Literature review > Issue_5 > Review on Yoshida et al. 

Culture of Mycoplasma and Ureaplasma species is rather technically demanding and for M. genitalium almost impossible. Furthermore, the discrimination between the two ureaplasma species U. urealyticum and U. parvum is generally only performed using molecular methods such as PCR. Consequently, there is a need for rapid methods that can improve detection.

This paper describes an approach based on PCR amplification of the 16S rRNA gene using Mollicutes generic primers and detection of the amplicons using hybridization with four species-specific oligonucleotide probes and an internal control probe immobilized in a microtiter plate. The idea of this approach is appealing, however, there are some major concerns regarding the clinical sensitivity of the method. The method was intended to be an improvement of the PCR phylogenetic assay previously described by the authors, and indeed this is probably true. Besides being technically demanding and rather expensive, the PCR phylogenetic assay suffered from a lack of sensitivity in cases of mixed infection. In order to detect a mixed infection, the minor population of DNA should be >30%. This was improved in the hybridization assay to approximately 1%, an improvement clearly demonstrated in the results, where the number of mixed infections is significantly higher. However, I would suspect that a proportion of patients, particularly those with M. genitalium infections, could remain undetected due to a concomitant ureaplasma infection. Many men may carry >10e4 viable ureaplasmas regardless of the presence of urethritis and at the same time many men with M. genitalium urethritis have DNA loads <10-50 genome equivalents.

Before the assay could be recommended for diagnostic use I would like to see an evaluation comparing clinical specimens tested separately for the each of the bacteria in question using one of the many species specific assays available.

A minor point of criticism concerns the use of purified plasmid DNA for determination of the limit of detection. Since the complexity of the DNA in a plasmid is much less than that found in a clinical specimen, such an approach will tend to underestimate the limit of detection, i.e. making the assay look better than it is. Considering the problems of getting purified mycoplasma DNA, an alternative approach could be to spike the diluted plasmid into a number of negative clinical specimens, preferably with many replicates in the range near the limit of detection in order to determine the DNA concentration yielding a positive result in 95% of the runs. This has become a widely accepted method for such estimates.

It is not clear whether the internal control plasmid was used in the determination of the limit of detection. Apparently it was not since it is stated that the limit of detection was the same on ethidium bromide stained gels. Although the use of internal controls should be strongly encouraged in all clinical assays, the evaluation of the assay should be carried out in a way mimicking the future use as closely as possible.

The fact that the authors have used the assay for clinical studies in subsequent publications apparently with results in accordance with previous studies performed using other methods may indicate that my concerns about the method are more of theoretical than of practical importance.

about SDI | newsletters | grants | publications | literature reviews

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