Bacterial translocation in an experimental intestinal obstruction model . C-reactive protein reliability ?

Background: Bacterial translocation occurs in preseptic conditions such as intestinal obstruction through unclear mechanism. The C-reactive protein is an acute phase reactant and a marker of ischemia. Methods: 45 albino male rats were divided into 3 groups each 15 rats. GI control, GII simple intestinal-obstruction and GIII strangulated obstruction. Outcome measures were: (1) Bacteriologic count and typing for intestinal contents, intestinal wall, liver, mesenteric lymph nodes and blood (cardiac and portal) (2) Histopathologic: mucosal injury score, inflammatory cell infiltrate in the wall, MLN, liver, (3) Biochemical: serum CRP, IL-10, mucosal stress pattern (glutathione peroxidase-malonyldialdhyde tissue levels). Results: (1) Intestinal obstruction associates with BT precursors (Bact-overgrowth, mucosal-acidosis, immuno-incomptence), (2) Bacterial translocation (frequency and density) was found higher in strangulated I.O, that was mainly enteric (aerobic and anaerobic) and mostly E.coli, (3) The pathogen commonality supports the gut origin hypothesis but the systemic inflammatory response goes with the cytokine generating one. (4) The CRP median values for GI, II, III were 0.5, 6.9, 8.5 mg/L, for BT +ve 8 mg/L and 0.75 mg/L for BT –ve rats. Conclusion: Bacterial translocation occurs bi-directional (systemic-portal) in intestinal obstruction and the resultant inflammatory response pathogenesis is mostly 3 hit model. The CRP is a non selective marker of suspected I.O cases. However, it is a reliable marker of BT, BT density and vascular compromise during I.O.


Introduction
Intestinal obstruction (IO) is a common lethal abdominal emergency resulting in high mortality, mostly due to multiorgan dysfunction syndrome (MODS) 1 , significantly bacterial translocation (BT) together with septic peritonitis are the major contributors of MODS in IO 2 .
Early studies focused on BT as a unifying mechanism to explain MODS but recently other specific mechanisms are operational (immuno inflammatory) 3 .
Bacterial translocation is precipitated by bacterial overgrowth disturbing the normal ecologic balance 4,5 , host immunedysfunction inciting pro and anti-inflammatory cytokines balance 6 , and mucosal barrier dysfunction, favoring oxidants release 7 .
Apart from computerized tomography, no reliable diagnostic test for intestinal strangulation is currently available, that is costy and not reproducible 8 .
Lastly the C-reactive protein (CRP) which is an inflammatory marker, is considered a marker of ischemia and neovascularization 9,10 , hence validated in this study for detection of IO, IO subtypes and BT.

Aim of work
Is to study the possible effects of I.O on the microbiologic ecologic balance, chemical and immunologic barriers that bring protection against B.T., to demonstrate BT during I.O (pathogen typing, routes and commonality) in addition to the BT local and systemic inflammatory response as well as CRP reliability in studying BT and I.O.

Methods
This study was conducted in Mansoura Faculty of Medicine, Histology experimental laboratory unit from June 2003 to June 2006 that entails 45 albino male rats divided into 3 group each 15 rats.Group I as a control, Group II represents simple I.O group (simple ileal ligation 5 cm proximal to caecum) and group III comprises strangulated I.O group (symmetrical ligation of a 5 cm ileal loop with its mesentry, 5 cm proximal to caecum).

Surgical maneuver
Anaesthesia with intramuscular ketamine, 5 mg/kg body weight for GII and GIII, followed by midline laparotomy after sterilization to perform the intented I.O type and layered abdominal closure using vicryloo.
Subsequent to recovery all groups were held in a semi-acclimitized room at 23ºC (±2), both laboratory chow and tap water were allowed for 28 hours till relaparotomy.
In the 2 nd laparotomy (same anaesthetic procedure for the three groups).
Under complete sterile conditions a thoracoabdominal generous midline incision was performed and firstly direct cardiac blood sampling, second portal venous blood sampling, thirdly left hepatic lobe resection, fourthly multiple mesenteric lymph nodes excision, and fifthly ileal segment proximal to ligature i.e obstruction level in GII, strangulated ileal loop in GIII and ileal segment in GI resection together with their luminal contents.Lastly animals were sacrificed via cervical dislocation.

