Abstract
A quantitative microassay for bacterial endotoxin in whole blood, based on Limulus amebocyte lysate gelation and involving the use of a chromogenic substrate, was studied. The lower limit of detection was 10 pg/ml, and the mean level of recovery from plasma spiked with 1 ng of endotoxin/ml was 0.93 ng/ml. At clinically significant levels endotoxin was more easily recovered from plasma than from serum. Platelet binding had no adverse effect on the assay. Despite the use of conventional methods for the removal of nonspecific activators and inhibitors, simultaneous assay of six specimens spiked with the same amount of endotoxin resulted in variation in recovery by >100%. These experiments may explain in part the conflicting results of studies of endotoxemia in humans.