusside in a second control group studied to evaluate the effect of hypotension. No permeability changes were found in these animals which indicates that the increased permeability changes following reperfusion were not due to hypotension. The horseradish peroxidase technique is well established 14 and when instilled in an isotonic solution horseradish peroxidase penetration cannot occur in a healthy bowel. 18 It remains a qualitative test despite the pathologist being unaware of the origin of the sample and evidence of even one tight junction showing horseradish peroxidase penetration was considered abnormal. The mechanism for this increased gut permeability remains uncertain but the eicosanoids such as thromboxane A2 and leucotriene B4 and neutrophil-dependent free radical production may be important. 19-23 The gut may then precipitate or amplify MOF by the abnormal absorption of endogenous bacteria and/or endotoxin. 1° Indeed these factors have been implicated, for many years in the aetiology of MOF following a number of types of e x p e r i m e n t a l shock. 24' 25 Our results suggest that the increased gut permeability takes time to develop as there were no abnormalities after 15min of reperfusion. We are unable to say whether histological changes occur prior to the elevation in plasma endotoxin. If however endotoxin was absorbed prior to the demonstrable histological damage, endotoxin itself could be an important agent of increased permeability as it has been shown to alter mucosal permeability in normal bowel. 26 The use of anti-endotoxin monoclonal antibodies may help to resolve this question. 27 This finding that reperfusion of ischaemic tissue well away from the gut may damage the gut is new and points to the importance of the gut in the aetiololgy of MOF following major vascular surgery. The presence of this injury in humans requires to be confirmed and this is presently being studied in a prospective clinical trial in patients undergoing aortic aneurysm repair. Further animal studies to define the mechanism of this injury are also underway to investi_gate whether agents such as the thromboxane Eur J Vasc Surg Vol 6, March 1992
antagonists could have of any beneficial therapeutic effect.
Acknowledgements We would like to thank Mrs B. Crawley, and Dr B. Oppenheim of the Bacteriology Department, and Dr N. Y. Haboubi of the Pathology Department of the University Hospital of South Manchester for their help in performing this study. This research was funded by the BUPA foundation.
References 1 CARRICO CJ, MEAKINS JL, MARSHALLJ. Multi-organ failure syndrome. Arch Surg 1986; 121: 196-208. 2 GORIS RJA, BOEKHORSTTPA, NUYTINCKJKS, GIMBREREJS. Multiple organ failure. Generalised autodestructive inflammation. Arch Surg 1985; 120: 1109-1115. 3 KLAUSNERJM, ANNER H, KOBIZ L. Leg ischamia induced lung permeability requires thromboxane A2 and neutrophils. Fed Prod 1987; 44: 195. 4 KORTHIUSRJ, GRISHAM MB, GRANGERDN. Leukocyte depletion attenuates vascular injury in postischaemic muscle. Am J Physiol 1988; 254: H823-H827. 5 LELCUKS, ALEXANDERF, VALERICR, SHEPRO D, HECHTMAN HB. Thromboxane A2 moderates permeability after limb ischaemia. Ann Surg 1985; 202: 642-646. 6 AMBROSENS, JOHNSON DW, BURDONDW, KEIGHLEYMRB. Incidence of pathogenic bacteria from mesenteric lymph nodes and ileal serosa during Crohn's disease surgery. Brit J Surg 1984; 71: 623-625. 7 DRESS DL, GARRISON RN, FRY DE. Candida sepsis. Arch Surg 1985; 120: 345-348. 8 WILMORE DW, SMITH RJ, O'DWYER ST, JACOBSDO, ZEIGLERTR, WANG X-D. The gut: a central organ after surgical stress. Surgery 1988; 104: 917-923. 9 MAEJIMA K, DEITCH EA, BERGRD. Bacterial translocation from the gastrointestinal tracts of rats receiving thermal injury. Infect Immun 1984; 43: 6-10. 10 DEITCH EA, MA W-J, MA L, BERG R, SPECIAN RD. Endotoxininduced bacterial translocation: a study of mechanisms. Surgery 1989; 106: 292-300. 11 BORDER JR, HASSETT J, LADUCA J, et al. The gut origin septic states in blunt multiple trauma (ISS = 40) in the ICU. Ann Surg 1987; 206: 427-448. 12 FIDDIAN-GREEN RG. Splanchnic ischaemia and multi-organ failure in the critically ill. Ann R Coll Surg Eng 1988; 70: 128-134. 13 ROSENTHAL SM. Experimental chemotherapy of burns and shock. Pub Health Rep 1943; 58: 1429-1436. 14 RHODES RS, KARNOVSKY MJ. LOSS of macromolecular barrier function associated with surgical trauma to the intestine. Lab Invest 1971; 25: 220-229. 15 DEITCH EA, BRIDGES W, BAKER J, et al. Hemorrhagic shockinduced bacterial translocation is reduced by xanthine oxidase inhibition or inactivation. Surgery 1988; 104: 191-198. 16 FLYNN WJ, CRYER HG, GARRISON RN. Pentoxifylline restores intestinal microvascular blood flow during resuscitated hemorrhagic shock. Surgery 1991; 110: 350-356. 17 REOFERN WS, LITTLE RA, STONER HB, MARSHALLHW. Effect of limb ischaemia on blood pressure and the blood pressure-heart rate reflex in the rat. Q J Exp Physiol 1984; 69: 763-779. 18 MARCIALMA, MADARAJL. Analysis of absorptive cell occluding junction structure-function relationships in a state of enhanced junctional permeability. Lab Invest 1987; 56: 424-434.
