Performance of Targeted Fungal Sequencing for Culture-Independent Diagnosis of Invasive Fungal Disease.

Abstract

Identification of fungi causing invasive fungal disease (IFD) is critical for guiding antifungal therapy. We describe the performance and clinical impact of a targeted panfungal polymerase chain reaction (PCR) amplicon sequencing assay for culture-independent diagnosis of IFD. Between January 2009 and September 2016, 233 specimens, consisting of fresh and formalin-fixed, paraffin-embedded (FFPE) tissues and sterile body fluids with known diagnosis of IFD based on reference method results (n = 117), and specimens with negative fungal culture, but with microscopic and ancillary findings indicative of IFD (n = 116), were included. PCR amplicons from the internal transcribed spacer 2 and the D2 region of 28S ribosomal RNA gene were sequenced and fungi identified. Sensitivity and specificity of fungal sequencing in specimens with known diagnosis were 96.6% (95% confidence interval [CI], 87.4%-99.4%; 58/60) and 98.2% (95% CI, 89.4%-99.9%; 56/57). In patients with suspected IFD, the diagnostic yield of fungal sequencing was 62.9% (73/116) overall and 71.3% (57/80) in patients classified with proven IFD based on the European Organization for Research and Treatment of Cancer/Invasive Fungal Infections Cooperative Group and Mycoses Study Group (EORTC/MSG) criteria. Samples obtained by open biopsy had a significantly higher diagnostic yield (71.5% [40/56]) compared with core-needle biopsy (50% [17/34] P = .04) and fine needle aspiration (0% [0/2]; P = .009). Additionally, D2 sequencing diagnosed 5 cases of invasive protozoal infections due to Toxoplasma gondii (n = 3), Trypanosoma cruzi, and Leishmania species. Sequencing results altered patient management in the majority of suspected cases. The targeted fungal sequencing assay allowed accurate identification of fungi causing IFD and additionally provided partial-protozoal coverage. The diagnostic yield was dependent on the amount of tissue available for testing.