Background: The potential of next-generation sequencing (NGS) for hypothesis-free pathogen diagnosis from (poly-)microbially contaminated, formalin-fixed, paraffin embedded tissue samples from patients with invasive fungal infections and amebiasis was investigated. Samples from patients with chromoblastomycosis (n = 3), coccidioidomycosis (n = 2), histoplasmosis (n = 4), histoplasmosis or cryptococcosis with poor histological discriminability (n = 1), mucormycosis (n = 2), mycetoma (n = 3), rhinosporidiosis (n = 2), and invasive Entamoeba histolytica infections (n = 6) were analyzed by NGS (each one Illumina v3 run per sample). To discriminate contamination from putative infections in NGS analysis, mean and standard deviation of the number of specific sequence fragments (paired reads) were determined and compared in all samples examined for the pathogens in question.
Results: For matches between NGS results and histological diagnoses, a percentage of species-specific reads greater than the 4th standard deviation above the mean value of all 23 assessed sample materials was required. Potentially etiologically relevant pathogens could be identified by NGS in 5 out of 17 samples of patients with invasive mycoses and in 1 out of 6 samples of patients with amebiasis.
Conclusions: The use of NGS for hypothesis-free pathogen diagnosis from contamination-prone formalin-fixed, paraffin-embedded tissue requires further standardization.
Keywords: FFPE, formalin-fixed, paraffin-embedded samples; Hypothesis-free diagnosis of infection; Invasive amebiasis; Invasive fungal infections; Metagenome; Molecular diagnostics; NGS; Next-generation sequencing; Tropical infectious diseases.