Introduction
Soon after the discovery of Mycobacterium tuberculosis in 1882 by Robert Koch, several other species of mycobacteria were described. It was not until half a century later, however, that these mycobacteria were recognized to cause disease in humans, and by the 1980s they were known to cause a broad spectrum of disease. Over the years, these organisms have gone by many names, including “mycobacteria other than tuberculosis,” “environmental mycobacteria,” “anonymous or atypical mycobacteria,” and “ nontuberculous mycobacteria ” (NTM), the preferred term in the United States. The epidemic of the human immunodeficiency virus (HIV) heralded a new awareness of NTM infections because of the high rates of disseminated infections due to Mycobacterium avium complex (or, as commonly abbreviated, MAC) and other species. Disseminated NTM infections have declined significantly in HIV-infected populations following the advent of antiretroviral drugs whereas rates of NTM disease in HIV-uninfected populations appear to be increasing.
NTM represent a broad array of organisms that are ubiquitous in our environment. They have been isolated from natural and drinking waters as well as soil, and exposure to these reservoirs is thought to be the source of human infection. At least 160 species of NTM have been identified, and many of these have been reported to cause disease in both immunocompetent and immunocompromised patients. Unlike Mycobacterium tuberculosis, NTM do not appear to be transmitted from human to human in the absence of extraordinary circumstances, they vary greatly in their ability to cause disease, and evidence for latency is lacking. Unfortunately, NTM are difficult to treat because of high levels of in vitro resistance to antimicrobial drugs, which requires long courses of therapy with relatively poor outcomes when compared with tuberculosis. Not surprisingly, these factors result in a treatment cost comparable to other chronic infectious diseases such as HIV infection and acquired immunodeficiency syndrome (AIDS). To date, our lack of understanding of the transmission and pathogenesis of these increasingly important infections has limited our ability to develop public health measures aimed at preventing infection.
Microbiology and Taxonomy
The genus Mycobacterium consists of organisms within the M. tuberculosis complex, Mycobacterium leprae , and NTM. The latter were classified in 1959 by Runyon into additional groups based on pigmentation in the presence or absence of light (photochromogens, scotochromogens, nonchromogenic) and growth characteristics (slow versus rapid). All mycobacteria are characterized by their slow rate of growth when compared with other bacterial species, and NTM are further characterized into rapidly growing and slowly growing organisms: rapid growers can be distinguished by growth in subculture in less than 7 days and slow growers by growth within 2 to 3 weeks. NTM, like all mycobacteria, are also characterized by a thin peptidoglycan layer surrounded by a thick lipid-rich outer membrane. The lipid-rich outer membrane results in a number of properties that allow the organisms to survive in diverse environments. For example, the hydrophobic cell surface allows surface attachment, resistance to disinfectants and antibiotics, slow growth, and tolerance to heat. Additional important properties that allow survival in the environment are the ability to grow in low carbon concentrations (oligotroph) and oxygen concentrations.
The availability of molecular methods has rendered the Runyon classification obsolete for clinical purposes and resulted in a marked increase in the identification of new species. In 1997, approximately 50 species of NTM had been identified, with only 13 described as respiratory pathogens. Currently, there are more than 160 identified NTM species with at least 50 that can be associated with lung infection ( Table 36-1 ). Thus, it is reasonable to ask why there appears to be such a profusion of NTM species in general and those associated with lung disease in particular.