II. Histopathologic study
(1) Ileal segments were examined to score the mucosal injury 13 .
(1) Luminal contents were homogenized in sterile isotonic saline, plated on McConkay and Columbia blood agar (aerobic and anaerobic) (Oxid-Germany), (2) For the intestinal wall, liver, MLN.The tissues were ground in phosphate buffered saline, then plated as the contents and incubated in the Gas Park system for anaerobic culture, (3) For the blood samples they are centrifuged at 3000 rpm for 30 min and the sediments were plated as the tissues.

Statistical analysis
• The CRP values are expressed as median, other variables as mean ± SD and the CFU are logarithmically converted.
• For comparison the Chi-square test, Mann Whitney U test and one way ANOVA test are used when applicable.
• The pearson correlation test was used to detect CRP levels correlation with CFU density within different tissues, meanwhile the logistic regression was used to detect the significant predictors.

Results
Intestinal obstruction mainly the strangulated type is significantly associated with bacterial overgrowth, oxidative stress pattern (disproportionate MDA and GPx increase) and deranged IL-10 response (significant decrement of IL-10 levels) (Table 1).The frequency and density of BT were higher in GIII than in GII but not detected in GI, their distribution was centrifugal (lumen -wall -MLN -liver -blood) (Table 2).The bacterial translocation was commonly polymicrobial (49% in GII -60% in GIII), mostly enteric with substantial anaerobic ratio and E.coli predominance (42.2% in GII and 45.5% in GIII) (Table 3).The pathogen commonality in the simple I.O was highest lumen to wall, lowest lumen to blood (Figure 1A) and in the strangulated group was highest lumen to wall, blood, MLN and modest lumen to blood (Figure 1B).The mucosal injury score and inflammatory cell infiltrate were significantly higher in GIII (Figure 2).Also the local and remote inflammatory response (liver + MLN inflammatory cell infiltrates) was higher in GIII (Table 4).The CRP median value was significantly higher in I.O and its subtypes than control (Table 5) with higher 75% percentile ratio in GIII than in GII (Figure 3).

GII I.O GIII I.O
The CRP median value was significantly higher in cases of BT during I.O compared to -ve BT (Table 6) and its cutoff level significantly predicted to -ve cases resulting in good exclusion power on validation (sensitivity: 77.7 %, specificity: 33.3 %, accuracy: (66.6 %), also the logistic regression defined CRP as a significant predictor of BT during I.O [(P = 0.002*** OR = 3.074 and 95% CI (6.3 -1.492)].The frequency and density of BT were parallel to CRP levels in the studied tissues whether simple or strangulated obstruction but their correlations were insignificant (Figure 4).

