Microbiology of nontuberculous mycobacteria

INTRODUCTION — Mycobacterial species other than Mycobacterium tuberculosis and Mycobacterium leprae are generally free-living organisms that are ubiquitous in the environment. They have been recovered from water, soil, domestic and wild animals, milk, and food [1-4]. As the incidence of tuberculosis (TB) declined in areas of the world where socioeconomic conditions were rapidly advancing, the frequency of isolating nontuberculous mycobacteria (NTM) began to increase, and their relevance to human disease became apparent [5].

In 1959, Runyon proposed the first classification system for these organisms that divided human isolates of NTM into four groups on the basis of growth rates, colony morphology, and pigmentation in the presence and absence of light (table 1) [6]. This classification allowed mycobacterial laboratories to more readily identify individual species of NTM, resulting in clearer characterization of distinct diseases or syndromes associated with these organisms. This system is outdated and no longer used, but the terms may be encountered in older NTM literature.

With the availability of 16S ribosomal DNA sequencing and high-performance liquid chromatography (HPLC), polymerase chain reaction-restriction length polymorphism analysis (PRA), and multi-gene and whole-genome sequencing, the number of new species of NTM has risen dramatically with the naming of species such as Mycobacterium genavense, Mycobacterium interjectum, Mycobacterium triplex, Mycobacterium celatum, and Mycobacterium lentiflavum. Over 180 species have been recognized in the genus Mycobacterium [7].

The microbiology of NTM will be reviewed here. Other issues related to NTM are discussed separately. (See "Diagnosis of nontuberculous mycobacterial infections of the lungs" and "Pathogenesis of nontuberculous mycobacterial infections" and "Treatment of Mycobacterium avium complex pulmonary infection in adults".)

CLASSIFICATION — Within the genus Mycobacterium, four groups of human pathogens can be delineated on the basis of microbiologic, clinical, and epidemiologic characteristics (table 1):

●The M. tuberculosis complex

●M. leprae

●Slowly growing nontuberculous mycobacteria

●Rapidly growing mycobacteria

The most common nontuberculous species causing human disease in the United States are the slowly growing species of the Mycobacterium avium complex (MAC) and Mycobacterium kansasii and the rapidly growing species in the Mycobacterium abscessus group (M. abscessus subsp abscessus, subsp massiliense, and subsp bolletii). Less common human pathogens include the slowly growing species Mycobacterium marinum, Mycobacterium xenopi, Mycobacterium simiae, Mycobacterium malmoense, and Mycobacterium ulcerans, and the rapidly growing species Mycobacterium fortuitum and Mycobacterium chelonae. Certain relatively common laboratory isolates, such as Mycobacterium gordonae, are important to clinicians because they are almost always contaminants and uncommonly are true pathogens. (See "Buruli ulcer (Mycobacterium ulcerans infection)".)

Microscopy — Microscopic examination after staining and culture using specific media are the cornerstones of the identification of mycobacteria. All mycobacteria share the characteristic of "acid-fastness," ie, after staining with carbol-fuchsin or auramine-rhodamine, they do not decolorize with acidified alcohol. Thus, the common term acid-fast bacilli (AFB) is essentially synonymous with mycobacteria. Nocardia, the main exception, is weakly or variably acid-fast, but often a modified acid fast stain is necessary to distinguish this characteristic in Nocardia species. (See "Nocardia infections: Clinical microbiology and pathogenesis".)

Specimens may be stained with the Ziehl-Neelsen stain or one of its modifications, such as the Kinyoun stain, and examined by routine light microscopy. However, microscopy is relatively insensitive, since at least 10,000 organisms per milliliter of sputum are required for smear positivity. Thus, other procedures are often performed in order to increase the sensitivity of direct microscopy of clinical specimens.

●Most laboratories use a fluorochrome stain such as auramine-O or auramine-rhodamine and examine specimens by fluorescence microscopy.

●Liquid specimens may be centrifuged first and the sediment stained.

●Experienced microscopists may also detect mycobacteria in Gram stained specimens in which they appear as refractile gram positive or gram neutral rods. Under a 100x oil immersion objective, mycobacteria are slightly bent, often beaded rods 2 to 4 microns in length and 0.2 to 0.5 microns in width.

Culture — Confirmation of the presence or absence of mycobacteria in clinical specimens has traditionally required culture, because of the relative insensitivity of direct microscopy. In general, clinical specimens that are normally sterile, such as blood, cerebrospinal fluid, or serous fluids, can be inoculated directly onto media. In comparison, nonsterile specimens, such as sputum or pus, must be chemically decontaminated first in order to eliminate common bacteria and fungi that would overwhelm the culture. However, decontamination procedures may inhibit the growth of mycobacteria, especially NTM, as well.

