Question
What is the performance of a multiplex PCR assay compared with culture for the detection of three genital mollicutes?

Design
A multiplex PCR assay that detected Ureaplasma spp., Mycoplasma hominis, and M. genitalium in a single amplification reaction was compared with culture for the detection of these organisms in urogenital specimens from men and women.  

Participants
Eighty-four specimens, which were sent for Ureaplasma culture, from patients seen at a fertility clinic, were also tested by PCR.   The specimens included 26 cervical swabs, 2 vaginal swabs, 4 female urines, 49 semen samples, 2 male urines, and 1 nonspecified sample.  Also tested was DNA from ATCC strains of U. urealyticum, M. hominis, and 
M. genitalium, 6 other mycoplasma species, and 11 other urogenital and respiratory microorganisms.  

Description of Tests and Diagnostic Standard
Swabs specimens were placed into 2 mL of 2SP medium.  Urine samples were concentrated 10-fold by centrifugation.  Specimens were cultured upon receipt in the laboratory and the remaining material was frozen at –70^oC for PCR testing.  Specimens were inoculated onto A7 agar and incubated at 37^oC in 5% CO₂ for 5 days.  Cultures were examined microscopically daily for the appearance of typical mycoplasma colonies.  Ureaplasma were identified by the urease test.   DNA was isolated by lysis in NucliSens lysis buffer (Organon Teknika), extraction with phenol-chloroform, precipitation with isopropanol, and suspension in water.  Multiplex PCR was performed with biotinylated primers for the highly conserved regions of the urease gene of Ureaplasma spp., the 140-kD adhesion protein gene of M. genitalium, and the 16S rRNA gene of M. hominis.  Ureaplasma and M. genitalium amplicons were detected by separate enzyme-linked oligosorbent assays (ELOSA).  Amplicons were hybridized to enzyme-labeled probes and captured onto streptavidin-coated microtiter wells.  A colorimetric substrate was added and the absorbance measured at 490 nm.  Specimens were considered positive if the optical density (OD) was greater than the mean OD of the negative amplification control plus a cutoff factor derived from the mean OD of 50 negative specimens plus two standard deviations.  M. hominis amplicons were detected by electrophoresis through a 7% acrylamide gel stained with ethidium bromide.  M. hominis amplicons produced bands of 334 bp before, and 62 and 272 bp after, digestion with NarI.  

The sensitivity of the PCR assay was determined by amplification of two-fold serial dilutions of bacterial DNA ranging from 3.13 to 100 cfu.  The lower limit of detection was the last positive sample in the dilution series.  All samples were tested by a second PCR assay that targeted the multiple-banded antigen gene of ureaplasma.  Amplicons were visualized by gel electrophoresis.  A true positive case was defined as a patient having a positive result by culture or by both PCR assays.  Samples positive by culture only were investigated for inhibitors of PCR by spiking them with about 50 cfu of U. urealyticum and analyzing by PCR.  

Main Outcome Measures
The sensitivity, specificity, positive predictive value (PPV), and the negative predictive value (NPV) of multiplex PCR and culture for the detection of genital mycoplasmas as determined by the expanded gold standard described above were calculated.  

Main Results
The multiplex PCR assay was sensitive for Ureaplasma spp., M. hominis, and M. genitalium at a limit of detection less than 12.5 cfu.  All other organisms tested produced no bands on gel electrophoresis and yielded negative results by ELOSA.  The results of the 84 specimens analyzed by multiplex PCR and culture are shown in Table 1.  Of the 21 culture positive specimens, Ureaplasma spp. were isolated from 17 and Mycoplasma spp. from 4.  Of the 28 PCR positive specimens, 23, 3, and 2 were positive for Ureaplasma spp., M. hominis, and both, respectively.  No specimens were shown to contain M. genitalium.   One of the 3 organisms was detected by PCR in 15 and by culture in 8 of 28 cervical or vaginal specimens; by PCR in 10 and by culture in 12 of 49 semen specimens.   The 4 specimens that were culture positive and PCR negative were all semen specimens.  Ureaplasma was detected in 23 specimens by the PCR for the multiple-banded antigen gene of ureaplasma.  Confirmatory testing did not result in additional discordant results.  Thirty patients (36%) were determined true positives for mycoplasma.  The spiked DNA was amplified in all 13 specimens analyzed for the presence of PCR inhibitors.  The performances of the multiplex PCR and culture for the detection of genital mycoplasma as determined by the expanded gold standard are shown in Table 2.   

Table 1.  Results of 84 specimens analyzed by multiplex PCR and culture for Ureaplasma spp., M. hominis, and M. genitalium

Table 2.  Performances of the multiplex PCR and culture for the detection of genital mycoplasma as determined by an expanded gold standard

Authors’ Conclusions
The multiplex PCR was more sensitive than culture and enhanced the detection of ureaplasma.   The increased sensitivity was seen primarily in female specimens.  The use of multiplex PCR did not increase the detection of Mycoplasma spp. 

Source of funding:  None given.

For correspondence:  Kathleen A. Stellrecht, Department of Pathology and Laboratory Medicine, Division of Clinical Micribiology, MC-22, Albany Medical Center, 43 New Scotland Ave., albaany, NY 12208.  E-mail address:   HYPERLINK "mailto:stellrk@mail.amc.edu" stellrk@mail.amc.edu.