Mycobacterium spp.

Associated infections

• Pulmonary disease, particularly M. avium complex and M. abscessus infection
• Skin, soft tissue and surgical wound infection
• Catheter-related bloodstream infection
• Post-procedural and post-injection abscess
• Disseminated infection in immunocompromised patients
• Lymphadenitis in children
• Post-surgical infection following cardiothoracic and cosmetic procedures
• Keratitis following ocular surgery or trauma

Transmission route

Non-tuberculous mycobacteria are acquired from environmental sources rather than by person-to-person spread, principally through inhalation of aerosols, direct inoculation of tissue during a procedure, or contact between a colonised device and a mucosal or sterile site. Healthcare-associated cases have been linked to contaminated water used in clinical procedures and to inadequately reprocessed devices. Because acquisition is environmental, cases connected to a single water reservoir may be separated by long intervals and may not present as a recognisable cluster; the slow onset of NTM disease, which can follow exposure by weeks to months, further obscures the link between an exposure and a presentation.

The aerosol route is the most important for respiratory disease. Any process that atomises water containing NTM, including showers, spray outlets, humidifiers, cooling systems and air-driven drying equipment, generates droplet nuclei in the size range capable of reaching the distal airway, and the hydrophobic mycobacterial surface causes preferential enrichment in the aerosol phase relative to the bulk water. Direct inoculation accounts for the second major group of presentations, in which water or a water-wetted instrument contacts a surgical site, an injection site or a mucosal surface. Rapidly growing species predominate in this group.

In endoscopy, both true infection and pseudo-infection are recognised. Pseudo-outbreaks, in which mycobacteria recovered from patient specimens originate from the device or its rinse water rather than from the patient, have been reported repeatedly and are the more common presentation. A pseudo-outbreak of M. chelonae together with Methylobacterium mesophilicum was traced to contamination of an automated endoscopy washer, and pseudo-outbreaks involving M. avium complex, M. fortuitum and M. abscessus have been described in bronchoscopy units, in several instances attributable to contaminated reprocessor water lines, filters or the incoming supply. The clinical cost of a pseudo-outbreak is not trivial: patients may receive prolonged multi-agent antimycobacterial therapy, undergo further invasive investigation, or have an unrelated diagnosis delayed.

Outside endoscopy, the same reservoirs are relevant to CSD and dental water. NTM are among the organisms consistently recovered from untreated dental unit waterlines, where the small-bore, low-flow, intermittently stagnant tubing supports dense biofilm and where treatment water is both directly applied to oral tissue and extensively aerosolised. In CSD, mycobacteria in rinse water are of concern for devices that receive high-level disinfection rather than terminal steam sterilisation, since steam sterilisation is fully effective against them.

Relevance in endoscopy and reprocessing

The relevance of NTM to flexible endoscope reprocessing is well established and rests on two properties. First, the mycobacterial cell wall confers substantially greater tolerance of aldehyde and oxidising disinfectants than that of vegetative Gram-negative bacteria; glutaraldehyde-tolerant strains of M. chelonae have been isolated directly from endoscope washer-disinfectors. Mycobactericidal activity is consequently one of the defining performance requirements for a high-level disinfectant, and a process validated only against vegetative bacteria cannot be assumed adequate. Second, rinse water applied after high-level disinfection is not subsequently disinfected, so any mycobacteria present recontaminate the device at the final stage of the cycle. Environmental mycobacteria are for this reason included as a named indicator in final rinse water specifications for flexible endoscope reprocessing under AS 5369:2023, with absence required in a 100 mL sample.

Biofilm behaviour compounds both problems. NTM incorporate readily into mixed biofilm within reprocessor water lines, filter housings, storage vessels and distribution pipework, and their hydrophobic surface favours firm attachment to the polymeric materials used throughout these circuits. Once established, they are protected both by the matrix and, in many systems, within amoebal hosts, and chemical sanitisation regimes that suppress Gram-negative counts may leave the mycobacterial population substantially intact. Published bronchoscopy pseudo-outbreaks illustrate the practical pattern: in one, resolution required increasing the frequency of reprocessor filter changes from quarterly to monthly; in others, the incoming hospital water supply itself was shown to carry the outbreak species, so machine-level remediation alone was insufficient.

