The slowly growing NTM, including those species usually associated with healthcare-associated diseases, grow well on the same types of media used for cultivation of M. tuberculosis complex (MTBC). Optimal incubation temperatures vary from 28°C (for species typically associated with cutaneous infections such as M. marinum, M. haemophilum, and M. chelonae) to 35°C to 37°C for most of the slowly growing NTM.

M. xenopi grows optimally at 42°C to 43°C but, with prolonged incubation, will grow at 37°C. Most species produce visible colonies on solid agar within 7 to 14 days. More than 60 species of rapidly growing mycobacteria (RGM) are recognized, with this number increasing rapidly with the primary use and availability of 16S ribosomal RNA (rRNA) gene sequencing for identification. The RGM are susceptible to the NaOH decontamination process performed on sputum to facilitate isolation of MTBC. They are nonfastidious microorganisms that produce mature, visible colonies on solid agar in 3 to 7 days. The microorganisms grow well at 30°C to 37°C on standard bacterial media, including 5% sheep’s blood and chocolate agar, and on media specifically formulated for mycobacterial species. Isolation of M. chelonae is optimal at an incubation temperature of 28°C to 32°C.

The reemergence of tuberculosis including multidrugresistant strains (MDR and XDR) as well as the heightened awareness of NTM as human pathogens has fueled an intense effort to develop rapid and accurate methods of identifying mycobacteria at the species level. Traditional identification schemes for slowly growing species utilize growth rates, pigment production, and biochemicals such as niacin production, urease, and catalase. Biochemical tests traditionally used in identifying the pathogenic RGM include the rapid (3-day) arylsulfatase reaction, iron uptake, nitrate reduction, and the ability to utilize mannitol, inositol, and/or citrate as carbon sources (32). Although these biochemical methods (utilized throughout the 1980s) have been replaced by molecular methods and high-performance liquid chromatography (HPLC), some laboratories report the NTM only to a group or complex level.

The decade of the 1990s saw identification of NTM by HPLC (33,34), while the last 10 years have seen a revolution of molecular-based tests, including commercial DNA probes, polymerase chain reaction (PCR) amplification followed by PCR restriction-enzyme analysis (PRA), 16S rRNA rpoB, hsp65, and other gene sequencing (35, 36, 37, 38, 39 and 40). HPLC methods for separation of mycolic acids have allowed the identification of some slowly growing mycobacterial species and some RGM groups or complexes, with a greater specificity and speed than traditional biochemical methods (33,34). Nonradioactive commercial probes are available (Accu-Probe, Gen-Probe) and are routinely utilized for identifying isolates of MTBC, MAC, M. kansasii, and M. gordonae. These probes offer excellent sensitivity and specificity, and because they can be used directly on broth
cultures (usually the first medium to show growth), they have significantly reduced the time for final reporting of results (41,42). Currently, no commercial probes are available for identification of the RGM. Newest among the modern identification methods are adaptations of PCR technology for detection and identification of mycobacteria in clinical samples (43). Two different nucleic acid amplification techniques for assaying directly from sputum are now commercially available (Amplicor, Roche; and MTD test, Gen-Probe) and are approved for detection of M. tuberculosis. Currently, no systems are commercially available for direct detection of NTM from clinical specimens or for species identification of pure cultures, although several novel approaches have been published. One approach utilizes PCR to amplify the gene encoding the 16S rRNA. The amplified fragment is then analyzed by species-specific probes or partial nucleotide sequencing (39) for speciation. A second approach capitalizes on species-specific restriction fragment length polymorphisms (RFLPs) in a PCR-amplified segment of the 65 kDa heat shock protein gene (44, 45 and 46).