Clinical specimens for mycobacterial cultures should be inoculated onto one or more solid media (eg, Middlebrook 7H11 media and/or Lowenstein-Jensen media, the former of which is the preferred medium for NTM) and into a liquid medium such as Mycobacteria growth indicator tube [MGIT] broth. Growth of visible colonies on solid media typically requires two to four weeks generally at 35 to 37°C for slowly growing NTM and 30°C for rapidly growing NTM and M. marinum. Primary cultures in modern liquid media, such as the radiometric BACTEC system, usually produce results within 10 to 14 days. However, this method is not 100 percent sensitive; as a result, it supplements but does not replace traditional solid media.

Traditional methods of speciating mycobacterial isolates were based upon growth characteristics on solid media and subsequent biochemical tests, requiring additional weeks for subcultures. These time-consuming methods were replaced with more rapid techniques, including specific nucleic acid probes, which are now available for the most common isolates (M. tuberculosis complex, M. kansasii, M. avium complex, and M. gordonae), and HPLC, which evaluates mycolic acid patterns in AFB smear-positive sputum samples (M. tuberculosis and MAC) and in primary positive cultures and permits identification of some species within hours. HPLC is, however, not sufficient for definitive species identification for many NTM species, including the rapidly growing mycobacteria.

Other newer techniques, which include polymerase chain reaction-restriction length polymorphism analysis (PRA), 16S ribosomal DNA sequencing, multi-gene sequencing, and whole genomic sequencing, also allow NTM to be identified and speciated more reliably and rapidly from clinical specimens [8]. Cost considerations limit widespread adoption of these techniques. However, the recognition that M. abscessus can be separated into more than one subspecies and that there are important prognostic implications of that separation lends urgency to the broader adoption of newer molecular techniques in the mycobacteriology laboratory [9].

Susceptibility testing — Traditional methods of testing mycobacteria for susceptibility to antituberculosis drugs were developed for M. tuberculosis but may not apply directly to NTM (see below). The most common, called the proportion method, compares growth in media containing a specified concentration of the drug to growth in the absence of the drug. Colony growth on drug-containing media that exceeds one percent of the number of colonies growing in the absence of the drug indicates resistance to that drug. For liquid media, such as the BACTEC system, this procedure has been modified to yield comparable results.

Only one "critical concentration" of drug is usually tested in contrast to other common bacteria where many drug concentrations are tested, yielding the familiar "MIC" or minimum inhibitory concentration. For rapidly growing mycobacteria, a broth microdilution system similar to that used for bacterial species is the preferred method. The broth microdilution method for susceptibility testing is also preferred for M. marinum, M. kansasii, and M. avium complex (MAC) [10].

MYCOBACTERIUM AVIUM COMPLEX — Traditionally, M. avium complex was thought to include two species, M. avium and Mycobacterium intracellulare. Due to advances in molecular identification, MAC is actually composed of several different species including M. avium, M. intracellulare, Mycobacterium indicus pranii, Mycobacterium chimaera, Mycobacterium arosiense, Mycobacterium vulneris, Mycobacterium bouchedurhonense, Mycobacterium colombiense, Mycobacterium marseillense, Mycobacterium yongonense, and Mycobacterium timonense [11]. M. avium and M. intracellulare are also separate species; distinguishing between them has been of uncertain clinical value for individual patients, but there is a growing consensus that this distinction is both meaningful and necessary [12]. Nevertheless, most commercial labs do not identify the species within the complex.

Mycobacterium scrofulaceum was once grouped with these organisms as the "MAIS complex" (M. avium-intracellulare-scrofulaceum) but this antedated modern diagnostic methods which readily separate M. scrofulaceum.

MAC are slowly growing organisms that are readily detected by acid fast smear and culture techniques. The organisms grow well in standard media such as MGIT broth and on Middlebrook 7H10 and 7H11 agar. In broth, the organisms do not show clustering or "cording," and, on agar, MAC typically produces small, flat, translucent, smooth colonies that occasionally exhibit a pale yellow color. These colony morphologies differ from M. tuberculosis, which typically shows cording in broth and appears as rough, buff colored colonies on agar. Experienced mycobacterial laboratorians can accurately identify MAC prior to confirmation with more sophisticated techniques.

