Mycobacterium mucogenicum and Mycobacterium neoaurum Bloodstream Infection in Immunocompromised Hosts

Background Mycobacterium mucogenicum and Mycobacterium neoaurum are an infrequent cause of bloodstream infection in humans. When they are isolated from it is not them to be contaminants We identied 28 patients who had bloodstream infection caused by these organisms. These patients were immunocompromised hosts, patients with chronically indwelling vascular access devices or injection drug users. In 28.5% of the patients the isolation of these mycobacteria was considered colonization, 82% of the cases received antimicrobial therapy and in 95% of the patients the central line was removed.


Abstract
Background Mycobacterium mucogenicum and Mycobacterium neoaurum are an infrequent cause of bloodstream infection in humans. When they are isolated from blood, it is not unusual for them to be considered to be contaminants initially.

Methods
A retrospective chart review of patients diagnosed with M. mucogenicum and M. neoaurum bacteremia was conducted from January 1998 to December 2018 at the University Health Network, Toronto, Canada.

Results
We identi ed 28 patients who had bloodstream infection caused by these organisms. These patients were immunocompromised hosts, patients with chronically indwelling vascular access devices or injection drug users. In 28.5% of the patients the isolation of these mycobacteria was considered colonization, 82% of the cases received antimicrobial therapy and in 95% of the patients the central line was removed.

Conclusions
The removal of the vascular access device and treatment with either single or combination antimicrobial therapy produced successful outcomes for these infections.

Background
With the expansion of the immunocompromised patient population, infections due to nontuberculous mycobacteria (NTM) are increasing in frequency 1 , and their impact is signi cant as they may produce considerable patient morbidity. As a result, more familiarity with these organisms and the infections that they may produce is required. There are currently more than 150 species of NTM that are divided into rapidly, intermediate and slowly growing species 2 .
Rapidly growing mycobacteria produce mature colonies on solid media within 7 days. There are about 20 species of rapidly growing NTM that are capable of infecting human beings 3 . Of interest among the rapidly growing NTM, are two less commonly isolated mycobacteria, Mycobacterium mucogenicum and Mycobacterium neoaurum. They are ubiquitous in the environment, including household water, potting soil, animals, birds and vegetables. These organisms have been increasingly noted to be the cause of signi cant infections in immunocompromised hosts.
The current literature highlights that the most common infection caused by M. mucogenicum and M. neoarum is catheter-related bloodstream infection (CRBSI) [4][5][6][7] . The frequent use of indwelling intravascular catheters in immunocompromised individuals has been accompanied by NTM CRBSI, especially in patients with underlying malignancies 9 .
These organisms have the ability to form a protective bio lm, which plays a role in CRBSI. Bio lms in turn confer decreased penetration to antimicrobials, enhance the potential of resistance and are integral to device-related infections 9 . These organisms are also able to tolerate disinfectants, chlorination and extreme temperatures.
M. mucogenicum and M. neoaurum infections are seen more often in long-term central intravenous catheters causing catheter-related sepsis, but they may also occur with peritoneal or shunt catheters. Central nervous system infections involving Mycobacterium mucogenicum are rare but serious, particularly for immunocompromised patients resulting in meningitis due to this organism 5 . Skin and soft tissue infections caused by these organisms have also been reported 4,10 .
In immunocompromised patients, M. mucogenicum or M. neoaurum isolated from the bloodstream should be considered as true pathogens. The treatment of these NTM CRBSI involves catheter removal combined with antibiotic therapy 3 . These species are usually susceptible to multiple antimicrobial agents including aminoglycosides, cefoxitin, clarithromycin, minocycline, doxycycline, quinolones, trimethoprim/sulfamethoxazole, and imipenem 11 . Currently, there is no guidance or consensus on the appropriate antimicrobial therapy (whether single agent or combination therapy) for these infections or the duration of We identi ed patients with infection due to M. mucogenicum and M. neoaurum cultures through a search of the Microbiology Laboratory database of the Mount Sinai Hospital. The microbiological isolates were con rmed with Geno Type Mycobacterium Assay for identi cation of Mycobacterial Species or by matrix assisted laser desorption ionization (MALDI-TOF) if available. Only two isolates of M. mucogenicum had in vitro susceptibility testing, that was done by a microdilution minimum inhibitory concentration (MIC) method. Susceptibility testing for the mycobacterial isolates was performed at the Public Health Ontario Laboratories. We excluded patients who were younger than 18 years old or whose source of the culture was other than the bloodstream.
Data collection included patient demographics, comorbidities, major underlying predisposing condition, origin of the positive blood culture for M. mucogenicum and M. neoaurum, microorganism susceptibility patterns if available, presence and site of intravascular catheters, clinical manifestations, treatment regimen and duration, as well as outcome.
We used the Centers for Disease Control de nition of central line-associated bloodstream infection. 12 This was de ned as a patient who had at least one of the following signs or symptoms: fever (> 38ºC), chills or hypotension, had a microorganism identi ed in blood not related to an infection at another site and whose microorganism was identi ed by culture or non-culture based microbiologic testing method, from two or more blood specimens collected on separate occasions. We employed the differential time to positivity i.e. the growth of organism from blood samples drawn from a catheter hub at least 2 hours before microbial growth is detected in a blood sample obtained from a peripheral vein, in order to de ne CRBSI 13 . Catheter-related infection was also determined by the presence of ≥ 15 colony forming units on the vascular access device tip according to the semiquantitative method of Maki 14 .
Descriptive statistics were employed as appropriate to compare patients with successful outcomes to those with unfavourable outcomes.

