New RAPMYCOI SensititreTM Antimicrobial Susceptibility Test for Atypical Rapidly Growing Mycobacteria (RGM)

Rapidly growing mycobacteria (RGM) cause an increasing international concern, mainly due to their natural resistance to many antibiotics. The aim of this study was to conduct species identification and determine the antimicrobial susceptibility profiles of RGM isolated in Poland. Antimicrobial susceptibility was tested using broth microdilution and the RAPMYCOI panel. A total of 60 strains were analysed, including the following species: M. fortuitum complex (30), M. abscessus subsp. abscessus (16), M. abscessus subsp. massiliense (7), M. chelonae (5), and M. mucogenicum (2). For 12 M. abscessus subsp. abscessus strains, the presence of the erm 41T28 genotype associated with inducible macrolide resistance and a functional erm gene was confirmed. A MUT2 mutation in the rrl gene (constitutive resistance) was identified for two strains from the subtype M. abscessus subsp. massiliense. Among the 15 tested antibiotics, amikacin and linezolid had the strongest antimycobacterial activity. Most of the tested strains were resistant to doxycycline and trimethoprim/sulfamethoxazole. Tigecycline MICs were low for all tested strains. Findings from our study highlight the importance of correct identification of clinical isolates and antimicrobial susceptibility testing.


