Mycobacterium gordonae

Associated infections

• Pseudo-infection and specimen pseudo-outbreaks (the predominant scenario)
• Disseminated infection in severely immunocompromised patients
• Prosthetic device, catheter and implant-associated infection
• Rare pulmonary infection in patients with underlying lung disease
• Rare peritonitis in patients receiving peritoneal dialysis
• Rare keratitis and cutaneous infection following inoculation

Transmission route

Acquisition is environmental, from water and water-derived aerosols; person-to-person transmission is not described. In healthcare settings, the dominant mechanism is not infection but contamination — of clinical specimens rinsed or gargled with tap water, of laboratory reagents, and of endoscopes and bronchoscopes exposed to non-sterile water after high-level disinfection. A French investigation linked M. gordonae isolates from five patients to heavily contaminated refrigerated drinking fountains, with pulsed-field gel electrophoresis matching patient and fountain strains; four patients received unnecessary anti-tuberculous therapy before the species was identified. This illustrates the principal clinical harm of the organism: diagnostic confusion, unwarranted treatment with hepatotoxic drugs, and extended admission.

The contamination pathways are worth naming individually because each has a specific control. Patients asked to gargle or rinse with tap water before sputum collection introduce the organism into the specimen. Ice from a colonised machine used to chill or transport specimens does the same. Laboratory water used to prepare reagents or to rinse glassware carries it into the processing stream. Tap water used to irrigate during a procedure, or to flush a bronchoscope suction channel, delivers it directly to the sampling site. In each case the organism reaching the culture plate never came from the patient, and in each case the fix is to remove non-sterile water from a step where sterile or filtered water is required.

For endoscope reprocessing, M. gordonae is highly relevant. It colonises the same plumbing biofilm that supplies washer-disinfectors, survives chlorinated water, and can breach or bypass terminal filtration if filters are incorrectly rated, incorrectly fitted or beyond service life. Recovery from final rinse water should be treated as a genuine reprocessing water failure and not dismissed on the basis of the organism's low virulence. Under AS/NZS 5369 the requirement applies to non-tuberculous mycobacteria as a class, so a positive M. gordonae result triggers the same escalation as any other NTM: quarantine of affected instruments, review of the water treatment train and filtration, inspection of the reprocessor for biofilm, and repeat sampling only after remediation.

The same logic applies in CSD and dental water. Instrument rinse water in a sterile services department, and dental unit waterline output, both derive from building plumbing and both are subject to stagnation in narrow-bore tubing. M. gordonae recovered from either indicates that mycobacteria are established in the line, and the remedy is line disinfection and barrier verification rather than reassurance drawn from the species identification.

Relevance in endoscopy and reprocessing

The specific value of M. gordonae in an endoscopy service is as an early-warning organism. Because it is the most abundant mycobacterium in most potable supplies, it is statistically the one most likely to appear first when a barrier begins to fail. A filter approaching the end of its service life, a housing seal that has started to weep, a storage vessel that has developed biofilm, or a reprocessor rinse line that has not been reaching its self-disinfection temperature will usually announce itself with M. gordonae before it announces itself with M. avium. Treating that first result as actionable is what prevents the second one.

Against high-level disinfection, M. gordonae behaves as a typical slow-growing mycobacterium. Validated glutaraldehyde, ortho-phthalaldehyde and peracetic acid cycles are mycobactericidal and will inactivate it when the instrument has been properly cleaned and all channels are perfused. As with other NTM, the risk is not that the disinfectant fails on a clean surface but that the organism is protected — by residual organic soil, by biofilm within a channel, or by a scratched or delaminated channel lining — or that it is reintroduced after disinfection by the rinse. Once established in an endoscope channel, mycobacterial biofilm is markedly harder to eradicate than a planktonic contamination, and repeat cycles are unlikely to resolve it; borescope inspection and, where damage is found, instrument withdrawal are the appropriate responses.

Drying and storage matter as much here as for the rapid growers. Residual moisture in a channel allows any surviving cells to persist and any biofilm present to remain metabolically active, and a channel stored wet for a weekend provides ample opportunity for growth from a very small inoculum. Alcohol flush followed by verified channel-perfusing forced-air drying, and storage that keeps channels dry, are the controls. In dental practice the equivalent exposure is continuous rather than episodic: waterlines that sit stagnant overnight and between patients are a standing mycobacterial habitat, and recovery of M. gordonae from a dental unit is a direct indication for line disinfection and for independent-reservoir or point-of-use filtration rather than reliance on mains water quality.

