Immunoglobulin M (IgM) antibodies were detected by a commercially available enzyme-Linked Immunosorbent Assay (ELISA) in 36 of 49 (73%) pregnant women with primary cytomegalovirus (CMV) infection. A positive ELISA-IgM result occurred in 10 of 13 patients (77%) assessed within 8 weeks of seroconversion. The sensitivity of the radioimmunoassay (RIA) to identify primary infection in pregnant women was comparable, 78% in general and 86% for women tested within 16 weeks of seroconversion. Of the 36 women with primary infection who had detectable IgM antibodies by ELISA, 25 (69%) were delivered of congenitally infected infants, whereas of the 13 with undetectable IgM antibodies, 7 (54%) transmitted the infection in utero. IgM antibodies were detected by ELISA in only 5 of 43 (11%) women who experienced a recurrence of CMV which either did or did not result in congenital infection. RIA was less likely to measure CMV-specific IgM in recurrent infection, inasmuch as 1 of 19 (5.2%) women with proven recurrent infection had detectable IgM antibody, giving RIA a better specificity for primary infection. Specific IgM antibodies were detected by ELISA in 42 of 61 (69%) babies congenitally infected with CMV and in 4 of 70 (5.7%) uninfected control newborn infants. The RIA was superior for diagnosis of congenital CMV infection, with a sensitivity of 89% and a specificity of 100%. The lower sensitivity of the ELISA-IgM occurred in the category of congenitally infected infants born to mothers with recurrent infection (43%), a group that is at the lowest risk of disease or to develop sequalae. This commercially available ELISA-IgM could be used in combination with a CMV-specific IgG test for monitoring women during pregnancy for primary infection.

During the past 10 to 15 years immunoassays have gained acceptance as the methods ofchoice in the diagnosis of a number of disease states. At present the immunodiagnostic techniques employed range from radioimmunoassay for haptens through immunofluorescence for autoimmune diseases to complement fixation for viral infections. All of these assays have their own individual limitations such as safety, short sheiflife and sensitivity. The development ofenzyme immunoassays, in particular enzyme linked immunosorbent assay (ELISA), has led to a substantial literature which offers the view that This type of variation was dramatically pointed out in the most recent Center for Disease Control proficiency study (183).

An enzyme-linked immunosorbent assay (ELISA) was developed for the detection of measles immunoglobulin G antibody (MEASELISA). This assay was found to be comparable to the measles hemagglutination inhibition (HAI) test. Approximately 500 sera from three centers were tested by MEASELISA and the HAI test. MEASELISA demonstrated values of greater than 99% for sensitivity, specificity, and accuracy. Values were very precise, with a mean coefficient of variation of 5.4%. MEASELISA values were shown by linear regression analysis to increase as HAI titers increased. A coefficient of determination of 1.00 was obtained from test center three. MEASELISA values were found to be linearly related (r2 greater than 0.97) to MEASELISA titers, thus enabling quantitation of measles antibody from a single value. Also, data are presented that show MEASELISA to be equivalent to complement fixation for evaluating paired sera for the presence of a significant increase in antibody levels to measles virus.

The level of humoral and cell mediated immunity in persons with low or undetectable hemagglutination-inhibition (HAI) rubella titers was investigated by ELISA, IgG presence in sucrose gradient separated serum fractions and lymphocyte transformation. The study population consisted of persons with stated history of natural rubella infection and rubella vaccinees. Persons with natural rubella infection and HAI titers of 1:8 or +/- 1:8 (i.e., incomplete HAI at serum dilution of 1:8) were all ELISA positive and the stimulation index (SI) of specific lymphocyte transformation was higher than 2.5. Among the 20 persons with HAI titers of < 1:8, 8 were found to be ELISA positive and their SI was also > 2 and IgG was detected in their serum. Rubella vaccinees with HAI titers of 1:8 or +/- 1:8 were likewise ELISA positive. Their SI was lower: none higher than 3, but none lower than 1.5. Among 23 HAI negative vaccinees, 14 were found ELISA positive. This serum fraction contained IgG and the SI was > 1.5. It appears that ELISA test is able to detect antibodies where the HAI test fails. The positive outcome of ELISA test in this case was confirmed by the presence of IgG in serum fractions and by the lymphocyte response to rubella specific stimulation.

An enzyme-linked immunosorbent assay was developed for the detection of antibodies to murine hepatitis virus. A high prevalence of antibody to murine hepatitis virus was found by the enzyme-linked immunosorbent assay in colonies with a low prevalence of complement-fixing antibodies. Murine hepatitis virus strain A59 was found to be broadly reactive as an enzyme-linked immunosorbent assay antigen.