Abstract
Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) systems are now widely implemented in routine microbiology laboratories. However, new solutions are emerging that show promising analytical performance. This study aims to compare the analytical and practical performance of the Autof MS2600 (Autobio) with that of the market-leading Biotyper Sirius (Bruker). A total of 1006 clinical isolates, representative of the bacterial strains routinely encountered in a university clinical microbiology laboratory, were analysed in parallel on the Autof MS2600 and the Biotyper Sirius systems. According to the manufacturers’ classification criteria, 91.0% and 95.5% of isolates were correctly identified at the species level by Autof MS2600 and Biotyper Sirius, respectively. This difference was statistically significant for Gram-positive but not for Gram-negative isolates. The rate of non-identification was very low and similar between the two systems. Apart from two discordant results, no additional identification errors were detected, and these had no clinical consequences. Overall, both systems showed comparable analytical and practical performance. Lowering the identification threshold for the Autof MS2600 enabled a species-level identification rate for Gram-positive bacteria statistically equivalent to that of the Biotyper Sirius. However, these thresholds should be validated beforehand. Anaerobic bacteria, filamentous fungi and yeasts also need to be validated in larger cohorts. This new MALDI-TOF MS system is suitable for implementation in routine clinical microbiology laboratories.
Supplementary Information
The online version contains supplementary material available at 10.1007/s10096-026-05497-2.
Keywords: MALDI-TOF MS, Autof MS2600, Autobio, Biotyper Sirius, Bruker, Analytical performance
Introduction
For over fifteen years, matrix-assisted laser desorption/ionisation mass spectrometry (MALDI-TOF MS) has been widely used in routine clinical microbiology laboratories [1, 2]. From one small colony combined with a matrix, a microorganism can be identified within minutes, unlike older biochemical and phenotypic identification methods, which may take up to 24 h. In addition to being cost-effective with a rapid return on investment [3, 4], MALDI-TOF MS accelerates crucial therapeutic decisions, thereby contributing to reduced hospital stays [5].
The sensitivity of the MALDI-TOF MS system increases over time thanks to the continuous integration of protein spectra from microorganisms into its database. This has strengthened its identification capabilities, particularly for uncommon pathogenic microorganisms. For example, Staphylococcus lugdunensis (involved in skin infections) and Aerococcus urinae (responsible for urinary tract infections) can now be more easily differentiated from commensal or non-pathogenic bacteria [6]. Moreover, the inclusion of a representative number of isolates per species has increased the robustness of the system against intra-species variability, facilitating the identification of fastidious organisms such as those from the HACEK group. These were previously identified using morphological and biochemical analyses, and, if unsuccessful, required 16 S rRNA gene sequencing [7, 8]. This has resulted in a reduction in the number of additional molecular biology analyses required, which are often very costly [9–11].
One of the well-known and still unresolved limitations of MALDI-TOF MS systems is its inability to reliably distinguish between certain groups of bacteria such as Escherichia coli/Shigella sp. or the Enterobacter cloacae complex group due to their very similar protein composition. Another limitation of MALDI-TOF MS is that accurate identification requires pure cultures. In mixed cultures, overlapping peaks generate complex spectra, preventing reliable identification [1, 12].
Currently, the MALDI-TOF MS Biotyper Sirius (Bruker Daltonics, Bremen, Germany) and VITEK MS (BioMérieux, Marcy l’Étoile, France) systems, both IVD-certified systems, dominate the global MALDI-TOF MS market [13]. In recent years, other MALDI-TOF MS systems such as ASTA MicroIDSys (ASTA, Suwon, Republic of Korea), Axima (Shimadzu, Kyoto, Japan) and EXS3000 (Zybio, Chongqing, China) have emerged [1, 14, 15]. Recently, the Autobio company (Zhengzhou, China) released a new MALDI-TOF MS in early 2025: Autof MS2600. Several comparative studies have assessed the performance of these systems, generally reporting similar analytical accuracy [13, 16, 17]. The choice of one system over another depends mainly on practical considerations and agreements with suppliers.
To our knowledge, only one study (Elena de Carolis et al.) conducted outside China has looked at the performance of the Autof MS2600. The comparison involved 535 strains (434 bacteria and 101 yeasts) analysed using the Autof MS2600 system (Autof V2.0.196 software, Autobio library V1120) and the MBT Smart system (Bruker Daltonics, Bremen, Germany; MALDI Biotyper software version 4.1.14, Bruker Biotyper library V11.0.0). Both systems showed good performance [18]. Autobio has also recently obtained CE IVD certification for its Autof series in December 2023. Although the Autof MS2600 has shown promising performance compared with the older MBT Smart system, no study has assessed its performance against the latest Bruker Biotyper Sirius. Furthermore, limited data are available regarding its usability in routine workflows.
The objective of this study is to compare the analytical performance and practicality of two MALDI-TOF MS systems: the new Autof MS2600 (Autobio) and the Biotyper Sirius (Bruker), for the purpose of use in a routine microbiology laboratory.
Materials and methods
Population and bacterial isolates
Samples were collected at the Brussels University Hospital Laboratory (LHUB-ULB), serving the main hospitals in Brussels. Between March and May 2025, at least one MALDI-TOF MS target plate per day, randomly selected from routine laboratory workflow, was included in the evaluation to reflect the typical bacteriological activity of the LHUB-ULB. Clinical isolates originated from a wide range of sample types—including urine, genital swabs, wound swabs, respiratory screening swabs, ophthalmic swabs, and catheter tips—all processed through the laboratory’s automated Kiestra workflow system.
Samples were cultured according to LHUB-ULB’s standardized procedures, with medium selection and incubation conditions adapted to the specimen type. Incubation was usually performed for 18–24 h at 36 ± 1 °C under atmospheric O₂ or with added CO₂, with some plates re-examined after 48 h. The culture media tested included Columbia agar (Becton Dickinson, Franklin Lakes, NJ, USA), Mannitol salt agar (Becton Dickinson), MacConkey agar (Becton Dickinson), Cystine–Lactose–Electrolyte-Deficient (CLED) agar (Becton Dickinson), Columbia CNA/MacConkey (CAP/Mac) agar (Thermo Fisher Scientific, Waltham, MA, USA), methicillin-resistant Staphylococcus aureus (MRSA) screening agar (Biomérieux, Marcy-L’Etoile, France), vancomycin-resistant Enterococcus (VRE) agar (Biomérieux), extended-spectrum β-lactamase (ESBL) and carbapenemase-producing Enterobacteriaceae (Carboxa) agar (Biomérieux), and Chocolate agar (Biomérieux).
MALDI-TOF MS analysis
To compare the two automated identification systems, routine technologists deposited a single colony per isolate on the target plate for analysis with the Bruker Biotyper Sirius, following LHUB-ULB’s standard procedure, while for the Autof MS2600 evaluation, the same isolates were additionally spotted in duplicate by an investigator (a student).
MALDI-TOF MS Biotyper Sirius analysis
The analyses were carried out in accordance with the manufacturer’s protocols. Using a toothpick, colonies on plates were taken and spotted on a target plate. For each target plate, a bacterial test standard (BTS) strain of E. coli was applied as an internal control, in accordance with the manufacturer’s requirements; successful identification of this control was mandatory for the subsequent analysis of the routine isolates to proceed. Next, one microliter of a α-cyano-4-hydroxycinnamic acid (CHCA) matrix solution (ref. 8255344, Bruker) was added to each spot. The CHCA was reconstituted with an organic solvent (OS) solution (50% acetonitrile, 47.5% water, 2.5% trifluoroacetic acid) purchased from VWR International, Leuven, Belgium (ref. HONC 19182, 250 mL). Once the matrix was dry, the plate was loaded into the instrument. The system, equipped with a 355 nm laser (up to 500 million laser shots), acquired mass spectra in the 2–20 kDa range, the standard range for microorganism identification, using MBT Compass HT IVD software (version 5.2.320). Spectra were acquired in automatic mode using 240 laser shots per spot, with 40 shots collected from each of six different positions. Spectra were then compared with the MALDI Biotyper database (version 12) for microbial identification. The library contains thousands of spectra comprising 4,239 species and 686 genera.
