Comparison of MALDI-TOF and 16S rRNA methods in identification of viridans group streptococci

Abstract

Accurate identification of viridans group streptococci (VGS) frequently encountered as a causative agent of infective endocarditis is always a challenge for the clinical microbiology laboratory. Clinical microbiology laboratories generally use semi automatic/full automatic systems, molecular methods and also conventional methods for the identification of these bacteria. There are recent published studies that have used MALDI-TOF (Matrix Assisted Laser Ionization Mass Spectrometry-Time of Flight) systems in the identification of VGS. The aim of the study was to compare the performance of the conventional methods, semi automatic and MALDI-TOF MS system used in identification of VGS in oral microbiota of persons under the risk of infective endocarditis, with the gold standard method 16S rRNA sequence analysis and to create a diagnosis algorithm for the identification of VGS in clinical microbiology laboratories according to the obtained data.The study was conducted with 51 VGS strains isolated from oral microbiota of the patients with rheumatologic cardiac, valve and/or prosthetic valve diseases, under the risk of development of infective endocarditis, who have admitted to Ankara Numune Training and Research Hospital, Department of Cardiology, between February-June 2015. Standard microbiology procedures, optochin susceptibility and bile solubility tests were done for the isolation of bacteria. Bacteria were also identified with APISTREP (bioMérieux, France) and MALDI-TOF MS Bruker Microflex (Bruker Biotyper; Bruker Daltonics, Bremen, Germany) methods. BSF-8 (5´-AGAGTTTGATCCTGGCTCAG-3´) and BSR-534(5´-ATTACCGCGGCTGCTGGC-3´) primers were used in the 16S rRNA sequence analysis of bacteria. ABI PRISM 3100 Avan t Genetic Analyzer (Applied Biossytems, Foster City, CA, USA) were used for the sequence analysis. Electropherograms were analyzed in SeqScape Software (Applied Biosystems, Foster City, CA, USA) and compared with the reference sequences in GenBank with BLASTN (NCBI). According to the result of optochin and bile solubility tests, with API STREP system, 16 (31,37%) of the isolates were identified as Mitis group, 15 (29.41%) as Anginosus group, 9 (17.5%) as Salivarius group, 7 (13,73%) as Sanguinis group and 4 (7.84%) as Bovis group among optochin and bile resistant alpha hemolytic streptococci. Moreover, of the same isolates 20 (39.22%) were identified as Mitis group, 14 (27.45%) as Anginosus group, 13 (25.49%) as Salivarius group and 4 (7.84%) as Sanguinis group with MALDI-TOF system. In the identification with 16S rRNA, 25 (49.02%) of the isolates were identified as Mitis group, 13 (25.49%) as Anginosus group, 12 (23.53%) as Salivarius group and 1 (1.96%) as Sanguinis group. According to the results, it was determined that 33 (64.70%) of the isolates identified in MALDI-TOF MS system and 31 (60.78%) of the isolates identified in API STREP system were compatible with 16S rRNA sequence analysis method. For Mitis group, API STREP test sensitivity was 48.00% and specificity was 84.62% and MALDI-TOF system sensitivity was 80.00% and specificity was 100%. As VGS identification is a complicated process, we believe a single method will be insufficient for the identification of these isolates in clinical microbiology laboratories. We suggest that MALDI-TOF system can be used for VGS diagnosis, however, optochin test and/or molecular methods should also be included in the diagnosis algorithm when necessary.