Sara Posadas-Cantera ^1* Negin Mehrbarzin ¹ Simon Wetzel ¹ Hanna Goelz ^1,2 Lampros Kousoulas ³ Stefan Utzolino ³ Georg Häcker ^1,4 Mohamed Tarek Badr ^1,5*

• ¹ Institute of Medical Microbiology and Hygiene, Medical Center - University of Freiburg, Faculty of Medicine, Freiburg, Germany
• ² Institute of Clinical Chemistry and Laboratory Medicine, Medical Center - University of Freiburg, Faculty of Medicine, Freiburg, Germany
• ³ Department of General and Visceral Surgery, University of Freiburg Medical Center, Freiburg, Germany
• ⁴ BIOSS Centre for Biological Signaling Studies, University of Freiburg, Freiburg, Germany
• ⁵ IMM-PACT-Program, Faculty of Medicine, University of Freiburg, Freiburg, Germany

Objectives Ascites, often associated with critical pathologies such as liver cirrhosis or bowel perforation, can be complicated by fungal infection, increasing mortality especially in intensive care settings and demanding rapid diagnosis and adequate treatment. Traditional microbiological diagnostic methods have limited sensitivity in accurately identifying fungal pathogens in ascitic fluid. Alternative diagnostic methods may offer important insights to enable guiding of antifungal therapy and refining empirical treatment strategies. The objective of this study was to evaluate the potential of next-generation sequencing methods to identify specific fungal pathogens responsible for ascitic fluid infections. Methods We prospectively collected 50 ascitic fluid samples from ICU patients with suspected ascites infection. In addition to standard culture-based microbiological testing, an ascitic fluid aliquot underwent fungal DNA isolation and was analyzed by next-generation sequencing (NGS) methods for identification of fungal species. Results Of 50 ascitic samples collected, five samples showed growth of Candida spp. in culture. After DNA isolation and ITS2 PCR, detectable amplification was achieved in 10 samples. Sequencing of the 50 patients’ samples identified facultative pathogenic fungi in 19 patients. In 15 cases, culture alone would not have permitted the identification of all facultative pathogenic fungi. The identification of fungal DNA by sequencing was significantly associated with poor patient outcome and a number of clinical parameters. Conclusions Our results show a higher sensitivity for NGS-based diagnostic methods in the identification of ascitic fluid fungal infections compared to culture-based diagnostics. This may be beneficial especially for patients in a critical care setting, who have an increased prevalence of comorbidities and high mortality. The implementation of such methods in standard diagnosis will require increased standardization of the workflows and interpretation of the sequencing results with respect to patients’ clinical picture.