In many industrialized countries, the ability to rapidly distinguish between MTC and NTM is critical in clinical practice. Indeed, the anti-tuberculosis drug resistance survey in Japan revealed that 19.3% of all clinical mycobacterial isolates are NTM , underscoring the importance of rapid and accurate detection of MTC from acid-fast bacillus-positive culture. The immunochromatographic assay kit for the identification of MTC is now widely used in many countries. Capilia TB-Neo is the improved version of Capilia TB, and has been subjected to few clinical evaluations. Here, we report good overall performance of the kit but with several limitations.
In this study, the sensitivity of Capilia TB-Neo was 99.4% to clinical MTC isolates or 99.6% excluding M. bovis BCG, while the specificity of the kit tested to clinical NTM isolates was 100%. However, the isolation of BCG could present a practical problem. The M. bovis BCG Tokyo strain is sporadically isolated in Japan as a complex of vaccination or bladder cancer therapy, and is identified as MTC with the kit . Some BCG strains such as Connaught, Pasteur, and Tice lack RD2 including the mpt64 gene, but RD2 is conserved in others such as Tokyo, Moreau, and Russia . This issue should be properly addressed to avoid confusion. Although it is difficult to discriminate BCG Tokyo from MTC with mpt64/mpb64, their differentiation would be an important advance in the development of a future TLC product. The weak false-positive reaction to M. marinum that was reported using Capilia TB  was not observed in this study, and resulted in better specificity. The minimum detection concentration of M. tuberculosis for Capilia TB-Neo was 105 CFU/ml (data not shown), which was one-tenth than that for the previous kit. There was a report that Cpilia TB-Neo was higher sensitivity than Capilia TB . In summary, the overall performance of Capilia TB-Neo was better than Capilia TB in both sensitivity and specificity.
SD MPT64 and TBc ID were also tested with reference strains. Both SD MPT64 and TBc ID showed false-positive results against several NTM strains in this study. Kodama et al.  reported that no M. marinum strains grown on 2% Ogawa medium tested positive by using the Capilia TB, while all strains grown on 3 kinds of liquid medium, MGIT (Becton Dickinson, Japan), KRD medium (Japan BCG Laboratory, Japan) and Myco Acid (Kyokuto Pharmaceutical Industrial Co. Ltd., Japan), eventually displayed a positive reaction that intensified with time. Kodama et al. speculated that nonspecific antigen which could make complex with anti-MPB64 antibody may be produced in liquid mediums, but not on solid medium. Considering the effect of liquid culture, the original bacterial suspensions giving false-positive results, that were prepared from liquid and solid culture, were then re-tested before and after 10-fold dilution. Interestingly, none of these diluted strains tested positive in these kits, but bacterial concentrations were high enough for positive results in case of MTC. These results implied that a high concentration of bacterial antigens could induce non-specific reactions in SD MPT64 and TBc ID. The manufacturer’s instructions for the TBc ID indicate that this kit may be used up to 10 days after a positive MGIT alarm. This non-specific reaction should be properly addressed in clinical practice, and the users should perform morphological characterization with a microscope to identify cord formation.
Several mutations in the mpt64 gene produce a negative test result for M. tuberculosis isolates in the TLC assay using anti-MPB64 monoclonal antibodies. To date, these include a 63-bp deletion from nucleotide 196, a 1-bp deletion from nucleotide 266, a point mutation at position 388 or 402, IS6110 insertion mutation at position 177 or 501, a 176-bp deletion from nucleotide 512, and a 1-bp insertion at position 287 [10, 13, 21]. In our study, 2 M. tuberculosis isolates gave false-negative results by using the Capilia TB-Neo, SD MPT64, and TBc ID. One isolate had a deletion of 63 bp from nucleotide 196 in the mpt64 gene as reported previously, and the other isolate possessed a 3,659-bp deletion from nucleotide 874 in Rv1977 to 905 in Rv1981c, including the whole mpt64 gene. To the best of our knowledge, this is the first report of a large deletion in mpt64. A transposon site hybridization (TraSH) study  indicated that mpt64 is not essential for infection or in vitro growth of M. tuberculosis. This large deletion mutant supported the finding.
In summary, the TLC assay detecting MPB64 or MPT64 can be applied to specimens prepared from liquid and solid culture. It does not need special reagents, instruments, or complex techniques. Capilia TB-Neo tested in this study showed excellent sensitivity with perfect specificity.