Introduction
Nontuberculous mycobacteria (NTM), also known as mycobacteria other than tuberculosis (MOTT), are ubiquitous environmental microorganisms [1]. Currently, more than 150 species are known worldwide and many of them are increasingly recognized as important human pathogens [2,3]. Based on their growth rate, NTM species are divided into slowly growing mycobacteria (SGM) and rapidly growing mycobacteria (RGM) [4]. To date, more than 75 RGM species have been identified, which represents approximately 50% of all known mycobacterial species [5]. Due to advances in molecular research, the number of newly discovered species continues to increase.
Tuberculosis caused by Mycobacterium tuberculosis complex remains a serious global health problem in developing countries [13]. However, in recent decades, the number of Diagnostics 2022, 12,1976 2 of 13 reported cases of mycobacteriosis, a disease caused by atypical mycobacteria, has increased significantly [14]. Factors contributing to the increased incidence of NTM infections include: demographic changes, ageing of the population, increased incidence of comorbidities, and immunosuppression [15]. However, the epidemiology of NTM infections remains unknown as the reporting of mycobacterial cases to public health authorities is not mandatory in most countries [16]. Undoubtedly, the diagnosis of the disease is facilitated by recently improved testing methods [17].
Epidemiological statistics indicate that people living in Asia are particularly susceptible to NTM infections. In 2014, the incidence of NTM in the Japanese population was estimated at 14.7/100,000 [18][19][20]. In Great Britain, the incidence of NTM infection increased from 0.9 to 2.9/100,000 between 1995 and 2006 [21]. Studies from North America and Australia revealed that the annual incidence of NTM in these regions in 1997-2010 was 3.2-9.8/100,000 [14]. In Denmark, the incidence of NTM increased between 2003 and 2008 from 0.6 to 1.5/100,000 [22]. In Poland, statistics published by the National Institute of Public Health (PZH) and the Chief Sanitary Inspectorate show that the incidence rate of mycobacteriosis was 0.69 in 2017, 0.63 in 2018, and 0.61/100,000 in 2019 [23].
Atypical mycobacteria are opportunistic pathogens, ubiquitous in the environment, and are found in fresh and marine water, soil, and on biofilms [24]. Infections mainly concern the population of high-risk patients, which includes patients with cystic fibrosis (CF), bronchiectasis, emphysema, chronic obstructive pulmonary disease (COPD), and immunoincompetence (human immunodeficiency virus (HIV) infection, organ transplant, diabetes mellitus, renal failure). Among rapidly growing mycobacteria, the highly pathogenic non-pigmented species include M. fortuitum, M. abscessus, and M. chelonae, which are responsible for more than 80% of all clinical cases [25].
The clinical manifestations of RGM infections are very diverse. They most often concern the respiratory tract, skin, soft tissues, bones and joints, lymphadenitis, or disseminated infections [26]. Chronic lung infections are usually caused by M. abscessus subsp. abscessus and M. abscessus subsp. massiliense [25,27]. In patients with cystic fibrosis, these pulmonary infections are associated with a very high mortality. M. fortuitum complex is most frequently isolated from infected skin after accidental injuries, cosmetic procedures, and laser surgery. Reportedly, M. abscessus is responsible for 90% of respiratory diseases caused by RGM, and M. fortuitum is responsible for 60-80% of postsurgical and catheterrelated infections [26]. The most common symptoms of infection caused by M. chelonae are diseases of the skin, bones, and soft tissues as well as ophthalmic infections, including keratitis. Rapidly growing mycobacteria are also isolated from patients with catheter-related bloodstream infection. In this case, the causative pathogens are M. mucogenicum and M. fortuitum, but also M. neoaurum and M. bacteremicum [28][29][30][31].
To determine the etiological factors of mycobacteriosis, it is necessary to correctly identify atypical mycobacteria to the species level. This is due to the different antimicrobial susceptibilities of mycobacteria. The management of a wide spectrum of NTM infections is a serious challenge worldwide. The selection of the appropriate antibiotic therapy for the patient should be based on the results of in vitro antimicrobial susceptibility testing.
However, the suitability of antimicrobial susceptibility testing in the treatment of patients with mycobacteriosis remains controversial due to the discrepancy between test results and clinical response [27,32]. Good correlations demonstrated in the studies carried out to date have been shown for two groups of antibiotics: macrolides and aminoglycosides.
Macrolides (clarithromycin and azithromycin) are among the basic antibiotics used in the treatment of mycobacteriosis. All macrolides bind to the V domain in 23S rRNA on the 50S ribosome subunit [33]. Two mechanisms of resistance to this class of drugs have been identified so far among atypical mycobacteria. The first mechanism is the constitutive resistance associated with a point mutation at either the A2058 or A2059 position of the 23S rRNA (rrl gene). The second mechanism, defined as inducible macrolide resistance, is associated with functional erm genes encoding ribosomal methyltransferase. The erm genes have been identified in the following species: erm (41) [13,34,35].
The aminoglycosides (amikacin and tobramycin) act by binding stably to the 30S ribosomal subunit in bacterial cells, leading to misreading of the genetic code and inhibition of protein synthesis and consequently to cell death. Resistance to aminoglycosides in atypical mycobacteria is associated with single-point mutations in the 16S rRNA (rrs gene) [36].
According to the Clinical and Laboratory Standards Institute (CLSI), the broth microdilution method is considered the gold standard for testing the drug sensitivity of atypical RGM strains. Antimicrobial susceptibility testing should include the following antibiotics: clarithromycin, amikacin, moxifloxacin, linezolid, imipenem, cefoxitin, ciprofloxacin, doxycycline, trimethoprim/sulfamethoxazole, and tobramycin (only for M. chelonae). It is also recommended to determine the minimal inhibitory concentration (MIC) value for tigecycline, but to date there are no consensus breakpoints or guidelines for the interpretation of results [34,37].
There is a commercially available RAPMYCOI test for RGM from Thermo Fisher Scientific (Waltham, MA, USA) that includes all the antibiotics recommended for the treatment of RGM infections. M. fortuitum complex, M. abscessus subsp. abscessus, M. abscessus subsp. massiliense and M. chelonae are the most common rapidly growing mycobacteria (RGM) isolated in Poland.
In the presented study, the susceptibility of 60 RGM strains to 15 antibiotics was determined using the RAPMYCOI panels. The obtained results were compared with data published worldwide, which made it possible to obtain a complete picture of the drug resistance in this group of mycobacteria.

Bacterial Strains and Growth Conditions
The study was conducted on 60 strains of atypical mycobacteria (RGM) originally isolated from respiratory specimens (sputum, bronchial washings), in the period from 2019 to 2020 in mycobacterial laboratories in Poland.
The respiratory specimens were decontaminated with the sodium hydroxide and Nacetyl-L-cysteine (NaOH/NALC) (Chempur, Poland) method. The strains were cultured on solid media: egg-based Lowenstein-Jensen medium, Stonebrink medium, and in automated system MGIT (Becton Dickinson, Franklin Lakes, NJ, USA).

Strain Identification
For DNA extraction, the GenoLyse (Hain Lifescience, Nehren, Germany) kit was used according to protocol.
The strains were identified using the GenoType Mycobacterium CM assay ver. 2.0 (Hain Lifescience, Nehren, Germany) in accordance with the manufacturer's instructions.

