1. Introduction
Filamentous fungal infections are an emerging problem in clinical practice, ranging from mild onychomycosis to severe invasive disease in immunocompromised patients. The high mortality of systemic mould infections requires prompt initiation of specific antifungal therapy. Since some species are intrinsically resistant to certain antifungals and susceptibility testing is not widely available, rapid and accurate species identification is crucial for the selection of appropriate antifungal treatment [
1]. This diagnostic goal, however, is increasingly challenged due to the description and emergence of new opportunistic fungal species [
2,
3].
Traditional diagnostic techniques for filamentous fungi are based on morphological and physiological characteristics and, therefore, require experienced mycologists, a resource that is scarce in most microbiological laboratories [
4,
5]. Additionally, these methods are time-consuming and not always accurate [
6,
7,
8]. DNA amplification and sequencing is an additional diagnostic tool; however, this approach is not commercially available and mostly used in reference laboratories [
4].
Matrix-assisted laser desorption/ionization-time-of-flight (MALDI-TOF) mass spectrometry has revolutionized the diagnosis of bacteria and yeast, but its capacity to identify filamentous fungi has been less studied [
7]. Early reports showed poor performance due to non-optimized extraction procedures and limited pathogen databases [
9]. However, subsequent studies have demonstrated that with an appropriate reference database and extraction protocol, MALDI-TOF can provide rapid identification results for filamentous fungi and has the potential to become a standard identification method [
10,
11]. For example, the comparison of three pretreatment procedures applied to different MALDI-TOF instruments showed better diagnostic efficacy when cell wall disruption agents were used before analysis [
12].
In 2017, the VITEK MS 3.0 system received FDA approval for a Mould Kit (inactivation/extraction reagents] and a database version, which showed promising results in initial evaluations [
11]. An updated version of the database (Knowledge Base v3.2.0) was released in 2018, containing a broader spectrum of fungal pathogens (55 additional taxa). The present study evaluated the specificity and practicality of this new kit together with the updated database on the VITEK MS platform to identify filamentous fungi in a microbiology laboratory in Chile.
3. Results
The study included a total of 67 isolates of filamentous fungi; 33 derived from clinical samples and 34 from strain collections (including one ATCC strain). Strains belonged to 30 species and 5 species complexes within 15 genera;
Aspergillus was the predominant genus. Of these 30 species, 20 (66.7%) were correctly identified by VITEK MS; 3 (10%) (
Aspergillus flavus,
Rhizopus microspores, and
Sporothrix globosa) and 1 (
Coccidioides posadasii) (3.3%) were correctly classified within the species complex and genus, respectively (
Table 1). Six species were not identified, of which 4 (
Aspergillus tritici,
Penicillium canescens,
Rhizopus delemar, and
Sporothrix chilensis) were not included in the database;
Penicillium brevicompactum and
Penicillium expansum were not identified, albeit being part of the library database. All 5 species complexes were correctly diagnosed by VITEK MS (
Table 1).
Of the 67 isolates, 61 (91.0%) were correctly identified by VITEK MS; 39 (58.2%) to the species level, 21 (29.9%) to the species complex level, and 1 (1.5%) to the genus level (
Table 2). Six strains (9.0%) were not recognized by VITEK MS. Overall, in 55 isolates (82.0%; 39 species and 16 species complex) VITEK MS result coincided with the identification of the reference laboratories (
Table 2). No misidentifications were observed.
The performance of the VITEK MS system showed variations among the tested genera. The genus Aspergillus accounted for 31 (46.3%) of isolates, with Fumigati as the predominant section (n=12). All Aspergillus fumigatus isolates were identified to the species level by VITEK MS. Other clinically important, but less frequent members of the genus Aspergillus, such as Aspergillus flavus, Aspergillus niger, and Aspergillus terreus were identified to the species complex level (
Table 1). Fusarium was the second most frequent genus comprised in our study. All isolates of the 3 included Fusarium species or species complexes were correctly identified. Of the 5 Penicillium species, 2 were correctly identified; the other 3 were not recognized by VITEK MS, although 2 were present in the database. Most members of the Mucorales (i.e., Lichtheimia corymbifera, Mucor velutinosus, Rhizopus arrhizus complex, and Rhizopus microsporus) were identified and only R. delemar, which was not included in the library, failed to be recognized.
