Evaluation of Species Specific Score Cutoff Values of Routinely Isolated Clinically Relevant

Evaluation of Species Specific Score Cutoff Values of Routinely Isolated Clinically Relevant

Evaluation of species specific score cutoff values of routinely isolated clinically relevant bacteria using a direct smear preparation for matrix assisted laser desorption / ionization time of flight mass spectrometry-based bacterial identification.

Szabados Florian 1*, Tix Heike1,3, Anders Agnes1, Kaase Martin1 and Gatermann Sören G. 1 and Geis Gabriele1,2

*Corresponding author: F. Szabados

Tel : +49-(0)234-32-26467

Fax: +49-(0)234-32-14197

1Institute for Hygiene and Microbiology, Dept. for Medical Microbiology, University Bochum, National Reference Laboratory for multidrug-resistant gram-negative bacteria for Germany

2Institut für Medizinische Laboratoriumsmedizin (IML), Universitäts-Aninstitut, Ruhr-University Bochum

3 Hygiene Institut Gelsenkirchen, Gelsenkirchen

Detailed description of the strain collection in addition to the tables 3 to 6:

Gram positive facultatively anaerobic bacteria identified by the GP identification card:

Two isolates identified as Streptococcus dysgalactiae /Streptococcus pluranimalium group were correctly identified by MALDI TOF MS as Streptococcus constellatus and S. dysgalactiae spp. equisimilis. In one isolate the MALDI TOF MS-based identification was correct, in contrast to the GP-card, as confirmed by sodA amplification and sequencing. One isolate of Streptococcus anginosus was identified as S. anginosus/constellatus group by the GP-card and was correctly identified by MALDI TOF MS as S. anginosus. Two isolates were identified as S. anginosus/constellatus group by the GP-card and were also correctly identified by MALDI TOF MS as S. anginosus and S. constellatus. Other isolates of S. anginosus (n=2) , Streptococcus pyogenes (n=2), Streptococcus agalactiae (n=1), Streptococcus gallolyticus spp. gallolyticus (n=1) and Streptococcus salivarius (n=1) were concordantly identified by both methods. In one isolate of Streptococcus pneumoniae, the MALDI-TOF MS gained two times a “no peaks” result. This result was thus interpreted as an incorrect identification and as a putative discordant result by MALDI TOF MS. One isolate of Streptococcus gordonii was identified by the GP-card and was discordantly identified by MALDI TOF MS as S. anginosus. Two isolates classified as Streptococcus mitis/oralis group were identified correctly by MALDI TOF MS as Streptococcus australis and Streptococcus infantarius (with scores of 1.68 and 1.84), as confirmed by sodA amplification and sequencing. Two isolates classified as Streptococcus mitis/oralis group were misidentified by MALDI TOF MS as Streptococcus pneumoniae (score 1.93 and 2.24), as confirmed by bile solubility and optochin resistance. One isolate was not identified by the GP-card but was correctly identified as Streptococcus constellatus by MALDI TOF MS. One isolate of Staphylococcus capitis was identified correctly by MALDI TOF MS as Staphylococcus epidermidis, as confirmed by sodA amplification and sequencing. Three isolates of S. capitis initially identified as S. capitis / Staphylococcus caprae were all correctly identified with an arithmetic mean score value of 1.95 (min=1.79, max=2.18). One isolate of S. caprae was concordantly identified by MALDI-TOF MS. Another isolate of S. caprae was correctly identified as S. capitis by MALDI TOF MS, as confirmed by sodA amplification and sequencing. Other staphylococci were also concordantly identified as Staphylococcus saprophyticus (n=1, score 2.05), Staphylococcus warneri (n=1, score 2.19) and Staphylococcus intermedius (n=1, score 2.05). One isolate of S. warneri identified by GP-card was initially identified by MADLI TOF MS as Proteus mirabilis. This isolate was retested and correctly identified as Staphylococcus haemolyticus by MALDI TOF MS, as confirmed by sodA amplification and sequencing. One Enterococcus gallinarum and one Enterococcus hirae isolate were each concordantly identified. Four isolates, misidentified as Enterococcus durans/E. faecium group by the GP-card, were identified correctly as E. hirae (n=3, scores 2.2, 2.2 and 2.38) and E. faecium by MALDI TOF MS, as confirmed by sodA amplification and sequencing. One isolate identified as Enterococcus durans by the GP-card was discordantly identified as E. caccae by MALDI TOF MS. Other gram positive organisms, such as Listeria monocytonenes (n=1, score 2.29), Micrococcus luteus (n=2, scores 1.83 and 2.29) and Pediococcus pentosacceus (n=1, score 2.16) were concordantly identified. An isolate of Gemella spp. (n=1, score 1.54) and an isolate of Leuconostoc spp.. (n=1, score 1.54) were identified by the GP-card only to the genus level, the MALDI-TOF-based identification gained the same genus, but did so repeatedly with low score values. One isolate was identified by the GP-card as Granulicatella adiacens but was discordantly identified by the MALDI TOF MS as S. dysgalactiae spp. equisimilis. Three isolates were not identified by the GP-card but were correctly identified by MALDI TOF MS as Dermabacter hominins (n=2, scores 1.81 and 1.99) and S. constellatus.

