| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2004;77:612-618
© 2004 The Society of Thoracic Surgeons
a Department of Anesthesia, Unit of Cardiac Anesthesia, Rome, Italy
b Department of Surgery, Rome, Italy
c Department of Medicine, Rome, Italy
d ENEA, Section of Toxicology and Biomedicine, Rome, Italy
e Department of Microbiology, Rome, Italy
f Department of Nuclear Medicine, Rome, Italy
g Department of Cardiac Surgery, Università Cattolica del Sacro Cuore, Policlinico "A. Gemelli," Rome, Italy
Accepted for publication August 1, 2003.
* Address reprint requests to Dr Rossi, Department of Anesthesia, Università Cattolica del S. Cuore Policlinico A. Gemelli, largo A. Gemelli n.8, 00168 Rome, Italy
e-mail: cardioanucsc{at}rm.unicatt.it
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and methodsResultsCommentReferences |
|---|
METHODS: Eleven selected patients scheduled for elective coronary artery bypass graft surgery were enrolled in a open, prospective clinical study. Gastric tonometry, chromium-labeled test and double sugar intestinal absorption tests, polymerase chain reaction microbial DNA test, and measurement of cytokines and transcriptional factor (nuclear factor 
B) activation were performed.
RESULTS: During the postoperative period, gastric pH remained stable (range,7.2 to 7.3). The partial pressure for carbon dioxide gradient between the gastric mucosa and arterial blood increased significantly (from 1 to 23 mm Hg), peaking in the sixth postoperative hour. Interleukin 6 increased significantly over basal levels, peaking 3 hours after cardiopulmonary bypass (96.3 versus 24 pg/mL). Nuclear factor B never reached levels higher than those observed after lipopolysaccharide stimulation. Escherichia coli translocation was documented in 10 patients: in eight cases from removal of aortic cross-clamps and in two cases from the first postoperative hour. With respect to basal value (6.4%), the urine collection revealed a significant increase in excretion of the radioisotope during the first 24 hours after surgery (39.1%), although there were no significant variations with the double sugar test.
CONCLUSIONS: The results obtained showed a correlation between the damage of the gastrointestinal mucosa, subsequent increased permeability, E coli bacteremia, and the activation of a self-limited inflammatory response in the absence of significant macrocirculatory changes and postoperative complications.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentReferences |
|---|
Cardiopulmonary bypass (CPB), routinely used in cardiac surgery, also provokes a systemic inflammatory reaction, characterized by fever, leukocytosis, and tachycardia that, in 1% to 2% of all cases, leads to multiple organ dysfunction syndrome [4, 5]. The inflammatory stimulus is primarily related to the bioincompatibility of the materials used in the circuit, which activate various humoral cascades and blood elements, and to the maldistribution of blood flow, which can lead to ischemic damage in sensitive organs like the intestine [6]. Gut involvement and the possible translocation of bacteria into the bloodstream can amplify the inflammatory response.
In a highly selected population of uncomplicated patients undergoing elective cardiac surgery, CPB can provide a useful model for the study of the inflammatory reaction, and the pathogenetic events that cause it. We conducted a multivariable study of patients receiving CPB during elective coronary artery bypass graft operation in an attempt to identify elements characteristic of intestinal damage and the development of bacterial translocation during CPB as possible causes of bacteremia and intensification of the perioperative inflammatory response.
|
Material and methods |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentReferences |
|---|
Strict exclusion criteria for study enrollment were age more than 70 years, preoperative left ventricular ejection fraction less than 0.40, acute myocardial infarction less than a month before surgery, previous coronary artery bypass graft, inflammatory bowel disease, peptic ulcer, previous abdominal surgery, liver disease, preoperative serum creatinine more than 1.6 mg/dL, diabetes mellitus, peripheral vascular disease, previous cerebrovascular accidents, or preoperative antibiotic or corticosteroid therapy.
