Nontuberculous Mycobacteria in Respiratory Tract Infections, Eastern Asia

To characterize the distribution of nontuberculous mycobacteria (NTM) species isolated from pulmonary samples from persons in Asia and their association with pulmonary infections, we reviewed the literature. Mycobacterium avium complex bacteria were most frequently isolated (13%–81%) and were the most common cause of pulmonary NTM disease (43%–81%). Also pathogenic were rapidly growing mycobacteria (M. chelonae, M. fortuitum, M. abscessus). Among all NTM isolated from pulmonary samples, 31% (582/1,744) were considered clinically relevant according to American Thoracic Society diagnostic criteria. Most patients were male (79%) and had a history of tuberculosis (37%). In Asia, high prevalence of rapidly growing mycobacteria and a history of tuberculosis are distinct characteristics of pulmonary NTM disease. This geographic variation is not well reflected in the American Thoracic Society criteria for NTM infections and could be incorporated in future guidelines.

N ontuberculous mycobacteria (NTM) are common in the environment and have been isolated worldwide (1). They are increasingly recognized as pathogens in humans. Pulmonary disease is the most common manifestation (2) and is thought to result from aerosol inhalation. Because of their omnipresence in the environment, isolation of NTM from the respiratory tract does not, per se, indicate NTM disease. Therefore, the American Thoracic Society (ATS) has established diagnostic criteria to help distinguish between contamination and true NTM disease (1).
Although NTM are present worldwide, much of the literature on pulmonary NTM disease comes from industrialized countries, mainly Europe, North America, and Japan. Differences in the NTM species distribution in clinical samples and disease have been noted among these regions (3); by extrapolation, these characteristics in other parts of the world probably differ as well.
Asia has a long history of NTM research, both clinical and fundamental. Tsukamura et al. have reported on pulmonary NTM infections in Japan dating back to the early 1970s (4). Yet because many studies were not reported in English, knowledge of the distribution of NTM species in Asia is limited. However, these data will enhance our understanding of NTM diversity between and within species and their association with NTM disease in humans. For this reason, we searched the literature on clinical NTM isolation and disease from different regions in Asia and compared our fi ndings with previously published data from other regions.
We found 256 citations. If we considered the abstract to be relevant, we obtained a full copy of the article; we contacted authors if full-text articles could not be retrieved. Furthermore, the reference sections were screened for other eligible citations. We considered 67 articles to be relevant, of which 37 were excluded for the following reasons: 11 were case reports, 7 referred to disseminated NTM infections, 6 were not in English (5 in Japanese, 1 in Chinese), 6 did not concern pulmonary NTM, 4 were reviews, and 3 represented neither epidemiologic data nor clinical cases. From the remaining 30 articles (4-33), the following data were abstracted for this review: country, research setting, NTM species, clinical features of the patients, and radiographic data. All articles were screened to determine whether ATS diagnostic criteria for the determination of clinical relevance of NTM isolations applied (1). Cases consistent with the ATS diagnostic criteria were considered clinically relevant.

Clinical Relevance
The clinical relevance of pulmonary NTM isolates ( Table 2) was mentioned in 10 articles covering a total of 1,744 patients (6,16,21,23,25,(28)(29)(30)32,33). No articles covered the clinical relevance of NTM in China. Different criteria were used to defi ne clinical relevance, but all criteria comprised a combination of clinical, bacteriologic, and radiographic criteria. Only 4 studies reported a priori use of ATS criteria (1). According to ATS criteria, for 31%  (28); however, in a study of HIV-infected patients in Thailand, relevance rose to 76% (30). Figure 2 shows the distribution of mycobacterial species among patients with defi nite pulmonary NTM disease in Asia, according to ATS criteria. MAC was the most frequently reported (68% of all cases) cause of NTM disease; RGM were second (14% of all cases). Prevalence of RGM pulmonary infections ranged from 2.6% (Japan) to 44% (South Korea). M. malmoense and M. xenopi were not reported as causative species in any of these studies in Asia.
We found some discrepancies between isolation frequency ( Figure 1) and clinical relevance ( Figure 2

