Reduced tumor necrosis factor alpha production in lipopolysaccharide-treated whole blood from patients in the intensive care unit.

Objective: To determine the cytokine response to lipopolysaccharide in patients in the intensive care unit.

Patients: Patients in a mixed medical/surgical intensive care unit with fever and a de novo clinical dysfunction of at least one organ system.

Methods: Whole blood from patients and from laboratory controls was stimulated with 8 ng/mL of lipopolysaccharide (Escherichia coli 0111:B4) at 37 degrees C, and tumor necrosis factor alpha (TNF-alpha) was measured using enzyme linked immunosorbent assay at 4, 8, and 24 hours. The same subjects' purified monocytes were cultured with 8 ng/mL of lipopolysaccharide in the presence of autologous or pooled control plasma or cocultured with purified autologous polymorphonuclear leukocytes at a polymorphonuclear leukocyte-monocyte ratio of 10:1, and TNF-alpha was measured at 24 hours using the enzyme linked immunosorbent assay.

Results: We detected high (n = 5) and low (n = 5) TNF-alpha responders in whole blood producing a mean (+/- SEM) of 27.2 +/- 6.3 pg/mL per 1000 monocytes vs 0.0 +/- 2.4 pg/mL per 1000 monocytes, respectively (controls, 58.0 +/- 13.0 pg/mL per 1000 monocytes). The kinetics of TNF-alpha production in both groups were comparable. Purified monocytes from both groups of patients cultured with lipopolysaccharide alone produced equivalent TNF-alpha values (42.4 +/- 10.5 vs 40.8 +/- 12.5 pg/mL per 1000 monocytes). Assayable TNF-alpha was not different with autologous vs control serum but was markedly diminished by the presence of polymorphonuclear leukocytes in patients as well as in controls; the two groups of patients did not differ in this polymorphonuclear leukocyte effect.

Conclusion: Lipopolysaccharide stimulation of monocytes in the whole blood results in marked variation of TNF-alpha production. This phenomenon may account for the variable septic response to infection in patients in the intensive care unit.