There is growing recognition of the clinical importance of non-tuberculous mycobacteria (NTM), a group of versatile opportunistic bacterial pathogens. We describe the characteristics of NTM isolates in Scotland over an 11-year period using data held by the Scottish Mycobacteria Reference Laboratory. American Thoracic Society microbiological criteria were used to evaluate the clinical significance of isolates. Data presented include analysis of trends across time, species/body site associations, gender and age differences, geographical variations and the association between cystic fibrosis and Mycobacterium abscessus. We emphasise the need for standardised reporting criteria for NTM isolates to ensure optimal surveillance of NTM disease.
- Atypical Mycobacterial Infection
- Respiratory Infection
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
The non-tuberculous mycobacteria (NTM) are a group of opportunistic bacterial pathogens numbering >120 species. Infection is acquired from environmental reservoirs and usually manifests as pulmonary disease, although lymphadenitis, soft tissue infection and bacteraemia may also result.
Disease caused by NTM is not generally notifiable and, as a result, its epidemiology is not well described. The ubiquity of NTM in the environment makes contamination of samples a problem. In addition, distinguishing pulmonary colonisation from infection can be difficult. Recent work describes the epidemiology of NTM in England, Wales and Northern Ireland; however, such data are currently lacking for Scotland.1 In this paper we use data held by the Scottish Mycobacteria Reference Laboratory (SMRL) to describe the characteristics of NTM isolates in Scotland over an 11-year period.
Scottish hospital laboratories submit mycobacterial isolates to the SMRL for speciation, susceptibility testing and typing. Patient age, sex and sample details are routinely included with submitted isolates. Submitted clinical, radiological and HIV information is variable, although patients with cystic fibrosis (CF) are reliably identified. We analysed data from January 2000 to December 2010 inclusive.
We used the ‘microbiological’ criteria of the American Thoracic Society 2007 NTM guidelines to estimate the clinical significance of pulmonary isolates. These were therefore only included in the analysis if a patient had ≥2 positive sputum cultures of the same organism on separate days or a single positive culture from bronchial specimens. In the case of multiple pulmonary isolates of the same species, all of those within 12 months were regarded as representing the same episode of infection. All subsequent isolates within 12 months of the first were not considered further. Samples which grew Mycobacterium gordonae were excluded, given the high likelihood that these represented contamination or colonisation. M gordonae is usually a non-pathogenic coloniser, even in immunocompromised patients.2 For extrapulmonary sites, a single positive isolate was considered significant as these were generally from biopsies or sterile sites.
Analysis of data
When calculating population rates, population estimates and NHS Health Board sizes were obtained from the General Register Office for Scotland. A χ2 test was used to analyse the relationship between species and sample details.
Between 2000 and 2010 a total of 4193 positive NTM isolates from humans were recorded. There were 2176 clinically significant isolates from 1370 episodes of infection, with 2017 representing contamination or colonisation (see online supplementary figure S1); 806 were positive pulmonary isolates from within 12 months of a previous one from the same patient so were considered to represent the same episode of infection. Specimens came from 931 patients.
Trends across time
The rate of all NTM episodes varied year by year and no clear trend was seen (see online supplementary figure S2). A mean rate of 2.43 episodes per 100 000 of the population (range 2.06–2.71) was found. Mycobacterium avium complex was most frequently identified, accounting for 48.1% (n=659) of all episodes with a mean rate of 1.18 episodes per 100 000 each year. A 45% increase in the incidence of M avium complex was seen over the 11 years. Table 1 shows the most common species, sites and gender ratios of episodes.
Mean rates of NTM episodes for males and females were 2.53 and 2.33 per 100 000, respectively, with a male:female ratio of 1.08 : 1 (table 1). A ratio of 1.09 was seen when pulmonary episodes were considered alone. Mycobacterium abscessus episodes were more common in females (male:female ratio 0.7, p=0.004).
Site and mycobacterial species
The majority of NTM episodes were pulmonary (68.1%, n=933) (see online supplementary figures S3 and S4). These 933 pulmonary episodes occurred in 557 patients, with 209 patients having more than one episode of infection during the study period (94% with the same species, representing either relapse or reinfection) and an average of 1.6 episodes per patient. Renal, adenitis, bloodstream and cutaneous episodes contributed 8.7% (n=119), 6.7% (n=92) 3.4% (n=46) and 3.9% (n=53) of the total, respectively. M avium complex were the most frequently isolated organisms from pulmonary (44.8%), adenitis (75%) and bloodstream episodes (67.4%). Mycobacterium marinum was the most commonly isolated species from cutaneous episodes and was associated with this site (32.1% of episodes, p<0.0001); 81% of M marinum episodes came from cutaneous specimens. Mycobacterium chelonae accounted for 28.3% of cutaneous episodes and 15.2% of bloodstream episodes, being associated with both sites (p<0.0001 and p=0.0022, respectively). Mycobacterium mucogenicum was associated with bloodstream episodes (causing 10.9%, p<0.0001), with 45.5% of M mucogenicum episodes grown from blood cultures.