Discussion
NTM isolation and disease in Asia have several features. First, a substantial percentage (31%) of patients from whom pulmonary NTM were isolated had clinically relevant NTM disease. This fi nding is similar to the 33% and 25% found in studies in Canada and the Netherlands, respectively (2,3). In Asia, NTM may cause substantial pulmonary disease; differences in clinical relevance exist among species (Figure 3), as previously observed (2) (4,6,10,13,17,19,21,25,(27)(28)(29)(30)(31)(32)(33). study supports their conclusion of the predominance of MAC in Asia, which is consistent with its predominance in other parts of the world, namely, North America and most parts of Europe (3). Third, we found that in some regions in Asia, RGM are a major cause of pulmonary NTM disease. This fi nding contrasts with studies of NTM in other parts of the world (3). In a surveillance study from the Netherlands for instance, RGM caused only 3% of all pulmonary NTM infections (2). In the United States, this percentage is ≈5% (36). In the present review, RGM was found to generally cause 14% of pulmonary NTM infections, but in 3 countries (India, Taiwan, South Korea) this percentage rose to >30% of infections. The fact that RGM were frequently found in pulmonary samples (Figure 1) could refl ect higher environmental exposure of RGM in Asia and, hence, higher isolation frequency. The predominance of RGM species may be the result of laboratory practices as well. Ethnic factors may also contribute to susceptibility to different species; i.e., Asian persons could be more susceptible to RGM infection.
Contrary to the high frequency of isolation of RGM species, M. malmoense and M. xenopi, frequently seen in other parts of the world, were not seen as causative species in any of the studies from Asia. M. xenopi has been associated with hot water systems (1); as a result, it might be expected to be more rare in Asia, where the water delivery infrastructure is less developed than that in Europe and North America.
A fourth feature of pulmonary NTM disease in Asiacompared with Europe and North America-was the relatively high percentage of patients with a history of TB. This fi nding might merely refl ect the higher incidence of TB in Asia, or it could refl ect higher clinician awareness in Asia, such that physicians order Mycobacterium spp. cultures in former TB patients with coughing and hence fi nd a relatively higher number of NTM isolates. Alternatively, it could refl ect a true predilection of NTM for patients with structural lung disease (1) associated with a higher susceptibility to mycobacterial infection in general. The role of TB in the pathogenesis of pulmonary NTM disease is controversial; structural lung damage by a TB infection renders the host vulnerable to NTM disease (1), but there are also clues that exposure to TB infers cross-protection to NTM disease (37).
Our study has some limitations. The major limitation is the language restriction. The inclusion of languages other than English would probably have increased precision. For instance, during our literature search we came across 5 articles, published as abstracts in PubMed, on NTM infections in Japan. Although certain aspects of these data were already published in the English-language articles we included, we did not have the means to include these non-English-language studies. Nevertheless, our study illustrates the distribution of NTM infections from different geographic areas in Asia and will increase knowledge of the distribution and relevance of NTM species in Asia.
Another limitation is the long time span of the included studies. Because they ranged from 1969 to 2008, they used different decontamination, culture, and identifi cation methods. Data should therefore be considered with caution because of the variety of laboratory procedures used by the several authors to isolate and identify NTM. First, different sample decontamination protocols may determine the yield of NTM by selecting for certain NTM species and inhibiting others (38). Second, the introduction of more sensitive liquid media probably increased NTM isolation and perhaps selected for specifi c species (26). Finally, and foremost, the major differences in identifi cation methods used in the studies introduce important biases. Because the taxonomy of NTM has been changing in recent years (37,39,40), the different NTM identifi cation methods used in the various studies might have infl uenced our results; use of molecular tools to identify the historical isolates would probably result in different, more detailed, identifi cations, especially among the MAC, the M. simiae group, and the RGM (37,39,40). We did note increased use of molecular tools for NTM identifi cation over time (Table 1), which will, over the next few years, provide us with a more up-todate overview of NTM species distribution in Asia.  In conclusion, despite the limitations of language and species identifi cation methods, we have described the scale of human pulmonary NTM infections in eastern Asia. MAC bacteria were not only the most prevalent NTM found in pulmonary samples, but they were also the most common cause of pulmonary NTM disease in this geographic region. Distinct epidemiologic and clinical characteristics of pulmonary NTM disease in Asia were found: many patients had a history of TB, and RGM were a frequent cause of pulmonary NTM infections. These distinct characteristics of pulmonary NTM disease in this part of the world are not well refl ected in the current ATS criteria on NTM infections and could be incorporated in future guidelines.
Dr Simons is a resident in pulmonary medicine at the Radboud University Nijmegen Medical Centre. His primary research interest focuses on resistance mechanisms in multidrugresistant tuberculosis.