MALDI-TOF MS Autof MS2600 analysis
Isolate identification was performed according to the manufacturers’ protocol. Colonies were spotted on the target plate with a control spot based on the E. coli calibrator provided by Autobio. Each spot was overlaid with 1 µL of CHCA matrix solution (MSA03, Autobio), supplied ready to use and containing acetonitrile, water, formic acid, and a small amount of trifluoroacetic acid. Once dry, the plate was loaded into the automated system equipped with a 355 nm laser (up to 7 billion laser shots). Spectra were acquired in the 2 to 20 kDa, the range used for microorganism identification, in automatic mode using 240 laser shots per spot, with 40 shots collected from each of six different positions. An identification is provided by comparison with the spectra in the Autof Acquirer database (version V1.1.22). This database contains 5189 species, more than 1000 genera and 18375 strains.
Data analysis
The reliability thresholds provided by the two manufacturers, Autobio and Bruker (see Table 1), were used as criteria to identify germs. For each isolate, the highest identification score obtained from the two spots was used as the final result. For certain species with similar protein compositions, such as E. coli and Shigella sp. or the Enterobacter complex group, both systems flagged potential identification limitations by displaying comments indicating matches with other species (‘matching hints’), as specified by the manufacturers.
Table 1.
The reliability thresholds for species and genus identification provided by the two manufacturers, Autobio and Bruker
| Biotyper Sirius (Bruker) |
Autof MS2600 (Autobio) |
|
|---|---|---|
| Identification of genus and species (green) | ≥ 2 | ≥ 9 |
| Genus identification (yellow) | 1.7–1.99 | 6–8.999 |
| No reliable identification (red) | < 1.7 | < 6 |
Classification
The results of both automated systems were classified as follows: “correct identification of genus and species”, “correct identification of genus” (i.e., a score providing a reliable identification at the genus level but not at the species level, or cases in which the species was not specified), “no identification”, “incorrect identification at genus level” and “incorrect identification at species level”. Organisms that are known to be indistinguishable by MALDI-TOF MS and for which this limitation has no clinical implication were classified in the “correct identification of genus and species” category (e.g., Enterobacter cloacae complex isolates), as is customary in our routine practice. Isolates were also categorized as Gram-positive bacteria, Gram-negative bacteria, or yeasts to enable comparative analysis within each microbial group.
Discrepancies
During the first run, if the Biotyper Sirius failed to identify the isolate at the genus or species level, a second run was performed on both systems using duplicate spots deposited by the same investigator (a student), in order to ensure comparability and avoid bias linked to the initial double-spot protocol used exclusively for the Autof MS2600.
In the event of a discrepancy in genus or species between the Autof MS2600 and the Biotyper Sirius, a new double spot analysis was performed on each of the automated systems to eliminate any risk of contamination (e.g. incorrect colony picked). If discrepancies were confirmed, colonies were identified by partial 16 S rRNA gene sequencing using the Sanger method on an Applied Biosystems 3500 Series Genetic Analyzer (Thermo Fisher Scientific, Waltham, MA, USA). The protocol with the different stages of the study is presented in Fig. 1.
Fig. 1.
Summary of the study protocol comparing Autobio’s Autof MS2600 and Bruker’s Biotyper Sirius
Partial 16 S rRNA gene sequencing
DNA was extracted from pure strains using the EZ1&2 Virus Mini Kit on an EZ2® Connect MDx instrument (Qiagen, Germany). The full-length 16 S rRNA gene was amplified by PCR with Promega 2X Master Mix using primers TWACACATGCAAGTCGARCG (forward) and CCAACATCTCACGACACGAG (reverse), including a short pre-amplification step. Sequencing was performed with the BigDye Terminator v1.1 Kit on a 3500 Series Genetic Analyzer. Species and genus were assigned based on GenBank sequence similarity (≥ 99% and ≥ 97%, respectively).
Exclusion criteria
Strains for which a duplicate re-analysis was required, due to discrepancies or colony collection errors, but could not be technically spotted in duplicate (because they were dead or had been discarded by the automated line), were excluded from the study (n = 12).
Impact of the culture medium
The impact of the culture medium on identification by MALDI-TOF MS was assessed using data from our study, focusing on the two most frequently identified isolates: one Gram-positive (S. aureus) and one Gram-negative (E. coli). For each culture medium, species- and genus-level identification rates, as well as the proportion of unidentified results, were calculated according to each manufacturer’s classification thresholds (Table 1) for both strains on both automated systems. This approach allowed us to perform statistical comparisons between culture media and to determine whether culture media have an impact on MALDI-TOF MS identification performance.
Repeatability and reproducibility
To evaluate repeatability and reproducibility on the Autof MS2600, one strain of E. coli and one strain of S. aureus, representing the most frequently isolated Gram-negative and Gram-positive bacteria, were randomly selected from routine testing. Repeatability was assessed by measuring each strain 30 times on the same target plate. Reproducibility was evaluated by processing each strain daily over a 12-day period using duplicate spot deposition, with subculturing every 2 to 3 days to maintain freshness. For reproducibility, the spot with the highest identification score each day was selected. Species- and genus-level identification rates, as well as the proportion of unidentified results, were calculated according to the classification thresholds defined by the two manufacturers (Table 1).
Statistical methods
To compare identification results between Bruker’s Biotyper Sirius and Autobio’s Autof MS2600, the two-tailed McNemar test was used. Additionally, chi-square test was used to assess the effect of culture medium on species- and genus-level identification rates, as well as on unidentified results for both the Autof MS2600 and Biotyper systems.
Practical points
To complement the analytical comparison, both systems were evaluated on several practical aspects. These included, among others, software usability (assessed by four laboratory technologists/experimenters using a 5-point Likert scale), routine maintenance requirements, plate capacity, laser repetition frequency and theoretical throughput.
Results
Among the 1006 included isolates, a total of 95 species and 34 genera were identified, with four species belonging to the genus Candida. Species sharing the same matching hint were grouped together to account for the ambiguity inherent in MALDI-TOF MS identification. The study isolates originated from various sample types, distributed as follows: 32% screening tests, 24% urine samples, 20% samples from skin and soft tissue infections, 12% respiratory samples, 5% genital samples, and 7% from miscellaneous other sources. The distribution of routine isolates and the variety of samples are presented in Figs. 2 and 3.
Fig. 2.
Distribution of the 1006 included isolates
Fig. 3.
Overview of sample types processed in routine microbiology at LHUB-ULB (1006 Isolates). RTI, respiratory tract infection; UTI, urinary tract infections; GEN, genital tract swabs; SSI, skin and soft tissue infections; Screening, screening for methicillin-resistant S. aureus, vancomycin-resistant Enterococcus, Enterobacteriaceae carrying extended-spectrum beta-lactamases and carbapenemases; Others, other samples such as normally sterile fluids, ophthalmic samples and catheter
The results of strain identification at the genus or species level, obtained from the Biotyper and Autof databases, are presented in Table 2. A comparison of the performance of the MALDI-TOF MS systems is presented in Table 3.
