ABSTRACT
The IR Biotyper and matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS) using ClinProTools software (MALDI-TOF MS-ClinProTools) are two novel typing methods that rely on the analysis of carbohydrate and peptide residues in intact bacterial cells. These two methods have shown promising results in the rapid and accurate typing of bacteria. In this study, we evaluated these novel typing methods in comparison with genotypic typing for cluster analysis of Burkholderia cenocepacia epidemic strain ET12, isolated from adult cystic fibrosis patients. Sixty-six isolates of B. cenocepacia were used in this study, 35 of which were identified as the ET12 strain and 31 as non-ET12 strains by repetitive-element PCR (rep-PCR). Twelve isolates were used for the creation of typing models using IR Biotyper and MALDI-TOF MS-ClinProTools, and 54 isolates were used for external validation of the typing models. The IR Biotyper linear discriminant analysis (LDA) model had a diagnostic sensitivity of 84.6% for typing the epidemic strain, ET12. At a cutoff of 70%, MALDI-TOF MS-ClinProTools had 87.5% diagnostic sensitivity in detecting the ET12 strain (P = 1.00). Both methods had a diagnostic specificity of ≥80% for detecting the ET12 strain. In conclusion, IR Biotyper and MALDI-TOF MS-ClinProTools offer rapid typing using proteomics and analysis of small cellular molecules with a low running cost. Our pilot study showed suboptimal accuracy of both methods for typing outbreak strains of B. cenocepacia. Extending the spectral region analyzed by the IR Biotyper can improve the accuracy and has the potential of improving the generalizability of this technique for typing other organisms. IMPORTANCE Respiratory infections due to Burkholderia cenocepacia, particularly the ET12 epidemic strain, are considered sentinel events for persons with cystic fibrosis, as they are often associated with person-to-person transmission and accelerated decline in lung function and early mortality. Current typing methods are generally only available at reference centers, with long turn-around-times, which can affect the identification of outbreaks and critical patient triage. This pilot study aims to add to the growing literature illustrating the potential utility of Fourier transform infrared spectroscopy (FTIR), a novel rapid method, for the successful typing of clinically significant bacteria. In this study, we evaluated its utility to discriminate between the ET12 clone and non-ET12 isolates of B. cenocepacia and compared it to proteomics cluster analysis using MALDI-TOF MS and ClinProTools software. Both methods had encouraging but suboptimal accuracy (≥85% sensitivity and ≥83% specificity), which will likely be improved by extending the spectral region analyzed by the IR Biotyper with updated software.
