Development of a panfungal real-time PCR

Abstract number: o219

Hinrikson  H.P., Jaton  K., Ciardo  D., Altwegg  M., Bille  J., Hauser  P.M.

Objectives: 

To develop and evaluate a panfungal real-time PCR for early detection and presumptive identification of pathogenic fungi directly in clinical specimens.

Methods: 

Based on a comprehensive sequence alignment, we designed a panfungal PCR targeting the variable region 5 of the small-subunit (18S) ribosomal gene. The assay consisted of 4 forward primers, 8 reverse primers and 1 fluorescence-labeled probe to ensure efficient detection of ascomycetes, basidiomycetes and zygomycetes. The sensitivity was assessed by comparison to a quantitative Aspergillus fumigatus-specific PCR. The specificity was tested using reference DNAs of fungal, bacterial and human origins. A retrospective clinical evaluation was conducted using DNA extracts from clinical specimens of both patients with proven invasive fungal infections [N = 15] and negative controls [N = 18]. The identity of amplicons was determined by comparative sequence analysis.

Results: 

The analytical sensitivity of the panfungal PCR was found comparable to that of the A. fumigatus-specific approach, i.e. one target copy per reaction. However, in order to avoid false positive reactions due to minute fungal contaminations introduced occasionally while sampling or processing specimens, the diagnostic detection limit was considered 5 target copies per reaction. The specificity for fungal DNA was confirmed by amplification of all fungal reference DNAs investigated and by the absence of positive reactions with reference DNAs of bacterial and human origins. The clinical sensitivity and specificity were 87 and 100%, respectively, when considering a particular specimen as PCR positive if it exhibited at least 2 positive reactions per triplicate testing. False negative results were obtained with 2 specimens previously shown to contain only trace amounts of target DNA. Sequencing of the clinical amplicons allowed presumptive identification of yeasts (Candida [N = 5], Cryptococcus [N = 1]), molds (Absidia [N = 1], Aspergillus [N = 2], Fusarium [N = 1]), Rhizomucor [N = 2] and Pneumocystis [N = 1] to at least the species group level. For instance, the usually more susceptible Candida species (e.g. C. albicans, C. parapsilosis, and C. tropicalis) were clearly distinguished from the usually less susceptible species such as C. glabrata and C. krusei.

Conclusion: 

The present panfungal real-time PCR represents a promising tool for improved diagnosis and empirical treatment of invasive fungal infections thus warranting further clinical validation.