Table 2.
Results of identifications performed on the Autof MS2600 and Biotyper Sirius on 1006 isolates from routine samples
| Autof MS2600 (Autobio) | Biotyper Sirius (Bruker) | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Total | Correct genus and species ID a | Correct genus ID | No ID | Species-level misidentification | Genus-level misidentification | Correct genus and species ID | Correct genus ID | No ID | Species-level misidentification | Genus-level misidentification | |
| Total isolates | 1006 | 915 | 78 | 11 | 2 | 0 | 961 | 33 | 12 | 0 | 0 |
| Gram-positive | 411 (40.9%) | 357 | 52 | 2 | 0 | 0 | 397 | 11 | 3 | 0 | 0 |
| Actinobaculum sp. | 1 (0.1%) | 0 | 1 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 |
| Actinomyces turicensis | 1 (0.1%) | 1 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 |
| Anaerococcus sp. | 1 (0.1%) | 0 | 1 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 |
| Bacillus cereus thuringiensis | 3 (0.3%) | 2 | 1 | 0 | 0 | 0 | 3 | 0 | 0 | 0 | 0 |
| Bacillus sp. | 1 (0.1%) | 0 | 1 | 0 | 0 | 0 | 0 | 0 | 1 | 0 | 0 |
| Corynebacterium amycolatum | 2 (0.2%) | 2 | 0 | 0 | 0 | 0 | 2 | 0 | 0 | 0 | 0 |
| Corynebacterium aurimucosum | 2 (0.2%) | 1 | 1 | 0 | 0 | 0 | 1 | 1 | 0 | 0 | 0 |
| Corynebacterium diphtheriae | 1 (0.1%) | 1 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 |
| Corynebacterium otitidis | 1 (0.1%) | 0 | 1 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 |
| Corynebacterium phoceense | 1 (0.1%) | 0 | 1 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 |
| Corynebacterium propinquum | 2 (0.2%) | 2 | 0 | 0 | 0 | 0 | 1 | 0 | 1 | 0 | 0 |
| Corynebacterium pseudodiphtheriticum | 1 (0.1%) | 0 | 0 | 1 | 0 | 0 | 1 | 0 | 0 | 0 | 0 |
| Corynebacterium striatum | 4 (0.4%) | 4 | 0 | 0 | 0 | 0 | 4 | 0 | 0 | 0 | 0 |
| Enterococcus avium | 1 (0.1%) | 1 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 |
| Enterococcus faecalis | 31 (3.1%) | 30 | 1 | 0 | 0 | 0 | 30 | 1 | 0 | 0 | 0 |
| Enterococcus faecium | 32 (3.2%) | 32 | 0 | 0 | 0 | 0 | 32 | 0 | 0 | 0 | 0 |
| Gardnerella vaginalis | 4 (0.4%) | 2 | 2 | 0 | 0 | 0 | 3 | 1 | 0 | 0 | 0 |
| Gleimia europaea | 1 (0.1%) | 1 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 |
| Lactobacillus crispatus | 1 (0.1%) | 1 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 |
| Lactobacillus gasseri | 2 (0.2%) | 2 | 0 | 0 | 0 | 0 | 2 | 0 | 0 | 0 | 0 |
| Lactobacillus jensenii | 2 (0.2%) | 2 | 0 | 0 | 0 | 0 | 2 | 0 | 0 | 0 | 0 |
| Macrococcus caseolyticus | 1 (0.1%) | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0 | 0 |
| Paenibacillus amylolyticus | 1 (0.1%) | 1 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 |
| Pediococcus acidilactici | 3 (0.3%) | 3 | 0 | 0 | 0 | 0 | 3 | 0 | 0 | 0 | 0 |
| Pseudoglutamicibacter sp. | 1 (0.1%) | 0 | 1 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 |
| Staphylococcus aureus | 141 (14%) | 134 | 7 | 0 | 0 | 0 | 141 | 0 | 0 | 0 | 0 |
| Staphylococcus borealis | 2 (0.2%) | 1 | 1 | 0 | 0 | 0 | 2 | 0 | 0 | 0 | 0 |
| Staphylococcus capitis | 1 (0.1%) | 1 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 |
| Staphylococcus epidermidis | 40 (4.0%) | 28 | 12 | 0 | 0 | 0 | 39 | 1 | 0 | 0 | 0 |
| Staphylococcus haemolyticus | 41 (4.1%) | 33 | 8 | 0 | 0 | 0 | 41 | 0 | 0 | 0 | 0 |
| Staphylococcus hominis | 3 (0.3%) | 3 | 0 | 0 | 0 | 0 | 3 | 0 | 0 | 0 | 0 |
| Staphylococcus lugdunensis | 4 (0.4%) | 3 | 1 | 0 | 0 | 0 | 4 | 0 | 0 | 0 | 0 |
| Staphylococcus pasteuri | 2 (0.2%) | 1 | 1 | 0 | 0 | 0 | 1 | 1 | 0 | 0 | 0 |
| Staphylococcus pragensis | 1 (0.1%) | 0 | 0 | 1 | 0 | 0 | 1 | 0 | 0 | 0 | 0 |
| Staphylococcus saprophyticus | 1 (0.1%) | 0 | 1 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 |
| Staphylococcus simulans | 3 (0.3%) | 2 | 1 | 0 | 0 | 0 | 3 | 0 | 0 | 0 | 0 |
| Staphylococcus warneri | 1 (0.1%) | 0 | 1 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 |
| Staphylococcus xylosus | 1 (0.1%) | 0 | 1 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 |
| Streptococcus agalactiae | 42 (4.2%) | 40 | 2 | 0 | 0 | 0 | 42 | 0 | 0 | 0 | 0 |
| Streptococcus anginosus group | 8 (0.8%) | 7 | 1 | 0 | 0 | 0 | 7 | 1 | 0 | 0 | 0 |
| Streptococcus dysgalactiae | 3 (0.3%) | 2 | 1 | 0 | 0 | 0 | 3 | 0 | 0 | 0 | 0 |
| Streptococcus gordonii | 1 (0.1%) | 1 | 0 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 |
| Streptococcus mitis/oralis/pneumoniae | 3 (0.3%) | 2 | 1 | 0 | 0 | 0 | 3 | 0 | 0 | 0 | 0 |
| Streptococcus pyogenes | 10 (1%) | 10 | 0 | 0 | 0 | 0 | 10 | 0 | 0 | 0 | 0 |
| Streptococcus salivarius | 1 (0.1%) | 0 | 1 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 |
| Streptococcus vestibularis | 1 (0.1%) | 0 | 1 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 |
| Gram-negative | 586 (58.3%) | 554 | 22 | 8 | 2 | 0 | 559 | 20 | 7 | 0 | 0 |
| Achromobacter xylosoxidans | 1 (0.1%) | 1 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 |
| Acinetobacter baumannii group | 8 (0.8%) | 8 | 0 | 0 | 0 | 0 | 8 | 0 | 0 | 0 | 0 |
