Differentiation between Streptococcus pneumoniae and other viridans group streptococci by matrix-assisted laser desorption/ionization time of flight mass spectrometry

Matrix-assisted laser desorption/ionization-time of flight mass spectrometry (MALDI-TOF MS) is becoming the method of choice for bacterial identification. However, correct identification by MALDI-TOF of closely related microorganisms such as viridans streptococci is still cumbersome, especially in the identification of S. pneumoniae. By making use of additional spectra peaks for S. pneumoniae and other viridans group streptococci (VGS). We re-identified viridans streptococci that had been identified and characterized by molecular and phenotypic techniques by MALDI-TOF. VGS isolates (n = 579), 496 S. pneumoniae and 83 non-S. pneumoniae were analysed using MALDI-TOF MS and the sensitivity and specificity of MALDI-TOF MS was assessed. Hereafter, mass spectra analysis was performed. Presumptive identification of proteins represented by discriminatory peaks was performed by molecular weight matching and the corresponding nucleotides sequences against different protein databases. Using the Bruker reference library, 495 of 496 S. pneumoniae isolates were identified as S. pneumoniae and one isolate was identified as non-S. pneumoniae. Of the 83 non-S. pneumoniae isolates, 37 were correctly identified as non-S. pneumoniae, and 46 isolates as S. pneumoniae. The sensitivity of the MALDI-TOF MS was 99.8% (95% confidence interval (CI) 98.9–100) and the specificity was 44.6% (95% CI 33.7–55.9). Eight spectra peaks were mostly present in one category (S. pneumoniae or other VGS) and absent in the other category and inversely. Two spectra peaks of these (m/z 3420 and 3436) were selected by logistic regression to generate three identification profiles. These profiles could differentiate between S. pneumoniae and other VGS with high sensitivity and specificity (99.4% and 98.8%, respectively). Spectral peaks analysis based identification is a powerful tool to differentiate S. pneumoniae from other VGS species with high specificity and sensitivity and is a useful method for pneumococcal identification in carriage studies. More research is needed to further confirm our findings. Extrapolation of these results to clinical strains need to be deeply investigated.