Mycobacterium chelonae

Associated infections

• Post-procedural and post-injection skin, soft tissue and wound infection
• Surgical site and post-cosmetic-procedure infection
• Catheter-related bloodstream infection
• Keratitis, including post-LASIK keratitis
• Disseminated cutaneous infection in immunosuppressed patients
• Chronic pulmonary infection
• Otitis media following tympanostomy in contaminated-water exposure
• Pseudo-infection arising from contaminated bronchoscopy and endoscopy specimens

Transmission route

Acquisition is environmental, through direct inoculation of contaminated water or fluid into tissue, through contact with contaminated instruments or solutions, or through inhalation of aerosols. Documented sources across the wider literature include tap water used to rinse or dilute, ice, humidifier and nebuliser reservoirs, injectable and cosmetic preparations, tattoo ink, footbaths, surgical marking solutions and inadequately reprocessed instruments. The unifying feature is water that has contacted a device, a solution or a tissue surface at a point where sterility was assumed but not achieved.

In the endoscopy setting the organism has particular significance because it has been documented repeatedly as a contaminant of reprocessing equipment. M. chelonae has been isolated from the rinse water of endoscope automated disinfection systems, including systems subject to daily chlorine dioxide treatment, and from bronchoscopes linked to a contaminated automated bronchoscope disinfection machine. In the latter investigation, a phenotypically distinctive strain was recovered both from patient specimens and from the rinse water of the disinfecting machine over an extended period. A pseudo-outbreak involving M. chelonae together with Methylobacterium mesophilicum has likewise been traced to contamination of an automated endoscopy washer.

The route into the patient or the specimen is straightforward once the machine is colonised. Organisms in the final rinse water are deposited on internal and external instrument surfaces after the disinfection step has been completed, at the last point in the process where any contamination is introduced and at which nothing downstream will remove it. Where the instrument is a bronchoscope, the organisms are recovered in subsequently collected respiratory specimens; where it is a gastroscope or duodenoscope, they are deposited on mucosal surfaces or carried into tissue by accessories. Residual moisture retained in channels after an inadequate drying step allows any organisms deposited to persist and multiply during storage, so drying is a direct control on the size of the inoculum an instrument carries to the next patient.

Relevance in endoscopy and reprocessing

M. chelonae is one of the strongest examples of a genuinely endoscopy-relevant water organism, and the documented incidents define how it should be handled. In the reported bronchoscope investigation, a distinctive strain was recovered from both patient specimens and the rinse water of the automated bronchoscope disinfection machine over an extended period. Intensified measures applied to the affected machine, including the use of sterile water for wash and rinse cycles, longer glutaraldehyde exposure and more frequent replacement of disinfectant, failed to clear the organism. That failure was attributed to established biofilm within the machine itself, and it is the single most instructive fact on this page: once the reprocessor is colonised, escalating the chemistry does not resolve the problem. Separately, M. chelonae has been recovered from endoscope autodisinfector rinse water alongside Acremonium species despite daily chlorine dioxide treatment, again demonstrating persistence under an active biocide regime, and a documented pseudo-outbreak has been traced to contamination of an automated endoscopy washer.

This pattern has two consequences. First, contamination is typically machine-resident rather than transient, so remediation requires attention to biofilm in the reprocessor circuit, water lines, storage vessels and terminal filtration, and frequently disassembly, mechanical cleaning or component replacement rather than a change of disinfectant alone. Components with internal surfaces that cannot be reached, such as flexible tubing runs, check valves, dosing lines and filter housings, are the usual harbour points, and dead legs and low-flow branches in the supply loop are the usual upstream contributors. Verification after remediation requires repeat sampling over multiple rounds, because intermittent shedding from residual biofilm produces alternating negative and positive results.

Second, the clinical harm is often indirect. Contaminated rinse water transfers organisms to the instrument, they appear in subsequently collected bronchial or endoscopic specimens, and patients are misdiagnosed and treated for mycobacterial disease they do not have, with the exposure to prolonged multidrug antimycobacterial therapy that this entails. Recognising a pseudo-outbreak requires the endoscopy unit and the laboratory to be in communication, since the pattern is visible in the aggregate, as an unexplained cluster of acid-fast isolates from patients without compatible clinical or radiological findings, rather than in any single result. Drying and storage close the loop: an instrument that leaves the reprocessor wet retains water in which this organism grows, so validated forced-air drying of all channels and controlled storage are essential complements to water quality, not optional refinements.

