Mycobacterium avium

Associated infections

• Nodular-bronchiectatic and fibrocavitary pulmonary disease
• Disseminated MAC infection in advanced HIV and other severe immunosuppression
• Cervical lymphadenitis, particularly in young children
• Hypersensitivity pneumonitis associated with aerosolised water exposure
• Skin, soft tissue and bursal infection following inoculation
• Catheter-related and prosthetic device infection
• Pseudo-infection following contamination of respiratory specimens or bronchoscopes

Transmission route

M. avium is acquired from the environment, principally by inhalation of aerosols generated from colonised water at showers, taps and spa pools, and by ingestion or inoculation. Person-to-person spread is not a recognised feature of M. avium, and the organism is not transmitted by the respiratory route between patients. Its persistence depends on biofilm: cells attach to pipe and fixture surfaces, survive within free-living amoebae, and tolerate chlorine and chloramine concentrations that eliminate enteric bacteria, so the organism is common in treated, compliant potable supplies.

Exposure intensity is governed less by the concentration in bulk water than by the opportunity for aerosolisation and by conditions in distal plumbing. Warm water, stagnation, long runs of small-bore pipe, dead legs from decommissioned outlets, and low or absent disinfectant residual at the point of use all favour regrowth, and the amplification that matters happens in the last few metres of pipework rather than in the main. Hypersensitivity pneumonitis associated with heavily aerosolised water exposure represents an immunological response to inhaled antigen rather than invasive infection, and further illustrates that the aerosol route is the dominant one for this organism.

This makes M. avium directly relevant to endoscope reprocessing. Any final rinse drawn from building plumbing may carry MAC unless the water treatment train and terminal bacterial-retentive filtration are functioning correctly. Because reprocessing water contacts the interior channels of a device immediately after high-level disinfection, contamination at this step recontaminates the endoscope and can seed biofilm within channels and within the reprocessor itself, where the organism is protected from routine chemical decontamination. Bronchoscopes present the greatest clinical concern, since MAC recovered from a bronchial specimen after a contaminated procedure may be misinterpreted as genuine pulmonary infection and lead to prolonged, toxic multi-drug therapy. Detection of M. avium in final rinse water should be escalated immediately, with quarantine of affected instruments and investigation of the water system, filtration and reprocessor rather than repeat sampling alone.

In central sterile services the same water supply feeds instrument rinse stages, and in dental practice it feeds unit waterlines whose narrow bore, low flow and long idle periods make them an efficient mycobacterial habitat. Neither setting is exempt from the reasoning that applies to endoscopy: the organism is in the plumbing unless a validated barrier is excluding it.

Relevance in endoscopy and reprocessing

M. avium is the non-tuberculous mycobacterium with the clearest documented capacity to cause serious human disease and the best-characterised reservoir in building plumbing, which together make it the highest-consequence finding in reprocessing water monitoring. The pathway of concern is specific and short. Water enters the reprocessor from the facility supply; if the treatment train and terminal filter do not exclude MAC, the final rinse delivers viable organisms into every channel of an instrument that has just completed a valid high-level disinfection cycle. The disinfection step is not the failure point. Glutaraldehyde, ortho-phthalaldehyde and peracetic acid are mycobactericidal under validated conditions and will inactivate M. avium on a clean surface. The failure is that the last thing to touch the device is water that was never made safe.

Biofilm is what converts that event into an ongoing problem. M. avium attaches readily to the polymer and metal surfaces of rinse lines, tanks, connectors and endoscope channels, and within a matrix its susceptibility falls substantially below that of the same organism in suspension. A cycle validated against planktonic cells cannot be assumed to clear an established biofilm, and biofilm sheltered in a scratch, a crack or a delaminated channel lining is effectively beyond the reach of any chemistry. This is the basis for borescope inspection of instruments with a persistent positive history and for withdrawing instruments where internal damage is found: reprocessing an instrument with a compromised channel simply repeats the cycle around a protected reservoir. It is also why rinse water quality is treated as a control point in its own right rather than as an adjunct to disinfection — it is the mechanism by which those reservoirs are seeded.

Drying and storage determine the size of the inoculum a patient meets. Residual water in a channel is the growth medium; remove it and even a contaminated rinse leaves relatively little behind, retain it and a small number of cells becomes a large number over a weekend. Alcohol flush followed by forced-air drying that genuinely perfuses each channel, storage in a manner that keeps channels dry, and drying cabinets supplied with filtered air and verified for channel perfusion are the relevant controls. Cabinets that dry the exterior while leaving lumens wet give false assurance. For CSD, the equivalent risk is lumened instruments retaining rinse water before packaging — terminal sterilisation will kill the organism but will not remove deposited biofilm residue or endotoxin. For dental units, overnight stagnation in warm narrow-bore tubing is close to an ideal MAC habitat, and independent reservoirs, routine line disinfection and point-of-use filtration are the practical answers.