| Slowly Growing Mycobacteria | Rapidly Growing Mycobacteria | ||
|---|---|---|---|
| M. arupense | M. nebraskense | M. abscessus | M. holsaticum |
| M. avium | M. nonchromogenicum | M. alvei | M. houstonense |
| M. asiaticum | M. palustre | M. boenickei | M. mageritense |
| M. branderi | M. parascrofulaceum | M. bolletii * | M. massiliense * |
| M. chimaera | M. phlei | M. brumae | M. mucogenicum |
| M. celatum | M. riyadhense | M. chelonae | M. peregrinum |
| M. florentinum | M. saskatchewanse | M. confluentis | M. phocaicum |
| M. heckeshornense | M. scrofulaceum | M. elephantis | M. septicum |
| M. intermedium | M. senuense | M. fortuitum | M. thermoresistible |
| M. interjectum | M. shimodei | M. goodii | |
| M. intracellulare | M. simiae | ||
| M. iranicum | M. szulgai | ||
| M. kansasii | M. triviale | ||
| M. kubicae | M. triplex | ||
| M. lentiflavum | M. xenopi | ||
| M. malmoense | |||
* Taxonomy is in transition: consider these subspecies of M. abscessus .
The primary explanation can be found in the mycobacteriology laboratory. Before the era of high-resolution methods for organism identification, NTM were identified based on their morphologic and biochemical characteristics as well as their in vitro susceptibility patterns. The inadequacy or insensitivity of these techniques was reflected in the number of organisms that were grouped as “complexes,” such as the “Mycobacterium fortuitum complex,” which included multiple rapidly growing species such as M. fortuitum, Mycobacterium chelonae, and Mycobacterium abscessus. Although it was apparent that multiple species were present with similar growth and morphologic properties, standard laboratory techniques were not adequate to separate them reliably to the species level.
Currently, there are multiple methods for NTM species identification, including high-performance liquid chromatography (HPLC) and DNA probes for the most commonly isolated NTM species. HPLC has been especially useful in the identification of new NTM species but lacks the sensitivity of molecular techniques. The revolutionary change in NTM identification was the advent of readily available DNA sequencing, especially the sequencing of the 16S ribosomal RNA gene, which is highly conserved so that 1% or greater differences in this gene sequence can define an NTM species. Publicly available databases of 16S ribosomal RNA gene sequences allow rapid comparison of mycobacterial isolates to determine whether a new NTM species is present. Alternatively, sequences of the heat shock protein 65, rpoB , secA genes or the entire genomic DNA sequence of an NTM species can be compared to other publicly available gene sequence databases.
The expansion of new NTM species in the past 15 years is, therefore, primarily a consequence of higher resolution identification techniques capable of separating closely related NTM species, rather than a proliferation of new NTM species. Predictably, many of these newly identified species have microbiologic properties very similar to those of other closely related NTM, so that identification of the new NTM species may not be associated with any new or surprising clinical properties. It remains to be seen if this revolution in NTM species identification will have an equally profound impact on the understanding of clinical NTM disease. What is certain, however, is that clinicians will need to become familiar with many more names as new NTM species are identified.
Epidemiology
Incidence and Prevalence
The epidemiology of NTM disease has been difficult to determine because reporting is not mandatory in most countries and differentiation between infection and disease is often difficult. The incidence and prevalence of NTM infections have varied significantly across studies, partially because the studies have used different methodologies and studied different populations. Existing data suggest that the incidence and prevalence of NTM infections are increasing.
Epidemiologic studies of NTM infection and disease use one of three approaches: (1) studying the reactivity to delayed-typed hypersensitivity to mycobacterial antigens; (2) analyzing the reported isolation rate of NTM from laboratories; or (3) using both clinical and laboratory data to define disease more accurately. Studies using delayed-type hypersensitivity reaction to subcutaneously injected mycobacterial antigens have estimated that 11% to 33.5% of the population in the United States have skin test reactivity to NTM. A study used data from the National Health and Nutritional Examination Survey cohort to describe skin reactivity to purified protein derivative B during two time periods. For the years 1971–1972 and 1999–2000, rates of a positive skin test were 11.2% and 16.6%, respectively. Skin test reactivity was noted to increase between the two time periods in foreign-born but not in U.S.-born individuals, suggesting a higher rate of exposure and infection in the foreign-born individuals.
