RESUMEN
Nontuberculous mycobacteria (NTM) that cannot be identified to the species level by reverse line blot hybridization assays and sequencing of the 16S rRNA gene comprise a challenge for reference laboratories. However, the number of 16S rRNA gene sequences added to online public databases is growing rapidly, as is the number of Mycobacterium species. Therefore, we re-analysed 178 Mycobacterium isolates with 53 previously unmatched 16S rRNA gene sequences, submitted to our national reference laboratory in 19992007. All sequences were again compared with the GenBank database sequences and the isolates were re-identified using two commercially available identification kits, targeting separate genetic loci. Ninety-three out of 178 isolates (52%) with 20 different 16S rRNA gene sequences could be assigned to validly published species. The two reverse line blot assays provided false identifications for three recently described species and 22 discrepancies were recorded in the identification results between the two reverse line blot assays. Identification by reverse line blot assays underestimates the genetic heterogeneity among NTM. This heterogeneity can be clinically relevant because particular sub-groupings of species can cause specific disease types. Therefore, sequence-based identification is preferable, at least at the reference laboratory level, although the exact targets needed for clinically useful results remain to be established. The number of NTM species in the environment is probably so high that unidentifiable clinical isolates should be given a separate species status only if this is clinically meaningful.