Remote Ischaemia Reperfusion and Gut Damage
19 GRISHAMMB, HERNANDEZ LA, GRANGERDN. Xanthine oxidase and neutrophil infiltration in intestinal ischaemia. Am J Physiol 1986; 251: G567-G574. 20 PARKSDA, SHAN AK, GRANGERDN. Oxygen radicals: effects on intestinal vascular permeability. Am J Physiol 1984; 247: G167G170. 21 KLAUSNER JM, PATERSON IS, KOBZIK L, VALERI R, SHEPRO D, HECHTMAN HB. Leukotrienes but not complement mediate limb ischaemia induced lung injury. Ann Surg 1989; 209: 462-470. 22 PATERSON IS, KLAUSNERJM, GOLDMAN G, et al. Thromboxane mediates the ischaemia induced neutrophil oxidative burst. Surgery 1989; 106: 224-229. 23 KLAUSNER JM, PATERSON IS, KOBZIK L, VALERI R, SHEPRO D, HECHTMAN HB. Oxygen free radicals mediate ischaemiainduced tung injury. Surgery 1989; 105: 192-199. 24 SCHWINENBERGFB, DAVIDOFF D, KOVEN IH, FINE J. Host resis-
163
tence to bacteria in hemorrhagic shock. VI. Effect of endotoxin on antibacterial disease. Proc Soc Exp Biol Med 1956; 92: 662-667. 25 RAVIN HA, ROWEEVD, JENKINS C, FINE J. On the absorption of bacterial endotoxin from the gastro-intestinal tract of the normal and shocked animal. J Exp Med 1960; 112: 783-792. 26 O'DWYER ST, MICHIE HR, ZIEGLER TR, REVHAUGA, SMITH RJ, WILMORE DW. A single dose of endotoxin increases intestinal permeability in healthy humans. Arch Surg 1988; 123: 14591464. 27 ZIEGLER EJ, FISHER CJ, SPRUNG CL et al. Treatment of Gramnegative bacteraemia and septic shock with HA-1A h u m a n monoclonal antibody against endotoxin. N Engl J Med 1991; 324: 429-436.