The majority of all NTM episodes came from patients over the age of 50 (64.0%, n=875), and an even higher proportion of pulmonary episodes were from this age group (72.2%, n=672) (see online supplementary figure S5). However, 73.9% (n=68) of adenitis episodes came from children aged up to 5 years. Mycobacteraemias peak at 35 years (20%, n=46), being rare in later life. In contrast to other species, the incidence of M abscessus peaks between the ages of 11 and 25 years, with only 33.6% of episodes occurring in patients aged ≥25 years, probably due to patients with CF (see below).
Greater Glasgow and Clyde and Lothian Health Boards had average rates of NTM of 3.25 and 3.63 episodes per 100 000 Health Board population, respectively, which were consistently higher than the national average (ratios of 1.34 and 1.49, respectively, compared with the national average for the study period). From 2007 the Borders Health Board also reported rates above the national average (average of 4.64 episodes per 100 000 Health Board population; ratio 1.91). There were no geographical trends in the distribution of NTM species.
A total of 129 episodes came from patients with CF, and all of these were from the pulmonary site (see online supplementary figure S6). The mean age of the patients was 22 years and the median age was 19 years (range 6–52 years), with 63% (n=81) coming from women. CF was associated with the isolation of M abscessus (p<0.0001), which was isolated in 68.2% (n=88) of CF episodes compared with 5.0% (n=40) of pulmonary episodes from patients without CF; 75% (n=66) of M abscessus episodes in patients with CF occurred in females. In addition, 68.75% of our M abscessus pulmonary episodes occurred in patients with CF.
These data gathered over an 11-year period are the first to describe the epidemiology of NTM across the whole of Scotland. A previous retrospective study in south-east Scotland suggests that up to 75% of NTM isolates reported were associated with episodes of clinical disease, whereas this larger nationwide study found that figure to be 52%.3 The previous study included a review of available case notes, which the current report does not.
We found a mean annual rate of 2.43 NTM episodes per 100 000 of the population, with no clear trend visible between 2000 and 2010. Moore et al described similar data from England, Wales and Northern Ireland during the period 1995–2006. They found a 250% increase in NTM reports, from 0.9 per 100 000 in 1995 to 2.9 in 2006.1 This increase may have occurred earlier in Scotland, and may be related to different laboratory methods used in different centres at different times or heterogeneous case definitions.
Consistent with the work of others, we found the majority of episodes were pulmonary, M avium complex was the most common species and episodes were more likely to come from older patients. The clinical significance of NTM found in urine is unclear and may frequently represent urethral colonisation. M marinum is primarily a fish pathogen and, following contamination of trivial skin lesions, can cause skin and soft tissue infections in man (‘fish-tank’ granuloma). Indeed, we found M marinum was associated with cutaneous infection, and this was the species most often isolated from this site. M chelonae usually causes localised infection but may spread haematogenously, often in immunocompromised patients. Consistent with this, our study found M chelonae to be associated with both cutaneous and bloodstream episodes. There are some reports in the literature of M mucogenicum bloodstream infections, 4 and our study confirms the association of M mucogenicum with NTM bacteraemia.
Approximately 13% of patients with CF have NTM isolated from pulmonary specimens.5 We found an association between CF and the isolation of M abscessus, and this association has been described previously in both American and French cohorts.
Variable referral and reporting of isolates is a significant limitation of this work. We know that some laboratories submit all isolates to the SMRL whereas others use local experience to only submit ‘first patient isolates’ over a variable period. This practice not only confounds laboratory-based surveillance but relies on local expertise to differentiate between species from the same patient. This may occasionally be dangerous if two species are present in the same patient, especially if Mycobacterium tuberculosis complex is included. We believe that standardised referral and reporting criteria for isolates as well as clinical disease are required to ensure optimal patient care and surveillance of NTM disease. Relying on laboratory data alone is insufficient, and our data therefore probably underestimate the true incidence of NTM disease. Only with mandatory reporting of full clinical data can disease caused by NTM be clearly evaluated.
We would like to thank referring clinicians and laboratories in Scotland for the isolates and information referred to SMRL. We also thank SMRL colleagues for their support in this work. Culture reports are submitted to HPS and referring laboratories.
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Data supplement 1 - Online figures
Contributors PC, CD, A-LS, AR and IFL collected the data as part of SMRL activities and contributed to planning this retrospective review of the data. CDR and IFL performed the data analysis and wrote the manuscript. All authors then reviewed and approved the manuscript for publication. CDR and IFL assume overall responsibility for the content.
Funding This research received no specific funding. SMRL is funded via Health Protection Scotland by National Services Scotland and this work is part of its remit.
Competing interests None.
Provenance and peer review Not commissioned; internally peer reviewed.