Early laboratory-based studies used consecutive isolates from a well-defined catchment area to estimate the frequency of infection. Survey data from state laboratories in the United States in the early 1980s estimated a prevalence of NTM infection of 1 to 2 cases per 100,000 population. A similar follow-up survey from 1993 to 1996 reported an annual case rate in HIV-uninfected people of 7 to 8 per 100,000, documenting an increase in NTM isolation when compared to the previous survey. These laboratory-based surveys suffer from the fact that they provided isolate-based information but did not take into account the number of patients and whether or not the patients had disease. These early studies have largely been supplanted by better studies based on more rigorous methodologies.
Studies that combine culture and clinical data may be more useful for estimating disease incidence and prevalence. Studies since the 1950s from Czechoslovakia, Wales, Ireland, Australia, and the United States have reported increases in incidence or prevalence. In some reports, the increase in the rate of NTM infection was associated with a decline in the rate of tuberculosis. For example, in Japan from 1971 to 1984, the incidence of pulmonary tuberculosis decreased from 133.1 to 46.3 per 100,000 while the incidence of NTM pulmonary disease increased from 0.89 to 2.15. Similarly, in Switzerland, the incidence of pulmonary tuberculosis decreased from 16.2 to 13.2 per 100,000 over 6 years while the incidence of pulmonary NTM increased from 0.4 to 0.9 per 100,000. In Queensland, Australia, where NTM are reportable, the incidence of clinically significant pulmonary disease rose from 2.2 per 100,000 in 1999 to 3.2 per 100,000 in 2005.
The most significant increases in NTM prevalence have been reported from North America. Marras and coworkers reported an increase in the number of pulmonary NTM isolates in Ontario between 1997 and 2003; during this same time period, the rate of isolation of M. tuberculosis declined in that province. In a subsequent study, the 5-year prevalence of pulmonary NTM disease in Ontario was reported to increase from 29.3 cases per 100,000 in 1998–2002 to 41.3 cases per 100,000 persons in 2006–2010. A recent study reported a significant increase in the annual prevalence of NTM pulmonary disease in patients hospitalized in 11 states between 1998 and 2005 ; annual prevalence increased among men and women in Florida (by 3.2 per year and 6.5% per year, respectively) and among women in New York (by 4.6% per year) but there was no significant increase in California. The annual prevalence of NTM pulmonary disease in U.S. Medicare beneficiaries (all 65 years of age and older) increased from 20 per 100,000 in 1997 to 47 per 100,000 in 2007. The annual percent change for women was 9.1%, which was significantly higher than that for men (6.4%) ( Fig. 36-1 ). In one particularly rigorous study from Oregon, clinical and radiographic studies from patients with NTM respiratory isolates were evaluated. This study found an overall prevalence of 8.6 cases per 100,000 persons and 20.4 per 100,000 in those at least 50 years of age. While not all studies have documented an increase in NTM prevalence, the preponderance of evidence suggests that pulmonary disease due to NTM is increasing.
The reasons for the increase in incidence and prevalence have not been explained, although increased awareness of the disease and improved diagnostic techniques, especially the widespread application of high-resolution chest computed tomography (CT) scanning, could be a factor. A true increase in incidence could be related to changes in the host, such as an aging population, to an increased prevalence of chronic lung disease, or to an increase in the number of immunocompromised individuals. The observation of a decreased incidence of pulmonary tuberculosis and an increased incidence of pulmonary NTM, as noted previously, could be explained by cross-immunity between mycobacterial species. Finally, an increase in the prevalence or virulence of environmental organisms or changes in human behavior that would lead to increased exposure to organisms could be contributing factors.
Geographic Distribution and Variation
NTM have been reported to cause pulmonary disease around the globe, although there is marked variation in the prevalence of disease and the predominant species. In the United States, the southeastern region has long been considered to have the highest rates of infection. NTM have been recovered with higher frequency from water samples in the southeastern versus northern United States, so exposure is likely higher in these regions. However, a recent study reported that among Medicare patients, the states with the highest prevalences of pulmonary NTM were Hawaii (396/1000,000) followed by California (191/100/000). Florida, Louisiana, and Mississippi also had high prevalences ranging between 151/100,000 and 200/100,000.