| Acinetobacter beijerinckii | 2 (0.2%) | 2 | 0 | 0 | 0 | 0 | 2 | 0 | 0 | 0 | 0 |
| Acinetobacter junii | 1 (0.1%) | 1 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 |
| Acinetobacter ursingii | 2 (0.2%) | 2 | 0 | 0 | 0 | 0 | 2 | 0 | 0 | 0 | 0 |
| Alcaligenes faecalis | 2 (0.2%) | 1 | 1 | 0 | 0 | 0 | 1 | 1 | 0 | 0 | 0 |
| Citrobacter amalonaticus | 2 (0.2%) | 2 | 0 | 0 | 0 | 0 | 2 | 0 | 0 | 0 | 0 |
| Citrobacter farmeri | 2 (0.2%) | 2 | 0 | 0 | 0 | 0 | 2 | 0 | 0 | 0 | 0 |
| Citrobacter freundii/braakii | 14 (1.4%) | 14 | 0 | 0 | 0 | 0 | 14 | 0 | 0 | 0 | 0 |
| Citrobacter koseri | 6 (0.6%) | 6 | 0 | 0 | 0 | 0 | 6 | 0 | 0 | 0 | 0 |
| Comamonas kerstersii | 1 (0.1%) | 1 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 |
| E. coli | 179 (17.8%) | 173 | 4 | 2 | 0 | 0 | 175 | 2 | 2 | 0 | 0 |
| Enterobacter cloaecae complex | 32 (3.2%) | 31 | 1 | 0 | 0 | 0 | 30 | 2 | 0 | 0 | 0 |
| Haemophilus haemolyticus | 2 (0.2%) | 1 | 1 | 0 | 0 | 0 | 2 | 0 | 0 | 0 | 0 |
| Haemophilus influenzae | 19 (1.9%) | 19 | 0 | 0 | 0 | 0 | 19 | 0 | 0 | 0 | 0 |
| Haemophilus parahaemolyticus | 1 (0.1%) | 1 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 |
| Haemophilus paraphrohaemolyticus | 1 (0.1%) | 1 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 |
| Haemophilus sp. | 2 (0.2%) | 0 | 1 | 0 | 1 | 0 | 1 | 1 | 0 | 0 | 0 |
| Haemophilus sputorum | 1 (0.1%) | 0 | 1 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 |
| Klebsiella aerogenes | 12 (1.2%) | 12 | 0 | 0 | 0 | 0 | 12 | 0 | 0 | 0 | 0 |
| Klebsiella oxytoca | 15 (1.5%) | 14 | 0 | 1 | 0 | 0 | 13 | 1 | 1 | 0 | 0 |
| Klebsiella pneumoniae/varricola | 86 (8.5%) | 80 | 5 | 1 | 0 | 0 | 81 | 5 | 0 | 0 | 0 |
| Moraxella catarrhalis | 3 (0.3%) | 3 | 0 | 0 | 0 | 0 | 2 | 0 | 1 | 0 | 0 |
| Moraxella nonliquefaciens | 1 (0.1%) | 1 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 |
| Morganella morganii | 4 (0.4%) | 2 | 1 | 1 | 0 | 0 | 3 | 1 | 0 | 0 | 0 |
| Neisseria flavescens | 2 (0.2%) | 0 | 2 | 0 | 0 | 0 | 2 | 0 | 0 | 0 | 0 |
| Neisseria sp. | 1 (0.1%) | 0 | 1 | 0 | 0 | 0 | 0 | 0 | 1 | 0 | 0 |
| Ochrobactrum anthropi | 1 (0.1%) | 1 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 |
| Proteus mirabilis | 17 (1.7%) | 17 | 0 | 0 | 0 | 0 | 17 | 0 | 0 | 0 | 0 |
| Proteus penneri/vulgaris | 3 (0.3%) | 3 | 0 | 0 | 0 | 0 | 3 | 0 | 0 | 0 | 0 |
| Providencia rettgeri | 2 (0.2%) | 2 | 0 | 0 | 0 | 0 | 2 | 0 | 0 | 0 | 0 |
| Providencia stuartii | 1 (0.1%) | 1 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 |
| Pseudomonas aeruginosa | 125 (12.4%) | 121 | 2 | 2 | 0 | 0 | 120 | 3 | 2 | 0 | 0 |
| Pseudomonas alcaligenes | 1 (0.1%) | 1 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 |
| Pseudomonas citronellolis | 2 (0.2%) | 2 | 0 | 0 | 0 | 0 | 2 | 0 | 0 | 0 | 0 |
| Pseudomonas koreensis | 1 (0.1%) | 1 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 |
| Pseudomonas lundensis | 2 (0.2%) | 2 | 0 | 0 | 0 | 0 | 1 | 1 | 0 | 0 | 0 |
| Pseudomonas massiliensis | 2 (0.2%) | 2 | 0 | 0 | 0 | 0 | 2 | 0 | 0 | 0 | 0 |
| Pseudomonas putida group | 2 (0.2%) | 1 | 1 | 0 | 0 | 0 | 2 | 0 | 0 | 0 | 0 |
| Pseudomonas rhodesiae | 2 (0.2%) | 2 | 0 | 0 | 0 | 0 | 1 | 1 | 0 | 0 | 0 |
| Pseudomonas sp. | 1 (0.1%) | 0 | 0 | 1 | 0 | 0 | 0 | 1 | 0 | 0 | 0 |
| Pseudomonas taetrolens | 1 (0.1%) | 0 | 0 | 0 | 1 | 0 | 1 | 0 | 0 | 0 | 0 |
| Serratia marcescens | 5 (0.5%) | 5 | 0 | 0 | 0 | 0 | 5 | 0 | 0 | 0 | 0 |
| Stenotrophomonas maltophilia | 16 (1.6%) | 15 | 1 | 0 | 0 | 0 | 15 | 1 | 0 | 0 | 0 |
| Yeasts | 9 (0.9%) | 4 | 4 | 1 | 0 | 0 | 5 | 2 | 2 | 0 | 0 |
| Candida albicans | 4 (0.4%) | 2 | 2 | 0 | 0 | 0 | 3 | 1 | 0 | 0 | 0 |
| Candida glabrata | 2 (0.2%) | 2 | 0 | 0 | 0 | 0 | 1 | 0 | 1 | 0 | 0 |
| Candida paraspilosis | 1 (0.1%) | 0 | 1 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 |
| Candida sp. | 2 (0.2%) | 0 | 1 | 1 | 0 | 0 | 0 | 1 | 1 | 0 | 0 |
aID, identification
Table 3.
Statistical comparison of MALDI-TOF MS results between the Autof MS2600 and the Biotyper Sirius
| Type of isolates (number of isolates) | ID a | % Correctly identified (Autof MS2600 against Biotyper Sirius) | p value |
|---|---|---|---|
| Routine (n = 1006) | 1 | 91.0% VS 95.5% | < 0.0001 |
| 1 + 2 | 98.7% VS 98.8% | 1.000 | |
| 3 | 1.1% VS 1.2% | 1.000 | |
| Gram negative (n = 587) | 1 | 94.5 VS 95.4% | 0.4414 |
| Gram positive (n = 410) | 1 | 86.9% VS 96.6% | < 0.0001 |
| Yeasts (n = 9) | 1 | 44.4% VS 55.5% | 1.000 |
aID, identification; 1, correct identification of genus and species; 2, correct identification of genus; 3, no identification
Respectively, 91.0% and 95.5% of the 1006 isolates were identified at species level by Autobio’s Autof MS2600 and Bruker’s Biotyper Sirius according to the manufacturers’ classification, and there was a statistically significant difference between the two (p < 0.0001). Isolates correctly identified at species and those identified at genus together represented 98.7% and 98.8% for the Autof MS-2600 and the Biotyper Sirius, respectively. These two groups were statistically similar (p = 1.000).
