Characteristics of Nontuberculous Mycobacteria from a Municipal Water Distribution System and Their Relevance to Human Infections
Total Page:16
File Type:pdf, Size:1020Kb
CHARACTERISTICS OF NONTUBERCULOUS MYCOBACTERIA FROM A MUNICIPAL WATER DISTRIBUTION SYSTEM AND THEIR RELEVANCE TO HUMAN INFECTIONS. Rachel Thomson MBBS FRACP Grad Dip (Clin Epi) A thesis submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy School of Biomedical Sciences Faculty of Health Queensland University of Technology 2013 Principal Supervisor: Adjunct Assoc Prof Megan Hargreaves (QUT) Associate Supervisors: Assoc Prof Flavia Huygens (QUT) i ii KEYWORDS Nontuberculous mycobacteria Water Distribution systems Biofilm Aerosols Genotyping Environmental organisms Rep-PCR iii iv ABSTRACT Nontuberculous mycobacteria (NTM) are environmental organisms associated with pulmonary and soft tissue infections in humans, and a variety of diseases in animals. There are over 150 different species of NTM; not all have been associated with disease. In Queensland, M. intracellulare predominates, followed by M. avium, M. abscessus, M. kansasii, and M. fortuitum as the most common species associated with lung disease. M. ulcerans, M. marinum, M. fortuitum and M. abscessus are the most common associated with soft tissue (both community acquired and nosocomial) infections. The environmental source of these pathogens has not been well defined. There is some evidence that water (either naturally occurring water sources or treated water for human consumption) may be a source of pathogenic NTM. The aims of this investigation were to 1) document the species of NTM that are resident in the Brisbane municipal water distribution system, then 2) to compare the strains of NTM found in water, with those found in human clinical samples collected from Queensland patients. This would then help to prove or disprove whether treated water is likely to be a source of pathogenic strains of NTM for at risk patients. The third aim was to investigate whether shower aerosols contained NTM in respirable particles. As there are difficulties identifying NTM in water samples using culture methods, a pilot study was performed using spiked samples of sterile and tap water, comparing selected methodological variables. These included centrifugation vs filtration for concentration of samples, the addition of Sodium thiosulphate (routinely used to neutralize chlorine), the addition of Cetylpiridium chloride (CPC) for decontamination, different culture media – including Löwenstein-Jensen (L-J), Middlebrook 7H10 and 7H11, and MGIT Bactec with and without PANTA, and incubation temperature (32°C/35°C/32°C with carbon dioxide (CO2)). The results of this study revealed that sodium thiosulphate was not necessary and might inhibit mycobacterial growth, also that filtration was more effective than centrifugation, v Middlebrook media were favoured over L-J, and there was no difference in growth at different incubation temperatures. The addition of CPC 0.005% decreased contamination rates, but also significantly reduced the mycobacterial numbers. From these results a study was designed to test Brisbane water distribution samples for NTM. This is the first Australian study of its kind and one of the largest internationally. Multiple pathogenic and nonpathogenic NTM were identified in 40.2% of sites sampled in summer and 82% sites in winter. The summer samples were marred by bacterial contamination and overgrowth. Factors associated with the isolation of NTM included site variables such as distance from water treatment plants, elevation of sample site above sea level, and certain pipe materials. The species documented to cause disease in humans residing in Brisbane that were also found in water include M. gordonae, M. kansasii, M. abscessus, M. chelonae, M. fortuitum complex, M. intracellulare, M. avium complex, M. flavescens, M. interjectum, M. lentiflavum, M. mucogenicum, M. simiae, M. szulgai, M. terrae. M. kansasii was frequently isolated, but M. avium and M. intracellulare (the main pathogens responsible for disease in QLD) were isolated infrequently. The main pathogenic species identified in water included M. lentiflavum, M. abscessus, M. kansasii and M. fortuitum. Very few studies have documented M. abscessus in water. The incidence of disease due to this pathogen is increasing in many parts of the world, including QLD. It is particularly difficult to treat, and a significant cause of morbidity and mortality. Patient isolates of M. abscessus referred to the Mycobacterium Reference Laboratory during the same time period as the water sampling, were compared to the water isolates using repetitive unit PCR (Diversilab). Analysis of the water isolates revealed 5 different clusters (Water Patterns (WP) 1-5). Groups of patient isolates clustered with each of these patterns except WP3. A smaller number