| Abstract: | Mycobacterium abscessus is an important cause of water-related nosocomial outbreaks or pseudo-outbreaks. Strain comparison has relied on pulsed-field gel electrophoresis (PFGE). Unfortunately, almost 50% of strains cannot be assessed by this method. We studied 118 strains of M. abscessus previously studied by PFGE by randomly amplified polymorphic DNA (RAPD) PCR, including isolates from eight nosocomial outbreaks. Ten random primers were evaluated by using DNA prepared by boiling or phenol-chloroform extraction. Both DNA preparations gave the same grouping of isolates for three outbreaks compared to the groupings obtained by PFGE. Five outbreaks due to M. abscessus which gave broken DNA by PFGE gave evaluable patterns when studied by RAPD-PCR, with isolate clustering being consistent with that from other laboratory and epidemiologic data. The patterns were highly method dependent, strain comparison required the use of multiple primers, and the method worked best with purified DNA and by using strains for comparison on the same gel. We propose categories of indistinguishable, different, and inconclusive when comparing strains by RAPD-PCR. This study demonstrates that RAPD-PCR can be used for genetic comparison of M. abscessus strains, including strains which cannot be compared by PFGE, but the potential for misinterpretation is greater than that by PFGE. |