The reasons for such geographic variation are not well understood. Pulmonary NTM cases were identified from a nationally representative sample of Medicare Part B beneficiaries and their counties of residence were divided into low- and high-risk counties in order to identify potential sociodemographic and environmental risk factors. The investigators identified seven clusters of pulmonary NTM cases within high-risk areas constituting 55 counties. The high-risk counties were then compared with 746 low-risk counties. High-risk counties were larger, had greater population densities, and higher education and income levels than the low-risk counties. In addition, high-risk counties had higher mean daily potential for evapotranspiration (the sum of evaporation and plant transpiration of water into the atmosphere) and areas covered by surface water. They were also more likely to have higher copper and sodium and lower manganese levels in the soil. Thus, specific environmental factors correlate with the rates of pulmonary NTM infection.
The predominant species in North America have been MAC, followed by Mycobacterium kansasii, M. abscessus, M. fortuitum, M. chelonae, and M. scrofulaceum. MAC has also been reported as the predominant species in Central and South America, Europe, and Asia. The predominant species of MAC also varies from region to region. In Europe and South America, M. avium is the predominant species whereas M. intracellulare is the predominant species in Australia and South Africa. Mycobacterium xenopi is common in Europe and Canada, whereas Mycobacterium malmoense is more common in Northern than Southern Europe. Populations of miners in Czechoslovakia and South Africa have been reported to have high rates of M. kansasii infection.
Transmission and Pathogenesis
Transmission
NTM are ubiquitous in the environment and have been found in natural and drinking waters, biofilms, soil, and aerosols. The presumed source of infection is exposure to these environmental reservoirs because human-to-human transmission has been documented only under extraordinary circumstances. Studies using genotyping techniques, such as pulsed-field gel electrophoresis, variable number tandem repeat (VNTR) analysis, and whole genome sequencing have been able to isolate the same strains of NTM from patients and their environments. Although the mechanism by which an environmental exposure eventually leads to pulmonary infection is poorly understood, it has been hypothesized that exposure to infected aerosols can lead to infection and possibly disease. In a study to determine whether household plumbing could serve as a source for a patient’s NTM isolate, water samples were obtained from the households of 37 patients across the United States. Seventeen (46%) of the households yielded at least one mycobacterial isolate of the same species found in the patient and in seven patients, the isolate had the same genotype. In a recent report from Australia, 35% of patients with NTM lung disease had the same species isolated from their home water supply. In other reports, there are at least two cases that had the same strain of MAC isolated from the patients’ respiratory specimens and water from their shower head. In another report, the patient’s strain of MAC was isolated from potting soil in the home. There have also been reports of outbreaks in which patients’ isolates matched environmental isolates of NTM. In a recent study using interviews of HIV-negative adults with MAC, aerosol-generating activities and home and garden water supply features were not found to be risk factors but prior lung disease and immune-suppressing drugs were.
Until recently, person-to-person transmission was considered an unlikely mode of transmission. However, two recent studies have described possible transmission in cystic fibrosis (CF) clinics. The first report described a potential outbreak of M. abscessus subsp massiliense in a CF clinic in Seattle. A smear-positive patient may have transmitted infection to four other patients with CF at the same clinic; the infecting strain was indistinguishable by pulsed-field electrophoresis and polymerase chain reaction analysis. An additional report from the United Kingdom also described transmission of M. abscessus subsp massiliense at a CF center. The authors noted that direct person-to-person transmission was unlikely but that cross infection in the hospital setting was likely.
Extrapulmonary NTM infections in immunocompetent patients are usually the result of a puncture wound or surgery. A recent outbreak of skin and soft tissue infections due M. abscessus subspecies massiliense involved more than 2000 patients who had undergone invasive procedures such as laparoscopic, arthroscopic, plastic, or cosmetic surgery.