Our laboratory personnel categorized the Mould Kit as easy-to-use, which permitted an easy workflow integration. Since all reagents were provided as ready-to-use components, hands-on time of the inactivation/extraction steps was short (30 to 60 minutes). The smear preparation was the same as for bacteria or yeasts, including the calibration step using Escherichia coli ATCC 8739 strain. All VITEK MS results were unambigous; all given identifications were provided with the maximum discrimination value (99.9%). In case of endemic fungi identification (Coccidioides posadasii), the system gave a biohazard warning, which helped to rapidly implement laboratory security measures.
4. Discussion
The present study examined a broad spectrum of filamentous fungi, including various clinically relevant species such as
Aspergillus spp.,
Fusarium spp. and species of Mucorales and dematiaceous fungi. Timely and correct identification of such fungal isolates is of high clinical priority, allowing a prompt adaptation of antimycotic therapy. This goal is often not achievable using traditional methods, which often require weeks until the necessary growth of mature fungal colonies [
6]. Besides, the shortage of well-trained and experienced mycologists further prolongs the time to diagnosis, since difficult-to-identify isolates must be sent to reference centers [
4]. Under these aspects, the reduction of time to reliably identify mould species in a routine laboratory is potentially the major advantage of commercial MALDI-TOF platforms.
Since its implementation, VITEK MS has increased the diagnostic speed and accuracy of our laboratory, especially in cases of difficult-to-identify microorganisms [
17,
18]. The present study showed similar results for filamentous fungi, which were processed and identified after 48-72 hours of incubation, with the initial growth of mycelia. This is in accordance with previous studies, in which identification was achieved after 48 hours [
7,
8,
10]. In early reports, the use of intact fungal cells hindered mass spectrometry processing due to the rigid cell walls of moulds [
19]. Protocols that use an on-plate formic acid extraction method have shown insufficient results [
12,
20]. However, the development of specific extraction solutions, such as the here tested VITEK MS Mould Kit, aimed to overcome this limitation [
19,
21].
Another crucial factor for the identification of less frequent pathogens is the system´s database. Previous diagnostic studies of filamentous fungi using the former database version (v3.0) showed high rates of misidentifications [
11,
22,
23] or identification only to the species complex level [
24]. The v3.2.0 database utilized in our study exhibited a better performance and coincided with the reference identification in over 80% of isolates.
Fusarium spp., for example, were correctly identified in the present study (n = 7), while in studies, using the older library version, only 65%-70% were accurately diagnosed [
22,
24]. Furthermore, we correctly detected cryptic species such as
Aspergillus sydowii and
Aspergillus versicolor, which pose a treatment challenge due to their azole resistance [
25]. Of notice, we did not observe any misidentification, which is of clinical importance, avoiding incorrect or sub-optimal treatment, which is in accordance with previous data [
12].
Aspergillus fumigatus, grouped in the
Fumigati section, is the most prevalent cause of invasive mold infection [
26]. This species was the most common in our study and correctly identified in all cases. Accurate diagnosis of this species has become crucial due to the emergence of azole-resistant cryptic species within this section, such as
Aspergillus lentulus, Aspergillus novofumigatus, Aspergillus fumigatiaffinis, Aspergillus thermomutatus (
Neosartorya pseudofischeri), and
Aspergillus viridinutans [
27]. These and other members of this section associated with invasive fungal diseases were not reliably identified by the earlier version (v3.0) of the platform [
27]. Those species are now included in the new database version. However, our study did not include such cryptic
Fumigati species. Isolates belonging to the
Aspergillus niger and
Aspergillus terreus complexes were correctly identified to the complex level, as with the older library (v3.0) [
24]. The v3.2.0 database now also includes some cryptic
A. niger complex species such as
A. tubingensis. Our work included cryptic species of other sections (
Aspergillus calidoustus and
Aspergillus sydowii), which were correctly identified.
