Literature review > Issue_3 > Review on Zariffa/Obata-Yasuoka. 

Bacterial vaginosis (BV) is a vaginal infection characterized by complex changes within the vaginal ecosystem. The organism(s), which actually causes BV, is unknown. There may be a pathogenic agent, or BV may represent an "overgrowth" phenomena of normally present bacteria. This lack of knowledge regarding the pathogenesis of this syndrome leads to problems with diagnosis. The two most commonly used methods of diagnosis are 1) the clinical criteria of Amsel based on a combination of vaginal pH, appearance of discharge, microscopic detection of clue cells, and a positive "whiff" test and 2) the Gram stain scoring method of Nugent et al. which is based on the detection and quantification of morphotypes consistent with lactobacilli, Gardnerella, and Mobiluncus.

The study by Zariffard et al. was designed to determine the feasibility of using PCR to detect vaginal bacteria in cervicovaginal lavage samples. As the authors point out, this qualitative and quantitative technique could be useful to look at differences in patterns of bacterial flora between women with differing clinical outcomes, to detect organisms such as mycoplasma which are not easily cultured and not detectable on Gram stain, and to examine bacterial flora in stored specimens when Gram stain or clinical criteria are not available.

The results supported the author's hypothesis that this type of testing is feasible for stored specimens. Women without BV had higher concentrations of lactobacilli than women with, while the inverse was true for Gardnerella. No differences were found between groups with respect to M. hominis.

The study is hampered, however, by small numbers of patients. Only 31 patients were included in the study and of these only 5 met the clinical criteria for BV. An additional issue to be dealt with in the future is choice of primers, especially for lactobacilli. In this present study only two species were sought and the vaginal ecosystem is much more complex. Nonetheless, the technique holds promise for investigating the vaginal flora associated with BV and is clearly more sensitive than culture or Gram stain.

The study by Obata-Yasuoka et al. examines the sensitivity and specificity of a PCR assay on vaginal specimens to diagnose BV. Four organisms are sought - Mobiluncus mulieris, M. curtissii, Bacteroides fragilis and G. vaginalis. Results of the assay were compared to Nugent Gram stain results from 173 women attending a clinic for obstetrics and gynecology in Japan. Vaginal swab samples were considered diagnostic for BV when one or more species of the targeted bacteria were detected using the multiplex PCR. The sensitivity of the multiplex PCR assay was 6 X 103 to 2 X 104 colony-forming units per 2 mL for the four target organisms. No amplified products were detected in samples containing other bacteria or yeast. The sensitivity, specificity, PPV, and NPV of the multiplex PCR assay compared to Gram stain examination for the diagnosis of BV were 78.4%, 95.6%, 82.9%, and 94.2%, respectively. The authors conclude that the multiplex PCR assay may be useful as a diagnostic test for BV.

Both of these articles confirm that PCR is feasible for the detection of various bacteria in the vaginal ecosystem, which is not surprising. The qualitative aspects depend on the various primers used. Tools such as these could prove quite productive in the research setting but seem unlikely to supplant currently available diagnostic methods.

about SDI | newsletters | grants | publications | literature reviews

WHO Home - WHO Search - TDR Home - SDI Home - SDI Contact us
(c) WHO/OMS 2001