Résumé
The objective of this study was to provide standardized antibiotic susceptibility testing (AST) for Mycobacterium abscessus with regard to subspecies.
One hundred and sixty-five clinical isolates were tested for susceptibility to 15 antibiotics using a commercial microdilution method, at two reading times: (i) early reading time (ERT), when the growth control was first positive; and (ii) late reading time (LRT), of 14 days, for detecting inducible resistance. In addition, genes or mutations involved in resistance were studied [erm(41), rrl and rrs].
Three patterns were observed for clarithromycin: (i) MIC >16 mg/L at ERT (median 5 days) for 15 isolates [10 subsp. abscessus erm(41) sequevar T28, 3 subsp. bolletii and 2 subsp. massiliense] among which 9 harboured an a2058g/c rrl mutation; (ii) MIC ≤16 mg/L at ERT, but >16 mg/L at LRT, for 106 isolates [84 abscessus erm(41) T28 and 22 bolletii] showing intrinsic inducible resistance; and (iii) MIC ≤4 mg/L at ERT and LRT for 44 isolates [18 abscessus erm(41) C28 and 26 massiliense]. Amikacin MIC was >64 mg/L for eight isolates [five abscessus erm(41) T28, two massiliense and one bolletii] among which seven harboured the a1408g rrs mutation, but ≤64 mg/L for the remaining isolates without mutation. For the other antibiotics, only one WT pattern was observed, with cefoxitin, tigecycline and linezolid showing MIC values compatible with susceptibility.
Standard AST can predict clarithromycin and amikacin resistance using interpretation rules with regard to subspecies. For other antibiotics, since only one pattern is observed, there is no need for systematic phenotypic or genotypic testing.