Molecular Determination of Antimicrobial Susceptibility to Macrolides and Aminoglycosides
GenoType NTM-DR assay enabled the detection of resistance to macrolides (erm (41) and rrl genes) and aminoglycosides (rrs genes).
Erm (41) gene was only detected in members of the M. abscessus complex. The above test detected mutations at position 28 of the erm (41) gene:

•
If the strain had a genotype in which C was at position 28 it meant that the tested strain was sensitive to macrolides.

•
If the strain had a genotype in which T was at position 28 it meant that the tested strain was resistant to macrolides.
In the Tables 1 and 2 below, the mutations detected by the applied test was shown. Table 1. Mutations determining resistance to macrolides detected using the GenoType NTM-DR assay within the rrl gene.

Absence of Wild-Type Band
Analysed Nucleic Acid Positions

Phenotypic Determination of Antimicrobial Susceptibility Profile
Antimicrobial susceptibility was tested using broth microdilution. For this purpose, 96-well RAPMYCOI Sensititre ™ titration plates (Thermo Fisher Scientific, Waltham, MA, USA) were used, which allow for the simultaneous determination of susceptibility to 15 antibiotics.
RAPMYCOI plates contain freeze-dried antibiotics in a range of concentrations (µg/mL). The plate design and the tested antibiotic concentrations are presented in Figure 1.
At the first stage of the test, an inoculum of a mycobacterial suspension at the optical density of 0.5 McFarland scale was prepared. A total of 50 µL of inoculum was transferred to 10 mL of CAMHB medium (cation-supplemented Mueller-Hinton broth and TES buffer) (Thermo Fisher Scientific, Waltham, MA, USA). The 100 µL suspension prepared according to this protocol was pipetted onto a 96-well titration plate and incubated at 30 • C ± 2 • C. Plates with RGM were incubated for 3 to 5 days. Only for clarithromycin, the incubation period was prolonged to 14 days in order to detect inducible resistance associated with the presence of the erm genes. If microbial growth in the positive control sample was sufficient, MICs were measured. In cases of difficulties with visual reading, 10 µL of Alamar Blue (BIO-RAD, Hercules, CA, USA) reagent and 25 µL of 5% Tween 80 (Fisher Scientific, Hampton, NH, USA) were added. A colour change from blue to pink indicated the growth of a strain. Measured MICs were interpreted and each strain was classified into one of three groups (sensitive (S), intermediate (I), and resistant (R)) in accordance with the CLSI guidelines (document M62, 1st edition) (37) ( Table 3).
RAPMYCOI plates contain freeze-dried antibiotics in a range of concentrations (μg/mL). The plate design and the tested antibiotic concentrations are presented in Figure 1. At the first stage of the test, an inoculum of a mycobacterial suspension at the optical density of 0.5 McFarland scale was prepared. A total of 50 µL of inoculum was transferred to 10 mL of CAMHB medium (cation-supplemented Mueller-Hinton broth and TES buffer) (Thermo Fisher Scientific, Waltham, MA, USA). The 100 μL suspension prepared according to this protocol was pipetted onto a 96-well titration plate and incubated at 30 °C ± 2 °C. Plates with RGM were incubated for 3 to 5 days. Only for clarithromycin, the incubation period was prolonged to 14 days in order to detect inducible resistance associated with the presence of the erm genes. If microbial growth in the positive control sample was sufficient, MICs were measured. In cases of difficulties with visual reading, 10 µL of Alamar Blue (BIO-RAD, Hercules, CA, USA) reagent and 25 µL of 5% Tween 80 (Fisher Scientific, Hampton, NH, USA) were added. A colour change from blue to pink indicated the growth of a strain. Measured MICs were interpreted and each strain was classified into one of three groups (sensitive (S), intermediate (I), and resistant (R)) in accordance with the CLSI guidelines (document M62, 1st edition) (37) ( Table 3).     The data obtained in the performed antimycobacterial susceptibility test showed that amikacin and linezolid had the strongest antituberculotic activity against RGM. Most of the analysed strains were resistant to doxycycline and trimethoprim/sulfamethoxazole.

Results
The following tables (Tables 5-9) present the obtained results separately for each RGM species.
In the analysed collection of RGM, all 60 (100%) strains had low MIC values (from 0.06 to 1 µg/mL) for tigecycline, an antibiotic considered as a potential therapeutic agent and a drug of last resort in the treatment of severe cases of mycobacteriosis.

