The long-term objective of this project is the development of a system based on high performance liquid chromatography (HPLC) for rapid and accurate antibiotic susceptibility testing of Mycobacterium tuberculosis (MTB). Susceptible and resistant strains of MTB, exposed to anti-MTB antibiotics, exhibit pronounced differences in mycolic acid content. Initial research will focus on adapting the HPLC-based susceptibility testing to current commercially available growth/susceptibility systems. The MIDI Sherlock(r) Mycobacteria Identification System is an HPLC-based system that has FDA 510(k) clearance for identification of MTB. The MIDI Sherlock software will be modified to allow for automated calculation of ratios of mycolic acid content of MTB strains from """"""""control"""""""" or from antibiotic-containing media. Although isoniazid causes 35-40% loss of mycolic acids, in susceptible strains, with as little as 20 minutes exposure to the inhibitor, rifampin, ethambutol and pyrazinamide may require up to 72 h for optimal determination. This is in contrast to the typical 6-9 days required for other automated susceptibility test systems. As the highly specific pattern of mycolic acids of MTB is seen in the analysis, contamination by non-mycobacterial species do not cause false resistance evaluations, thus solving a serious problem in other automated systems. A """"""""mixed culture"""""""" algorithm will be evaluated for effectiveness in identifying contamination by non-TB mycobacteria. Initial research will use commercially available media and anti-MTB antibiotics to determine resistance or susceptibility by mycolic acid content analysis. This will facilitate adoption of the mycolic acid test by laboratories having the growth-based systems. Although not crucial to the success of the project, direct-from-sputum analysis for MTB identification and susceptibility testing will also be evaluated. ? ? ?
