A CONTROLLED TRIAL OF HA-1A IN A CANINE MODEL OF GRAM-NEGATIVE SEPTICSHOCK

Objective.-To investigate the therapeutic efficacy and microbiological and physiological effects of a human IgM monoclonal antibody (HA-1A) directed against the lipid A component of endotoxin in a canine model of sepsis that simulates the cardiovascular abnormalities of human sep tic shock. Design.-Blinded, placebo-controlled 28-day trial. Intervent ions.-Purpose-bred beagles were implanted with an intraperitoneal clot infected with Escherichia coli O111:B4. At clot placement, animals re ceived HA-1A (10 mg.kg-1), control human IgM antibody (10 mg.kg-1), or control human serum albumin intravenously. All animals were given ant ibiotic and fluid therapy. Measures.-Survival and microbiological and physiological events. Results.-Only two (15%) of 13 animals in the HA- 1A group, compared with eight (57%) of 14 control animals (combined co ntrol human IgM antibody and control human serum albumin groups) (P=.0 5), survived 28 days. At 24 hours, the HA-1A group had lower mean arte rial pressure (P=.04) and cardiac index (P=.004) and higher lactate le vels (P=.05) compared with the combined-controls group. In addition, t hese parameters in the HA-1A group were significantly more predictive of death. The HA-1A and combined-controls groups had similar significa nt increases in the level of endotoxemia and bacteremia. Studies of to xic effects showed no harmful effects of control human IgM antibody in infected animals or HA-1A in non-infected animals. Conclusion.-In a c anine model of E coli sepsis, HA-1 A did not alter levels of bacteremi a or endotoxemia and actually decreased survival. If these data are re levant to human septic shock, HA-1A therapy should be limited until th e conditions under which this monoclonal antibody has beneficial or de leterious effects are more completely defined.