Drying and storage are decisive controls. Residual moisture in endoscope channels permits mycobacterial persistence, and the hydrophobic cell surface means NTM adhere to channel walls and are not readily displaced by flushing alone. Verified forced-air drying of every channel, functioning drying cabinets and defined storage limits are the practical measures. Bronchoscopes warrant particular attention because the respiratory tract is the site at which NTM most readily establish disease, because bronchoscopy specimens are the specimens most often affected by pseudo-outbreaks, and because bronchoscope channels are narrow and difficult to dry. In CSD and dental settings, NTM in the water system indicate a persistent biofilm reservoir; for dental units this is directly relevant because treatment water is aerosolised and applied to oral tissue, and for CSD it is relevant chiefly for devices not terminally steam sterilised.

Interpreting a detection

A detection of environmental mycobacteria in final rinse water is a significant finding and is not readily explained as a sampling artefact. NTM are not common incidental laboratory contaminants of water cultures, and their recovery requires specific media and extended incubation, so a positive result generally means the laboratory looked for them deliberately and found them. Under the Healthcare Infection Society action framework, the presence of an organism of significance in final rinse water requires that no endoscopes be reprocessed in the affected washer-disinfector until satisfactory results are obtained; this applies on a single isolate and does not require corroboration by a trend. Conversely, a compliant heterotrophic plate count is not evidence of mycobacterial absence, because the short incubation and general-purpose media used for total viable counts will not recover them. Where NTM are a specific concern, they must be requested as a separate test.

The investigation should establish the extent of the reservoir rather than stopping at the failing outlet. Sampling should extend to the incoming supply, post-treatment product water, storage, the distribution loop at near and far points, the connecting hose, the reprocessor water path and the rinse-water filter, using mycobacterial culture at each point. Because NTM concentrate in biofilm rather than in bulk water, surface or swab sampling of accessible wetted components, and sampling of the first draw after a period of stagnation, are more informative than a flushed bulk sample. Points to review include filter type, integrity and change frequency, since filter housings are a recognised site of mycobacterial accumulation and an extended change interval has been the identified root cause in reported pseudo-outbreaks; loop temperature and stagnation; dead legs; storage vessel turnover; the sanitisation agent, concentration and contact time and whether it has demonstrated mycobactericidal activity; and the compatibility of the high-level disinfectant in use with a mycobacterial challenge.

Escalation should occur on first detection. In addition to withdrawing the affected reprocessor, the investigation should consider whether any recent unexplained mycobacterial isolates from patient specimens processed through the affected pathway might represent pseudo-infection, since recognising a pseudo-outbreak promptly prevents unnecessary and prolonged patient treatment. Liaison with the clinical microbiology laboratory to compare environmental and patient isolates, including by molecular typing where available, is the definitive step and should be initiated early. Verification of remediation requires repeat mycobacterial culture rather than total viable count alone, and should be repeated after an interval sufficient for regrowth to become detectable, because the slow growth of these organisms makes a single early clear result particularly unreliable. Where the incoming supply is implicated, remediation must address the building water system and not only the reprocessing equipment.

Antimicrobial resistance

Non-tuberculous mycobacteria are intrinsically resistant to most conventional antibacterial agents as a consequence of their impermeable cell envelope and efflux activity, and treatment typically requires prolonged multi-agent regimens guided by species-level identification and susceptibility testing. Therapy commonly extends over many months and carries substantial toxicity, which is the reason a pseudo-outbreak has real clinical consequences even in the absence of true infection. M. abscessus is particularly refractory, with inducible macrolide resistance mediated by the erm(41) gene in many isolates, meaning that an isolate reported as macrolide-susceptible on initial testing may fail clinically unless extended incubation is used to detect inducible resistance.

Of specific importance to reprocessing, resistance to chemical disinfectants is a separate phenomenon from antibiotic resistance and is governed by different mechanisms. The mycolic acid-rich, highly hydrophobic cell wall restricts the entry of aqueous biocides, and this intrinsic barrier can be further enhanced in strains adapted to repeated sublethal exposure. Strains showing marked tolerance of 2% alkaline glutaraldehyde have been recovered from endoscope washer-disinfectors, while alternative agents such as ortho-phthalaldehyde and peracetic acid have retained mycobactericidal activity against those strains. This has a direct operational implication: where a glutaraldehyde-based process is in use and mycobacteria are recovered from the reprocessing environment, the choice of disinfectant chemistry should form part of the review, not only the water supply.

Biofilm and amoebal association add a third layer of protection that is independent of both antibiotic and intrinsic biocide resistance. Mycobacteria residing within amoebal cysts or deep within a mixed biofilm may survive sanitant concentrations that would be lethal to the same organism in suspension, and may emerge with enhanced tolerance following amoebal passage. Remediation of a mycobacterial water system problem therefore has to address the physical biofilm reservoir, not only the choice of chemical agent.