Although endotoxemia has been observed during cardiac surgery, the identity
of endotoxins to which patients are exposed is unknown. We tested the hypo
thesis that antibodies to Bacteroides fragilis tan anaerobic gut commensal
and a common pathogen) decrease during cardiac surgery, thereby reflecting
systemic exposure to this type of endotoxin. Serum antiendotoxin antibody l
evels were measured in 55 patients during routine cardiac surgery at the fo
llowing times: Preoperatively, Pre-CPB (immediately before initiation of ca
rdiopulmonary bypass [CPB]), Pre-CPB +5 (5 min after initiation of CPB), an
d End tend of surgery). Antiendotoxin antibody levels were determined by us
ing enzyme-linked immunosorbent assay. Total immunoglobulin M (IgM) levels
were measured by using laser nephelometry and decreases in total IgM levels
were used to control changes in antiendotoxin antibody levels attributable
to hemodilution. Median (interquartile range) hemodilution corrected IgM a
nti-B fragilis antibody levels decreased by 12% (5%-20%) from Preoperativel
y to End of surgery (P < 0.001). In contrast, median hemodilution corrected
anti-B fragilis antibody levels did not change significantly from Pre-CPB
to Pre-CPB +5, validating the correction for hemodilution. Immunoglobulin G
anti-B fragilis antibody levels and IgM and immunoglobulin G anticore anti
body levels decreased similarly during surgery. Intraoperatively, levels of
anti-B fragilis endotoxin antibodies decreased significantly out of propor
tion to hemodilution. These results suggest that cardiac surgical patients
are exposed to B fragilis endotoxin. implications: We prospectively measure
d hemodilution-corrected antiendotoxin antibody levels in 55 cardiac surgic
al patients. We observed significant decreases in hemodilution-corrected le
vels of antibody to both Bacteroides fragilis and the core of endotoxin.