Discussion
With advances in molecular techniques and genetic tools, including whole genome sequencing (WGS), knowledge about the genetic diversity of NTM species and genes determining resistance to antibiotics continues to grow. Long treatment (18 to 24 months on average) and the need to use a combination of antibiotics with multiple side effects increase the importance of drug resistance testing, especially in RGM strains naturally resistant to first-line antituberculotic drugs.
Guidelines on antimicrobial susceptibility testing (AST) of atypical mycobacteria were developed by the CLSI and last updated in December 2018. Currently, CLSI M24 (3rd edition) provides recommendations on AST for slowly growing non-tuberculous mycobacteria, including M. avium complex (MAC), M. kansasii, and M. marinum, as well as rapidly growing mycobacteria (RGM) [34]. Since atypical mycobacteria may colonize the respiratory tract, their isolation from clinical specimens does not always correlate with the identification of an etiological factor responsible for the observed changes. This primarily refers to single sputum cultures. A negative sputum smear indicates a small number of microorganisms that are unlikely to be clinically significant, i.e., insufficient to establish a diagnosis of NTM. Therefore, detailed criteria for the diagnosis of mycobacteriosis have been developed for clinically significant isolates from the respiratory tract [27,38]. These criteria include the following: • at least two NTM culture-positive sputa or one bronchial wash or lavage sample, • a transbronchial or lung biopsy specimen with supporting mycobacterial histopathology and a positive NTM culture.
According to the current CLSI recommendations, AST includes antimicrobial agents for RGM such as amikacin, cefoxitin, ciprofloxacin, clarithromycin, doxycycline (or minocycline), imipenem, linezolid, moxifloxacin, trimethoprim-sulfamethoxazole, and tobramycin (for M. chelonae only) ( Table 3). Worth noting is the fact that there are insufficient data to establish MIC breakpoints for tigecycline and clofazimine, and therefore for these agents a MIC without interpretation should be given [34].
The results of AST with selected drugs may concern specific species of atypical mycobacteria. For this reason, CLSI and most experts in RGM recommend identifying RGM strains at the species or even subspecies level (Table 10), especially for the M. abscessus complex, before performing a new AST RAPMYCOI and initiating treatment [5,34,39,40]. Because the incubation period for most RGM species ranges from 2 to 5 days, the final MIC reading in the RAPMYCOI test should be performed <5 days. This is mainly due to the instability of some drugs, including carbapenems and tetracyclines. There are only two exceptions where this incubation time should be extended when performing the RAPMYCOI test. The first case concerns strains representing M. abscessus complex isolated from patients who had a history of long-term treatment, including patients with cystic fibrosis. Mycobacterial strains isolated from this population of patients need a longer incubation period; therefore, in some cases it may be helpful to change the incubation temperature or to establish a culture in a shaking incubator. However, if the culture incubation period is longer than 5 days, results are only reliable for AST related to two drugs: clarithromycin and amikacin. The CLSI recommends a comment on the AST report such as: this NTM strain required extended incubation and results for only clarithromycin and amikacin are reliable after incubation for >5 days (Table 10) [34].
The second exception in the RAPMYCOI test to the incubation period longer than 5 days is clarithromycin. Phenotypical detection of inducible resistance to macrolides is achieved by extending the incubation of clarithromycin to 14 days unless the MIC is ≥16 µg/mL at an earlier time point. If the clarithromycin MIC is 4 or 8 µg/mL after 14 days of incubation, the test should be repeated. If the MIC is 4 or 8 µg/mL in the retest, sequencing of the erm gene for the given strain is recommended.
Worth noting is the fact that several RGM species have a non-functional or absent erm gene and are naturally sensitive to clarithromycin [41]. Therefore, sensitivity to clarithromycin can be reported at the initial MIC reading as no prolonged incubation is required for these specific species [42].
In the presented study, we identified 60 rapidly growing mycobacterial strains and determined their antimicrobial susceptibility in accordance with CLSI guidelines.
The most frequently isolated species was Mycobacterium fortuitum complex, which accounted for 50% (30/60) of all identified strains. The tests demonstrated that among all RGM species this group is characterized by high sensitivity to antibiotics. Our study confirmed this thesis and showed that 100% (30/30) of the strains from this group were sensitive to amikacin and moxifloxacin, 97% (29/30) were also sensitive to ciprofloxacin, and 93% (28/30) were sensitive to linezolid. In contrast, tests with clarithromycin showed a high level of resistance for 77% (23/30) of the strains. This resistance is higher compared to that reported by Sriram [43,44].
