Evaluation of MALDI-TOF MS as an alternative method for epidemiological cluster analysis of Mycobacterium tuberculosis isolates

Background: Identification of Mycobacterium tuberculosis (MTB) transmission, including strain typing and cluster analysis is recommended for infection control. Existing molecular methods require complex laboratory infrastructure: variable number of tandem repeats (VNTR) typing is considered the gold standard, but is only available in a limited number of laboratories worldwide. Recently, Matrix-Assisted Laser Desorption/Ionisation Time of Flight Mass Spectrometry (MALDI-TOF MS) has been implemented for rapid bacterial identification in many clinical laboratories. Additionally, its application for epidemiological strain analysis has been evaluated for a wide range of pathogens. In this preliminary study, we evaluated the technical constraints and the performance of MALDI-TOF MS for typing of MTB isolates. Material/methods: 22 MTB lysates previously characterised with VNTR typing were analysed with MALDI-TOF MS. Measurements were realized on a microflex LT (Bruker Daltonik, Germany). Ten spots of each extracted strain were measured 3 times using MBT_FC.par flexControl and MBT-autoX.axe autoExecute methods. The resulting 30 spectra for each MTB strain were downloaded in the MALDI BioTyper software for manual review. Only strains which yielded more than 20 valid spectra were considered for further analysis. For each strain, a consensus main spectrum (MSP) was generated using the BioNumerics software (Applied Maths, Belgium). An MSP dendrogram was generated using ranked Pearson correlation. VNTR analysis was performed with the VNTRplus web platform. Results: A minimum of 20/30 valid spectra was obtained for 7 out of 22 samples. The dendrogram of the 7 strains is presented in Figure 1. It shows > 98% homology between two strains (9 and 10). This cluster was closely related to a third strain (11). The NJ tree obtained by MIRU-VNTR analysis confirmed the detection of a cluster for strains 9 and 10, and a close homology was furthermore found between this cluster and the strain 11 (figure 2). Conclusions: MALDI-TOF MS must be further evaluated for its utility in rapid identification of MTB clusters. Nevertheless, three major concerns were identified needing to be solved before this technique may be more widely implemented as a rapid and low-cost typing technique. First, the performance of MADLI-TOF MS for Mycobacterial applications is still imperfect. Second, the encouraging results obtained in this early cluster analysis need to be confirmed by a much wider evaluation. Last, although MALDI-TOF MS is widely available, the typing process used in this analysis is still complex and time-consuming. In order to be implementable in routine conditions, the process should be simplified and the comparison should be made using a pooled database of MSP, continuously alimented by users.