| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
Departments of
*Anesthesiology and
Biomathematics, The Mount Sinai School of Medicine, New York, New York;
Scottish National Blood Transfusion Service; and
§Department of Medical Microbiology, University of Edinburgh Medical School, Edinburgh, United Kingdom
Address correspondence and reprint requests to Elliott Bennett-Guerrero, MD, Department of Anesthesiology, Columbia University College of Physicians & Surgeons, 630 W. 168th St., New York, NY 10032-3784. Address e-mail to eb413{at}columbia.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Implications: We prospectively measured hemodilution-corrected antiendotoxin antibody levels in 55 cardiac surgical patients. We observed significant decreases in hemodilution-corrected levels of antibody to both Bacteroides fragilis and the core of endotoxin.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Bacteroides fragilis is an anaerobic gut commensal and a common pathogen. Anaerobic bacteria such as Bacteroides constitute up to 30% of the colonic microbial flora and greatly outnumber enterobacterial species, such as Escherichia coli, by up to 1000 fold (3). B fragilis contains biologically active endotoxin (4,5), although there is still controversy as to the pathogenic potential of this endotoxin (i.e., some investigators have found this endotoxin to be less toxic than that from E coli).
Despite the ubiquitous nature of these bacteria and their potential toxicity (4,5), it is not known if endotoxins from B fragilis bacteria are responsible for endotoxemia during cardiac surgery. It should not be assumed that all intestinal contents, including endotoxins from B fragilis, leak into the systemic circulation during cardiac surgery. Leakage of intestinal contents is not "nonspecific," and the magnitude of leakage can vary markedly for different compounds (6,7). This is a clinically relevant issue, because several antiendotoxin strategies under development (8,9) do not focus on B fragilis as a potential source of endotoxemia, yet B fragilis bacteria are known to produce significant amounts of biologically toxic endotoxin (4,5).
Hence, our study was designed to test the hypothesis that patients undergoing cardiac surgery are exposed to endotoxin from B fragilis. Because there are no known methods available to measure the presence of B fragilis endotoxin in blood, we used a decrease in hemodilution-corrected serum anti-B fragilis antibody levels during cardiac surgery as a surrogate endpoint for systemic exposure to B fragilis endotoxin. A secondary aim of the study was to compare differences between anti-B fragilis antibodies and anti-E coli core antibodies (EndoCAb).
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Exclusion criteria included age less than 18 yr, emergency surgery, any religious/ethical prohibition from receiving blood products, concomitant surgical procedures (e.g., aortic surgery), and procedures requiring circulatory arrest.
After oral benzodiazepine premedication, venous and radial arterial catheters were inserted. A pulmonary artery catheter was inserted after the anesthetic induction. Maintenance of general anesthesia was accomplished by using a balanced technique of midazolam hydrochloride, fentanyl citrate, and isoflurane. Patients underwent standard nonpulsatile hypothermic (23°–32°C) cardiopulmonary bypass (CPB) with a membrane oxygenator and hemodilution by using a crystalloid CPB prime. Porcine heparin was administered as a bolus of 300 U/kg and supplemented as necessary to maintain a kaolin-activated coagulation time of >450 s during CPB. Heparin was neutralized with 1 mg protamine/100 U heparin. After CPB, fluids were administered to optimize intravascular volume, inotropic drugs administered to maintain an adequate cardiac index, and vasoactive drugs administered to maintain a mean systemic arterial blood pressure between 50 and 80 mm Hg.
For each sample, 3 mL of blood was withdrawn from a freshly placed intraarterial catheter, placed into a plain nonadditive glass tube (VacutainerTM; Becton Dickinson, Rutherford, NJ), centrifuged for 10 min at 2000g, and serum stored at -70°C until assayed. Blood samples were obtained at the following time points: Preoperative (before the induction of general anesthesia), Pre-CPB (immediately before the initiation of CPB), Pre-CPB+5 (5 minutes after the initiation of CPB), and End (end of surgery).