Gram negative facultatively aerobic bacteria identified by the GN identification card

All isolates of Citrobacter freundii were concordantly identified with a high mean score value. Three out of four isolates of Citrobacter brakii (n=4) were concordantly identified. One isolate was misidentified as C. freundii by MALDI TOF MS, as confirmed by 16S rDNA amplification and sequencing. Eight isolates were identified by the GN-card as Citrobacter koseri/diversus group and were identified as C. koseri with high mean score values of 2.43. Echerichia coli were concordantly identified with high mean score values of 2.24. Enterobacter aerogenes (n=6) were concordantly identified with an arithmetic mean score value of 2.35. One isolate of Enterobacter asburiae was concordantly identified by both methods. Isolates of Enterobacter cloacae were identified by the GN-card. Thirty-five of them were identified as E. cloacae with a mean score value of 2.23. Other isolates were identified by MALDI TOF as Enterobacter hormaechii (n=1, score 2.25), Enterobacter kobei (n=2, scores 2.05 and 2.28), Enterobacter ludwigii (n=1, score 2.3), Enterobacter asburiae (n=3, scores 2.03, 2.16 and 2.14) and two as Enterobacter spp./E. cloacae (scores 2.27 and 2.28). Isolates of Hafnia alvei (n=4) were concordantly identified with an arithmetic mean score value of 2.16.

All but one isolates of Klebsiella pneumoniae were concordantly identified with an arithmetic mean score value of 2.24 (min=1.90, max=2.41). One isolate of K. pneumoniae was misidentified by MALDI TOF MS as Klebsiella oxytoca with a score of 1.78, as confirmed by indol testing and also by 16S rDNA amplification and sequencing. All isolates of K. pneumoniae spp. pneumoniae (n=35) were concordantly identified to species level with a mean score value of 2.28. One isolate of K. pneumoniae spp. ozaenae was correctly identified as K. oxytoca by MALDI TOF, as confirmed by 16S amplification and sequencing. All isolates of Klebsiella oxytoca were concordantly identified to species level with a mean score value of 2.29. Two isolates identified as Raoultella planticola by the GN-card were identified as Raoultella ornitholytica by MALDI TOF MS. Unfortunately, these two isolates were discarded without further confirmation of the identification. Isolates of Morganella morganii were correctly identified to species level with a high mean score value of 2.44. Two isolates were identified by the GN-card as Moraxella spp. and were identified correctly as Moraxella lacunata and M. nonliquefaciens, as confirmed by 16S.

In one sample initially a Proteus vulgaris was detected, leading to a discrepant result at the genus level. A mixture of two species was suspected and a concordant identification of both strains was obtained after isolation of the strains and repeated measurement in both methods (Proteus vulgaris and Serratia marcescens). One isolate each of Providentia rettgeri and Providentia stuartii was also concordantly identified to species level. Isolates of Proteus mirabilis and P. vulgaris were concordantly identified. All Serratia marcescens identified by the GN-card were concordantly identified with a mean score value of 2.26. One isolate was identified as S. marcescens spp. marcescens by the GN-card and was correctly identified as Cedecia lapagei. One isolate was initially misidentified as Serratia ficaria by the GN-card. This first lead to a discrepant result on genus level. This result was corrected after isolation of two isolates and repeated measurement in both methods. Species identification was later concordantly confirmed as E. coli and K. pneumoniae in both methods. One isolate of Serratia rubidaea was concordantly identified in both methods. Four isolates were not identified by the GN-card but were correctly identified by MALDI-TOF MS as E. coli, K. pneumoniae, C. freundii and E. cloacae. These were confirmed by amplification and sequencing of the 16S rDNA. Isolates of Salmonella enterica (n=4) were identified as Salmonella group by the GN-card and as Salmonella enterica by MALDI TOF MS.

Isolates of Pseudomonas aeruginosa and Pseudomonas fluorescens identified by the GN-card were concordantly identified by MALDI TOF MS. Two isolates identified as Pseudomonas putida/fluorescens group by the GN-card were correctly identified as P. putida and P. fluorescens by MALDI TOF MS as verified by 16S rDNA amplification and sequencing. One isolate, identified as P. aeruginosa or P. putida/fluorescens by the GN-card, was identified as Pseudomonas citronellosis by MALDI TOF MS. P. citronellosis is a member of the P. aeruginosa group. Other isolates of Pseudomonas, such as Pseudomonas oryzihabitans (n=1, score 2.15), and Pseudomonas stutzeri (n=2, scores 1.98 and 1.89) were concordantly identified by MALDI TOF MS. Three isolates identified as P. putida by the GN-card were correctly identified as Pseudomonas fulva, a member of the Pseudomonas putida group.