Accepted 9 October i991
Eur J Vasc Surg Vol 6, March 1992
Lower Limb Ischaemia and Reperfusion Alters Gut Permeability* R. J. C o r s o n 1, I. S. P a t e r s o n 1 , S. T. O ' D w y e r 1 , P. R o w l a n d 2, E. K i r k m a n 3, R. A. Little 3, C. N. M c C o l l u m 1 University Departments of lSurgery and 2pathology, University Hospital of South Manchester, U.K. and 3North Western Injury Research Centre, University of Manchester, U.K. The gut may be important in the aetiology of multiple organfailure (MOF) by amplifying of the inflammatory response to trauma. We investigated the effects of the reperfusion of ischaemic lower limbs on gut permeability. Male Wistar rats (n = 30) were randomised to (group 1) controls; (group 2) 3 h bilateral hind-limb ischaemia alone; or 3 h ischaemiafollowed by (group 3) 15 min reperfusion or (group 4) 2 h reperfusion. Gut permeability and plasma endotoxin were measured prior to tourniquet application, immediately before tourniquet release, and following reperfusion. To evaluate the effect of the hypotension that follows tourniquet release, (group 5) sodium nitroprusside was infused &further controls to maintain mean arterial pressure (MAP) at 75 mmHg for 2 h. Horseradish peroxidase was instilled into the isolated ileo-caecal loop 15 min before the animals were killed to measure permeability of horseradish peroxidase through mucosal intercellular tight junctions by electron microscopy. Mean arterial pressure increased from 105 + 5mmHg to 136 + 4mmHg on tourniquet application and fell to 79 ± 7mmHg following reperfysion (p < 0.05). In group i (controls), group 2 (ischaemia alone animals) and group 5 (ischaemia and nitroprusside) one animal out of six demonstrated permeability to horseradish peroxidase. Following reperfusion, horseradish peroxidase permeability had not developed by 15 min (group 3) but was present in all animals by_2 h (group 4) (p = 0.015 Fisher's exact test). Plasma endotoxin increased from 21.8 + 2.0pgmI -I to 30.7 ++_2.6pgm1-1 following 2h reperfusion (p
Introduction
Multiple organ failure (MOF) frequently occurs in the critically ill without evidence of sepsis. 1'2 Activated inflammatory mediators such as those liberated u p o n the reperfusion of ischaernic tissue m a y be important in the aetiology. 3-5 Loss of mucosal barrier function in the gut m a y result in bacterial translocation and/or endotoxin absorption in the critically ill6-8 and m a y amplify this inflammatory response particularly following trauma. 9-11 Gut damage has been demonstrated following aortic surgery 12 but the mechanism by which this promotes MOF remains unclear. We studied the effects of lower limb ischaemia and reperfusion on gut mucosal barrier function and examined the role of the hypotension that follows reperfusion.
Methods
Male Wistar rats weighing 300 + 30 g were housed in a standardised environment of 12 h light per day at a constant 20°C with free access to a standard rat diet (CRM, Special Diet Services, U.K.) and water until the night before each experiment w h e n food but not water was withheld. Animals were anaesthetised with 5% isoflurane (Abbot, U.K.) in oxygen and sterile polythene cannulae were inserted into (i) a tail vein for the intermittent intravenous administration of sodium pentobarbitone (RMB Animal Health U.K.) at a rate of 1 8 m g k g lh-1 to maintain general anaesthesia once isoflurane had ceased, (ii) the ventral tail artery to monitor m e a n arterial pressure and (iii) the right external jugular vein to allow hydration * Presented at the 5th Annual Meeting of the European Society for using an infusion of 0.9% saline at the rate of Vascular Surgery, Warsaw, September 1991. 3 . 3 m l k g - l h -1 using a Fresenius Injectomat 30 inPlease address all correspondence to: R. J. Corson, University fusion p u m p (Runcorn, U.K.) and to allow venous Hospital of South Manchester, M20 8LR, U.K. sampling.
0950-821X/92/020158+06 $03.00/0© 1992Grune & StrattonLtd.
Remote Ischaemia Reperfusion and Gut Damage
Maintenance of endotoxin-free conditions All infusion fluids were pyrogen-free; cannulae were inserted under aseptic conditions and all disposables were sterile and pyrogen-free. Non-disposable equipment was washed in laboratory detergent, rinsed with deionised water, flushed with methylated spirits and allowed to dry before re-use. Blood was obtained and cultured aerobically and anaerobically immediately prior to laparotomy. Soda-glass blood collection bottles were autoclaved twice and heparinised with pyrogen-free heparin (Kabivitrum U.K.).
Experimental design Thirty animals were randomised to five groups: group 1 (controls) were maintained under general anaesthesia for 5 h. Bilateral hind-limb ischaemia was produced by the application of rubber band tourniquets as described by Rosenthal. ~3 Six animals underwent: (group 2) 3h ischaemia without reperfusion; (group 3) 3 h ischaemia followed by 15 rain of reperfusion; (group 4) 3 h ischaemia followed by 2 h reperfusion and (group 5) 3 h of ischaemia followed by an infusion of sodium nitroprusside (Sigma, U.K.; 6 txg ~1 ~ diluted in 5% glucose) infused at the rate of 28 txg min -1 kg -1 for 2 h with the tourniquets remaining in situ.
