Comparison of culture, PCR, and different serologic tests for detection of Mycoplasma gallisepticum and Mycoplasma synoviae infections.

In this study, the technical performance of culture, two commercially available polymerase chain reaction (PCR) tests, rapid plate agglutination (RPA) test, hemagglutination inhibition (HI) test, and eight commercially available enzyme-linked immunosorbent assays (ELISAs) were compared for the detection of avian mycoplasma infections from 3 days postinfection (d.p.i.) through 35 d.p.i. The tests were carried out on samples from specified pathogen-free layers that were infected at 66 wk of age with recent Mycoplasma synoviae (MS) and Mycoplasma gallisepticum (MG) field strains, MS and MG ATCC strains, and Mycoplasma imitans (MIM), respectively. Results showed a high percentage of positive samples in the homologous infected groups and a high percentage of negative samples (100%) in the uninfected and heterologous infected groups during 35 d.p.i. of both culture and PCR tests. For the group infected with the MG 15302 ATCC strain, serology was more sensitive than bacteriology. All MG and MS tests, with the exception of MG ELISA kit D showed a lower percentage of positive samples during 35 d.p.i. for the detection of the MG and MS ATCC strain infection compared with that of the field strains. Also, the number of cross-reactions (false positives) in the serologic tests was lower after infection with an ATCC strain than after an infection with the MG or MS field strain. Contradictory to other studies, the ELISAs and the RPA test using undiluted serum showed a relatively high number of false-positive results. The MG ELISAs (except ELISA kit D) showed more false-positive results (up to 37%) in the MIM-infected group than in the MS-infected groups. This was not unexpected, as MIM and MG have a close antigenic relationship. The results of the serologic tests in this study showed that a certain level of false-positive results can be expected in about any serologic test. Although the level of false-positive results varied between several serologic tests, this study showed that it is not advisable to rely completely on one test (system) only.