Unidentified strains represented 1.1% and 1.2% of cases on the Autof MS2600 and Biotyper Sirius, respectively. Five strains were not identified by either the Autof MS2600 or the Biotyper Sirius, while six strains were identified by the Sirius but not by the Autof MS2600, and seven strains were identified by the Autof MS2600 but not by the Biotyper Sirius, making this category statistically equivalent (p = 1.000). The species not identified by either system were two Pseudomonas aeruginosa, one E. coli, one Candida sp., and one Klebsiella oxytoca.
Presented in Table 4, two discrepancies at the species level emerged during the study. In order to determine exactly which identification was correct, 16 S rRNA gene sequencing was performed. In this study, we arbitrarily considered the Biotyper Sirius results as correct at the species level, as it represents our routine technique and is widely implemented in many laboratories (Table 2). Importantly, this choice does not affect the conclusions of the statistical analyses.
Table 4.
Differences in species identification between the Autof MS2600 and Biotyper Sirius with sequencing results
| Types of sample | Autof MS2600 (Autobio) | Score | Biotype Sirus (Bruker) | Score | Sequencing | |
|---|---|---|---|---|---|---|
| Expectoration (Chocolate medium) | Haemophilus sputorum | 9.303 | Haemophilus parainfluenzae | 2.16 | /(Result unusable for identification) | |
| Screening (ESBL) | Pseudomonas lundensis | 9.625 | Pseudomonas taetrolens | 2.13 | Pseudomonas Sp. | |
During the study, a strain of Pasteurellaceae (Table 5) showed identification discordance on the first run on both MALDI-TOF MS, despite identical colonies. On the second run with double spotting, the best scores no longer showed discordance. However, the second spot on the Autof MS2600 indicated a different genus, with a reliable genus score.
Table 5.
Specific case of a genus discrepancy on the first run between the Autof MS2600 and the Biotyper Sirius
| Types of sample | Deposit | Autof MS2600 (Autobio) | Score | Deposit | Biotype Sirus (Bruker) | Score | |
|---|---|---|---|---|---|---|---|
| 1st run | Expectoration (Chocolat medium) | 1st | Pasteurella pneumotropica | 9.101 | One | Haemophilus influenzae | 2.21 |
| 2nd | Haemophilus influenzae | 8,999 | |||||
| 2nd run | 1st | Haemophilus influenzae | 9,506 | 1st | Haemophilus influenzae | 2,37 | |
| 2nd | Pasteurella pneumotropica | 8,999 | 2nd | Haemophilus influenzae | 2,34 |
Within the Gram-negative group (n = 587), respectively 94.5% and 95.4% of isolates were correctly identified to species by the Autof MS2600 and Biotyper Sirius, with no statistically significant difference (p = 0.4414). For Gram-positive bacteria (n = 410), 86.9% (Autof MS2600) and 96.6% (Biotyper Sirius) were reliably identified at species level, with a statistically significant difference (p < 0.0001). For Candida, statistical equivalence was obtained between the machines, but this should be viewed in the context of the small number of strains (n = 9).
E. coli (n = 179) and S. aureus (n = 141) were the two most frequently encountered species in the study. As shown in Fig. 4, the percentage of correct species-level identifications for E. coli using the Biotyper Sirius ranged from 80% on MacConkey agar to 95% on ESBL agar and reached 100% on the other media, showing a significant difference according to the culture medium (p = 0.036). For E. coli identified with the Autof MS2600, correct species-level identification rates ranged from 89% on ESBL agar to 100% on the other media, with no significant difference observed between culture media (p = 0.325). For S. aureus, the percentage of correct identifications was 100% across all media when using the Biotyper Sirius, with no significant difference between culture media (p = 1.000). Using the Autof MS2600, identification rates ranged from 92.5% on columbia agar to 95.5% on mannitol salt agar and reached 100% on the other media, again with no significant difference between culture media (p = 0.710).
Fig. 4.
Impact of the growth medium on correct identification rates of the two most common species. Cap/mac, biplate agar for selecting gram-positive and gram-negative bacteria; CLED, cystine lactose electrolyte-deficient agar; Col, Columbia agar; ESBL, chromogenic screening plate for the detection of extended-spectrum β-lactamase-producing organisms; MacConkey, MacConkey agar; Mannitol, mannitol agar; MRSA, screening agar for detecting methicillin-resistant S. aureus; PVX, blood agar for the growth of fastidious bacteria
The repeatability and reproducibility results of Autof MS2600 are presented in Supplementary Tables 1 and 2. Regarding repeatability, all spots (n = 60; 30 per species) were correctly identified as E. coli and S. aureus. However, for S. aureus, 53.4% of the identifications had reliable scores only at the genus level. For reproducibility testing of E. coli and S. aureus, correct identification was obtained every day. For E. coli, 11 out of 12 spots were reliably identified at the species level, and 1 at the genus level. For S. aureus, 10 spots were reliably identified at the species level, and 2 at the genus level.
The MALDI-TOF MS Biotyper Sirius and Autof MS2600 systems are also compared in terms of practicality, and the main observations are presented in Table 6. On the Biotyper Sirius, an E. coli BTS serves as both a calibrator and an internal quality control, and must be run prior to sample analysis. On the Autof MS2600, an E. coli–based calibrator is also used as both calibrator and internal quality control, but its use is not mandatory before sample analysis. The Autof MS2600 also allows modification of acquisition parameters, which correspond to the search mode used by the Biotyper Sirius.
Table 6.
Comparison of practical parameters between Autof MS2600 (Autobio) and Biotyper Sirius (Bruker)
| Properties |
Autof MS2600
(Autobio) |
Biotyper Sirius
(Bruker) |
|
|---|---|---|---|
| User-friendliness | Software | Identical | |
| Facility of preparing smear | Identical | ||
| Reusable plates | Yes | Yes | |
| Ready-to-use matrix | Yes | No | |
| Capacity | 96 wells | 96 wells | |
| Theoretical throughput | Approximately 960 tests per hour | Approximately 600 tests per hour | |
| Number of shots per spot | 240 (40 shots per position) | 240 (40 shots per position) | |
| Laser frequency |
1–1200 Hz Default setting at 300 Hz |
200 Hz | |
| Calibrator and internal control | Not necessary to start the analyses | Necessary to start the analyses | |
| Lysate | To help detect more difficult bacteria and yeasts a | Use formic acid for bacteria that are more difficult to identify and yeasts | |
| Source cleaning | Recommended every 1 million laser shots | A frequency of at least every two weeks is recommended to ensure good performance. | |
| Spot on the plate | No indication of good spot deposit quality | Indication of good spot deposit quality (orange or green) | |
| Quality | IVD | IVD | |
| Connectivity | Via LIS | Via LIS | |
| Power consumption (Instrument) | 320 VA | 250-350VA | |
a Not used during the study
Discussion
In recent years, MALDI-TOF MS systems have become essential in clinical microbiology laboratories for the identification of microorganisms, due to their sensitivity and speed of results [1, 16]. The Biotyper Sirius and VITEK MS systems are widely used in laboratories around the world, and new MALDI-TOF MS systems are being developed. The aim of this study was to compare the analytical performance and practical usability of the Autof MS2600 system with those of the Biotyper Sirius system in routine clinical microbiology practice.
Among the 1006 isolates identified during the study, the various sample types were representative of the routine bacteriological testing performed at LHUB-ULB. Most of these are urine samples, skin and soft tissue infection samples, and screenings which are the three specimen types for which MALDI-TOF MS identification is most commonly applied. The three most frequently identified bacterial species in this study (E. coli, S. aureus, and Pseudomonas aeruginosa) were consistent with the predominant species observed in the routine microbiological workflow at LHUB-ULB.
