ObjectiveTo evaluate in Gram-negative sepsis patients the human monoclonal immunoglobulin M antibody (MAB-T88) directed at the enterobacterial common antigen which is a specific surface antigen closely linked to lipopolysaccharide and shared by all members of the Enterobacteriaceae family of Gram-negative bacteria.DesignProspective, randomized, double-blinded, placebo-controlled, multicenter trial.SettingThirty-three academic medical centers in the United States.PatientsPatients were entered with a clinical diagnosis of sepsis, the presence of either shock or multiple organ dysfunction, and presumptive evidence for Gram-negative infection.InterventionsPatients received a single intravenous infusion, over 30 mins, of either 300 mg of MAB-T88 formulated in albumin, or placebo (albumin).Measurements and Main ResultsThe primary analysis group was prospectively identified as those patients with documented evidence of an infection with bacteria of the family Enterobacteriaceae at any site. The primary end point was survival within the first 28 days. A total of 826 patients were enrolled with 55% (n = 455) in the primary analysis group. There were no significant differences between the intervention and control primary analysis group study groups for sites of infection, severity of illness, underlying medical conditions, adequacy of antibiotic or surgical treatment, or other baseline variables except for a higher frequency of chronic renal failure in the MAB-T88 group (4.4% vs. 1.3%,p= .051). The average Acute Physiology and Chronic Health Evaluation II scores were 26.8 ± 8.6 (mean ± sd) in the MAB-T88-treated group and 26.5 ± 8.3 in the placebo-treated group (p= .72). There was no significant difference between MAB-T88- and placebo-treated groups during the first 28-day all-cause mortality in the primary analysis group (34.2% vs. 30.8%,p= .44) or in all 826 patients enrolled (37.0% vs. 34.0%,p= .36). On subset analysis, the use of MAB-T88 was not associated with significant mortality trends. More adverse events were seen with the use of MAB-T88 in the bacteremic enterobacterial common antigen group (p< .05).ConclusionsUse of the human monoclonal antibody, MAB-T88, did not improve the mortality in patients with presumed Gram-negative sepsis or in those patients with proven enterobacterial common antigen infections. No subset trends were identified that would support further investigation of this agent in sepsis.