A balanced anesthesia with sufentanil, propofol, and sevoflurane was used in all cases. Hemodynamic variables were continuously monitored by means of an 8F Intellicath catheter positioned in the pulmonary artery and connected to a Vigilance monitor (Baxter Healthcare Corp, Irvine CA). The end point was the perioperative maintenance of a cardiac index more than 2.2 L · min-1 · m-2 by volemia optimization or inotropic support if the volumetric challenge failed.
All patients underwent normothermic CPB with nonpulsatile flows of at least 2.4 L · min-1 · m-2 and hollow-fiber membrane oxygenators (Sorin Monolyth, Mirandola, Italy); cardiac arrest was achieved with normothermic blood potassium cardioplegia. Hematocrit was kept not less than 24% on CPB and 28% in the intensive care unit.
Table 1 shows the various data collection points used in the study.
|
Intestinal permeability
Intestinal absorption was evaluated by means of the chromium-labeled test (51Cr-EDTA) and the double sugar (lactulose and mannitol) absorption test. Both tests had been used to determine basal permeability 72 hours before the day of surgery. Radioisotope studies were performed after oral administration of an aqueous solution (10 mL) containing 2.96 MBq of 51Cr-EDTA, as previously described [7]. The 51Cr-EDTA solution was readministered the day of surgery, 1 hour before the induction of anesthesia, and a fractioned urine collection was performed every 6 hours through the 48th hour after ingestion. Urinary excretion of 51Cr-EDTA was expressed as a fraction of the orally administered dose. Normal values of urinary excretion are less than 3% of the amount of orally administered 51Cr-EDTA.
The sugar absorption test was performed at the end of CPB, and on the 2nd and 5th postoperative days. After oral administration of a solution containing 5 g of mannitol and 10 g of lactulose in 150 mL of distilled water, a 5-hour urine collection was obtained for measurement of mannitol and lactulose levels, as described elsewhere [8]. The results were expressed as ratios of urinary lactulose to urinary mannitol.
Bacterial translocation
Polymerase chain reaction (PCR) positivity was used as a qualitative marker of bacterial translocation. Several microbial strains of intestinal origin detectable with the PCR method were tested: Citrobacter freundii, Enterobacter aerogenes, Enterobacter cloacae, Escherichia coli, Hafnia alvei, Klebsiella pneumoniae, Proteus vulgaris, Salmonella typhi, Salmonella typhimurium, Serratia marcescens, Shigella dysenteriae, Shigella boydii, Yersinia enterocolitica. The extraction of DNA from blood samples and the PCR experiments were performed using two different sets of primers: the first (BG-1 and BG-4) was specific only for E coli, the second (p16SrRNA+ and p16SrRNA-) was specific for various gram-negative and gram-positive bacterial species, as previously described [9].
The samples that were negative in the PCR were subjected to amplification with primer for the human ß-globin gene to detect the presence of PCR inhibitors [10].
Inflammatory activation markers
At t1, t8, t15, t16, t17, and t18 C-reactive protein levels were recorded. Arterial blood samples for determination of nuclear factor B (NF-B), tumor necrosis factor (TNF)-
, and interleukin (IL) 6 levels were drawn at t1, t3, t6, t9, t10, t11, and t15. Serum concentrations of TNF and IL-6 were measured using quantitative sandwich enzyme immunoassay techniques (Endogen, Woburn, MA, for TNF-, and Quantikine, R&D Systems, Abingdon, UK, for IL-6). According to information provided by the manufacturers, assay sensitivities were as follows: TNF- less than 2 pg/mL and IL-6 less than 0.70 pg/mL.
Activation of NF-B (detected in the nucleus by radioactive probe binding) was analyzed with an electrophoretic mobility shift assay after isolation of peripheral blood monocyte cells from collected blood samples, as described elsewhere [11]. To determine the basal level for each patient and the potential for additional activation of the transcription factor before CPB, the peripheral blood monocyte cells were cultured in medium with and without lipopolysaccharides for 1 hour, lysed, and analyzed by electrophoretic mobility shift assay. Gels were analyzed by STORM 840, and the intensity of the bands was directly quantified by Image QuaNT software (Molecular Dynamics, Sunnyvale, CA), which furnishes a volume report based on integration of band area and density. A positive constant control, which was assigned a score of 1, was used in each electrophoretic mobility shift assay to normalize sample values from different assays.
