Background

Non-tuberculous mycobacteria (NTM) has been increasingly implicated in a broad range of infectious diseases worldwide [1,2,3,4]. These environmental microbes are found primarily in water and soil. Possible route of transmission includes direct exposure to aerosolized water or soil containing NTM species [2,3,4,5]. The interaction between host immune system and the pathogenicity of the organisms plays a key role in disease susceptibility [6]. Well-defined host risk factors include advanced age, male gender, slender and older Caucasian women, immune defects, structural pulmonary diseases, alpha1-antitrypsin deficiency [7,8,9,10,11]. Urban living, especially in wet area with dense population, also increases the susceptibility to NTM infections [12, 13]. Other predisposing factors include low body-mass index, skeletal abnormalities and gastroesophageal reflux [11, 14].

Pulmonary, lymphadenitis, skin/soft tissue and dissemination are the four major clinical manifestations of NTM infections. Pulmonary NTM is the commonest clinical syndrome [15, 16] and it is associated with significant morbidity among older adults [7, 17]. Structural pulmonary diseases such as cystic fibrosis, chronic obstructive pulmonary disease (COPD) bronchiectasis, and tuberculosis (TB) predispose individuals to NTM lung infections. In the United States (US), an estimation of 86,000 NTM pulmonary cases were reported in 2010 with the prevalence of 8.6 per 100, 000 [18], and the number was growing by approximately 8% every year [14]. In Canada, the prevalence of NTM pulmonary disease was estimated to be 9.8 per 100,000 in 2010 [19]. Asians and Pacific Islanders appear to be more susceptible to NTM diseases [14]. Several studies have reported the increasing number of annual NTM isolates and the significant clinical relevance [20,21,22]. In Korea, the incidence of pulmonary NTM disease has doubled in 2015 compared to the number in 2009 [23, 24]. Japan revealed that NTM disease has reached 14.7 cases per 100, 000 people-years [25] . Pulmonary NTM disease has been described common in Taiwan, with the average incidence rate of 46.0 episodes per 100,000 hospital-based patient-years from 2010 to 2014 [26]. The epidemiology of NTM pulmonary disease differs in various geographical regions. The data remains limited in Southeast Asian countries including Singapore.

Singapore is a highly urbanized city-state with a tropical climate. It is the world’s third most densely populated country with a total population of 5.6 million. The percentage of the population aged ≥65 years is over 10% and is projected to grow to nearly 20% by 2030 [27]. In addition, a large proportion of residents aged over 50 years have a history of TB as TB was prevalent in Singapore until the 1970s [28]. Collectively, high population density, a wet tropical urban environment, and a large aging population with history of TB make the residents of Singapore particularly susceptible to NTM diseases. In this study, we evaluated the epidemiological and clinical characteristics of NTM diseases in Singapore, with an emphasis on pulmonary infection.

Methods

Study setting and study subjects

Singapore General Hospital (SGH) is a 1700-bedder tertiary hospital located within the Outram Campus in Singapore with 5 other National Specialty Centres, facilitating a comprehensive range of medical services for over 1 million patients annually. SGH CTBL is one of the two centralized laboratories accredited for mycobacterial testing in Singapore and is responsible for about 80% of the country’s mycobacterial culture testing volumes. All NTM isolates identified in SGH CTBL from 1st January 2012 to 31st December 2016 were included in the study. These were speciated by a combination of methods including high-performance liquid chromatography (HPLC; Beckman Instruments, USA), INNO-LiPA Mycobacteria v2 (Innogenetics, Belgium) and/or DNA probes (AccuProbes; Gen-Probe, USA) and correlated with the colonial morphology, pigmentation and growth rate. For cases with multiple isolates of one or more species, only the first isolate of the same species was included for analysis. M. gordonae, a well-known environmental contaminant, was excluded from the analysis [29,30,31]. Isolates were cultivated from various anatomical sites which included pulmonary source (sputum, bronchoalveolar lavage (BAL), lung biopsy, endotracheal tube aspirate (ETTA) and pleural fluid) and extrapulmonary sources such as skin and soft tissue, blood, lymph node and other sterile sites.

Data collection

Medical records of included cases were reviewed from the electronic medical record system by trained research coordinators with nursing or medical background. The following information were collected: 1) Patients’ demographic information, 2) medical comorbidities, 3) clinical presentations, 4) radiology findings of nodular, cavitary opacities or bronchiectasis on chest X-ray (CXR); or bronchiectasis with multiple small nodules on high resolution computed tomography (HRCT) scan, and 5) microbiology results. In addition, history of tuberculosis, body-mass index (BMI) were also recorded.