159
the ileal loop. After 15min the gut was removed, irrigated with 2.5% glutaraldehyde, 2% formaldehyde and 0.4% CaC12 in 0.1M sodium cacodylate buffer (pH 7.4) and fixed for 15rain. The loop was removed from fixative and cut into 0.4-cm transverse sections which were re-immersed in fresh fixative for 24 h, after which they were cut into 1-mm rings and bisected. The tissue was then rinsed in 0.1 M sodium cacodylate buffer (pH 7.4) for l h at room temperature, treated with 0.055M 3,3-diaminobenzidine tetrachloride in Tris pH 7.6 in the dark at 4°C for i h, then with added 0.1% H202 for I h, rinsed with 0.1 M sodium cacodylate buffer (pH 7.4) and finally postfixed in reduced 1% osmium tetroxide with 2% potassium ferrocyanide for l h at room temperature. Following a rinse with buffer and dehydration through a graded series of alcohols and propylene oxide the tissue was embedded in Agar 100 epoxy resin and 1 txm traverse sections were cut and stained with toluidine blue to check tissue orientation by light microscopy. If orientation was satisfactory ultra-thin 90-nm sections were cut, collected onto copper 200 mesh grids and stained with uranyl acetate and lead citrate. -Over 30 fields per animal were examined, each containing three to four intercellular junctions, by a pathologist who had no knowledge of the experimental group using transmission electron microscopy (Associated Electrical Industries U.K., El 801). Gross mucosal disruption and/or evidence of penetration of horseradish peroxidase beyond intercellular tight junctions indicated abnormal permeability.
Measurements Endotoxin measurements Physiological monitoring Arterial blood pressure was recorded continuously from the ventral tail artery using an Elcomatic blood pressure transducer, connected to a Lectromed chart recorder. Core temperature was monitored using a thermistor inserted into the rectum, attached to an ELLAB du3s monitor (Kings Lynn, U.K.) and maintained at 37.5 + 0.5°C by external heating as required.
Intestinal permeability Abnormal permeability was detected by the method originally described by Rhodes. 14 Fifteen minutes prior to killing, the distal 10 cm of ileum and caecum was isolated in continuity by silk ligatures. Using a 23 G needle, 5 mg of type VIA horseradish peroxidase (Sigma, U.K.) in I ml 0.9% saline was instilled into
Plasma endotoxin was measured before tourniquet application, immediately prior to tourniquet release and prior to laparotomy. Blood (0.5ml) was Withdrawn from the external jugular vein into a sterile soda-glass bottle containing preservative-free heparin and placed on ice. To maintain isovolaemia 0.5ml sterile 0.9% saline was administered. The blood was spun at 850 x g for 10min at 4°C and the resultant platelet-rich plasma was stored at -20°C in a Falcon test tube. Endotoxin was measured in pg m1-1 using the Coatest chromogenic Limulus amoebocyte lysate assay (Kabivitrum, U.K.).
Statistical analysis Data were expressed as means + s.E. Analysis of variance using the Scheffe F-test and Fisher's exact Eur J VascSurg Vol 6, March 1992
160
R.J. Corson et aL
150
test for small numbers were used as appropriate and values deemed statistically significant at a p-value of <0.05.
I10
Results CGI
Blood pressure M e a n a r t e r i a l p r e s s u r e r e m a i n e d u n a l t e r e d in c o n t r o l a n i m a l s ( g r o u p 1) t h r o u g h o u t t h e 5 h of a n a e s t h e s i a at 107 + 2 m m H g . O n a p p l i c a t i o n of t o u r n i q u e t s m e a n a r t e r i a l p r e s s u r e ( M A P ) i n c r e a s e d s i g n i f i c a n t l y to 133 + 3 m m H g b y I h of i s c h a e m i a a n d r e m a i n e d signific a n t l y e l e v a t e d at 118 + 3 m m H g b y 3 h (Fig. 1). 15o
7(
:50 0
I I
I 2
I 5
I 4
I 5
Time (h)
Fig. 2. Mean + s.F. arterial pressure throughout the experiment. Mean arterial pressure increased following tourniquet application in the nitroprusside animals (group 5). Mean arterial pressure fell after commencing the infusion of nitroprusside at 3 h. There was no significant difference when compared with the 2 h reperfusion animals. D, 3h ischaemia + 2h reperfusion (group 4); B, 3h ischaemia + 2 h nitroprusside (group 5).