Of all isolates (n = 1006), respectively 91.0% and 95.5% of strains in Autof MS2600 and Biotyper Sirius were reliably identified to species, showing a significant difference (p < 0.0001). Biotyper Sirius appeared to have better analytical performance than Autof MS2600. In 2021, Park Jae Hyeon et al. [17] evaluated the previous-generation Autof MS2100 (Autobio) and the Microflex LT (Bruker), reporting excellent species-level identification rates. While higher performance might have been expected with the newer Autof MS2600 and updated databases, differences in isolates diversity and the smaller number of strains in their study likely explain the discrepancy. As this study involved an earlier Autof system, we focus primarily on the comparison with the study by De Carolis et al., which evaluated the same instrument as ours. The study conducted by Elena De Carolis et al. (2024) [18] reported a 99.31% species-level identification rate on the Autof MS2600 when compared with the MALDI-TOF MS MBT smart (98,39%). In our study, the identification rate on Autof MS2600 reached only 91.0%. Several factors may explain this difference. In our study, the strains were deposited on the Autof MS2600 target plate by an assistant with limited experience in bacteriology. Spot quality, which depends on the amount of microorganisms deposited, is highly operator-dependent and may have affected identification scores [19, 20]. It is well established that deposit quality improves with operator experience [2, 21]. However, in cases of discordance, spots on both systems were performed by the same operator, limiting operator-related effects. Second, the study by Elena de Carolis et al. included fewer sample types and culture medium (3 vs. 9). Additionally, 21% of their isolates originated from blood cultures, which were excluded from our analysis. However, those two differences are negligible and unlikely to have affected identification rates. Finally, while both studies analysed a diverse bacterial panel, our study encompassed a broader range of bacterial isolates (over 95), whereas De Carolis et al. identified 80 species, which may partially explain the observed differences in identification rates between the two MALDI-TOF MS systems.
The lower performance observed for Autof MS2600 compared to Biotyper Sirius may, in part, account for the influence of reliability scores used by Autobio for both genus- and species-level identification. It is possible that Autobio uses thresholds that are too strict compared to those proposed by Biotyper Sirius. Seventy-eight strains were reliably identified to genus by Autof MS2600. If a lower threshold of 8/10 had been used instead of the recommended 9 by Autobio, 34 additional isolates out of 1006 would have been classified as correctly identified at the species level. Except for two strains showing species-level discordance and four isolates where Autof MS2600 did not provide species identification, there were no species-level errors. Specifically, Autof MS2600 identified Haemophilus influenzae (8.999) and Staphylococcus haemolyticus (8.397), whereas Biotyper Sirius identified Haemophilus haemolyticus (2.02) and Staphylococcus borealis (2.22). The first discrepancy may be clinically relevant but the second is not, as this was a screening test. With this change in threshold, no statistically significant difference would persist between the Autof MS2600 and Biotyper Sirius systems in terms of species-level identification. A few studies have validated a score of 1.9 (instead of 2.0 out of 3.0) as the species identification threshold while maintaining a very good correct identification rate on the Biotyper Sirius system [22]. Therefore, accepting a score of 8–8.5.5 would be feasible, although further studies should be conducted to validate such thresholds. Among these strains, 36 strains (62%) are staphylococci and streptococci. Gram-positive bacteria are more difficult to identify than Gram-negative bacteria by MALDI-TOF MS due to a thick layer of peptidoglycans, which explains the lower scores observed in MALDI_TOF MS [23]. In cases of failed or low-confidence species identification, the addition of formic acid for cell lysis facilitates protein extraction, resulting in higher-quality MALDI-TOF spectra and improved identification accuracy [24, 25]. But this procedure is not routinely performed from the outset.
Four strains were identified at the species level by the Biotyper Sirius but not by the Autof MS2600, even though the latter assigned a score above 9, a threshold that typically supports species-level identification (more details in the Supplementary Table 3). The Biotyper Sirius system opted to report a species-level identification, although not certain, based on matches with one or more closely related species. In contrast, the Autof MS2600 system reported only the genus level, considering that the similarity of protein profiles between related species does not allow for a reliable decision. Ultimately, these cases do not represent real identification discrepancies; they reflect differing reporting strategies adopted by Autobio and Bruker.
The Autof MS2600 showed lower sensitivity for species identification, particularly in the Gram-positive bacteria group due to their thick layer of peptidoglycans. However, no significant difference (p = 1.000) was demonstrated between the two systems, Autof MS-2600 and Biotyper Sirius, in terms of unidentified strain rates, which were 1.1% and 1.2%, respectively.
All strains that were not identified by Autof MS2600 and Biotyper Sirius during the study were present in their respective databases. Several factors may explain these identification failures, such as poor deposit quality, an old strain [19], the culture medium [26], and more mucoid bacteria, which are more difficult to identify. Of the five strains not identified on the two MALDI-TOF MS systems, four were enterobacteria, which can sometimes be highly mucoid. Two of these were Pseudomonas aeruginosa, whose mucoid phenotype is explained by the synthesis of an exopolysaccharide matrix [27].
Out of the 1006 strains analysed, only two discrepancies (0.2%) were observed. In both cases, 16 S rRNA gene sequencing was unable to achieve species-level identification, making it impossible to disqualify one of the results provided by Autof MS2600 and Biotyper Sirius. The Haemophilus strain yielded no usable results and the Pseudomonas strain was only identified at the genus level, reflecting a probable genetic proximity between species. This limitation is inherent to 16 S rRNA gene sequencing, and it is therefore expected that MALDI-TOF MS may face similar challenges. In any case, the distinction between Pseudomonas lundensis (identified by Autof MS-2600) and Pseudomonas taetrolens (identified by Biotyper Sirius), as well as between Haemophilus sputorum (Autof MS-2600) and Haemophilus parainfluenzae (Biotyper Sirius), has no bearing on clinical management or interpretation.
Regarding the Pasteurellaceae strain mentioned above, Autof MS2600 gave Pasteurella pneumotropica (recently reclassified as Rodentibacter pneumotropicus) and Biotyper Sirius gave Haemophilus influenzae on the first run. The top 10 Autof MS2600 matches alternated between Pasteurella pneumotropica and Haemophilus influenzae, suggesting the hypothesis of a mixture of the two species in the colony or a similar protein composition preventing reliable differentiation by MALDI-TOF MS. The Biotyper Sirius database includes Rodentibacter pneumotropicus; it was not proposed among the top 10 matches in either run. In the second run, Autof MS2600 again identified H. influenzae with an acceptable species-level score, thus no formal discrepancy was recorded. We had considered performing 16 S rRNA sequencing to confirm the identification, but the isolate was not viable. As the isolate originated from a respiratory sample, H. influenzae infection appears more likely than Rodentibacter pneumotropicus, a rodent commensal rarely associated with human disease [28]. Given the clinical importance of H. influenzae, particular attention should be paid to this finding, and feedback will be provided to the manufacturer.
It is also useful to remember the importance of looking at the 10 best identification suggestions provided by a MALDI-TOF MS system. The Bruker Biotyper database assigns a similarity score between the 10 suggestions given by the system, ranging from A to C [22]. To our knowledge, the Autof MS2600 system does not provide a match index for the first 10 identification suggestions.