NTM diseases were defined using the criteria proposed by 2007 American Thoracic Society and the infectious Diseases Society of America (ATS/IDSA) and were further categorized as pulmonary disease, skin/soft tissue disease, disseminated disease (blood/other sterile sites infection), lymphadenitis. NTM pulmonary disease was defined when patients had a combination of clinical, radiological, and microbiological features [32].

Statistical analysis

Categorical variables were expressed as counts (percentage). Differences in frequencies were compared using a X2 test or Fisher’s exact test. Continuous variables were expressed as median with 25-75th interquartile range (IQR), and the difference was assessed using Mann-Whitney-U test. Univariate analysis was performed to assess the association between risk factors and multi-species infection, and the results were presented as odds ratios (ORs) with 95% confidence intervals (95% CIs). All the analyses were done using STATA, version 13 for Windows.

Results

Specimen source and NTM species distribution

A total of 2026 NTM isolates from 852 patients were identified during the 5-year study period. The specimen types and detected species were shown in Tables 1 and 2 respectively. Respiratory specimens were predominant (1777, 87.7%). The top four prevalent species were M. abscessus-chelonae group (1010, 49.9%), M. fortuitum group (345, 17%), M. avium complex (MAC) (309, 15.3%), and M. kansasii (233, 11.5%).

Table 1 Sources of the specimens (N = 2026)
Table 2 NTM species distribution of all the investigated isolates (N = 2026)

NTM diseases

The study flowchart is shown in Fig. 1. Due to single sputum culture or the absence of characterized radiological changes as described in ATS/IDSA guideline, 387 patients with NTM isolate(s) from respiratory specimens (45.3%) were excluded from further analysis. Among the rest 465 patients (54.6%) who fulfilled ATS/IDSA diagnostic criteria, 460 patients had NTM isolated from one source, whereas 5 other patients had NTM isolated from ≥2 sources. The detailed disease distribution was illustrated in Fig. 2. Comorbidities of patients with NTM isolates from blood/other sterile sites mainly consisted of HIV/intro-venous drug abuse (11/49, 22.4%), malignancies (10/49, 20.4%), end stage renal failure (10/49, 20.4%), and ischemia heart disease /chronic heart failure (9/49, 18.4%). The species and isolates distribution of patients with NTM diseases were listed in Tables 3 and 4.

Fig. 1
figure 1

The study flowchart

Fig. 2
figure 2

The distribution of NTM diseases among the study subjects

Table 3 NTM species distribution among the patients diagnosed with NTM diseases (N = 460)
Table 4 Baseline characteristics of the patients with pulmonary NTM disease (N = 352)

NTM pulmonary diseases

Demographic information and clinical features

Of the 744 patients with NTM isolates from respiratory specimens, 352 patients (47.3%) met ATS/IDSA case criteria for pulmonary NTM disease. M. abscessus-chelonae group (197, 56%) was the predominately pathogenic species. The median age of the patients was 67 years (Table 4). Malignant diseases, COPD and bronchiectasis were the major comorbidities. Male patients were older, and more often had COPD; whereas female patients more often had bronchiectasis (Table 5).

Table 5 Comparison between male and female patients with NTM lung disease (N = 352)

Factors associated with multiple NTM species detection

Of the 352 patients with pulmonary NTM infection, 106 patients had ≥2 NTM species isolated either in the same setting (41/106) or at different time point (65/106). There were no species matching-pattern or sequential order observed. However, univariate analysis revealed that age ≥ 65 years and COPD were associated with multiple NTM species isolation with OR of 1.7 and 2.2 respectively (Table 6).

Table 6 Comparison of NTM lung disease characteristics between patients with 1 NTM species detected and patients with more than 1 NTM species detected

Discussion

In this study, we evaluated the clinical relevance of all NTM isolates cultured in CTBL in SGH from 2011 to 2015. We found that 54.6% of the patients with NTM isolates met ATS/IDSA diagnostic criteria of NTM diseases. Nearly half of the patients with pulmonary isolates may have active NTM lung infections, significantly higher than the rates reported in Hong Kong and Korea (47% vs. 20–25%) [20, 33]. This may suggest that pulmonary NTM isolates were more likely associated with pulmonary NTM infection rather than colonization among the patients in Singapore.

