Comparison of three serological methods--enzyme-linked immunosorbent assay, complement fixation, and microagglutination--in the diagnosis of human perinatal Listeria monocytogenes infection.

Serological diagnosis of human perinatal infection with Listeria monocytogenes is at present unsatisfactory. Difficulties in performing or interpreting serodiagnostic tests include weak or absent antibody responses to infection and extensive serologic cross-reactions with other organisms. In an attempt to overcome these problems, we developed an enzyme-linked immunosorbent assay to detect IgG (ELISA-G) and IgM (ELISA-M) antibody to L. monocytogenes and compared these tests with microagglutination and complement fixation tests. Studies were done using convalescent sera collected from 29 mothers approximately 1-2 months following recovery from Listeria infection and from 84 matched controls. Whole L. monocytogenes, serotype 4b, was used in the microagglutination and ELISA tests, while a commercial antigen preparation containing both serotype 1a and 4b antigens, was used in the complement fixation test. All tests showed significant differences between cases and controls (p less than 0.0005). The specificity of the tests was similar, ranging from 78% to 91%. However, the sensitivity of each test was more variable ranging from 56% for the microagglutination test to 78% for the ELISA-M. The complement fixation test was better than the other tests with a sensitivity of 78% and a specificity of 91%. The serotype 4b specific antibody response was predominantly of the IgM class based on differential reactivity between serotype 1a and 4b. Although these serologic tests are specific and may be predictive in ruling out infection, their sensitivity is poor and isolation of the organism remains the only method for diagnosing infection.