Highly accurate nucleic acid probes (Accuprobe, GenProbe, Inc) are commercially available that can identify MAC isolates within one day after recognizable growth is evident, a technique currently used by most reference laboratories. These organisms can also be rapidly identified by their mycolic acid patterns from the same samples by HPLC, although the capital investment required limits this methodology to high volume laboratories.

Susceptibility testing — The 2018 Clinical and Laboratory Standards Institute (CLSI) guidelines and the 2020 guidelines from the American Thoracic Society (ATS), European Respiratory Society (ERS), European Society of Clinical Microbiology and Infectious Diseases (ESCMID), and Infectious Diseases Society of America (IDSA) recommend macrolide and amikacin susceptibility testing for all clinically important MAC isolates [10,13]. For MAC, macrolides and amikacin are the only agents for which the MICs have been shown to correlate clinically with in vivo response [10,13]. MIC cut-offs for susceptibility and resistance to macrolides and amikacin are discussed elsewhere. (See "Treatment of Mycobacterium avium complex pulmonary infection in adults", section on 'Antimicrobial susceptibility testing'.)

Because macrolide and amikacin resistance represent acquired mutational resistance, similar to the mechanism of acquired mutational antibiotic resistance with tuberculosis, recommended treatment regimens include companion medications to prevent the emergence of macrolide- or amikacin-resistant organisms. However, routine in vitro susceptibility testing of MAC isolates is not recommended for antibiotics other than macrolides (clarithromycin is the class agent) and amikacin [10]. These include ethambutol, rifampin, rifabutin, clofazimine, moxifloxacin, and ciprofloxacin. For these agents, in vitro results have not been shown to predict clinical (in vivo) response. The value of expanded in vitro susceptibility testing for macrolide-resistant MAC isolates has not been demonstrated.

Susceptibility testing of MAC is difficult and controversial compared with M. tuberculosis [10]. The single critical concentrations of drugs used for the testing of M. tuberculosis were initially chosen to separate treated from untreated strains and provided reasonable predictive values of clinical susceptibility and resistance for tuberculosis. These same values for anti-tuberculosis drugs, such as isoniazid, rifampin, and ethambutol, have not been useful for MAC, since almost all isolates are resistant to these single concentrations [14].

MYCOBACTERIUM KANSASII — Unlike other NTM, M. kansasii has never been found in soil or natural water supplies, but has been recovered consistently from tap water in cities where M. kansasii is endemic [15]. M. kansasii is seen on smear and recovered in culture by techniques designed for M. tuberculosis. It grows readily in the BACTEC broth, as well as on Middlebrook 7H10 agar and Lowenstein-Jensen agar. The organism typically produces rough large colonies that turn bright yellow with exposure to light (photochromogen). A species-specific nucleic acid probe for the identification of M. kansasii is commercially available (Accuprobe, GenProbe, Inc); this assay can be used once growth has occurred in culture.

M. kansasii is readily identified by biochemical tests and by HPLC. On acid fast smears, M. kansasii typically appears as a large, long bacillus which has unusual beading when stained with Ziehl-Neelsen or Kinyoun stains. This finding suggests M. kansasii rather than M. tuberculosis.

The 2018 Clinical and Laboratory Standards Institute (CLSI) guidelines recommends initial susceptibility testing of M. kansasii for rifampin and clarithromycin only [10]. However, if the M. kansasii isolate is rifampin resistant (minimum inhibitory concentration [MIC] >2 mcg/mL), testing of other drugs, including amikacin, fluoroquinolones, doxycycline, rifabutin, linezolid, and sulfamethoxazole, is recommended. MIC testing for ethambutol is no longer recommended because technical difficulties in testing result in inconsistent MIC results [10]. Like all NTM species, M. kansasii isolates are resistant to pyrazinamide at all test concentrations in vitro.

Acceptable susceptibility testing methods include agar proportion and broth micro dilution.

Almost all isolates of M. kansasii are susceptible at to isoniazid at MICs ≤5 mcg/mL and to streptomycin at MICs ≤10 mcg/mL [16]. The single critical concentration cut-offs for isoniazid of 1 mcg/mL and for streptomycin of 2 mcg/mL were used to separate treated from untreated strains of M. tuberculosis and are not useful for M. kansasii. Thus, reported resistance to either isoniazid or streptomycin at these concentrations should be disregarded and not considered in therapeutic decision-making. This observation highlights an important aspect of NTM in vitro susceptibility testing, that the cut-off points for M. tuberculosis susceptibility and resistance rarely apply to NTM species. M. kansasii is an exception with regards to in vitro susceptibility to rifampin because it is the NTM most closely related to M. tuberculosis.