The distribution of NTM species varies geographically. M. avium complex (MAC) are the most commonly isolated and involved in pulmonary NTM infections in the US, Canada, Japan, Korea, Hong Kong and Taiwan [22, 32,33,34,35,36,37,38,39,40]. In contrast, our study showed that M. abscessus-chelonae group was the most prevalent and most implicated in pulmonary NTM disease among the patients in Singapore. Here we grouped M. chelonae and M. abscessus together as the two species could not be reliably differentiated by the line-probe assay (INNO-LiPA). However, an in-house review of isolates that were tested by both INNO-LiPA and HPLC revealed that all M. abscessus-chelonae group identified by INNO-LiPA were confirmed to be M. abscessus by HPLC method (Data not published). In view of this and the fact that M. chelonae rarely causes chronic lung disease [32], M. abscessus was likely to be the main pathogenic species in pulmonary NTM infections in our patient cohort. This observation was in line with another study in Singapore and a study conducted in southernmost region of Japan with subtropical climate [41, 42]. We postulated the higher prevalence of M. abscessus in our setting may be due to the tropical climate, densely populated urban living and history of TB infection. However, more studies are required to prove these hypotheses.

M. abscessues subsp. abscessus is often associated with poor clinical prognosis because of the possession of a functional erm [41] gene that confers macrolide resistance. M. abscessus subsp. massiliense, on the other hand, generally remains susceptible to macrolide due to the truncation of erm gene [43, 44]. Subspeciation of M. abscessus complex may provide important guidance in the clinical management of M. abscessus complex infection. Unfortunately, this data is not available during the study period as the test was not performed routinely. In addition, macrolide resistance is not routinely tested. If clinically indicated, the clinicians will order susceptibility testing for M abscessus and the microbroth dilution plate will be kept for a total of 14 days to detect inducible macrolide resistance. This is in accordance to CLSI (Clinical and Laboratory Standards Institute) M24-A2 [45].

In this study, M. fortuitum species group was detected in about 25% of the patients with pulmonary NTM diseases. The role of M. fortuitum group as the causative agent in infection can be controversial because it could be colonization or transient infection [46, 47]. However, some studies have reported that M. fortuitum is the pathogen implicated in pulmonary NTM infection and skin and bone/joint infection as well [32, 48,49,50]. Therefore, it should always be correlated with the clinical findings to determine its clinical significance and management. In addition, M. kansasii is a virulent but an uncommon species in Asia. However, it was surprisingly the fourth commonest NTM species implicated in pulmonary infection among the patients in Singapore [20, 22, 33]. Hence, it is important to understand the local epidemiology before applying guidelines elsewhere.

Structural lung diseases predispose people to pulmonary NTM infection [9, 14, 34, 51,52,53,54,55,56]. In this study, about 50% of patients with pulmonary NTM infection had COPD or bronchiectasis. Male patients had a higher incidence of COPD and more often presented with recurring cough and sputum production; in contrast, female patients had bronchiectasis and haemoptysis more frequently. These distinct features were consistent with the findings observed in other studies [34, 57, 58]. Pulmonary tuberculosis is another important pathophysiological process associated with severe pulmonary structural damage. A history of tuberculosis is a distinct characteristic among Asian patients with pulmonary NTM diseases [59]. In this study, one-third of the patients with pulmonary NTM infections had prior TB, which was not surprising as the incidence of TB in Singapore was about 307 cases per 100,000 population in 1960 [28]. By the way, the prevalence of pulmonary NTM disease was similar in male and female patients. Both male and female patients had slim figure with BMI close to the lower end of the normal range (18.9 kg/m2 and 18.6 kg/m2 respectively).

Isolation of multiple NTM species in the same setting or over a period of time is another feature of NTM lung infection and the underlying reason is unclear [60,61,62]. In this study, around 30% of the pulmonary NTM patients had ≥2 species identified. As clinical treatment data was not gathered in current study, we cannot tell if was related to clinical treatment as suggested by Lee et al. [63] However, we found that age ≥ 65 years and COPD were significantly associated with multispecies isolation with OR of 1.7 and 2.2 respectively. Chronic inflammation and remodelling of airways in COPD may play a role in the increased susceptibility to NTM infection.

This study has several limitations. Firstly, this is a retrospective study and the collected data may be incomplete. Secondly, the study was conducted in a single hospital not all adult tertiary-care hospitals in Singapore. Nonetheless, as patients from all over Singapore can seek care from any tertiary-care hospital in the country, any selection bias was likely to be minimal. Thirdly, there may be an underestimation in the prevalence of pulmonary NTM infections as approximately half of the cases with single NTM isolate from sputum or had no characteristic radiological findings were excluded from the analysis based on ATS/IDSA guideline.

Conclusions

About half of the patients in Singapore with NTM isolates met ATS/IDAS criteria of NTM diseases with highest prevalence of pulmonary NTM disease. Of the 16 NTM species (group) detected in this study, M. abscessus-chelonae group was the most frequently isolated and most involved in pulmonary NTM diseases. The patients with pulmonary NTM disease were characterized with advanced age, having history of TB and other chronic illnesses such as malignant diseases and structural lung diseases. Male patients were more associated with COPD, whereas female patients more often had bronchiectasis. Increasing age and COPD were associated with multiple NTM species detection per patient.