Isolates of Stenotrophomonas maltophilia (n=9) were identified as S. maltophilia (n=5) and Pseudomonas hibiscola (n=4) with an arithmetic mean score value of 2.15 (min=1.97, max=2.36). One isolate was identified as Shevanella algae by the GN-card and was discordantly identified as Shevanella putrefaciens by MALDI TOF MS. One isolate was identified as Sphingomonas paucimobilis by the GN-card and was discordantly identified as Pasteurella multoccida (score 1.61) by MALDI TOF MS. Eleven isolates were identified by the GN-card as Acinetobacter baumanii group. This Acinetobacter baumanii group contains the Acinetobacter baumanii, Acinetobacter calcoaceticus, Acinetobacter genomospecies 3, and the Acinetobacter genomospecies 13TU. Among these, six isolates of A. baumanii were correctly identified with high mean score values and three isolates of A. genomospecies 3 were correctly identified but with scores lower than that of A. baumanii. One isolate of Acinetobacter haemolyticus misidentified by the GN-card was correctly identified as Acinetobacter junii by MALDI TOF MS. One isolate of Acinetobacter radioresiseance was concordantly identified by the GN-card and by MALDI TOF MS.

Obligate anaerobic bacteria identified by the ANC-card

Bacteroides fragilis, Bacteroides ovatus and Bacteroides thetaiotaomicron were all concordantly identified with variable mean score values. One isolate was identified as B. thetaiotaomicron by the ANC-card but was discordantly identified as Bacteroides vulgatus by MALDI TOF MS (score 1.88). One isolate was identified as Bacteroides stercoris by the ANC-card but was identified only to the genus level as Bacteroides spp. by MALDI TOF MS (score 2.16). Isolates of Parabacteroides distasionis (n=3) were all concordantly identified with low mean score values. Isolates of Clostridium perfringens were all concordantly identified with a high mean score value of 2.22. Seven isolates of Propionibacterium acnes (n=10) were concordantly identified with a low mean score value of 1.77 (min=1.56, max=1.96). Three isolates of P. acnes were discordantly identified as Eubacterium brachii with scores of 1.8, 1.94 and 1.99 by MALDI TOF MS. Three isolates of Prevotella bivia (n=4) were concordantly identified with a low mean score value of 1.90 (min=1.81, max=2.0). One isolate of P. bivia was discrepantly identified by MALDI TOF MS (score 2.05). Other strict anaerobes, such as Prevotella buccae, Prevotella disiens, Prevotella melanogenica (n=1, score 1.38), Peptostreptococcus anaerobius (n=2, scores 1.93 and 1.96) were concordantly identified, except for low score values in the MALDI TOF MS-based identification. One isolate of Lactobacillus acidophilus was identified discrepantly as S. infantarius (n=1, score 1.88). One isolate of Peptostreptococcus anaerobius was misidentified as Raoultella ornitholytica (score 1.62) by MALDI TOF MS. R. ornitholytica was ruled out by gram staining. Two isolates of Peptostreptococcus anaerobius were identified discrepantly as Porphyromonas asaccharolytica (score 1.42 and 1.88). The gram stains were also not consistent with the MALDI-TOF MS-based species identification. Two isolates of Finegoldia magna were concordantly identified as Finegoldia magna (scores 1.51 and 1.6) and one isolate of Finegoldia magna was discordantly identified as Porphyromonas asaccharolytica (score 1.88). The gram stain was not consistent with the MALDI-TOF MS-based species identification. Other isolates, such as Fusobacterium necrophilum (n=1, score 2.13) and Fusobacterium nucleatum (n=2, scores 1.59 and 1.68) were concordantly identified in both methods but with low score values.

Five isolates - a Corynebacterium jekeium, a Corynebacterium afermentans, a Bacillus cereus, a Lactococcus lactis and a Bacteroides fragilis- were not identified in the ANC-card but were correctly identified by MALDI TOF MS to species level as confirmed by 16S rDNA amplification and sequencing. One isolate of Actinomyces naeslundii was correctly identified as Dermabacter hominis (score 1.99) by MALTI TOF MS. One isolate of Actinomyces naeslundii was identified as Lactobacillus paracasei (score 2.07). Acid fast and Gram staining, growth on Lactobacillus spp. selective agar (Rogosa) and resistance to vancomycin indicate that the MALDI TOF MS identification of Dermatobacter hominis was consistent with these tests. We showed that the mean score values of common and clinically relevant, strictly anaerobic bacteria significantly varied between species. Our study indicates that certain strict anaerobe organisms, such as member of the genus Peptostreptococcus spp. and Prevotella spp., and species such as Porphyromonas asaccharolytica and Finegoldia magna were concordantly identified but with low corresponding score values.

Yeasts

Candia albicans, Candia glabrata, Candia tropicalis, Candia krusei, Candia guillermondii (n=1, score 2.15), Candia lusitaniae (n=1, score 2.15) and Candia parapsilosis (n=2, scores 2.2 and 2.25) were also correctly identified with high score values. One yeast was identified by the YST-card only to Candia species level, two yeast isolates remained unidentified by the YST-card, and the MALDI-TOF MS identified all three yeasts correctly to the species level (C. albicans, C. krusei and C. glabrata, as confirmed by amplification and sequencing of the 26S).