Rapid Detection of Invasive Fungal Infections in Hemato-Oncological Patients.

Invasive fungal infections (IFI) play an increasingly important role as life-threatening complications in immunocompromised patients. Recent surveys show that annually 450.000 patients suffer from invasive mycoses worldwide. Early application of antimycotic agents is an essential prerequisite for successful therapy. However, standardized diagnostic techniques permitting rapid, and sensitive screening for the clinically relevant fungi have been lacking. Moreover, the increasing incidence of invasive infections caused by hitherto uncommon fungal species requires diagnostic tests with very broad specificity. We developed a panfungal real-time PCR (RQ-PCR) screening assay targeting the highly conserved region of the 28S ribosomal multicopy gene. Universal primers and probes for two test panels in two separate PCR reactions were established facilitating the detection and quantitative assessment of a broad range of pathogenic fungi. The reaction of test panel I covers a range of moulds including species from the fungal genera Aspergillus, Fusarium, Penicillium, Cladosporium, and Scedosporium. The test panel of reaction II facilitates the detection of yeasts from the genera Candida, Cryptococcus, Malassezia, and Trichosporon as well as Zygomycetes (Mucor, Rhizopus, Rhizomucor, Absidia), and other emerging fungal pathogens. In total, more than 80 human pathogenic fungal species can be detected and quantified. The detection limit of the panfungal RQ-PCR assay in clinical specimens was shown to be in the range of 1 fg fungal DNA, which corresponds to a fraction of a single fungal genome. To assess the clinical applicability of the technique, more than 1.000 clinical specimens derived from 144 well-documented pediatric hemato-oncological patients at high risk of invasive fungal infection were analyzed. The specimens investigated included predominantly serial plasma samples (n>950), and a small number of other materials including bronchotracheal secretion, bronchoalveolar lavage, lung biopsy, liver biopsy, and cerebrospinal fluid. In about 65% of the patients analyzed, evidence of fungemia was found in serial specimens. In a large proportion of these patients (31%), both reactions revealed positive results indicating mixed fungal infections. Isolated positivity of reaction I or II was observed in 29% and 5%, respectively. The high percentage of patients with PCR-detectable fungal DNA in normally sterile clinical specimens raises questions about the interpretation and clinical relevance of the findings. The possible occurrence of contamination was largely excluded by the implementation of multiple controls at all stages of the diagnostic process. The great sensitivity of RQ-PCR analysis may permit the detection of low-level fungal DNAemia which may not generally reflect an imminent risk of severe fungal disease. Interpretation of the molecular findings therefore has to be performed with caution. Current evaluation of the RQ-PCR data in relation to the EORTC definitions of proven, probable and possible invasive fungal infection, which are based on clinical criteria, host factors, and microbiological criteria, revealed a very good correlation. Indeed, most patients with repeatedly positive RQ-PCR results, which had been generated in a double-blind fashion, received antifungal treatment on the basis of other findings. The assay presented is a powerful tool for rapid and economic screening of IFIs and ongoing analyses are expected to reveal how the technique could be implemented in clinical diagnostics to support timely onset and management of antifungal treatment.