RAPIDLY GROWING MYCOBACTERIA — Rapidly growing mycobacteria (RGM) include numerous species, but the main clinically relevant organisms are the three M. abscessus subspecies (M. abscessus subsp abscessus, subsp massiliense, and subsp bolletii). M. fortuitum and M. chelonae are less frequently encountered as true pathogens [17]. They are generally considered to be environmental saprophytes which are widely distributed in nature; they have been isolated from soil, dust, natural surface and municipal water, wild and domestic animals, and fish [5,18]. The organisms grow readily on Middlebrook 7H10 or 7H11 agar and BACTEC 12B broth, as well as routine bacteriologic media such as 5 percent sheep blood agar or chocolate agar, within seven days. Temperature requirements may also be important since some strains of M. chelonae grow optimally at 30°C rather than 35 to 37°C.

Although an infrequent respiratory pathogen, M. fortuitum is the most common of the RGM encountered in the clinical microbiology laboratory. These organisms may not stain well with the Ziehl-Neelsen or Kinyoun method and may not be recognized readily with the fluorochrome method. RGM are also highly susceptible to the NaOH decontamination procedures which are used to remove other bacteria from specimens. Thus, quantitation of the amount of RGM organisms present by smear or culture can be difficult.

Separation of M. fortuitum from M. chelonae and M. abscessus group can be accomplished in the laboratory if at least two tests (usually iron uptake and nitrate reduction) are used [19]. Separation of M. chelonae and M. abscessus cannot reliably be accomplished with HPLC or commercial DNA probes, but can be accomplished biochemically (utilization of citrate), by drug susceptibility patterns and by polymerase chain reaction-restriction length polymorphism analysis (PRA). The clinical source of the mycobacterial isolate can be an important clue to the identification of the organism. For instance, those in M. abscessus group are common respiratory pathogens, while M. chelonae is an exceedingly rare respiratory pathogen. M. fortuitum is also rarely a true respiratory pathogen except under specific conditions, such as chronic aspiration of gastric contents.

Susceptibility testing — M. fortuitum, M. abscessus group, and M. chelonae are resistant to all of the antituberculosis drugs (rifampin, ethambutol and isoniazid), while M. chelonae and M. abscessus group are also usually resistant to a number of other oral agents (such as doxycycline and sulfonamides) used to treat M. fortuitum [20]. Because of these resistance patterns, routine susceptibility testing to antituberculosis drugs is not recommended for RGM [10,13]. Isolates should be tested against selected antibacterial agents such as clarithromycin, cefoxitin, imipenem, amikacin, linezolid, fluoroquinolones, doxycycline (or minocycline), trimethoprim-sulfamethoxazole, and tobramycin (tobramycin is the preferred aminoglycoside for M. chelonae, which is usually resistant in vitro to amikacin) [10,13]. Caution is necessary when interpreting in vitro susceptibility results for M. fortuitum and M. abscessus because of the presence of an inducible macrolide resistance (erm) gene in these organisms.

In the past, all isolates of M. abscessus group and M. chelonae, and approximately 80 percent of isolates of M. fortuitum, on initial testing were found to be susceptible in vitro to the macrolide, clarithromycin [21]. However, these results were obtained prior to the recognition that M. abscessus subsp abscessus and subsp bolletii (but not M. abscessus subsp massiliense) and M. fortuitum have an inducible macrolide resistance (erm) gene. Macrolide antimicrobial agents act by binding to the 50S ribosomal subunit and inhibiting peptide synthesis. Erythromycin methylase (erm) genes, a diverse collection of methylases that impair binding of macrolides to ribosomes, reduce the inhibitory activity of these agents. The primary mechanism of acquired clinically significant macrolide resistance for RGM is the presence of an inducible erm gene (erm 41) [22,23].

Thus, if an M. abscessus subsp abscessus or subsp bolletii or M. fortuitum isolate is exposed to a macrolide, the erm gene activity is induced with subsequent in vivo macrolide resistance that may not have been reflected by the initial in vitro MIC of the organism for the macrolide. Therefore, the organism may appear to be susceptible in vitro to the macrolide but will not respond to the macrolide in vivo. Detection of this in vivo macrolide resistance requires incubation of an NTM isolate with macrolide for approximately two weeks prior to determining an MIC for the macrolide.

This appears to be one mechanism for the discrepancy between in vitro susceptibility results and in vivo responses for M. abscessus. Certain M. abscessus subsp abscessus isolates, moreover, may also have a mutation inactivating the erm gene, resulting in retention of in vitro and in vivo macrolide susceptibility [24]. Thus, both molecular identification of the erm gene as well as phenotypic analysis of macrolide susceptibility are necessary.