Factors Associated with Infection
A prospective study using skin testing data from Palm Beach, Florida, reported that 32.9% of 447 participants in a population-based random household survey had a positive reaction to M. avium antigens. Independent predictors of a positive reaction included black race, birth outside the United States, and more than 6 years’ cumulative exposure to soil. Exposure to water, food, and pets was not associated with skin test reactivity. Using National Health and Nutritional Examination Survey data, investigators reported similar findings with regard to sensitization to Mycobacterium intracellulare : male gender, non-Hispanic black race, and birth outside the United States were each independently associated with sensitization. The highest rate of skin test reactivity was found in persons who were 20 to 39 years of age; among individuals 20 years of age and older, working in agriculture or construction was strongly associated with sensitization. These two studies are interesting in that skin test reactivity to either M. avium or M. intracellulare antigens was associated with factors likely associated with soil exposure and were more common in men and foreign-born individuals. However, disease seems to be more common in older women and, at least in the United States, in U.S.-born. Additionally, the source for M. avium infection, frequently associated with nodular/bronchiectatic MAC lung disease, appears to be primarily municipal (tap) water in the United States, whereas M. intracellulare infection is acquired through some other source, likely soil. Thus, it is likely that the risk factors for infection are different from those associated with disease.
Factors Associated with Disease
Most NTM are significantly less pathogenic than M. tuberculosis and likely require some degree of host impairment to establish disease. A number of risk factors for disease have been described, and these can be subdivided broadly into three groups: (1) factors impairing host immunity, (2) factors leading to impaired local (lung) immunity, and (3) factors relating to the species. These factors can be reduced further into the “unusual dose” model and “susceptible person” model. In the “unusual dose” model, it is postulated that individuals who become infected do so because of an unusually large exposure to NTM, whereas in the latter model, it is assumed that some susceptibility is necessary for infection to manifest. It is likely that, in most patients, both models of pathogenesis are at play to varying degrees.
Increasing age has been consistently described as a risk factor for pulmonary NTM, in patients both with and without underlying lung disease. The most consistently reported risk factor is male gender, though there is also a population of elderly females who are at risk for nodular bronchiectatic lung disease. Perhaps the most important risk factor for the development of pulmonary NTM disease is underlying chronic lung disease. NTM disease has been described in association with CF, chronic obstructive pulmonary disease (including alpha 1 -antitrypsin deficiency), cavitary lung disease, pneumoconiosis, bronchiectasis, prior tuberculosis, and pulmonary alveolar proteinosis. Studies have documented a high prevalence of NTM in sputum cultures from patients with CF, with estimates ranging from 3% to 19.5% prevalence of NTM, the majority of which were MAC.
Patients with NTM lung disease often have associated gastroesophageal reflux disease (GERD) or other esophageal disorders. Historically, it has been reported that patients with rapidly growing mycobacteria (RGM) often have associated esophageal disorders ; recent reports have reported that patients with lung disease due to MAC, a slow grower, also have a high frequency of GERD. In a report from South Korea that used pH probes to diagnose GERD, the authors reported that the prevalence of GERD was 26%, although only one fourth of these patients had typical GERD symptoms. Patients with M. abscessus infection were more likely to have GERD than those with MAC, although the difference did not reach statistical significance.
An interesting patient population with NTM pulmonary disease consists of postmenopausal women who often have certain associated morphologic features. This constellation was first described by Prince and colleagues in 1989 and later referred to as the “Lady Windermere syndrome,” after the lead in Oscar Wilde’s play Lady Windermere’s Fan, who was characterized by her fastidious behavior in actively suppressing a chronic cough. The morphologic features such as pectus excavatum, scoliosis, thin body habitus, and mitral valve prolapse were described subsequently by Iseman and others. Recently, investigators reported that women with pulmonary NTM disease were taller and thinner and weighed less than matched control subjects ( Fig. 36-2 ): 51% had scoliosis, 11% pectus excavatum, and 9% mitral valve prolapse, all significantly more common than in reference populations. Another feature of this syndrome is the predilection for right middle lobe and lingular bronchiectasis. To date, extensive evaluation of the immune system of these patients has shown mixed results, but mutations in the CF transmembrane conductance regulator (CFTR) gene are common. Among 103 women with pulmonary NTM, some intriguing cytokine relationships were found to be abnormal: compared with the findings in uninfected control subjects, in those with NTM, the normal relationship between the adipokines, leptin and adiponectin, and body fat was lost and interferon-γ (IFN-γ) and interleukin (IL)-10 levels were significantly suppressed in stimulated whole blood. These intriguing in vivo and in vitro findings suggest a predisposing immunophenotype for NTM.