A low rate of drug resistance according to CLSI was also found for Mycobacterium chelonae. In our study, 100% of the strains (5/5) were sensitive to amikacin, clarithromycin, linezolid, and tobramycin. Our findings are consistent with those reported by Bhalla et al.
In the cited study, no resistance to the four above-mentioned antibiotics was found for the three tested isolates [43].
Mycobacterium chelonae and Mycobacterium mucogenicum are classified into the group of species lacking functional erm genes. In our study, all strains of M. chelonae (5/5) and M. mucogenicum (2/2) were sensitive to clarithromycin and no erm genes were detected. However, Esteban et al. detected resistance to clarithromycin associated with the presence of erm genes in two strains of M. chelonae [45]. In a study by Davalos et al., 100% (2/2) of M. chelonae strains were sensitive to clarithromycin. However, one strain (25%) of M. mucogenicum resistant to this antibiotic was detected [46]. In our study, two strains representing M. mucogenicum were sensitive to most of the tested antibiotics. Only one strain was resistant to doxycycline and minocycline. A different antimicrobial susceptibility profile for this species was found by Faridah et al., who reported resistance to ciprofloxacin, doxycycline, clarithromycin, and tobramycin in a strain isolated from blood [47].
Isolates representing the Mycobacterium abscessus complex accounted for 38% (23/60) of all identified strains and it was the second largest group. Most strains (16) represented M. abscessus subsp. abscessus subtype, while M. abscessus subsp. massiliense subtype was less frequently identified (seven strains). We did not identify M. abscessus subsp. bolletii. In our study, most strains of M. abscessus subsp. abscessus (75%) were resistant to clarithromycin (MIC >16 µg/mL). This resistance was associated with the presence of a functional erm gene. The situation was different for M. abscessus subsp. massiliense. Only 28% of strains representing this subtype were clarithromycin-resistant, and the MUT2 mutation in the rrl gene was responsible for the resistance mechanism. Our findings confirm the worldwide reports on the more frequent resistance of M. abscessus subsp. abscessus to clarithromycin compared to M. abscessus subsp. massiliense [48][49][50]. Considering amikacin, the vast majority of strains were sensitive to this antibiotic. Only one strain (4%) was resistant to amikacin (MIC > 64 µg/mL) and had the MUT1 mutation in the rrs gene. Similar findings were reported by Bhalla et al., who found 92.3% of sensitive strains [43].
Among the three tested tetracyclines (doxycycline, minocycline, and tigecycline), the lowest MICs (from 0.06 to 1 µg/mL) were found for tigecycline. If we assume the criteria for interpretation proposed by Wallace et al. (resistant strain when MIC ≥ 8 µg/mL), all tested strains (60/60) were sensitive to tigecycline [51]. Similar relationships were observed by Pang et al.: sensitivity to tigecycline was found for 96% (53/55) of strains from the M. abscessus complex, 91% (10/11) of M. fortuitum strains, and 100% (3/3) of M. chelonae strains [52]. Similarly, in a study by Comba et al., the MIC value was <0.25 µg/mL for 45.7% of the strains (16/35), and from 0.25 µg/mL to 0.5 µg for 54.3% of the strains (19/35) [53]. According to worldwide reports, tigecycline is used in the treatment of the most severe infections with RGM mycobacteria, but to date there are no CLSI guidelines for the interpretation of MIC values in the AST.

Conclusions
The new RAPMYCOI test is a rapid tool for the determination of drug resistance profile in RGM. The obtained results are reliable and reproducible, and the test setup is not time-consuming. The broth microdilution method on which the test is based and the selection of antibiotics are consistent with the CLSI guidelines.
Taken together, the findings from the presented study highlight the importance of a correct identification of clinical isolates to the species and subtype level and the role of antimicrobial susceptibility testing, especially for highly resistant rapidly growing mycobacteria (RGM). The obtained results confirm previous assumptions published worldwide according to which there are predictable drug resistance profiles depending on the identified mycobacterial species. However, there are some exceptions to this rule, and therefore the drug resistance of individual strains should be tested as standard practice. The correlation between data obtained from AST with clinical findings proving the effectiveness of treatment will enable the development of new therapeutic regimens. As a result, effective drugs can be selected and the patient's treatment optimized at an early stage.

Conflicts of Interest:
The authors declare no conflict of interest.