& 7~ c~
70
o
I
I
I
2
1
I
I
3 4 5 Time (h) Fig. 1. Mean arterial pressure throughout the experiment. Following tourniquet application MAP increased significantly when compared with controls (p < 0.05) and remained elevated until tourniquet release when it fell and remained low during the 2h of reperfusion (p<0.05 compared with controls). Q - - - O , control (group 1); G G, 3h ischaemia (group 2); O O, 3h ischaemia + 15 min reperfusion (group 3); [2[B, 3 h ischaemia + 2 h reperfusion (group 4). R e l e a s e of t h e t o u r n i q u e t s a n d l o w e r l i m b r e p e r f u s i o n r e s u l t e d in a d r o p in b l o o d p r e s s u r e to 81 + 2 m m H g b y 15 m i n a n d 76 + 4 m m H g b y 2 h o f r e p e r f u s i o n . F o l l o w i n g 3 h of i s c h a e m i a n i t r o p r u s s i d e w a s t i t r a t e d to a c h i e v e a s i m i l a r r e d u c t i o n in M A P to 75 m m H g for 2 h (Fig. 2).
Permeability change Electron microscopic examination demonstrated horseradish peroxidase penetration from the intestinal l u m e n i n t o t h e i n t e r c e l l u l a r s p a c e (Figs 3 a n d 4). Eur J Vasc Surg Vol 6, March 1992
Fig. 3. Transmission electron micrograph (x30000) of the intercellular junction between two normal enterocytes showing normal slender microvilli (A), an intact tight junction and desmosome (B) and normal intercellular space (C).
Remote Ischaemia Reperfusion and Gut Damage
161
Table 1. H o r s e r a d i s h peroxidase permeability according to group
Group
Negative
Positive
Controls
5
1
2 h nitroprusside
5
1
3 h ischaemia
5
1
15 min reperfusion
6
0
2 h reperfusion
0
6*
*p = 0.015, Fisher's exact test, relative to controls (n = 6 all groups).
1.3pgml 1 increased minimally to 15.7 + 1.4pgm1-1 and 17.5 + 1.8pgml i after 3 and 5h respectively but this increase was not significant. Ischaemia alone (group 2) did not influence endotoxin at 16.3 + 1.7 pg ml- 1 initially and 18.0 + 1.6 pg ml - 1 by 3 h. The initial endotoxin level of 24.0 + 3.2pgm1-1 in the animals subjected to 15min reperfusion (group 3) was increased by ischaemia to 26.8 + 4.2pgm1-1 and fell following 15 min reperfusion 23.5 + 2.1 pg m1-1 but neither of these changes were significant. Following 2 h reperfusion (group 4) there was a significant increase in endotoxin to 30.7 + 2.6 pg m1-1 compared with the baseline value of 21.8 + 2.0 pg m1-1 and 25.5 + 1.m1-1 before tourniquet release (p < 0.05; Fig. 5). Fig. 4. Transmission electron micrograph (x30000) of the intercellular space between two damaged enterocytes showing shortened microvilli (A), damaged tight junction and desmosome (B) and horseradish peroxidase penetration into the intercellular space signifying disruption of the mucosal barrier (C).
Horseradish peroxidase penetration was evident in one animal out of six in group 1 (controls), group 2 (ischaemia alone) and group 5 (ischaemia plus nitroprusside), respectively. Despite the marked fall in arterial pressure on tourniquet removal there was no evidence of abnormal permeability by 15rain of reperfusion. However by 2h following tourniquet release there was clear evidence of gut damage with mucosal disruption, stunting of microvilli and horseradish peroxidase penetration beyond intercellular tight junctions in all six animals (p = 0.015; Table 1).
10[
E
£
8 6 4
E o
O
Control
3 h ischaemia
(group I)
(group 2)
3 h ischaemia + 5 h ischaemia + 15 min reperfusion 2 h reperfusion (group 3) (group 4)
Fig. 5. Changes in endotoxin levels w h e n compared to baseline. Increase in endotoxin during ischaemia (N.S.). No significant increase following 15min reperfusion; however, significant increase in endotoxin following 2 h reperfusion w h e n compared to baseline (* p < 0.05). II, ischaemia; D, reperfusion.