Typing systems for RGM have utilized a number of phenotypic and genotypic methods, including detailed species identification, heavy metal and antimicrobial susceptibility patterns, plasmid profiles, multilocus enzyme electrophoresis (MEE), pulsed-field gel electrophoresis (PFGE), and, more recently, random amplified polymorphic DNA (RAPD) PCR (47,48,49,50, 51 and 52). PFGE has proven to be a highly useful tool for strain typing of RGM. This method utilizes restriction endonucleases with rare recognition sites such as XbaI, DraI, and AsnI to generate a small series of large genomic restriction fragments (LRFs), the pattern of which is strain specific. Wallace et al. (48) described the use of PFGE to type M. chelonae and the M. abscessus group with three reference strains, 28 sporadic isolates, and 62 healthcare-associated isolates from 10 healthcare-associated outbreaks. LRF patterns satisfactory for comparison were achieved in 54% of the M. abscessus group and 90% of M. chelonae isolates by using the restriction endonucleases DraI, AsnI, XbaI, and SpeI. The sporadic isolates were all highly variable. Isolates from 5 of 10 outbreaks that gave satisfactory LRF patterns were identical. Strains that had been repetitively isolated from patients over periods of time ranging from 2 to 11 years demonstrated that LRF patterns were highly stable. Previous studies with M. fortuitum with this technique showed similar results, except that satisfactory LRF patterns were obtained with all strains studied (49,53). Environmental water isolates were identical (clonal) to some outbreak strains, indicating that water was the likely source of these past outbreaks. No human carrier or environmental nonwater sources have been identified as an outbreak source by this technique. PFGE is currently the most definitive epidemiologic tool available for comparing suspected outbreak strains of most isolates and species of RGM.

More recently, methods have been introduced for DNA stabilization, which prevent the DNA denaturation seen with PFGE with select strains of bacterial species. Application of one of these methods, the use of hydroxyurea in the running buffer, allows for quality PFGE patterns with the approximate 50% of the M. abscessus strains that produced broken DNA (50).

Another nucleic acid amplification technique, RAPD-PCR or arbitrary primer -PCR, has also been applied to the investigation of outbreaks of the M. abscessus group (50, 51 and 52). This technique offers the advantage of being simpler and unaffected by spontaneous lysis of the DNA sample during preparation for PFGE, as occurred previously with 50% of isolates of M. abscessus (48). Its major disadvantage is that fewer patterns are produced with each primer; hence, at least three primers that produce quality patterns are needed—a finding that likely reflects the closely related character of these strains (50,52).

PFGE has also been used to study clustering or pseudooutbreaks of slowly growing mycobacteria, including MTBC, M. xenopi (53,54), M. kansasii (55,56), M. simiae, (57, 58 and 59), and MAC (60, 61 and 62). Other fingerprint techniques used for slowly growing species include MEE with M. fortuitum (47), the M. abscessus group (63), M. simiae (64), and the use of hybridization with repetitive insertional elements for M. xenopi (53), M. kansasii (55), and M. avium (60). Previously, serotyping of MAC was used for strain typing (62) but currently has been replaced by molecular strain typing.

Recently, a commercial system from DiversiLab system (BioMerieux, Durham, NC), using repetitive elements interspersed throughout the genome, was introduced for strain typing of microorganisms including mycobacteria. The method was reported to be more rapid, required less sample size, and provided equivalent or better than standard RFLP for some species of mycobacteria (65).

Another method that has been evaluated recently with isolates of M. fortuitum and the M. abscessus/chelonae group is the enterobacterial repetitive intergenic consensus (ERIC) PCR. In an outbreak of the M. abscessus subsp. bolletii in Brazil recently, the ERIC PCR showed higher discrimination than PFGE for the M. abscessus group but less discriminatory power among isolates of M. fortuitum (66, 67 and 68).

Finally, multigene sequencing has recently been used to characterize isolates of Mycobacterium in hemodialysis water and may provide a reliable method of DNA strain typing (69).

NTM are not reportable by law in most states, and thus, precise estimates of their incidence and prevalence are not available. Most NTM species, with the exception of MAC, are found in specific geographic areas. Overall, MAC is the most common NTM species recovered in the United States, followed by M. kansasii and the M. abscessus group (1). Although less frequently recovered in the United States, M. xenopi is the second most commonly isolated NTM species in England and Canada (70). In some areas of northern Europe, M. malmoense is second only to MAC (7). Although this species is rare in the United States, M. simiae is second to MAC in some cities in the southwestern United States (57,58,71,72). Tap water and biofilms in the pipes appear to be the major reservoirs for M. kansasii, M. xenopi, and M. simiae, and a reservoir for M. avium and M. intracellulare (73).