NTM in AIDS most commonly causes disseminated disease that presents with nonspecific symptoms such as fever, night sweats, diarrhea, abdominal pain, and lymphadenopathy, with MAC being the most frequent isolate (also see Chapter 90 ). Despite isolation of MAC from the sputum in up to 10% of AIDS patients with CD4 T-cell counts less than 50 cells/µL, pulmonary disease due to MAC is uncommon. Pulmonary disease has been reported in 2.5% to 8% of patients with disseminated MAC and has rarely been reported in the absence of dissemination. Patients with pulmonary disease tend to have higher CD4 counts and focal alveolar opacities, which rarely cavitate. Mycobacterium kansasii can also cause disease in patients with HIV and AIDS. In one study, the risk for infection with M. kansasii was increased 150-fold in HIV-infected patients and 900-fold in those with AIDS. In contrast to MAC, M. kansasii should always be treated as a potential pathogen.
Pulmonary disease due to NTM has been described in several other immunocompromised populations, including transplant recipients. Rates may be as high as 6.5% following heart or lung transplantation, and 2.9% following bone marrow transplantation, but are probably much lower in patients following liver or kidney transplantation. M. abscessus infection in patients with CF who have undergone lung transplantation has been associated with severe and sometimes fatal disease.
Multiple cytokines are necessary for containment of mycobacterial infections. The most important among these are IFN-γ, IL-12, and tumor necrosis factor. Patients with defects in any of these cytokines are susceptible to infection with NTM. Mutations in the genes that code for IFN-γ receptor 1 (IFN-γR1), IFN-γ receptor 2, IL-12 p40, and the IL-12 receptor have all been shown to lead to human disease. The dominant and recessive IFN-γR1 deficiencies have distinct clinical presentations. The adult onset immunodeficiency that can result from the development of autoantibodies against IFN-γ is associated with severe disseminated opportunistic infections including NTM. Mutations in GATA2 are associated with autosomal dominant and sporadic monocytopenia and disseminated mycobacterial infections called the MonoMAC syndrome. Large numbers of patients with inflammatory bowel disease, rheumatoid arthritis, or psoriatic arthritis are treated with tumor necrosis factor inhibitors. Although initially shown to predispose to tuberculosis, recent reports have linked these therapies to NTM infections as well. Patients with suspected or known mycobacterial infections should not receive these medications without proper antimycobacterial therapy. A potential risk factor for NTM lung infection in CF patients is the prolonged use of azithromycin, which inhibits autophagy, an intracellular mechanism that constrains mycobacterial infection. Prolonged exposure to azithromycin in this context can also select for macrolide resistance in MAC.
Although most studies have focused on evaluating possible host abnormalities that could lead to NTM disease, microbial factors are likely to be important as well. Recent studies have identified isolates and phenotypes associated with increased virulence in in vitro models. M. abscessus is known to exist in a smooth and rough phenotype. Limited clinical data as well as data from mouse models suggest that the rough may be more virulent than the smooth phenotype. It has been reported that the presence of glycopeptidolipids on the smooth strains mask agonists of Toll-like receptor 2 and prevent induction of the proinflammatory cytokine, tumor necrosis factor, allowing the organisms to colonize and form biofilms in the lung airways. If the organism then switches to a rough phenotype, the glycopeptidolipids are lost, allowing the strain to elicit an inflammatory response that paradoxically promotes invasive infection. Moreover, a recent population-based assessment of the clinical significance of NTM in a region of The Netherlands demonstrated a wide range of pathogenicity among different NTM species.