All laboratory measurements were performed on coded samples to insure blinding. Serum was tested for total immunoglobulin G (IgG) antibody and total immunoglobulin M (IgM) antibody concentrations by laser nephelometry (10). IgM and IgG anti-B fragilis antibody levels were determined by an enzyme-linked immunosorbent assay (ELISA) as described previously (11). The preparation of the antigens and their purity were described in detail previously (5). Results from test sera were expressed as a percentage of the reference serum pool. IgG EndoCAb and IgM EndoCAb levels were measured by using ELISA as described previously (12–15). Results for EndoCAb were expressed in EndoCAb median units (MU) as is customary, where 100 is the median value for 1000 healthy adults’ IgG or IgM, respectively (15). Binding of endotoxin to antibody makes these bound antibodies unassayable in the ELISAs described above (G. R. Barclay, written communication, June, 1999).
Over a given time period, the percent change in an individual subject’s total IgM level was used to control that patient’s percent change in anti-B fragilis antibody level as well EndoCAb level. This was done by subtracting the percent change in the total IgM level from the percent change in the antiendotoxin level in question.
The IgM class of total immunoglobulins was chosen as a reference for the correction for hemodilution because it remains almost exclusively in the intravascular space and is not known to undergo significant redistribution between the intra- and extravascular compartments. Total IgG levels were not used to correct for hemodilution because total IgG antibodies move between the intravascular and extravascular compartments. Furthermore, hematocrit was not used to control hemodilution, as both bleeding and red blood cell transfusions, which are common during cardiac surgery, would interfere with the hemodilution correction.
A predefined control period of approximately 10 min at the start of CPB was added to the protocol to test the feasibility of correcting for hemodilution. Each patient undergoes acute hemodilution attributable to the CPB prime fluid, and there should be no decrease in hemodilution-corrected antibody levels during this period.
Statistical calculations and analyses were performed using the SAS software system version 6.12 (SAS Institute Inc, Cary, NC). Statistical significance (P) was set at 0.05. A cohort size of at least 50 patients was estimated to provide sufficient statistical power to detect at least a 10% decrease in hemodilution-corrected antibody levels. Wilcoxon’s signed rank test was used to compare the percent decrease in a specific antiendotoxin antibody level with the corresponding percent decrease in the total IgM level for each operative period.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
|
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
There are no available methods for assaying the quantities of B fragilis endotoxin in our surgical patients’ blood. Hence, we used a decrease in hemodilution-corrected serum anti-B fragilis antibody levels during cardiac surgery as a surrogate marker for systemic exposure to B fragilis endotoxin. We have demonstrated that antibodies to B fragilis endotoxin decrease during cardiac surgery, over and above the effects of hemodilution. This finding may have implications for those interested in developing antiendotoxin strategies for potential use in surgical patients (8,9).
Several antiendotoxin agents under development focus on protecting patients from E coli endotoxin (8,9). For example, the novel monoclonal antibody WN1 222-5 has excellent binding characteristics to endotoxin from E coli (8). In another example, IgG prepared from rabbits immunized with a J5 E coli vaccine protected neutropenic rats against lethal challenge with Pseudomonas aeruginosa (9). Although these agents show promise for protection against endotoxin-mediated illness, our findings suggest that their overall clinical efficacy may be limited because they do not appear to confer protection to B fragilis endotoxin.
As part of a secondary aim of this study, we also observed decreases in the levels of EndoCAb. EndoCAbs are cross-reactive with endotoxins from E coli but do not appear to cross-react with endotoxin from B fragilis (11–13). EndoCAbs are an independent predictor of complications after cardiac surgery (14,15) and vary by up to 1000 fold in both healthy volunteers and surgical patients (14–17). These interpatient differences may be caused by variability in exposure to endotoxin, as well as to inherent genetic differences between individuals. In our patients, preoperative levels of IgG EndoCAb were significantly higher than those observed in healthy Scottish volunteers. This difference may be caused by a difference in the degree of exposure to endotoxin between the two populations and warrants further study.