With respect to the effect of the culture medium on identification of the two most prevalent species, E. coli and S. aureus, high-confidence identification with the Autof MS2600 was not affected by the type of medium, as confirmed by chi-square analysis. In contrast, for the Biotyper Sirius, chi-square testing revealed a significant effect of the culture medium on E. coli identification. This last result should be qualified, as the statistical difference stems from a sample that was not identified with certainty and was found on an underrepresented medium (MacConkey medium n = 5). After excluding the underrepresented media, no statistically significant difference was observed based on the type of medium for E. coli identification on Biotyper Sirius. Studies show that the culture medium influences protein composition and, consequently, MALDI-TOF MS spectra. For example, the presence of salts can inhibit certain protein peaks, but many genes remain constitutively expressed, regardless of culture conditions [26]. The impact may also be species-specific rather than uniform across all bacteria [27]. Consequently, such variations can lower identification scores, although they are generally insufficient to cause misidentifications [29].
Regarding repeatability, the Autof MS2600 demonstrated excellent performance for E. coli, with all isolates correctly identified, whereas more than half of S. aureus isolates were not identified to the species level. As observed in our comparative study, Gram-positive bacteria were generally less accurately identified than Gram-negative bacteria.
Regarding reproducibility, the Autof MS2600 performed well overall, although identification was less reliable for both E. coli and S. aureus when strains were older than 72 h. This effect was particularly pronounced for Gram-positive bacteria, which are more challenging to identify than Gram-negative species.
However, this study has several limitations. Firstly, the dual spot deposition strategy on Autof MS2600, considering the best score, may have led to a slight overestimation, introducing a methodological bias. An analysis of only the first spots on Autof MS2600 and Biotyper Sirius showed that the identification rate decreases slightly on Autof MS2600, with 13 additional samples identified only to the genus level and 17 additional samples remaining unidentified. These findings highlight the importance of deposit quality. However, statistical results were similar between the first analysis (best of the 2 spots) and the second analysis (first spot only) (Supplementary Table 4).
Secondly, it is well recognised that operator bias can affect identification performance [2, 21], potentially leading to under- or overestimation. The Autof MS2600 was operated by a student with limited experience, whereas the Biotyper Sirius was handled by experienced technologists. However, the experimental setting may have allowed the student to focus on spot quality; technologists working under routine clinical workflow and time pressure may have experienced factors affecting spot quality. Importantly, this bias was mitigated by repeating discordant analyses in duplicate spot, performed by the same operator.
Thirdly, the lack of identification of anaerobic bacteria, mycobacteria and filamentous fungi. Although a few Candida isolates (n = 9) were included, their number was insufficient to reliably assess analytical performance. These gaps can be partially filled by the following works. The study by Elena De Carolis et al. identified around 100 yeasts and demonstrated very good analytical performance (98.02% reliable species-level identification) [18].With regard to the identification of filamentous fungi, Autof MS-1000 appears promising according to studies by Keping Ao et al. and Yue Sun et al. [30, 31]. The identification of less common organisms, such as mycobacteria, anaerobic bacteria, filamentous fungi, and yeasts, remains challenging using MALDI-TOF mass spectrometry and requires further evaluation of the Autof MS2600 in more challenging diagnostic situations.
The Autof MS2600 and the Biotyper Sirius do not differ significantly in terms of practicality, each presenting advantages and limitations. Both MS systems feature user-friendly software and require minimal auxiliary equipment. The Autof MS2600 does not require routine cleaning of the ion source whereas the Biotyper Sirius provides a good spot quality score for colony deposition. According to data provided by the manufacturers, the Autof MS2600 operates at a higher laser frequency than the Biotyper Sirius while using the same number of laser shots per spot, which could theoretically result in faster acquisition. However, due to the different ages and usage histories of the instruments, it was not possible to reliably determine which system is faster in practice. What may raise questions is that the Autof MS2600 obtained IVD certification despite the fact that the use of an internal control is not mandatory prior to the analysis of bacterial colony spots and acquisition parameters can be adjusted if needed.
Conclusion
Although only a few MALDI-TOF MS systems have traditionally dominated the clinical microbiology market, our comparative evaluation shows that other commercially available instruments can achieve equivalent performance levels. Despite a statistically significant difference in the reliable species identification rate for Gram-positive bacteria between the two systems, the analytical performance is similar. The rates of correct species identification are similar, as are the rates of non-identification. The Autof MS2600 is suitable for use in a routine microbiology laboratory. The choice between different MALDI-TOF MS systems depends on practical considerations and agreements with the manufacturer.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
We would like to thank Sébastien Bussi from Analis, in collaboration with Autobio, for lending the Autof MS2600 instrument and the associated reagents, which enabled its evaluation. We are also grateful to the bacteriology technologists of the LHUB-ULB for their essential contribution to this work. We also thank Charlotte Nachtegael for her assistance with the statistical analyses.
Author contributions
Marine Piron, Chimène Nyamessameye, Souad Mohamed, Burak Musellim, Sukriye Aksoy and Aissatou Barry collected the data. Marine Piron and Chimène Nyamessameye analysed the collected data. Ethel Seyll acted as the contact point with the companies for inquiries and Delphine Martiny supervised the study. The manuscript was drafted by Marine Piron and critically revised by all authors. All authors contributed to study design and approved the final version.
Funding
This study was supported by Analis through the lending of the MALDI-TOF MS instrument and related reagents used for the evaluation. The company had no involvement in the study design, data collection, data analysis, or manuscript preparation.
Data availability
The datasets generated or analysed during this study are available from the corresponding author upon reasonable request.
Declarations
Ethics approval
Ethical approval was not required for this study.
Consent to participate
No consent to participate was required for this study.