Diagnosis and Management of Specific Pathogens
NTM lung infection presents with diverse manifestations ( Table 36-2 ). This chapter focuses on the most common clinical presentations in the immunocompetent hosts: chiefly, tuberculosis-like (cavitary) disease and disease associated with nodules and bronchiectasis (nodular/bronchiectatic disease). The diagnostic evaluation usually consists of (1) assessment for the presence of one or more compatible symptoms, which are usually insidious in onset (cough, sputum production, fatigue, weight loss, fever, hemoptysis); (2) radiographic evaluation, which frequently includes high-resolution computed tomography (HRCT) scans of the chest; and (3) microbiologic evaluation, which usually consists of three or more sputum specimens for microscopy and mycobacterial culture and/or collection of specimens bronchoscopically. The most important diagnosis to exclude is tuberculosis. Additional laboratory evaluation aimed at identifying potential reasons for the underlying infection may include testing for CF, alpha 1 -antitrypsin deficiency, primary ciliary dyskinesia, or esophageal disorders in patients with pulmonary disease and, in the case of disseminated disease, the presence of cell-mediated immunodeficiency or anti-IFN-γ autoantibodies.
|
The diagnosis of NTM lung disease can be challenging. Unlike pulmonary tuberculosis, in which, barring laboratory contamination, a single positive culture of sputum or of a bronchoscopic specimen establishes the diagnosis, confirmation of the diagnosis of NTM lung disease usually requires repeated isolation of a particular NTM species. Diagnostic criteria have been developed to aid the clinician in the evaluation of persons suspected of having pulmonary NTM disease. The NTM diagnostic criteria outlined in Table 36-3 are based on experience with common and well-described respiratory pathogens such as MAC, M. kansasii, and M. abscessus. However, it is important to note that, because of the varying pathogenicity among the many NTM species, no single diagnostic approach will work for all cases. Diagnostic criteria that are too sensitive promote overdiagnosis and the unnecessary exposure of patients to potentially toxic antimicrobial medications. By contrast, overly rigid diagnostic criteria put patients at risk for undertreatment and progressive NTM disease. Fortunately, NTM lung disease is usually sufficiently indolent that there is usually time for a careful assessment to determine with confidence the presence or absence of significant disease.
| Specimen | Results |
|---|---|
| At least three sputum results available with: or | Two positive cultures regardless of the results of AFB smear |
| Single available bronchial wash or lavage with: or | One positive culture regardless of the results of AFB smear |
| Tissue biopsy with: | Compatible histopathology (granulomatous inflammation) and a positive biopsy culture for NTM |
| Compatible histopathology (granulomatous inflammation) and a positive sputum or bronchial wash culture for NTM |
The difficulty in determining the clinical significance of an NTM respiratory isolate results from several factors. NTM isolates can contaminate clinical specimens, usually from environmental sources. Many NTM species, such as Mycobacterium gordonae, are generally nonpathogenic and almost invariably represent specimen contamination rather than true infection. NTM can be found in potable (municipal or tap) water (M. kansasii, MAC, Mycobacterium simiae, M. abscessus, M. xenopi ) , so that their presence in a clinical specimen can be due to waterborne contamination of the specimen, even if the isolated species is one that is capable of causing clinical disease. The recovery of even potentially virulent and pathogenic NTM may be associated with the absence of active or progressive disease, for reasons that are not understood. Unfortunately, no easily applied algorithm addresses all NTM species in all clinical circumstances. The clinician must have some knowledge of the disease-producing potential of NTM species and also an awareness of the circumstances associated with the isolation of the NTM species.
