Abstract
The recent discovery of a non-endotoxic reactive pathway in the conventional Limulus test has indicated that serum endotoxin (Et), being derived from the intestinal bacterial flora, may become pathogenic (endogenous entotoxemia) under some conditions. Revaluation of postoperative endogenous entotoxemia was done by an endotoxin-specific chromogenic assay (Endospecy, Seikagaku Kogyo, Tokyo, Japan) using 32 patients with esophageal varices and liver disease (cirrhosis; 26, idiopathic portal hypertension: 6). In addition, the effects of oral polymxin-B sulfate (PL-B) and/or tindiazol (TDZ) administration prior to surgery (esophageal transcetion: 27, Hassab's operation: 5) were estimated on the basis of changes in bacterial flora in the rectum after administration and the postoperative serum Et level. Et was increased on the 1st postoperative day (POD) (7.4±5.3pg/ml) compared with the preoperative period (5.8±4.6pg/ml). Thereafter, it was decreased on the 3rd and 5th POD. PL-B administration (300million units/day) reduced the increase in Et level. However, administration of a large quantity of PL-B (600million units/day) or adding TDZ (1000mg/day) to PL-B failed to cause a further reduction in the Et level. Based on the change in the bacterial flora of the rectum, an increase in anaerobic bacteria (in PL-B administration) or E. coli (in PL-B + TDZ administration) may be the cause of the failure. No apparent septic manifestion occurred during the postoperative course. Surgical stress on patients with liver impairment caused the serum Et level to increase up to within the normal limit. If additional postoperative morbidity occurred, the endogenous endotoxin level would probably increase enough to be pathogenic. Oral PL-B administration was effective in reducting the Et level.