Interpreting a detection

Recovery of M. chelonae from final rinse or supply water is not a marginal finding and should not be managed as a routine exceedance. Unlike skin and oral organisms, this species is a true water organism with an established mechanism for reaching the patient at the last uncontrolled step of the process, and it is a documented cause of both real infection and pseudo-outbreaks in endoscopy. A confirmed isolate should be treated as evidence that the water pathway or the reprocessor is colonised until demonstrated otherwise, rather than as a probable sampling artefact. Contamination at the point of collection is possible but is a much less likely explanation here than for a skin or oral organism, because the source of the organism is precisely the water being sampled.

Immediate actions on a confirmed result are quarantine of the affected reprocessor and of the instruments processed through it, suspension of release from that machine, and notification of infection prevention and the responsible clinical lead. The investigation should then localise the source by sampling at defined points along the pathway rather than only at the final rinse: incoming supply, post-treatment, the distribution loop, the reprocessor water inlet, the internal circuit, and a processed endoscope channel sample. Terminal filter integrity, installation date and change interval should be verified, along with the reprocessor self-disinfection cycle and its documented verification, the sanitisation regime and records for the loop, water temperature and flow, and the presence of dead legs, low-flow branches or recently disturbed plumbing. Concurrently, the recent laboratory record should be reviewed for mycobacterial isolates from patients who underwent procedures on the affected instruments, since a pseudo-outbreak may already be in progress and unrecognised. Isolates from both water and patients should be retained for identification to species level and typing, as strain matching is what converts a suspicion into a demonstrated link.

Single isolate versus trend is interpreted differently for this organism than for most others. Because biofilm sheds intermittently, a single positive followed by negative repeats does not clear the system, and a facility should not close the investigation on a clean repeat sample alone. Sustained clearance should be demonstrated over multiple sampling rounds using the same method and sampling points, with volumes and incubation conditions appropriate to mycobacterial recovery, since routine heterotrophic plate count conditions will under-detect the organism and a normal total viable count is not reassurance. Any recurrence after remediation indicates that the biofilm reservoir has not been eliminated and that component replacement or engineering intervention, rather than a further chemical cycle, is the required response. Escalation beyond the unit is warranted at the first confirmed isolate, not at the second, and any patient link identified during the record review should trigger a formal look-back under infection prevention direction.

Antimicrobial resistance

Mycobacterium chelonae is intrinsically resistant to the antimycobacterial agents used for tuberculosis and to most conventional antibacterial classes, a consequence of its impermeable cell wall combined with efflux systems and modifying enzymes. Treatment must be guided by species-level identification and broth microdilution susceptibility testing, and empirical therapy based on the presumption of a rapidly growing mycobacterium without species assignment is unreliable. Reported susceptibility patterns show clarithromycin activity against effectively all isolates, distinguishing M. chelonae from Mycobacteroides abscessus, in which inducible macrolide resistance mediated by erm genes is common; extended incubation of susceptibility plates is nonetheless used to detect inducible resistance, and macrolide monotherapy is avoided because acquired resistance emerges readily.

Tobramycin is the aminoglycoside of choice for this species, in contrast to amikacin for other rapidly growing mycobacteria, and linezolid, tigecycline, imipenem, moxifloxacin and clofazimine also show useful activity in reported series. Combination therapy for extended periods is standard, and management of localised disease frequently depends as much on surgical debridement or removal of an infected device as on the antimicrobial regimen, because the organism persists in devitalised tissue and on prosthetic surfaces.

Of equal importance in the reprocessing context is biocide tolerance, which is the property that actually determines control. The organism withstands chlorine, chloramine, chlorine dioxide, ozone and glutaraldehyde at concentrations effective against most vegetative bacteria, with reported tolerance to free chlorine well above that of enteric organisms, and within biofilm this tolerance increases further by orders of magnitude. Glutaraldehyde tolerance is of direct practical concern given its continued use in endoscope reprocessing, and glutaraldehyde-tolerant mycobacterial isolates have been described in association with reprocessing equipment. The operational conclusion is that chemical control alone should not be relied upon. Effective management combines physical removal through validated terminal filtration, elimination of stagnation and dead legs, mechanical disruption or replacement of biofilm-bearing components, thermal sanitisation where the system design permits it, and thorough drying of instruments, since a dry channel does not support growth regardless of what the organism tolerates chemically.