Introduction: New fungal species are increasingly reported in immunocompromised patients. Saprochaete clavata (S. clavata), an ascomycetous fungus formerly called Geotrichum clavatum, is intrinsically resistant to echinocandins and is often misidentified.
Objective: We describe a cluster of seven S. clavata infections in hospitalized hematology patients who developed this rare fungemia within a span of 11 months. Three of the seven patients died. Identification of the isolates was determined only with the Saramis database of VitekMS system and sequencing of the internal transcribed spacer (ITS) region. Clonal relatedness of the isolates was determined by matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF) analysis; clonal correlation between the strains was investigated by means of phylogenetic analysis, based on single-nucleotide variants (SNPs). Clinical presentation, 1-3 β-D-glucan (BG) and galactomannan (GM) antigen results and analysis of possible sources of contamination are also described with a prospective case-control study of the outbreak.
Results: MALDI-TOF MS-Vitek (bioMerieux, Marcy l'Etoile, France) failed to identify the six isolates, while SARAMIS (bioMerieux, Marcy l'Etoile, France) identified the isolates as S. clavata. Initially, Vitek 2 identified the strains as Geotrichum capitatum in two of the seven cases. Molecular identification gave 99% homology with S. clavata. BG was positive in three out of six patients (range 159 to >523 pg/ml), GM results were always negative. All the isolates were resistant to echinocandins (anidulafungin, micafungin, and caspofungin) and Fluconazole, but susceptible to Flucytosine and Voriconazole. One isolate showed acquired resistance to Flucytosine and Amphotericin B during treatment. Both the correlation-based dendrograms obtained by MALDI-TOF MS (Bruker Daltonics) and MS-Vitek not only clustered six of the seven bloodstream infection (BSI) isolates in the same group, but also showed their strong relatedness. Phylogenetic analysis using SNPrelate showed that the seven samples recorded during the investigation period clustered together. We observed a split between one case and the remainder with a node supported by a z-score of 2.3 (p-value = 0.021) and 16 mutations unique to each branch.
Conclusion: The use of proteomics for identification and evaluation of strain clonality in outbreaks of rare pathogens is a promising alternative to laborious and time-consuming molecular methods, even if molecular whole-genome sequencing (WGS) typing will still remain the reference method for rare emergent pathogens.