The recommended susceptibility testing method for RGM is the broth microdilution technique with MIC determinations and resistance breakpoints similar to those used for other bacterial species.

INFREQUENT OR DIFFICULT TO CULTURE NTM — A number of infrequently encountered species have been recently described or have become clinically more important. These include M. xenopi, M. simiae, M. malmoense, M. genavense, Mycobacterium haemophilum, and Mycobacterium smegmatis. Susceptibility testing is generally performed on these species, but what drugs to test and how useful they are is just beginning to be understood. Some species require special conditions for growth, such as M. haemophilum (iron or hemin supplementation, 30°C incubation) and M. genavense (broth media with six weeks incubation). The recommended susceptibility methods for M. marinum include agar proportion, and microdilution broth MIC at 30°C. However, Routine MICs are not recommended because all untreated strains have the same susceptibility pattern. Drug testing is necessary only in special circumstances such as therapy failure [10].

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Nontuberculous mycobacteria".)

SUMMARY

●With the availability of 16S ribosomal DNA sequencing and high-performance liquid chromatography (HPLC), polymerase chain reaction-restriction length polymorphism analysis (PRA), and multi-gene and whole-genome sequencing, the number of new species of nontuberculous mycobacteria (NTM) has risen dramatically with the naming of species, such as Mycobacterium genavense, Mycobacterium interjectum, Mycobacterium triplex, Mycobacterium celatum, and Mycobacterium lentiflavum. Approximately 200 species have been recognized in the genus Mycobacterium. (See 'Introduction' above.)

●Within the genus Mycobacterium, four groups of human pathogens can be delineated on the basis of microbiologic, clinical, and epidemiologic characteristics (table 1):

•The Mycobacterium tuberculosis complex

•Mycobacterium leprae

•Slowly growing nontuberculous mycobacteria

•Rapidly growing mycobacteria (see 'Classification' above)

●The most common nontuberculous species causing human disease in the United States are the slowly growing species of the Mycobacterium avium complex (MAC) and Mycobacterium kansasii and the rapidly growing species in the Mycobacterium abscessus group. Less common human pathogens include the slowly growing species Mycobacterium marinum, Mycobacterium xenopi, Mycobacterium simiae, Mycobacterium malmoense, and Mycobacterium ulcerans, and the rapidly growing species Mycobacterium fortuitum and Mycobacterium chelonae. (See 'Classification' above.)

●Microscopic examination after staining and culture using specific media are the cornerstones of the identification of mycobacteria. All mycobacteria share the characteristic of "acid-fastness," ie, after staining with carbol-fuchsin or auramine-rhodamine, they do not decolorize with acidified alcohol. Confirmation of the presence or absence of mycobacteria in clinical specimens has traditionally required culture, because of the relative insensitivity of direct microscopy. (See 'Microscopy' above and 'Culture' above.)

●Highly accurate nucleic acid probes are commercially available that can identify MAC isolates within one day after recognizable growth is evident, a technique currently used by most reference laboratories. A similar nucleic acid probe is also available for the identification of M. kansasii. (See 'Mycobacterium avium complex' above and 'Mycobacterium kansasii' above.)

●Routine in vitro susceptibility testing for clarithromycin and amikacin is recommended for all clinically important MAC isolates. Clarithromycin resistance is associated with a mutation in the 23S rRNA gene, and amikacin resistance is associated with a mutation in the 16S rRNA gene. (See 'Susceptibility testing' above and "Treatment of Mycobacterium avium complex pulmonary infection in adults", section on 'Antimicrobial susceptibility testing'.)

●Routine in vitro susceptibility testing for rifampin and clarithromycin is recommended for M. kansasii isolates. Susceptibility to other drugs should be tested only if the isolate is resistant to rifampin (minimum inhibitory concentration [MIC] >2 mcg/mL). (See 'Mycobacterium kansasii' above.)

●Rapidly growing mycobacterial isolates should be tested against selected antibacterial agents, such as clarithromycin, cefoxitin, imipenem, amikacin, linezolid, fluoroquinolones, doxycycline (or minocycline), trimethoprim-sulfamethoxazole, and tobramycin (tobramycin is the preferred aminoglycoside for M. chelonae, which is usually resistant in vitro to amikacin). Caution is necessary when interpreting in vitro susceptibility results for M. fortuitum and M. abscessus subsp abscessus and subsp bolletii because of the presence of an inducible macrolide resistance (erm) gene in these organisms. (See 'Susceptibility testing' above.)