Statistical analysis
Group means and standard deviations were calculated for all data collected. Time trends for each of the variables considered were assessed by means of one-way analysis of variance for repeated measures. The level of significance was considered p less than 0.05. When analysis of variance revealed significant variation as a function of time, the differences between single time points were analyzed with the Newman-Keuls test (level of significance, p < 0.05).
|
Results |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentReferences |
|---|
As for tonometric data (Table 2), the pHi decreased significantly as a function of time (from 7.4 to 7.2 in the intensive care unit); during the postoperative period, gastric pH remained stable (range, 7.2 to 7.3). The PgCO2 and Pg - Pa increased significantly and progressively (from 38 to 56 mm Hg and from 1 to 21 mm Hg, respectively), peaking in the sixth postoperative hour (55 and 23 mm Hg, respectively).
|
|
|
|
Concentrations of TNF- did not vary significantly during the study period, although a peak was observed at t10 (21.7 versus 17.5 pg/mL at t1). In contrast, IL-6 increased progressively and significantly (p < 0.001) over basal levels with a peak at t10 (96.3 versus 24 pg/mL) and subsequent decline to 55.4 pg/mL on the first postoperative day (Table 4).
|
B basal activity could be further stimulated by the addition of lipopolysaccharide. But NF-B activation never reached the postlipopolysaccharide stimulation levels and did not vary significantly during the study with respect to the basal level (Fig 3).
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentReferences |
|---|
The reduction of the pHi after CPB suggests an ischemic damage of the gastric mucosa, in agreement with previous reports in cardiac surgery [12, 13]. Postoperative increase of gastric Pg - Pa is even more significant because it is considered more reliable than pHi to assess gastrointestinal perfusion and oxygenation and is well correlated to mortality and complications after coronary artery bypass graft and valve replacement [14]. It is well known that gastric tonometry does not reflect global hepatosplanchnic perfusion, but for clinical practice no other noninvasive technique seems to be easily usable. Thus, it remains a valuable monitoring tool, above all when its data are acquired in association with other observations, as increased gut permeability and E coli appearance in the bloodstream.
The prolonged and marked suffering of the intestinal mucosa is in fact suggested by the increased 51Cr-EDTA urinary excretion for the first 24 perioperative hours. Also this finding is in agreement with the study of Riddington and coworkers [13]. The present paper validates the sensitivity of the radioisotope method in cardiac surgery and the reproducibility of the results obtained. Some aspects differ from the previous study; namely, we combined another test of intestinal permeability with different characteristics and normothermic CPB was performed because normothermia can mimic better what happens in other pathologic conditions in which the protective effect of hypothermia does not exist.
In conditions of increased permeability, 51Cr-EDTA passes through defective tight junctions of epithelial cells lining the mucosa: urinary 51Cr-EDTA levels are thus a marker of the permeability of the small intestine as well as of the colon [15]. The double sugar test is able to assess the permeability of the small intestine only if it is sterile: an increased ratio of the sugars indicates enhanced permeability of small intestine lacking the saccharolytic flora able to break down the sugars, as usually happens in the colon. It is therefore likely that the discrepancy between the two tests indicates that the increase in permeability observed in our patients is localized in the colon, in accordance with other reports that have shown that the small intestine is more resistant to hypoxic-reperfusion insult [16, 17]. These findings and tonometric results show the existence of "patchy" splanchnic perfusion, with the stomach and colon representing the weak links in the chain. This assumption supports the hypothesis of the presence of heterogeneous blood flow distribution within the splanchnic region, as the reported lack of correlation between the changes in hepatosplanchnic blood flow and intramucosal pH during a dopexamine infusion in patients after cardiac surgery suggests [18]. It has been reported that nonsteroidal antiinflammatory drugs can alter 51Cr-EDTA absorption [19]. All our patients took salicylate preoperatively, and this could explain the higher than normal basal excretion of the radioisotope (6.4%).