Endotoxin concentration
These levels were compared with the baseline level of each group and the normal limit for systemic endotoxin in this laboratory of 22.5 pgmin -1. In control animals (group 1) initial endotoxin levels of 15.3 +
Discussion
These studies demonstrate that reperfusion of ischaemic lower limbs is associated with a subsequent Eur J Vasc Surg Vol 6, March 1992
162
R.J. Corson et aL
increase in gut mucosal permeability and systemic endotoxaemia. This mucosal disruption was almost certainly the consequence of reperfusion as there were no important changes with ischaemia that was not reversed. We were concerned that the persistent hypotension following tourniquet release might have directly caused this injury. 15"16 For this reason MAP was monitored throughout the experiment to quantify the predicted changes in blood pressure following tourniquet application and release. 17This allowed us to mimic these changes with an infusion of sodium nitroprusside in a second control group studied to evaluate the effect of hypotension. No permeability changes were found in these animals which indicates that the increased permeability changes following reperfusion were not due to hypotension. The horseradish peroxidase technique is well established 14 and when instilled in an isotonic solution horseradish peroxidase penetration cannot occur in a healthy bowel. 18 It remains a qualitative test despite the pathologist being unaware of the origin of the sample and evidence of even one tight junction showing horseradish peroxidase penetration was considered abnormal. The mechanism for this increased gut permeability remains uncertain but the eicosanoids such as thromboxane A2 and leucotriene B4 and neutrophil-dependent free radical production may be important. 19-23 The gut may then precipitate or amplify MOF by the abnormal absorption of endogenous bacteria and/or endotoxin. 1° Indeed these factors have been implicated, for many years in the aetiology of MOF following a number of types of e x p e r i m e n t a l shock. 24' 25 Our results suggest that the increased gut permeability takes time to develop as there were no abnormalities after 15min of reperfusion. We are unable to say whether histological changes occur prior to the elevation in plasma endotoxin. If however endotoxin was absorbed prior to the demonstrable histological damage, endotoxin itself could be an important agent of increased permeability as it has been shown to alter mucosal permeability in normal bowel. 26 The use of anti-endotoxin monoclonal antibodies may help to resolve this question. 27 This finding that reperfusion of ischaemic tissue well away from the gut may damage the gut is new and points to the importance of the gut in the aetiololgy of MOF following major vascular surgery. The presence of this injury in humans requires to be confirmed and this is presently being studied in a prospective clinical trial in patients undergoing aortic aneurysm repair. Further animal studies to define the mechanism of this injury are also underway to investi_gate whether agents such as the thromboxane Eur J Vasc Surg Vol 6, March 1992
antagonists could have of any beneficial therapeutic effect.
Acknowledgements We would like to thank Mrs B. Crawley, and Dr B. Oppenheim of the Bacteriology Department, and Dr N. Y. Haboubi of the Pathology Department of the University Hospital of South Manchester for their help in performing this study. This research was funded by the BUPA foundation.
