Abstract
The Streptococcus bovis group (previously group D streptococci) consists of seven distinct species and subspecies. Definitive identification within the group is important, as certain organisms have been associated with gastrointestinal carcinoma, bacteremia, infective endocarditis, meningitis, biliary tract disease, and carcinoma, among others. Definitive identification, however, remains elusive due to limitations and inconsistencies across commonly used identification platforms in the United States. Here, we compared the performance of standard biochemical (Trek Gram-positive identification [GPID] plate, Vitek 2 GPID), sequencing (16S rDNA, sodA) databases (NCBI, RDP, CDC MicrobeNet), and matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS) platforms (Vitek MS, Bruker Biotyper MS) using a set of eight type strains representing all seven strains within the S. bovis group. Despite the evaluation of contemporary methods, no single platform was able to definitively identify all type strains within the S. bovis group. Vitek MS (85.7%, 7/8) provided the most accurate definitive identifications, followed by sodA sequencing (75%, 6/8). Vitek 2 and Bruker Biotyper RUO platforms performed the next best (62.5%, 5/8). All remaining platforms failed to adequately differentiate type strains within the S. bovis group (range, 0 to 37.5%). Laboratorians and clinicians should be aware of the identification limitations of routine testing algorithms and incorporate reflex testing, when appropriate, to platforms such as Vitek MS and/or sodA sequencing that are more able to definitively identify S. bovis group organisms. Further clinical evaluation was conducted using 65 clinical isolates from three geographically distinct U.S. institutions. Future improvements in identification platforms may reveal new clinical and epidemiological trends for members of the S. bovis group.