Discussion
Likely to the disturbed dynamic milieu of the intestinal tract during intestinal obstruction, this study defined bacterial overgrowth association with I.O in line with Akin et al. 15 and Sagar et al. 16 , also the oxidative stress pattern during I.O produce ATP depletion 17 , cytoskeleton disruption 18,19 , neutrophil priming 20 resulting in mechanical gut barrier dysfunction, and significantly the I.O association with impaired IL-10 response as reported by Zingarelli et al. 21and Souza et al. 22 declare the immune incompetence status.Although O'Boyle 23 deny those associations.
In accordance with Akcay et al. 24 ; Antequera et al. 25 ; Kocdor et al. 26 and Souza et al. 22 BT occurred during I.O as reported here, and the centripetal decrement in the frequency and density of the pathogens supports tissue colonization is gut derived not blood derived.This study also defined BT is bidirectional during I.O specially in the ischemic variant as reported by Wells et al. 27 and Mainous et al. 28 and that is related to direct intestinal wall insult.
Further support for the gut origin hypothesis during I.O from this study is the enteric bacteria predominance as detected by Brooks et al. 29 and MacFie et al. 30 .Moreover, E.coli outnumber other pathogens and this is mostly related to its facultative nature and its fimbriated surface i.e. colonizing factor thus supporting lymphatic route for BT.Furthermore, the obligate anaerobic organisms isolates detection as reported by Boedeker 31 ; O'Boyle et al. 32 and Cevikel et al. 1 define the colonization resistance failure in the ischemic intestinal obstruction.So bacterial overgrowth, bacterial virulence and wall integrity (structure and function) work together.
In this study the commonality of pathogens define the gut-origin hypothesis, provides additional support for local transmural route (Phagocytes or enterocytes) in line with Deitch 4 ; Brooks et al. 29 and MacFie et al. 30 who defined lymphatic route predominance in the simple I.O and venous portal predominance in the ischemic variant as found by; Moore et al. 33 ; Lemaire et al. 34 ; Adams et al. 35 and Kocdor et al. 26 .
The observed local and systemic immunoinflamamtory histopathologic changes as Akcay et al. 24 reported, might be related to cytokines release producing inflammatory cell influx resulting in tissue injury supporting the cytokine generating hypothesis 36 .Consequently we believe the 3 hit model 2 during I.O 1 st (increased intestinal pressure or ischemia), 2 nd (reperfusion injury) (increased secretion and decreased absorptionprostaglandin release -collaterals opening -distension) resulting gut barrier failure and 3 rd bacterial and cytokine translocation.
Despite CRP surge during I.O, its cutoff value didn't predict to any subtypes with consequent average accuracy in detection of I.O subtypes, so CRP is a non selective marker in suspected cases and confirming information from medical history and physical examination must be scrutinized to see if it support or contradicts CRP +ve cases.But significantly once I.O is diagnosed the CRP is a significant predictor of the ischemic variant as reported by Willet et al. 37 .
The high CRP level was associated with B.T during I.O and its subtypes.The ability of its cutoff level to define BT with high sensitivity, specificity and accuracy was associated with a statistically significant predictor value in accordance with Cevikel et al. 1 making CRP a reliable test to detect BT during I.O.
Complementary to previously found the parallel relation between CRP level and BT (frequency and density) specially the ischemic variant is mostly related to the cascades of systemic inflammatory response mediators as described by Moore 38 .So, CRP can be considered a predictor of vascular compromise and BT severity.
Conclusively, intestinal obstruction by its types is associated with BT precursors.BT is functioning during I.O, has bidirectional routes most pathogens are enteric, specially E.coli with obligate anaerobe occasionally in the ischemic variant.Bacterial overgrowth, virulence and wall structure and function work together.Both the BT and cytokine generating hypothesis are operational during I.O, and the 3 hit model is the appropriate model.The CRP is a non selective diagnostic marker in suspected cases of I.O but once diagnosed is a significant predictor of its subtypes.The CRP is a reliable test of BT during I.O.The CRP is a predictor of vascular compromise and BT severity.

FIGURE 2 -
FIGURE 2 -Photomicrographs of the local inflammatory response in the studied groups (100X)

2 FIGURE 3 -
FIGURE 3 -Box blot of CRP level in simple and strangulated intestinal obstruction

FIGURE 4 -
FIGURE 4 -Scatter plots showing relation between logarithmic BT density and CRP level in both simple and strangulated I.O within the wall (a), MLN (B), liver (C) and systemic blood (D)

TABLE 1 -
Bacterial translocation predisposing factors among the studied groups

TABLE 2 -
Frequency and density of bacterial translocation in the tissues among the studied groups

TABLE 4 -
The local and remote inflammatory response

TABLE 5 -
CRP median value and frequency of positive cases (cutoff level 0.5 mg%)

TABLE 6 -
CRP relation with bacterial translocation in intestinal obstruction and frequency of positive cases