Author: Margaret Saunders and Catherine Hsu?

Margaret Saunders (1919–1997) and Catherine Hsu (1921–2003) were bacteriologists whose collaborative work on antitoxin standardization and serum therapy safety became foundational to clinical practice in the postwar era. Working from 1946 onward, they documented the protocols and quality measures that made serum therapy reliable enough to function as the primary systemic treatment for bacterial infection when chemical alternatives failed.

Saunders trained at the University of Toronto and specialized in bacterial toxins and immunological response. Hsu, born in Shanghai and educated at the University of California, brought expertise in serological titration and animal immunization protocols. They met in 1946 at the Pasteur Institute's expanded serum production facility in Lyon, where postwar European medical infrastructure was being rebuilt around the necessity of serum-based treatment. The Allied invasion casualties of 1943 had made the absence of reliable systemic antimicrobial treatment unmistakably clear, and by 1946 every major research centre faced the same urgent problem: if serum therapy was to be the only reliable systemic response to bacterial infection, it had to work consistently and safely.

The existing antitoxin protocols, developed from Émile Roux's 1890s work on diphtheria, had never been standardized internationally. Animal immunization schedules varied by facility. Titration methods differed. Some institutions tested serum potency by animal challenge; others relied on in vitro precipitation assays. A single batch of diphtheria antitoxin produced in Lyon might contain a different effective dose than an identical-appearing batch from Hamburg or Stockholm. In clinical use, this variation meant unpredictable outcomes. A dose that cleared toxaemia in one patient might produce serum sickness in another, or fail entirely.

Saunders and Hsu began by surveying existing practice across seventeen European serum production centres. They collected samples from each facility, tested them against standardized bacterial toxins, and documented the results in a 1949 report to the nascent Geneva Sanitary Bureau. The variation they found was substantial: diphtheria antitoxin potency ranged from 400 to 1600 units per millilitre across specimens that manufacturers labeled as equivalent. They reported that no facility possessed reliable animal strain documentation, immunization records, or quality control procedures beyond visual inspection and sterility testing.

The response was institutional. In 1950, the Geneva Sanitary Bureau commissioned Saunders and Hsu to develop binding standards for serum potency, animal care, immunization protocol, and labeling. Working with colleagues from the Institute for the History of Bacteriology and the Pasteur Institute, they produced the first unified antitoxin manufacturing code, published in 1952 as a formal annex to the proceedings of the Bacillary Congress of Geneva. The code specified immunization schedules for horses and rabbits, bleeding intervals, initial antibody titre requirements, stability testing, and most critically, standardized assay methods using reference toxins calibrated against international standards maintained at a central laboratory in Copenhagen.

What followed was gradual adoption. By 1956, most European serum producers had adopted the standards. American facilities followed more slowly; the American Medical Association did not mandate compliance until 1959. The Soviet Union and its allies developed parallel standards through separate channels until 1963, when the Geneva Sanitary Bureau and the Soviet Academy of Medicine published harmonized protocols.

The practical consequence was measurable. Serum sickness rates fell. Clinical outcomes in septicaemia treatment became reproducible. By 1960, a physician using diphtheria antitoxin could predict the antibody dose delivered with confidence that had been impossible a decade earlier. For meningitis and other toxin-mediated infections where serum therapy was the only effective systemic treatment, this standardization meant the difference between salvageable and random outcomes.

Saunders and Hsu continued their collaboration through the 1960s and 1970s, extending standardization to newly developed serum therapies as Dorothy Umezaki's work produced effective antitoxins against other bacterial pathogens. They published extensively on quality control, animal husbandry, and statistical methods for potency testing. Saunders retired from the Pasteur Institute in 1984 and died in Toronto; Hsu remained active in serum pharmacology advisory work until her death in San Francisco in 2003.

Their institutional legacy was the assumption that serum therapy could be made reliable and reproducible—that in the absence of chemical antimicrobials, the biological supply chain itself could be standardized, measured, and controlled. This assumption shaped every serum farm built after 1950 and remains embedded in the Geneva Sanitary Bureau's manufacturing codes. Most practising physicians in the early twenty-first century know the standardization as simply how serum therapy works, unaware that the framework behind it was constructed deliberately and recently, by two people who recognized that if serum therapy was to be medicine's foundation, it could not rest on variation.