A. tritici was not included in the database and, therefore, not identified.
The genus
Fusarium is mainly described as a phytopathogenic fungi. However,
Fusarium oxysporum complex and
Fusarium solani complex are able to cause local (e.g., ocular) to life-threatening infections in immunocompromised patients [
28].
Fusarium solani complex was most virulent in animal models; furthermore it exhibits amphotericin B, voriconazole, and posaconazole resistance [
24,
29]. The rapid and accurate identification of this complex is therefore clinically relevant. This and previous studies demonstrated that the newer VITEK MS system exhibits a better performance than other MALDI-TOF instruments [
20].
Penicillium spp. are of low human pathogenicity.
Penicillium chrysogenum is the main species isolated from dwellings and considered an important cause of allergic rhinitis and asthma [
30]. Of the
Penicillium species included in this study, the VITEK MS only identified
P. chrysogenum, while
P. brevicompactum (agent of hypersensitivity pneumonitis for wood workers) and
P. expansum were not identified, although within the database.
Penicillium canescens was not part of the database, so no identification was obtained. These limitations within the
Penicillium genus have been described previously and might require further updates of the database [
23].
The most common Mucorales involved in human disease are
Rhizopus, Mucor and
Rhizomucor [
31]. An accurate identification of Mucorales is clinically relevant due to species-specific clinical presentations (e.g.
Rhizopus arrhizus causing rhinocerebral invasion). VITEK MS correctly identified the included Mucorales to the species/species complex level [
32]. However, only a low number of Mucorales strains were included, so further studies are warranted. This was also the case with dematiaceous fungi.
Coccidioides posadasii, a dimorphic fungus, was correctly identified as
C. immitis/posadasii and signaled by the device as a biosafety risk. The capability of MALDI-TOF to identify this fungus permitted timely measures for the prevention and control of laboratory acquired infections, as reported previously [
16]. This is an important add-on for laboratories in non-endemic countries, since clinical information leading to the suspicion of endemic mycosis, are often not available. Though dimorphic fungi are identifiable by VITEK MS and it is the only FDA approved database for their diagnosis,, the need for an inactivation step in a biosafety level 3 (BSL-3) facility and the requirement of a minimal amount of culture biomass for identification, will probably preclude its use in a routine BSL-2 facility [
33].
Other advantages of fungal identification by MALDI-TOF include its technical simplicity, short time for sample preparation and data analysis, and economical savings in laboratory costs [
4,
34]. Until recently, sequencing was the only alternative for unidentifiable mould strains, resulting in much higher expenses and time losses compared to MALDI-TOF [
35]. A major drawback to the implementation of the VITEK MS Mould Kit is the large package size (100 tests per box) together with a short expiry period after opening (4 weeks).
Main limitations of our study were the limited sample size and the incomplete spectrum of species, especially among cryptic Aspergillus species, Mucorales, dematiaceous molds, and dermatophytes.
Author Contributions
Conceptualization, L.P. and T.W; methodology, L.P., T.W. and R.C.; validation, T.W., L.P., C.D. and P.L.; formal analysis, T.W, L.P., R.C., C.D., I.P., P.G., P.L., F.V. and T.W; investigation, C.V. and I.P.; resources, C.V., R.C., C.D., P.G. and F.V.; data curation, I.P., R.C. and L.P.; writing—original draft preparation, L.P., R.C. and I.P.; writing—review and editing, L.P., R.C., I.P., C.V., P.G., P.L., F.V. and T.W; supervision, L.P., R.C., P.G., F.V. and T.W.; project administration, L.P. All authors have read and agreed to the published version of the manuscript.