![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
Journal of Clinical Microbiology, November 2007, p. 3856-3858, Vol. 45, No. 11
0095-1137/07/$08.00+0 doi:10.1128/JCM.01008-07
Copyright © 2007, American Society for Microbiology. All Rights Reserved.
| LETTERS TO THE EDITOR |
 |
LETTER |
|---|
 Top LETTER REFERENCESLETTER{nbsp}REFERENCES{nbsp} |
|---|
In the May 2007 issue of the Journal of Clinical Microbiology, Mangold et al. (2) presented a new confirmation assay which is capable of distinguishing between N. gonorrhoeae and Neisseria spp. in comparison to confirmation assays targeting the porA pseudogene and the cppB gene. They obtained positive results in their in-house assay for 102/181 (56%) COBAS Amplicor-positive samples; the porA confirmation assay was positive for 69/181 (38%) samples, and the cppB assay detected 115/181 (64%) positive samples. Additionally, Mangold et al. reported 15%, 2%, and 4% positive results for COBAS Amplicor-negative samples with the cppB, porA, and in-house assays, respectively. The presented data are discrepant from observations at our institution. Upon optimization of published detection protocols, we have compared the performances of three assays targeting the porA pseudogene (5), the cppB gene (6), and the 16S rRNA gene (in-house) for quality assurance. In samples positive for all three targets, the cppB confirmation assay was the most sensitive (mean crossing point [Cp], 25.9), followed by the porA assay (mean Cp, 26.2) and the 16S rRNA gene assay (mean Cp, 32.8).
Over a period of 6 months, 398 samples yielded a positive COBAS Amplicor result (optical density [OD] of >0.2, as recommended by the manufacturer). Of these 398 samples, 258 samples (64.8%) were negative in all three confirmation assays. The remaining 140 samples (35.2%) yielded a positive result in at least two of the confirmation assays, with the majority being positive in all three assays (128/140 [91.4%]). Eight of 140 samples (5.7%) were negative in the 16S rRNA gene assay, probably due to small DNA amounts in the sample, as this assay is less sensitive than the other two confirmation assays, which is corroborated by the fact that 2/8 samples showed an OD value lower than 2.0 when initially tested in the COBAS Amplicor system. Four specimens were positive in the 16S rRNA gene and porA detection assays only and were interpreted as gonococci lacking the cppB-harboring plasmid (2.9%); the COBAS Amplicor value was above 2.0 OD in all four samples. Taken together for 136/140 (97.1%) specimens, the porA and cppB confirmation assays were concordantly positive.
In contrast to the findings of Mangold et al., we did not obtain positive results for the confirmation assays when applied to COBAS Amplicor-negative samples.
The conflicting results might in part be explained by an epidemiological distinct pool of N. gonorrhoeae isolates examined and an unfortunate study design, as modified interpretation guidelines for the COBAS Amplicor assay were used by Mangold et al. Samples were considered positive if two out of three runs gave a positive signal above an OD of 2.0, as opposed to one sample above an OD of 0.2 being sufficient. This decreases the number of positive samples and inevitably has a negative impact on the sensitivity of the screening assay. Presumably, this protocol accounts for the "false"-positive results of the cppB plasmid assay (33/227). It would be interesting if these samples were initially positive in the COBAS Amplicor system (OD of >0.2) by use of the manufacturer's interpretation criteria. Our analysis further cannot confirm the low sensitivity of the porA detection assay, as the majority of our specimens yielded positive results in all confirmation assays. Regarding the in-house NsppID assay published by Mangold et al., the design of the test itself is not completely conclusive: N. gonorrhoeae displays 100% homology with strains of Neisseria polysaccharea at the binding site of the detection probe (referred to as Npolysaccharea1 in the publication). N. polysaccharea was not tested for cross-reactivity in the NsppID assay. By use of an NCBI BLAST search, strains of Neisseria meningitidis also display 100% homology in this region of the 16S rRNA gene. Additionally, coamplification in PCRs should be avoided if no important information is obtained by the coamplified product, as it is likely to decrease sensitivity. A confirmation assay for N. gonorrhoeae that coamplifies apathogenic Neisseria spp. does not provide any additional information of clinical relevance.