Results
Between January 1998 to December 2018, we identi ed 81 isolates of M. mucogenicum, 24 of which were blood culture isolates, 3 from dialysis uid, and 54 respiratory isolates from bronchoalveolar lavage uid or sputum. M. mucogenicum was not considered to be a signi cant pathogen if the origin was the respiratory tract. We also identi ed 4 blood culture isolates of M. neoaurum. Therefore, we assessed a total of 28 episodes of bacteremia.
The patient characteristics are shown in Table 1.    Table 3. Therapy with a single antimicrobial agent was undertaken in only 35% (8/23); while combination therapy was initiated in 65% (15/23 episodes). In addition, moxi oxacin was part of the combination regimen in all 15 courses of combination therapy. Most commonly, moxi oxacin was paired with amikacin and/or a carbapenem in 6 episodes (40%) each respectively (Table 4).  1-year survival 21 (87.5%) 3 (75%) The outcomes for these cases of bacteremia were varied (Table 4). In 9 (32%) patients, the organisms were initially considered colonization/contaminant. In the M. mucogenicum group, the organism isolated was initially considered to be a contaminant in 4 patients, but all of them had repeated blood cultures that were also positive and were treated. In addition, 3 patients with positive blood cultures were discharged and there were no available records to assess whether the infection was ever treated. In one patient, the rst blood culture was considered a contaminant and because the repeat blood cultures were negative no treatment was provided. In the M. neoaurum group, only one patient did not receive treatment, as he had coinfection with methicillin-susceptible Staphylococcus aureus (MSSA) and only received therapy with cefazolin.
None of the patients had disseminated disease or another infection site caused by the mycobacterial infection. The 30-day survival was 100% in both groups and one-year survival was 87.5% in the M. mucogenicum and 75% in the M. neoaurum 75% groups, respectively.

Discussion
This case series highlights that M. mucogenicum and M. neoaurum may produce signi cant infection in immunocompromised hosts. Commonly, these pathogens produce morbidity by causing CRBSI in this population.
Previous reports of bacteremia caused by these rapidly growing mycobacteria have focused in immunocompromised hosts. However, in this case series we found that all the immunocompetent patients, had an intravascular indwelling catheter at the time of diagnosis, this association has also been reported in the past 16 . We also found that 5 patients were chronic injection drug users. Therefore, the use of intravenous drugs should be recognized as a risk factor for the development of infections caused by these organisms.
Of note lymphocytopenia, was present in 68% of our patients, rather than neutropenia and may have predisposed our patients to these infections. Furthermore, these organisms have a predilection for causing CRBSI as was seen in our series (68% of the patients). Of signi cance, M. mucogenicum can form a bio lm on venous access devices thus rendering it resistant to killing by antimicrobials. 3,6 As a result, it is strongly advised that vascular access devices be removed.
In this case series the central line was removed in 95% of the patients, very similar to what has been reported in other studies (96%) 3 .
With regard to the management of these bloodstream infections, in our series 5 of 28 episodes (18%) were not treated with antimicrobial therapy. Indeed, the isolation of these organisms in blood cultures may be somewhat delayed further casting doubt on their signi cance and delaying appropriate antimicrobial therapy. A testament to this issue was that It was surprising that 5 patients didn't receive any antimicrobial therapy with purported activity against M. mucogenicum in our case series but they recovered by means of venous access device removal. Thus, underscoring the need for removal of the venous access as a key component for successful treatment of these bacteremias. Combination antimicrobial therapy involving moxi oxacin with amikacin or a carbapenem may be preferred. Sixty-ve per cent of the patients in our series received a combination of antimicrobial agents involving moxi oxacin with another agent. Nevertheless, one active antimicrobial agent may be adequate. This would require veri cation in a randomized clinical trial.
These organisms have often been considered to be contaminants. Indeed, their clinical signi cance was not appreciated until persistent bacteremia was noted with resolution of the infection only after venous access removal with or without antimicrobial therapy as noted in 9 cases in our series. It should be noted that co-infections with Staphylococcus aureus may occur with M. neoaurum as noted in 2 of our patients.
As this was a retrospective review, our analysis was limited by the availability of the data. Thus, we did encounter missing information. Moreover, management of these bacteremias was also hampered by the lack antimicrobial susceptibility in all isolates. However, the data do underscore the morbidity caused by M. mucogenicum and M. neoaurum in immunocompromised hosts.

Conclusions
In summary, this report focuses on the morbidity of bacteremia caused by M. mucogenicum and M. neoaurum in immunocompromised patients that is often related to venous access devices and use of IV drugs. It highlights the need for venous access device removal and therapy with antibiotics that should include moxi oxacin.