In contrast to surveys done in the late 1970s and early 1980s (1,70,74), more recent studies show that there
are now more laboratory isolates of NTM in developed countries, especially MAC, than isolates of M. tuberculosis. The epidemiology of disease due to NTM has changed because of the improvement in laboratory recovery and identification of these species and the increased awareness of the clinician of these species as potential pathogens. The emergence of better antiretroviral therapies for human immunodeficiency virus (HIV) infection (acquired immunodeficiency syndrome [AIDS]) has resulted in a dramatic decline in the incidence of NTM disease in patients with far advanced disease. Among AIDS patients, MAC had been a common mycobacterial cause of opportunistic infection and a frequent cause of disseminated disease.

The RGM are the most commonly described and the most significant NTM for healthcare-associated epidemiology (75). Of the human diseases attributable to this group of microorganisms, over 90% are due to M. fortuitum, the M. abscessus group, and M. chelonae (2,13,21,75). These species readily survive nutritional deprivation and extremes of temperature. For example, most pathogenic species have been shown to grow and survive in distilled water, and they have been identified from soil, dust, domestic animals, and marine life (3,5,73,76,77). Multiple water sources have been identified, including tap water, municipal water, and aquariums (2,3,5,10). Mycobacteria have also been found in high numbers in biofilms on water-delivery devices, such as dental hand pieces (78). Similar biofilms may exist within bronchoscope channels, endoscope washers, ice machines, and water tanks, explaining the tendency of these devices to become colonized with mycobacteria. Biofilms are important not only because they enable bacteria to adhere and persist on artificial surfaces but also because they provide protection from the action of disinfectants (79, 80, 81 and 82).

The pathology of NTM infection can be identical to that of M. tuberculosis. Chronic inflammation, acute suppuration, nonnecrotic epithelioid tubercles, and caseation are all seen on histopathology. The coexistence of granulomatous and acute inflammation (so-called dimorphic inflammatory response) is not seen with tuberculosis but is commonly seen in cervical lymph nodes (83) and cutaneous disease (2,21,84) due to the NTM. Animal models to study the pathology of NTM have been difficult to develop, even when the animals are immunosuppressed (85,86).

Isolation of NTM in the laboratory may represent an environmental or laboratory contaminant, transient patient colonization, or true disease. In the absence of known environmental contamination, isolation of any NTM from a normally sterile site should be considered significant. Contamination of a skin wound with these microorganisms is rare, and even a single positive culture from this site generally indicates disease. Similarly, recovery of these microorganisms from cultures of lymph node specimens or blood is sufficient for establishing the diagnosis of nontuberculous lymphadenitis or disseminated disease, respectively.

In contrast, isolation of NTM from pulmonary specimens can be particularly difficult to evaluate. The American Thoracic Society last published criteria for the diagnosis of NTM pulmonary disease in 2007 (7). According to these criteria, a definitive diagnosis requires compatible clinical symptoms along with characteristic radiographic abnormalities, which are not attributable to any other cause. Multiple cultures of respiratory specimens are required to demonstrate persistent culture positivity. The microorganism must be grown from two acid-fast bacilli (AFB) specimens of sputum or from at least one specimen obtained from a normally sterile site such as a bronchial wash or bronchopulmonary tissue. The clinical syndromes most commonly associated with NTM infections and the microorganisms usually responsible are summarized in Table 39-1. The major risk factor for pulmonary NTM disease appears to be underlying bronchiectasis.

Healthcare-associated mycobacterial infections (almost exclusively due to RGM) have been recognized for more than 25 years and remain relatively common. They have been most often recognized primarily as causes of surgical site infections and postinjection abscesses; however, they have also been reported to cause catheter-related infections, dialysis-related infections, bronchoscope and endoscope contamination, and most recently, infections resulting from cosmetic procedures including plastic surgery, liposuction, and a healthcare-related practice involving subcutaneous injections of minute quantities of various drugs called mesotherapy (47,48,111,112,113,114,115). For outbreaks and sporadic reports of infections due to RGM, there is a strong geographic relationship to the Gulf Coast and southeastern states in the United States. Figure 39-1 demonstrates the focal geography of some of the early outbreaks reported in the United States. Recently, several outbreaks of NTM have occurred in South and Central America (66,116,117,118, 119, 120 and 121).