Interpreting a detection

Recovery of M. avium from a final rinse or reprocessing water sample is a critical finding and should be escalated on the day it is reported. Operationally it means that MAC has traversed every barrier between the facility supply and the instrument. The immediate checks are the terminal bacterial-retentive filter — rating, service life, installation date, housing seal integrity, evidence of drying and rewetting, and whether any bypass path exists — followed by the treatment train upstream, where carbon beds, softeners, reverse osmosis membranes, ultraviolet units and storage vessels each provide surface area for biofilm and each remove the residual disinfectant that suppressed growth in the incoming supply. The reprocessor is then examined internally: rinse water lines, tanks, dosing lines, non-return valves, and the record of self-disinfection cycles including whether specified temperatures were reached and held. Sampling technique is reviewed but should not be the first explanation reached for, since a slow-growing mycobacterium recovered on selective media after weeks of incubation is unlikely to be a laboratory artefact.

A single isolate warrants full escalation. Given the clinical consequences of MAC exposure through a bronchoscope, the appropriate posture on a first positive is to treat it as real: identify and quarantine instruments processed since the last satisfactory result, review the procedure log for bronchoscopies performed in that window, take a defined remediation action, and resample only afterwards so the repeat result tests the fix. Repeat positives, positives at more than one sample point, or a positive recurring after filter replacement and line disinfection indicate established biofilm in a component that cannot be cleaned in situ, and the response should move from cleaning to replacement of hose, housings, storage vessels or reprocessor plumbing. Results should be read alongside heterotrophic plate counts, Pseudomonas detections and any trend in filter differential pressure, because all track the same loss of hydraulic and microbiological control.

The most important surveillance point is that routine total viable counts will never detect this organism. TVC methods use general-purpose media, short incubation and temperatures optimised for fast-growing waterborne bacteria; M. avium grows over weeks and requires sample decontamination to suppress competitors plus selective mycobacterial media. NTM culture is a separately specified test, and a service that has not requested it has no data on mycobacteria regardless of how many satisfactory TVC results it holds. Escalation should convene infection prevention, the water safety group and facility engineering together, because remediation is engineering work. Where a bronchoscope is implicated, patient look-back is appropriate and molecular typing comparing patient and water isolates should be considered — without it, MAC recovered from a bronchial specimen after a contaminated procedure is likely to be treated as genuine pulmonary disease, committing the patient to a year or more of multi-drug therapy for an infection they do not have.

Antimicrobial resistance

M. avium is intrinsically resistant to standard anti-tuberculosis therapy and requires prolonged multi-drug treatment. Current ATS/ERS/ESCMID/IDSA guidance recommends susceptibility-based treatment for macrolides and amikacin, and a three-drug macrolide-containing regimen for macrolide-susceptible disease, specifically to prevent emergence of macrolide resistance — the single most important determinant of treatment failure and mortality in MAC pulmonary disease. Macrolide monotherapy, or the addition of a macrolide to a failing regimen, selects for 23S rRNA mutants and should be avoided. Treatment is measured in months to years, adherence is difficult, and adverse effects are common, which is why an unnecessary course initiated on the strength of a contaminated specimen represents a genuine harm rather than a merely academic error.

Environmentally, M. avium tolerates chlorine-based disinfection at potable concentrations, and biofilm-associated and amoeba-internalised cells are more resistant again. The underlying mechanism is the mycolic-acid cell wall — thick, hydrophobic and poorly permeable to hydrophilic oxidants — which slows disinfectant access to intracellular targets and simultaneously confers tolerance to desiccation, to low-nutrient conditions and to low-level disinfectants generally. Survival within free-living amoebae adds a further barrier, since the amoebal cell and cyst wall shield ingested mycobacteria from concentrations that would kill free cells outright. Growth at moderately elevated temperatures also means that hot water services set at reduced temperatures for scald prevention can act as amplifiers rather than as controls.

The operational conclusion is that indicator-based water quality assurance does not apply to this organism. A satisfactory disinfectant residual and absence of indicator organisms do not demonstrate control of MAC in a reprocessing water supply. Assurance rests on physical exclusion at the point of use through correctly specified, correctly installed and in-date bacterial-retentive filtration, on distribution design that avoids stagnation and dead legs, on temperature control in hot water services, and on verification by mycobacteria-specific culture rather than by coliform or plate-count testing.