Correct implementation of matrix-assisted laser desorption ionization-time of flight mass spectrometry in routine clinical microbiology.

As we have been users of matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF MS) since January 2009, we read with much interest the paper of van Veen et al. (9) in a recent issue of the Journal of Clinical Microbiology. We would like to respond to two issues addressed in this article as well as in that of Seng et al. (5) published in Clinical Infectious Diseases and in that of Eigner et al. (2) published in Clinical Laboratory. First, van Veen et al. (9) and Seng et al. (5) state that a major error was made by MALDI Biotyper, a frequently used MALDI-TOF MS instrument from Bruker, because some Stenotrophomonas maltophilia isolates were erroneously identified as Pseudomonas beteli, P. geniculata, or P. hibiscicola. They hypothesize that this discordance resulted from a mislabeling of bacterial species in the MALDI Biotyper database, as they consider P. beteli and P. hibiscicola invalid names for S. maltophilia. Two cited references (1, 8) in these papers indicate the very high level of homology between these species and S. maltophilia. However, the extensive study required for definite taxonomic conclusion was never performed, and official nomenclature change was never implemented. Therefore, the latter statement is not correct. However, we feel that the evidence for identification of these species is strong enough for them to be considered, in routine clinical biology practice, synonyms of S. maltophilia, and we do not experience this issue as a major identification error. Second, van Veen et al. (9) and others (2, 7) report that some viridans streptococci were erroneously identified as Streptococcus pneumoniae. These viridans streptococci were identified as Streptococcus mitis group strains (2, 7) but were reported without precise species identification in the study of van Veen et al. (9). In contrast, Seng et al. (5) report that nearly 50% of S. pneumoniae isolates were misidentified as Streptococcus parasanguinis, a member of the S. mitis group. Our own experience confirms the first finding: half of the S. mitis group isolates are misidentified as S. pneumoniae with MALDI Biotyper. This misidentification is caused by the fact that S. pneumoniae is strikingly similar to S. mitis (3), and the distinctive power of MALDI-TOF MS is insufficient to resolve this identification problem (7). Incorporation of additional spectra of these species into the database, as suggested by Seng et al. (5) and van Veen et al. (9), will not solve the problem. Additional tests, i.e., for optochin or bile solubility, are still needed. Analogously, the reported misidentification of Shigella species as Escherichia coli (2, 5) is also due to the close relation of these 2 organisms (4, 6). In fact, these 2 organisms are 1 species, but historical and clinical context prevented nomenclature change. So, this problem is not caused by lack of database entry—Shigella species are included in the separated Security Relevant (SR) database—and is also technically not solvable. In conclusion, MALDI-TOF MS is a revolution in the identification of pathogens in clinical laboratories, but when this technique is implemented in a routine clinical biology laboratory, a good knowledge of taxonomy and its limitations is needed. Closely related microorganisms, like E. coli and Shigella spp., and some streptococci, like S. mitis and S. pneumoniae, cannot be distinguished without additional tests.