of patient isolates were unrelated to the water isolates. A high degree of similarity between patient and water isolates, suggests that water may be the source of infection for these patients. This has important implications for those patients at risk for these infections, and for the risk of reinfection following treatment. vi Similarity between water and patient isolates was also demonstrated for the less common slow growing M. lentiflavum. This species can cause a variety of infections. Four cases of significant disease due to this pathogen are reported and the international literature summarised in parallel to the molecular epidemiology of water and clinical strains. In contrast, M. kansasii was frequently isolated from water samples and has been lablelled the ‘tap water bacillus’. Strain types of M. kansasii have previously been typed according to ITS_RFLP patterns as groups I-VI, with group I the most common pathogenic strain in humans in most parts of the world including Brisbane. Very few water isolates were Type I. Types IV and V predominated in water, but were less commonly associated with human infections. Rep-PCR proved to be a more sensitive strain-typing tool for analysis of these isolates, though clusters correlated with ITS_RFLP groups. These results suggest that there may be an alternative source of Type I strains that cause human infection. Despite the high prevalence of other strain types, very few cases of human disease can be attributed to these strains. Hence municipal water may not be the source of infection in most instances. Strains of M. fortuitum associated with both nosocomial and community acquired infections were compared with municipal water isolates. There was significant strain type heterogeneity, and the clinical and water isolates were diverse. It is postulated then that if water is the vehicle for transmission of nosocomial infections, that point source contamination of water sources within hospitals likely occurs. It is also unlikely that community acquired infections result from exposure to municipal water, and perhaps soil is a more likely source of infection. To determine whether household water exposure including showering was a risk factor for the acquisition of infection, extensive home sampling was performed on the homes of 20 patients with NTM disease. Mycobacteria isolated from clinical samples from 20 patients included M. avium (5), M. intracellulare (12), M. abscessus (7), M. gordonae (1), M. lentiflavum (1), M. fortuitum (1), M. peregrinum (1), M. vii chelonae (1), M. triplex (1) and M. kansasii (1). Species identified in the home matched that found in the patient in seven (35%) cases – M. abscessus (3), M. avium (1), M. gordonae (1), M. lentiflavum (1), and M. kansasii (1). In an additional patient with M. abscessus infection, this species was isolated from the potable water supply in the vicinity of her home. NTM species grown from aerosols included M. abscessus (3 homes), M. gordonae (2), M. kansasii (1), M. fortuitum complex (4), M. mucogenicum (1) and M. wolinskyi (1). The evidence that patients acquire NTM from their homes appears species specific for M. avium, M. kansasii, M. lentiflavum and M. abscessus. Despite a predominance of disease due to M. intracellulare I found no evidence for acquisition of infection from household water for this species. In summary, the method for isolation of Mycobacteria from water has been further refined, contributing to the first large scale sampling of the Brisbane Municipal water distribution system. A wide variety of NTM species were found throughout the system, including species known to cause disease in QLD residents. However the strains of these pathogenic NTM were not always the same as the environmental strains. A detailed sampling of NTM patient homes, firstly demonstrated that NTM in water could be aerosolised during showering to a respirable particle size (previously not done). Secondly the finding of strains/species causing human infection within patient homes appears to be species specific. The combined findings of strain comparisons of city wide and patient home sampling, suggests that patients are at risk of infection due to M. abscessus, M. lentiflavum, M. avium and M. kansasii from exposure to municipal water sources and showers. However the divergent findings for M. kansasii and heterogeneity between human and water strains of M. fortuitum suggest there is an alternative environnmental niche for these pathogens. Similarly the striking paucity of M. intracellulare in Brisbane water and patient homes, in an area where disease due to this pathogen predominates, would support the proposal that patients acquire infection with this pathogen from soil/dust/other exposures. viii ix x TABLE OF CONTENTS KEYWORDS .........................................................................................................