The studies indicate that the use of molecular diagnostic assays for the detection of N. gonorrhoeae infection remains challenging and that the characterization of the prevalence of target genes in distinct epidemiological pools is important to ensure accurate diagnosis.
|
REFERENCES |
|---|
|
TopLETTERREFERENCESLETTER{nbsp}REFERENCES{nbsp} |
|---|
|
Silke Peter-Getzlaff* Jacqueline Luethy Burkhard Springer University of Zurich Institute of Medical Microbiology Zurich, Switzerland
|
||||||
| * E-mail: silke.peter{at}immv.uzh.ch |
|
LETTER |
|---|
|
TopLETTERREFERENCESLETTER{nbsp}REFERENCES{nbsp} |
|---|
First, there are significant differences in the interpretive guidelines specified for the COBAS Amplicor CT/NG assay between the United States and Europe. The European guidelines indicate that a single 
0.2 A450 result is a presumptive positive for N. gonorrhoeae, with a recommendation for retesting for confirmation. However, in the U.S. product insert, recommendations define an equivocal zone and call for repeated testing of specimens reading between 0.2 and 3.5 A450; at least two of three results (of the initial and the duplicate repeats) must measure 2.0 A450 in order to call the sample presumptively positive for N. gonorrhoeae. These were the guidelines used, appropriately, in our study (3). The U.S. recommendations were at least partially based on a study by Van Der Pol et al. (5), which showed that employing the equivocal zone retesting algorithm increased specificity to 98.6 to 99.9%, with little effect (0.1 to 4.9% decrease) on sensitivity in most specimen types. Our own data support this conclusion, as indicated by Table 1.
|
Although the equivocal zone retesting algorithm improves the specificity of the COBAS N. gonorrhoeae assay, significant numbers of false-positive results are still generated during testing of low-prevalence populations, leading to the recommendation by the Centers for Disease Control and Prevention that presumptive N. gonorrhoeae infection should be confirmed using independent testing for a different N. gonorrhoeae-specific target (2). We developed the NsppID assay to function as this confirmatory test. Once again, our results (as listed in Table 1) identifying nongonorrheal Neisseria spp. in the presumptive positive specimens after repeat testing reinforce the value of a second independent test.
In their letter, Peter-Getzlaff and colleagues indicated that the 16S assay used in their studies was less sensitive than either the cppB or the porA assay. Most published 16S assays (and possibly theirs as well) for the confirmation of N. gonorrhoeae were designed to amplify only N. gonorrhoeae, whereas our assay was designed to amplify many different Neisseria spp., with mismatches under the probe differentiating the various species. As described in our paper, the sensitivity of our assay was four genome copies of N. gonorrhoeae per reaction, even in the presence of 40-fold excess concentrations of the nongonorrheal Neisseria spp. Indeed, due to the melt curve analysis that differentiates between the Neisseria spp., we were able to detect dual infections in six clinical specimens. In comparison, the optimized porA assay had an analytical sensitivity of 20 genome copies of N. gonorrhoeae per reaction. In addition, although the optimized cppB assay detected more of the COBAS-positive samples than NsppID, we question the specificity of this assay, as a significant number of the discrepant specimens did amplify a nongonorrheal Neisseria sp. in the absence of any indication of N. gonorrhoeae in our NsppID assay. Several other laboratories have also found the cppB assay to be less specific than first thought (1, 4).
At the beginning of this project, there was only a single GenBank entry of a nongonorrheal Neisseria strain (one of several N. polysaccharea strains entered) that was identical in sequence to N. gonorrhoeae in the fluorescein-labeled probe region, and the assay was designed based on that information. Since that time, additional sequences have been entered into the GenBank database, and the list of entries with the 30-base sequence identical to that of N. gonorrhoeae in the fluorescein-labeled probe region includes some, but not all, strains of two additional Neisseria spp. (N. meningitidis and N. lactamica). Of these three nongonorrheal Neisseria spp., only N. lactamica has been shown to produce a false-positive result in the COBAS Amplicor CT/NG assay. Neither N. meningitidis nor N. polysaccharea has been shown to cross-react in a similar fashion. In our analytical validation, we tested a single strain of N. lactamica (ATCC 23970); the melting temperature peak observed in the melt curve analysis was over 6°C lower than that observed for all of the N. gonorrhoeae strains tested and was lower than that for any of the other Neisseria specimens tested as well. For these reasons, after analytical validation of the NsppID assay, we continued with the validation by using clinical specimens (endocervical swabs, urogenital swabs, or male urine). A specimen that was presumptive N. gonorrhoeae positive after repeat testing using the COBAS Amplicor CT/NG assay and that amplified in the NsppID assay with a resultant melting temperature of approximately 58°C was called N. gonorrhoeae positive. As pointed out by Peter-Getzlaff and colleagues, there remains the small possibility that N. polysaccharea, N. meningitidis, or N. lactamica could give the same NsppID result, but the likelihood of this occurring is very small if used as a confirmatory test of the COBAS Amplicor CT/NG assay with the more restrictive U.S. guidelines on the type of specimens described.
|
REFERENCES |
|---|
|
TopLETTERREFERENCESLETTER{nbsp}REFERENCES{nbsp} |
|---|
|
Karen L. Kaul* Kathy A. Mangold Evanston Northwestern Healthcare Department of Pathology and Laboratory Medicine Evanston, IL 60201
| ||||||
| * E-mail: k-kaul{at}northwestern.edu |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Antimicrob. Agents Chemother. | Clin. Microbiol. Rev. |
|---|---|
| Clin. Vaccine Immunol. | ALL ASM JOURNALS |
|---|