The bacterial translocation detected until the sixth postoperative hour in almost all the patients studied may be related to the altered intestinal permeability. The fact that the only microbial species detected was E coli, which is found predominantly in the lumen of the colon, adds support to the hypothesis that this segment of the intestine is the predominant site of perioperative hypoperfusion. The existence of microbial translocation is well documented in literature, but the clinical significance of this phenomenon is not clear at the moment. Data from recent clinical studies support the hypothesis that the translocation of enteric flora is a possible cause of bacteremia and predisposes to infectious complications, among them multiple organ failure [20].
It is noticeable that the systemic inflammatory reaction appears self-limited in comparison with the marked alterations of gut function. It is well established that an inflammatory response occurs after CPB and is associated with systemic release of proinflammatory cytokines, including TNF- and IL-6 [5]. The significant increase of IL-6 is consistent with previous reports and is thought to be the result of exposure of blood to artificial surfaces [21]. Interleukin 6 and TNF- are intimately connected and specific for cardiac surgery; IL-6 is induced by TNF- and is considered a reliable surrogate for localized TNF- activity, otherwise very difficult to detect [22]. The postoperative increase in C-reactive protein levels is related to the increase in IL-6, which induces hepatic synthesis of the acute-phase protein. Data from the literature report that C-reactive protein levels may be still elevated even when the inflammatory stimulus has stopped, affecting the reliability of the marker [23]. The behavior of NF-B is interesting because its functions regulate the expression of genes involved in the immunologic response and inflammatory reactions. Nuclear factor-B never reached levels significantly higher than the basal nonstimulated value, indicating that its activation was not induced during the perioperative period. The failed activation of NF-B suggests a mild inflammatory reaction to the stress induced by normothermic CPB, and this could correlate with the good outcome of our patients. Another possible explanation is that all the patients received salicylate preoperatively and high doses of heparin were administered intraoperatively, both drugs being able to limit NF-B activation [24, 25].
Based on our data, we can assume that the bacterial translocation is the result of ischemic-reperfusion damage to the gastrointestinal mucosa. The inflammatory reaction does not seem so strong as to induce the intestinal damage, and other factors are probably involved. In a different scenario the demonstrable bacterial translocation caused by the increased intestinal permeability could play a stimulating role in creating a vicious cycle of events that are able to amplify the inflammatory reaction and potentially to increase perioperative risks. The evolution toward multiple organ dysfunction syndrome requires a more substantial ischemic-reperfusion insult associated with more prolonged CPB, or perioperative hemodynamic complications, preexisting debilitating diseases, cardiovascular risk factors, or a "second hit," ie, a serious complication of any type during the postoperative period.
In conclusion, bacterial translocation and impaired intestinal permeability, joined with a controlled inflammatory response, appear clinically well sustained during elective, low-risk CPB without determining relevant perioperative complications. This scenario could be different if higher-risk patients were enrolled in the study, and other investigations should be carried out to define selected protective strategies and to ascertain a possible different behavior with alternative surgical approaches, such as off-pump procedures.
|
References |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentReferences |
|---|
B. Int J Biochem Cell Biol 1995;27:865-879.[Medline]
This article has been cited by other articles:
|
|
 |
|
  G. Zhang, N. Wu, H. Liu, H. Lv, Z. Yao, and J. Li Case control study of gastrointestinal complications after cardiopulmonary bypass heart surgery Perfusion, May 1, 2009; 24(3): 173 - 178. [Abstract] [PDF]
|
|
|
|
 |
|
 M. Bossola, M. Sanguinetti, D. Scribano, C. Zuppi, S. Giungi, G. Luciani, R. Torelli, B. Posteraro, G. Fadda, and L. Tazza Circulating Bacterial-Derived DNA Fragments and Markers of Inflammation in Chronic Hemodialysis Patients Clin. J. Am. Soc. Nephrol., February 1, 2009; 4(2): 379 - 385. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. A. Hessel II Abdominal Organ Injury After Cardiac Surgery Seminars in Cardiothoracic and Vascular Anesthesia, September 1, 2004; 8(3): 243 - 263. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
