Serum Therapy in Bacterial Bloodstream Infection: A Multicenter Trial?

Origins and Design

The trial emerged from coordination between Dorothy Umezaki's laboratory at the Infectious Disease Research Centre in Cambridge and the International Serum Registry, which had been established in 1975 to standardize treatment reporting across nations. Participating institutions included major teaching hospitals in Britain, France, West Germany, Italy, Japan, the United States, and Sweden. Each site agreed to enroll hospitalized patients with documented bacterial bloodstream infection who had either failed to respond to azo drug therapy or could not tolerate sulfonamide toxicity.

The trial's protocol required that all serum preparations meet standards established by the Geneva Sanitary Bureau for donor hyperimmunization, collection timing, and antibody titre. Patient enrollment was stratified by primary infection site—endocarditis, meningitis, septicaemia, and secondary bacteraemia following surgical intervention—to permit analysis of serum efficacy across different bacterial pathologies. Dosing and administration schedule were standardized across all sites, departing from prior protocols in which individual physicians had titrated serum therapy empirically.

The statistical framework for the trial was developed by Paul Kaplan and colleagues at the Biostatistics Institute in Boston, adapting the Kaplan-Meier estimator from industrial reliability testing to clinical serum therapy outcomes. Because serum therapy rarely produced rapid bacterial kill—unlike the imagined trajectory of chemical antimicrobials—the method measured time to clinical stabilization, fever resolution, and reduction in bacteraemia titre rather than time to confirmed microbiological cure. Survival at 30 days, 90 days, and one year became the primary endpoints, with secondary measures including length of hospital stay and recurrence of infection in the six months following discharge.

Enrollment and Outcomes

Between 1976 and 1980, the trial enrolled 847 patients meeting infection criteria. The median age was 54 years; 61 percent were male. Thirty-four percent of enrolled patients had underlying immunosuppression, renal disease, or prior surgical intervention. Dosing of serum therapy was calibrated to measured antibody titre against the identified causative organism, with retreatment permitted if bacteraemia titre did not decline after 48 hours.

One-year survival in the trial cohort was 67 percent, compared to historical mortality rates of 85 percent in septicaemia cases before serum therapy became available. Outcomes varied substantially by infection site: endocarditis cases achieved 52 percent one-year survival, while secondary bacteraemia following surgery achieved 78 percent. Meningitis outcomes (44 percent survival) remained poor despite serum therapy, reflecting the blood-brain barrier's interference with antibody penetration.

Twenty-three percent of enrolled patients experienced adverse reactions to serum, ranging from mild allergic phenomena to anaphylaxis requiring immediate intervention. The trial documented that repeated serum doses in the same patient increased the risk of adverse reaction substantially, a phenomenon consistent with prior observations of antitoxin use but quantified systematically here for the first time. Seven deaths directly attributable to serum administration were recorded.

Clinical Impact

Scholarly Interpretation