Interpreting a detection

A detection of M. gordonae in a final rinse, reprocessor or dental waterline sample means the water reaching the device is not mycobacteria-free, and the operational response is identical to that for any other NTM. The first item to check is the terminal bacterial-retentive filter: its rating, its installed date against service life, the integrity of the housing seal, whether it has been allowed to dry out and rewet, and whether any bypass exists around it. Next is the water treatment train upstream — carbon beds, softeners, reverse osmosis membranes, ultraviolet units and any storage vessel, all of which accumulate biofilm and all of which strip the disinfectant residual that was suppressing growth in the incoming main. Then the reprocessor itself: rinse water lines, tanks, valves, and the record of self-disinfection cycles including whether the cycle is reaching and holding its specified temperature. Finally, the sampling process is reviewed, including sample point preparation, container sterility and transport time.

The distinction between a single isolate and a trend shapes the depth of the response but not whether there is one. A first positive is treated as a real breach: instruments processed since the last satisfactory result are identified and quarantined pending assessment, a specific remediation action is taken, and resampling is performed afterwards so that the repeat result tests the intervention. A pattern of repeated positives, positives at multiple sample points, or a positive that recurs after filter change and line disinfection points to established biofilm in a component that cannot be cleaned in place, and the response should shift toward replacement of hose, housings, storage vessels or reprocessor plumbing. It is also worth reviewing whether M. gordonae results coincide with rising heterotrophic plate counts or with Pseudomonas detections, since these commonly share a root cause in stagnation and biofilm.

The reason this organism requires deliberate looking is that routine monitoring will not find it. Total viable count methods use general-purpose media, short incubation and temperatures suited to common waterborne bacteria; M. gordonae needs sample decontamination to suppress competitors, selective mycobacterial media, and incubation over weeks. A facility running only TVC and coliform testing has no visibility of mycobacteria at all, and a run of satisfactory TVC results provides no assurance on this point. When a result does come back positive, escalation should bring together infection prevention, the water safety group and facility engineering, because the corrective actions are engineering actions. Where the finding follows clinical specimens positive for M. gordonae, the priority is to establish quickly whether those represent pseudo-infection, since the documented harm of this organism is patients treated unnecessarily with prolonged anti-mycobacterial therapy on the strength of a contaminated specimen.

Antimicrobial resistance

M. gordonae is not intrinsically resistant to the degree seen in rapidly growing mycobacteria, and reported isolates are commonly susceptible to rifampicin, ethambutol, clarithromycin and amikacin; because disease is uncommon, comparative treatment data are limited and regimens are individualised with susceptibility testing. Where treatment is genuinely indicated — which is unusual, and should follow careful assessment that the isolate represents disease rather than contamination — prolonged multi-drug therapy is used, as for other slow-growing NTM. The more common clinical error is the opposite one: treating a contaminant. Because M. gordonae is acid-fast and may be reported before species identification is complete, patients have been started on anti-tuberculous regimens with significant hepatotoxicity on the basis of an isolate that was never theirs.

The more practically important resistance is environmental. Like other mycobacteria, M. gordonae tolerates free chlorine and chloramine at potable-supply concentrations far better than enteric indicator organisms, and cells embedded in biofilm are less susceptible again. The barrier is the mycolic-acid cell wall, which is thick, hydrophobic and poorly permeable to hydrophilic oxidants, slowing their access to intracellular targets. The same property confers tolerance to drying, to low-nutrient conditions and to low-level disinfectants generally, and allows the organism to survive within free-living amoebae, where it is further shielded.

The operational consequence is that conventional water-quality reassurance does not transfer. Absence of coliforms, a satisfactory chlorine residual, or an acceptable heterotrophic plate count therefore provides no assurance that mycobacteria are absent from a reprocessing water supply. Control rests on physical exclusion at the point of use through correctly specified and maintained bacterial-retentive filtration, on system design that minimises stagnation and dead legs, and on verification by mycobacteria-specific culture rather than by indicator organisms.