References 1 CARRICO CJ, MEAKINS JL, MARSHALLJ. Multi-organ failure syndrome. Arch Surg 1986; 121: 196-208. 2 GORIS RJA, BOEKHORSTTPA, NUYTINCKJKS, GIMBREREJS. Multiple organ failure. Generalised autodestructive inflammation. Arch Surg 1985; 120: 1109-1115. 3 KLAUSNERJM, ANNER H, KOBIZ L. Leg ischamia induced lung permeability requires thromboxane A2 and neutrophils. Fed Prod 1987; 44: 195. 4 KORTHIUSRJ, GRISHAM MB, GRANGERDN. Leukocyte depletion attenuates vascular injury in postischaemic muscle. Am J Physiol 1988; 254: H823-H827. 5 LELCUKS, ALEXANDERF, VALERICR, SHEPRO D, HECHTMAN HB. Thromboxane A2 moderates permeability after limb ischaemia. Ann Surg 1985; 202: 642-646. 6 AMBROSENS, JOHNSON DW, BURDONDW, KEIGHLEYMRB. Incidence of pathogenic bacteria from mesenteric lymph nodes and ileal serosa during Crohn's disease surgery. Brit J Surg 1984; 71: 623-625. 7 DRESS DL, GARRISON RN, FRY DE. Candida sepsis. Arch Surg 1985; 120: 345-348. 8 WILMORE DW, SMITH RJ, O'DWYER ST, JACOBSDO, ZEIGLERTR, WANG X-D. The gut: a central organ after surgical stress. Surgery 1988; 104: 917-923. 9 MAEJIMA K, DEITCH EA, BERGRD. Bacterial translocation from the gastrointestinal tracts of rats receiving thermal injury. Infect Immun 1984; 43: 6-10. 10 DEITCH EA, MA W-J, MA L, BERG R, SPECIAN RD. Endotoxininduced bacterial translocation: a study of mechanisms. Surgery 1989; 106: 292-300. 11 BORDER JR, HASSETT J, LADUCA J, et al. The gut origin septic states in blunt multiple trauma (ISS = 40) in the ICU. Ann Surg 1987; 206: 427-448. 12 FIDDIAN-GREEN RG. Splanchnic ischaemia and multi-organ failure in the critically ill. Ann R Coll Surg Eng 1988; 70: 128-134. 13 ROSENTHAL SM. Experimental chemotherapy of burns and shock. Pub Health Rep 1943; 58: 1429-1436. 14 RHODES RS, KARNOVSKY MJ. LOSS of macromolecular barrier function associated with surgical trauma to the intestine. Lab Invest 1971; 25: 220-229. 15 DEITCH EA, BRIDGES W, BAKER J, et al. Hemorrhagic shockinduced bacterial translocation is reduced by xanthine oxidase inhibition or inactivation. Surgery 1988; 104: 191-198. 16 FLYNN WJ, CRYER HG, GARRISON RN. Pentoxifylline restores intestinal microvascular blood flow during resuscitated hemorrhagic shock. Surgery 1991; 110: 350-356. 17 REOFERN WS, LITTLE RA, STONER HB, MARSHALLHW. Effect of limb ischaemia on blood pressure and the blood pressure-heart rate reflex in the rat. Q J Exp Physiol 1984; 69: 763-779. 18 MARCIALMA, MADARAJL. Analysis of absorptive cell occluding junction structure-function relationships in a state of enhanced junctional permeability. Lab Invest 1987; 56: 424-434.
Remote Ischaemia Reperfusion and Gut Damage
19 GRISHAMMB, HERNANDEZ LA, GRANGERDN. Xanthine oxidase and neutrophil infiltration in intestinal ischaemia. Am J Physiol 1986; 251: G567-G574. 20 PARKSDA, SHAN AK, GRANGERDN. Oxygen radicals: effects on intestinal vascular permeability. Am J Physiol 1984; 247: G167G170. 21 KLAUSNER JM, PATERSON IS, KOBZIK L, VALERI R, SHEPRO D, HECHTMAN HB. Leukotrienes but not complement mediate limb ischaemia induced lung injury. Ann Surg 1989; 209: 462-470. 22 PATERSON IS, KLAUSNERJM, GOLDMAN G, et al. Thromboxane mediates the ischaemia induced neutrophil oxidative burst. Surgery 1989; 106: 224-229. 23 KLAUSNER JM, PATERSON IS, KOBZIK L, VALERI R, SHEPRO D, HECHTMAN HB. Oxygen free radicals mediate ischaemiainduced tung injury. Surgery 1989; 105: 192-199. 24 SCHWINENBERGFB, DAVIDOFF D, KOVEN IH, FINE J. Host resis-
163
tence to bacteria in hemorrhagic shock. VI. Effect of endotoxin on antibacterial disease. Proc Soc Exp Biol Med 1956; 92: 662-667. 25 RAVIN HA, ROWEEVD, JENKINS C, FINE J. On the absorption of bacterial endotoxin from the gastro-intestinal tract of the normal and shocked animal. J Exp Med 1960; 112: 783-792. 26 O'DWYER ST, MICHIE HR, ZIEGLER TR, REVHAUGA, SMITH RJ, WILMORE DW. A single dose of endotoxin increases intestinal permeability in healthy humans. Arch Surg 1988; 123: 14591464. 27 ZIEGLER EJ, FISHER CJ, SPRUNG CL et al. Treatment of Gramnegative bacteraemia and septic shock with HA-1A h u m a n monoclonal antibody against endotoxin. N Engl J Med 1991; 324: 429-436.
Accepted 9 October i991
Eur J Vasc Surg Vol 6, March 1992