Totty, and P. Deol are employees of bioMrieux, Inc., manufacturer of the BacTALERT 3D, BacTALERT MP culture bottles, and VITEK MS used in this study.

Here, we describe a novel, cost-effective, and validated method that allows for direct and rapid identification of mycobacteria from a positive liquid culture using VITEK MS with a total process duration under 45 min.

From a liquid mycobacteria culture a 3.0 mL aliquot is removed 2472 h post positivity and centrifuged to create a pellet.

After decanting, the tube is blotted dry, the pellet is re-suspended in 0.5 mL of 70 ethanol and then transferred into a 2.0 mL tube containing glass beads.

Mycobacteria are disrupted mechanically followed by a 10 min.

Inactivated material is pelleted by centrifugation and then re-suspended in 10 L of 70 formic acid and 10 L of acetonitrile.

After centrifugation, 1 L of supernatant (protein extract) is deposited onto target slide, allowed to dry, and then 1 L CHCA matrix is added.

A seeded study was conducted to demonstrate the reliability of the method, a total of 251 culture samples obtained from automated culture systems (BacTALERT MP bottles, BACTEC MGIT 960 tubes, and VersaTREK Myco bottles), were tested and resulted in 98.8 correct identification.

Reproducibility was shown by testing three organisms across three reagent lots, between four laboratory technicians, over the course of five days for three liquid media systems resulting in a total of 180 deposits with an overall correct identification of 98.9 with the remaining results giving no identification.

Additional studies were performed including comparison of different mechanical disruption techniques, stability of frozen extracts, and stability of slide deposits to allow for flexibility in a routine clinical workflow.

The described method proved to be safe while providing consistent and reproducible results for different species of mycobacteria and is compatible with the three most widely used liquid media medium detection systems.

Introduction Mycobacteria infections are a significant threat to worldwide public health.

In 2015, an estimated 13 of the worlds population was estimated to be infected with Mycobacterium tuberculosis, and tuberculosis claimed the lives of an estimated 1.4 million people (World Health Organization, 2016).

Non-tuberculous mycobacteria (NTMs) are increasingly recognized as mostly opportunistic pathogens of humans with the most frequent disease manifestation being chronic pulmonary infection (Buckwalter et al., 2015).

The incidence of infection with NTM is rising due to an increased immunocompromised population (Marras et al., 2007).

Alongside the rise in NTM infections, the number of mycobacteria species being recognized has increased over recent decades, which can largely be attributed to advances in genotypic analysis techniques (Tortoli, 2003).

DNA sequencing improves the turnaround time but is costly, labor intensive, requires specific equipment and a high level of expertise.

Other molecular methods such as DNA probes provide fast and reproducible results for the more common mycobacteria species such as M.

Section snippets Mycobacterium isolates Clinically relevant mycobacteria species (rapid and slow growers) were selected for all studies (Clinical and Laboratory Standards Institute, 2008).

Mycobacteria strains used in the studies are listed in Table 1.

All strains were identified by 16S rDNA or rpoB prior to testing.

Isolates were cultured from frozen stocks onto Lowenstein-Jenson slants or 7H11 plates prior to inoculation into MP, MGIT, and VersaTREK bottles containing the system-specific nutrient and antimicrobial supplements, at Protein extraction and identification from different media types A total of 251 samples from 9 Mycobacterium species across all media systems were analyzed on the VITEK MS.

Of these samples, 248251 (98.8) were identified to the correct species level with 99.9 confidence or higher.

Three samples resulted in no identification; however, no incorrect identifications were observed.

Performance per medium type indicates 97 of isolates correctly identified to the species level (Table 2).

Reproducibility Correct identifications were achieved for 8990 (98.9) samples (data not Discussion Identifying mycobacteria from liquid media is challenging due to the sensitivity of the automated detection systems which are designed to detect growth of mycobacteria as quickly as possible.

As a result, a positive liquid sample contains less biomass making it difficult to consistently identify using MALDI-TOF.

The method presented here has been designed to compensate for the low biomass at time of positivity and is compatible with the three most widely used detection systems.

A prolonged Declaration of interest E.

Bact Alert Package Insert