Consent to publish
No consent to publish was required for this study.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Calderaro A, Chezzi C (2024) MALDI-TOF MS: A Reliable Tool in the Real Life of the Clinical Microbiology Laboratory. Microorganisms 12(2):322. 10.3390/microorganisms12020322 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Seng P et al (2009) Ongoing revolution in bacteriology: routine identification of bacteria by matrix-assisted laser desorption ionization time-of-flight mass spectrometry. Clinical Infectious Diseases 49(4):543–551. 10.1086/600885 [DOI] [PubMed] [Google Scholar]
- 3.Gaillot O et al (2011) Cost-Effectiveness of Switch to Matrix-Assisted Laser Desorption Ionization-Time of Flight Mass Spectrometry for Routine Bacterial Identification▿. Journal of Clinical Microbiology 49(12):4412. 10.1128/JCM.05429-11 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Tran A, Alby K, Kerr A, Jones M, Gilligan PH (2015) Cost Savings Realized by Implementation of Routine Microbiological Identification by Matrix-Assisted Laser Desorption Ionization-Time of Flight Mass Spectrometry. Journal of Clinical Microbiology 53(8):2473–2479. 10.1128/JCM.00833-15 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.de la Pedrosa EGG, Gimeno C, Soriano A, Cantón R (2016) Estudios de coste-efectividad con MALDI-TOF e impacto clínico. Enfermedades Infecciosas y Microbiología Clínica 34:47–52. 10.1016/S0213-005X(16)30191-4 [DOI] [PubMed] [Google Scholar]
- 6.Rychert J (2019) « Benefits and Limitations of MALDI-TOF Mass Spectrometry for the Identification of Microorganisms », Journal of Infectiology and Epidemiology, vol. 2, no 4, févr. Consulté le: 7 septembre 2025. [En ligne]. Disponible sur: https://www.infectiologyjournal.com/articles/benefits-and-limitations-of-malditof-mass-spectrometry-for-the-identification-of-microorganisms.html
- 7.Schulthess B et al (2016) Evaluation of the Bruker MALDI Biotyper for Identification of Fastidious Gram-Negative Rods. Journal of Clinical Microbiology 54(3):543–548. 10.1128/JCM.03107-15 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Powell EA, Blecker-Shelly D, Montgomery S, Mortensen JE (2013) Application of matrix-assisted laser desorption ionization-time of flight mass spectrometry for identification of the fastidious pediatric pathogens Aggregatibacter, Eikenella, Haemophilus, and Kingella. Journal of Clinical Microbiology 51(11):3862–3864. 10.1128/JCM.02233-13 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Clark AE, Kaleta EJ, Arora A, Wolk DM (2013) Matrix-Assisted Laser Desorption Ionization–Time of Flight Mass Spectrometry: a Fundamental Shift in the Routine Practice of Clinical Microbiology. Clinical Microbiology Reviews 26(3):547–603. 10.1128/CMR.00072-12 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Dichtl K et al (2023) A head-to-head comparison of three MALDI-TOF mass spectrometry systems with 16S rRNA gene sequencing. J Clin Microbiol 61(10):e01913-22. 10.1128/jcm.01913-22 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Bizzini A, Jaton K, Romo D, Bille J, Prod’hom G, Greub G (2011) Matrix-Assisted Laser Desorption Ionization–Time of Flight Mass Spectrometry as an Alternative to 16S rRNA Gene Sequencing for Identification of Difficult-To-Identify Bacterial Strains. Journal of Clinical Microbiology 49(2):693–696. 10.1128/jcm.01463-10 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Zhang L, Smart S, Sandrin TR (2015) Biomarker- and similarity coefficient-based approaches to bacterial mixture characterization using matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS). Sci Rep 5(1):15834. 10.1038/srep15834 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Chung Y, Han M, Kim J-S (2021) Comparative Evaluation of Bruker Biotyper and ASTA MicroIDSys for Species Identification in a Clinical Microbiology Laboratory. Diagnostics (Basel) 11(9):1683. 10.3390/diagnostics11091683 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Lee Y, Sung JY, Kim H, Yong D, Lee K (2017) Comparison of a New Matrix-Assisted Laser Desorption/Ionization Time-of-Flight Mass Spectrometry Platform, ASTA MicroIDSys, With Bruker Biotyper for Species Identification. Ann Lab Med 37(6):531–535. 10.3343/alm.2017.37.6.531 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Xiong L et al (2023) Comparison of Autof Ms1000 and EXS3000 MALDI-TOF MS platforms for routine identification of microorganisms. Infect Drug Resist 16:913–921. 10.2147/IDR.S352307 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Martiny D, Busson L, Wybo I, El Haj RA, Dediste A, Vandenberg O (2012) Comparison of the Microflex LT and Vitek MS Systems for Routine Identification of Bacteria by Matrix-Assisted Laser Desorption Ionization–Time of Flight Mass Spectrometry. J Clin Microbiol 50(4):1313–1325. 10.1128/JCM.05971-11 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Park JH, Jang Y, Yoon I, Kim TS, Park H (2021) Comparison of Autof ms1000 and Bruker Biotyper MALDI-TOF MS Platforms for Routine Identification of Clinical Microorganisms. Biomed Res Int 2021:6667623. 10.1155/2021/6667623 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.De Carolis E, Ivagnes V, Magrì C, Falasca B, Spanu T, Sanguinetti M (2024) Evaluation of Autof MS2600 and MBT Smart MALDI-TOF MS Systems for Routine Identification of Clinical Bacteria and Yeasts. Microorganisms 12(2):382. 10.3390/microorganisms12020382 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Cuénod A, Foucault F, Pflüger V, Egli A (2021) Factors Associated With MALDI-TOF Mass Spectral Quality of Species Identification in Clinical Routine Diagnostics. Front Cell Infect Microbiol 11:646648. 10.3389/fcimb.2021.646648 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Cherkaoui A, Riat A, Renzi G, Fischer A, Schrenzel J (2023) Diagnostic test accuracy of an automated device for the MALDI target preparation for microbial identification. Eur J Clin Microbiol Infect Dis 42(2):153–159. 10.1007/s10096-022-04531-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Van Herendael BH et al (2012) Validation of a modified algorithm for the identification of yeast isolates using matrix-assisted laser desorption/ionisation time-of-flight mass spectrometry (MALDI-TOF MS). Eur J Clin Microbiol Infect Dis 31(5):841–848. 10.1007/s10096-011-1383-y [DOI] [PubMed] [Google Scholar]
- 22.Khot PD, Couturier MR, Wilson A, Croft A, Fisher MA (2012) Optimization of Matrix-Assisted Laser Desorption Ionization–Time of Flight Mass Spectrometry Analysis for Bacterial Identification. Journal of Clinical Microbiology 50(12):3845–3852. 10.1128/jcm.00626-12 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Alatoom AA, Cunningham SA, Ihde SM, Mandrekar J, Patel R (2011) Comparison of Direct Colony Method versus Extraction Method for Identification of Gram-Positive Cocci by Use of Bruker Biotyper Matrix-Assisted Laser Desorption Ionization–Time of Flight Mass Spectrometry ▿. J Clin Microbiol 49(8):2868–2873. 10.1128/JCM.00506-11 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.McElvania TeKippe E, Shuey S, Winkler DW, Butler MA, Burnham C-A (2013) Optimizing Identification of Clinically Relevant Gram-Positive Organisms by Use of the Bruker Biotyper Matrix-Assisted Laser Desorption Ionization–Time of Flight Mass Spectrometry System. J Clin Microbiol 51(5):1421–1427. 10.1128/JCM.02680-12 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Haigh J, Degun A, Eydmann M, Millar M, Wilks M (2011) Improved Performance of Bacterium and Yeast Identification by a Commercial Matrix-Assisted Laser Desorption Ionization–Time of Flight Mass Spectrometry System in the Clinical Microbiology Laboratory ▿. J Clin Microbiol 49(9):3441. 10.1128/JCM.00576-11 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Valentine N, Wunschel S, Wunschel D, Petersen C, Wahl K (2005) Effect of culture conditions on microorganism identification by matrix-assisted laser desorption ionization mass spectrometry. Appl Environ Microbiol 71(1):58–64. 10.1128/AEM.71.1.58-64.2005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Topić Popović N, Kazazić SP, Bojanić K, Strunjak-Perović I, Čož-Rakovac R (2023) Sample preparation and culture condition effects on MALDI-TOF MS identification of bacteria: A review. Mass Spectrom Rev 42(5):1589–1603. 10.1002/mas.21739 [DOI] [PubMed] [Google Scholar]
- 28.Soto-Vega E, Bello-Rossette E, Cabrera-Martínez M, Kuribreña JCA (2025) Rare Urinary Tract Infection by Rodentibacter pneumotropicus (formerly, Pasteurella pneumotropica) in an Immunocompetent Patient. The American Journal of Case Reports 26:e947344. 10.12659/AJCR.947344 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Anderson NW, Buchan BW, Riebe KM, Parsons LN, Gnacinski S, Ledeboer NA (2012) Effects of Solid-Medium Type on Routine Identification of Bacterial Isolates by Use of Matrix-Assisted Laser Desorption Ionization–Time of Flight Mass Spectrometry. J Clin Microbiol 50(3):1008–1013. 10.1128/JCM.05209-11 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Ao K et al (2023) Evaluation of the Autof ms1000 mass spectrometry for rapid clinical identification of filamentous fungi. BMC Microbiol 23(1):228. 10.1186/s12866-023-02968-w [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Y S et al (2021) Multicenter evaluation of three different MALDI-TOF MS systems for identification of clinically relevant filamentous fungi. Medical mycology. 10.1093/mmy/myaa037 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets generated or analysed during this study are available from the corresponding author upon reasonable request.