Utilization of a chromogenic Limulus amebocyte lysate blood assay in a multi-center study of sepsis

We conducted a prospective study of a chromogenic LAL assay in 346 patients with sepsis syndrome, as defined by a modification of the Bone criteria, and 131 healthy volunteers at eight member centers of the Academic Medical College Consortium (AMCC). We identified patients with endotoxemia (> 0.40 EU/ml) by measuring LAL-reactive material in whole blood, extracted by the Tamura nitric acid method, with the chromogenic LAL (Pyrochrome(R)) assay. The mean result in sepsis patients with detectable endotoxemia (n = 241) was 1.07 +/- 1.57 EU/ml, and the mean result in 131 volunteers was 0.151 +/- 0.113 EU/ml, with 73% of the volunteers' results falling below the detectable limit. The average incidence of endotoxemia in sepsis patients was 33%, but varied 2.7-fold among the clinical centers (range 16-44%). Assay results were repeatable when samples tested frozen at the clinical sites were compared to results on frozen samples tested at Associates of Cape Cod, Inc. (ACC). Multiple samples were obtained from 40 patients at 18-24 h interval(s). Fourteen multidraw patients (35%) were endotoxemic at one or more draw(s) and eight of these patients had two or more draws with endotoxin levels > 1.0 EU/ml. The presence of sulfa drugs gave false positive results in two patient samples. A positive LAL test did not correlate with culture-proven bacterial infection and did not significantly correlate with mortality. There was a correlation (P=0.014) between a patient having a positive LAL test and the presence of a fungal infection when mixed fungal and bacterial infections were included. There was no correlation with a positive LAL test when only a fungal infection was present (P=0.425) or when only a fungal and a Gram-positive infection was present (P=0.087).