Background: The objective was to compare commercial assays on clinical specimens for Mycoplasma genitalium (MG) detection and macrolide resistance mutation (MRM) frequency.
Methods: Three self-collected vaginal swabs (VS) and a first-void urine (FVU) from 300 consented women were tested by Aptima MG (AMG), ResistancePlus MG (RPMG) and Seeplex STD6 ACE (STD6) for detection of MG. Aptima MG and STD6 MG positives were tested for MRM using MG 23S rRNA polymerase chain reaction with Sanger sequencing (23SMGSS) compared with MRM determination in the RPMG assay. Unique AMG positives were tested with confirmatory Aptima assays.
Results: M. genitalium prevalence ranged from 7.1% to 19.7%, influenced by the assay used and the specimen tested. Overall agreements for MG detection were 96.3% (κ = 0.91) for VS and 93.3% (κ = 0.72) for FVU between AMG and RPMG with lower agreements with STD6. Using a rotating reference standard, sensitivities on VS and FVU were 100% and 100% for AMG, 100% and 83.3% for RPMG, and 54.2% and 48.4% for STD6. Specificities were high for RPMG and STD6 and AMG detected extra positives, most of which were confirmed. Macrolide resistance mutation frequency rates testing VS and FVU were 50% (24/48) and 58.1% (18/31) by RPMG compared with 52.5% (31/59) and 23.5% (12/51) by 23SMGSS. MRM overall agreements between RPMG and 23SMGSS were 73.2% (κ = 0.41) for VS and 76.0% (κ = 0.52) for FVU.
Conclusions: Aptima MG detected more cases of MG infections. ResistancePlus MG detection was more effective on VS than on FVU. Seeplex STD6 ACE performance was inferior. The MRM detection component of RPMG agreed with results from 23SMGSS most of the time.