The decreases in EndoCAb we observed as a secondary study objective are consistent with the results of Hamilton-Davies et al. (15) who demonstrated decreases in median levels of IgG and IgM EndoCAb during cardiac surgery. They, however, did not measure antibody levels to B fragilis or validate their correction for hemodilution. We measured anti-B fragilis antibodies and verified our correction for hemodilution. This correction appeared to be appropriate, because we observed no change in hemodilution-corrected antibody levels during a predefined control period (Pre-CPB to Pre-CPB+5).
Anti-B fragilis antibody levels and EndoCAb levels decreased by a similar percentage in this study. We cannot, however, directly compare these antibody levels or their reduction because they are expressed in different units. That is, there might be much more anti-B fragilis antibody than EndoCAb in absolute amounts, thus the same percentage decrease may reflect a much larger or smaller difference in endotoxin exposure. This issue requires further study.
It is important to clarify that there was marked interpatient variability in the degree to which antiendotoxin antibody levels decreased during surgery. This variability suggests that patients are exposed to varying amounts of endotoxin during surgery, which may be a result of differences in intestinal translocation secondary to episodes of chronic or acute splanchnic hypoperfusion (2). This interpatient variability is consistent with a large number of studies that have demonstrated interpatient variability in the degree of the inflammatory response. It is unclear whether this variability in putative endotoxin exposure translates into interpatient variability in postoperative outcome observed in other studies (14,15,18). Our study was not designed to have sufficient power to test this hypothesis.
Our study has several potential limitations. It is possible that the observed decreases in antiendotoxin antibody levels are artifacts of the testing methodology and do not reflect true consumption of these antibodies attributable to exposure to endotoxin. It is reassuring, however, that hemodilution-corrected levels did not decrease during the control period of hemodilution. Other than binding to endotoxin, we are not aware of any other mechanism to account for reduced levels of antibodies in this short period of time. There cannot be any effect from increased production or lack of production of antibodies within this acute period of several hours. Another potential limitation relates to the biophysical and immunologic properties of the endotoxins used in the ELISA. The method of purification of the endotoxin from intact bacteria can theoretically affect the ability of serum antibodies to bind to these endotoxins. All endotoxins we used were purified by using the phenol chloroform petroleum spirits method (19), which has been extensively used to purify endotoxins for use in numerous biological assays (19). Also, the ELISA uses endotoxin complexed with polymyxin B as solid phase. A complex of endotoxin and polymyxin B has been shown to be superior to purified endotoxin alone in detecting "natural" antiendotoxin core antibodies (20) and may mimic the structures in endotoxin when associated with bacterial outer-membrane proteins.
Another potential limitation is that blood products that may be administered to patients contain varying amounts of antibodies. These products contain antiendotoxin antibodies, as well as other antibodies that can change the total immunoglobulin and antiendotoxin serum levels during the course of surgery (21). It is unusual, however, for patients to receive fresh-frozen plasma during routine elective cardiac surgery, which is consistent with the fact that only 7% of our patients received fresh-frozen plasma. In fact, in a separate analysis, in which four patients who received fresh-frozen plasma were excluded, decreases in antibody levels during surgery were almost identical to those observed in the overall study population (n = 55).
In summary, levels of anti-B fragilis antibodies decreased significantly during cardiac surgery out of proportion to hemodilution. This finding is consistent with the theory that intestinal translocation of endotoxin is common during cardiac surgery and suggests that Bacteroides species may be a source of endotoxin.
|
Acknowledgments |
|---|
We thank Carol A. Bodian, DrPH, for her assistance in reviewing the statistical methods.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  E. W. Moretti, M. F. Newman, L. H. Muhlbaier, D. Whellan, R. P. Petersen, D. Rossignol, C. B. McCants Jr, B. Phillips-Bute, and E. Bennett-Guerrero Effects of Decreased Preoperative Endotoxin Core Antibody Levels on Long-term Mortality After Coronary Artery Bypass Graft Surgery Arch Surg, July 1, 2006; 141(7): 637 - 641. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