A single positive sputum culture for NTM is usually regarded as indeterminate for the diagnosis of NTM lung disease. In contrast, when two or more sputum cultures are positive, the diagnosis of disease is more likely. For example, 98% of patients with two or more positive sputum cultures for MAC had evidence of progressive disease during follow-up. A single positive NTM culture from bronchoscopy can be diagnostic for disease (see Table 36-3 ), so repeated bronchoscopy is not appropriate if its only purpose would be to obtain multiple positive cultures. However, the clinician must keep in mind those NTM species that are usually respiratory contaminants (especially M. gordonae ) and those NTM species that can be found in tap water (discussed earlier) since “pseudo-outbreaks” of NTM disease can be the consequence of bronchoscopic equipment that has been inappropriately rinsed with tap water. Patients who are suspected of having NTM lung disease but do not meet the diagnostic criteria should be followed clinically until the diagnosis is either firmly established or excluded, a process that can take months to years.
Diagnosis of NTM lung disease does not necessitate the institution of therapy, which is a decision based on potential risks and benefits of therapy for individual patients. Factors that might influence the decision to treat NTM lung disease include the virulence of the NTM pathogen and the potential for disease progression, the severity (or lack of severity) of symptoms and radiographic findings, the presence of known indolent disease, and the presence of advanced age and/or severe comorbid conditions (with limited life expectancy); another factor is the inability to tolerate the prolonged and sometimes toxic antimicrobial regimens for NTM disease. Again, for optimal management of NTM lung disease, there is no substitute for physician familiarity with NTM pathogens and with the individual patient.
Other clinical manifestations of NTM infection include lymphadenitis, disseminated disease, and skin, soft tissue, and bone disease. Lymphadenitis is the most common NTM disease manifestation in children and is usually due to MAC or, less commonly, Mycobacterium haemophilum or M. scrofulaceum. The most important differential diagnosis is tuberculosis lymphadenitis, although NTM account for approximately 90% of mycobacterial lymphadenitis in American children (but only 10% in adults). Symptoms are usually minimal, with frequent unilateral involvement of submandibular, submaxillary, preauricular, or cervical lymph nodes. Skin and soft tissue infections are usually due to the “rapidly growing mycobacteria,” M. abscessus, M. fortuitum, M. chelonae, or M. marinum, and are the result of direct inoculation after either trauma or surgery. Dissemination of NTM pathogens is most often associated with the severe immunosuppression of advanced AIDS and is caused by MAC. Disseminated NTM infections can be seen in other immune-compromised states, such as organ transplantation, autoantibodies to IFN-γ, or in association with indwelling foreign bodies such as venous catheters, dialysis catheters, or other prosthetic devices.
Laboratory Diagnosis
The diagnosis of NTM disease is based on isolation of these organisms from clinical specimens. Cultures should include both solid and broth media for sensitive detection of mycobacteria and, ideally, a semiquantitative reporting of colony counts. The optimal temperature for growth of most NTM species is between 28°C and 37°C, although some species require either higher or lower temperatures for optimal growth and some species require special supplementation for recovery from culture. Most NTM grow within 2 to 3 weeks of subculture but the group of rapidly growing NTM appear within 7 days of subculture.
Identification of specific species can be based on phenotypic, chemotaxonomic, and molecular methods. NTM can be categorized by their growth rate and pigmentation, although these characteristics are not specific enough for final speciation. In addition, biochemical testing can be used to help speciate the NTM. However, none of these procedures is sufficient to differentiate all NTM, particularly some of the newly identified species.
Additional techniques for speciation include HPLC, nucleic acid probes, polymerase chain reaction and other amplification methods, and nucleic acid sequencing. HPLC, which analyzes the chromatographic profile of the mycolic acids extracted from the bacterial cell wall, can be used to speciate a large number of NTM. Identification of mycobacteria by 16S ribosomal DNA sequencing provides more accurate determination of the species, although neither HPLC nor 16S ribosomal DNA sequencing can differentiate all species of NTM. The commercially available AccuProbe technology (Gen-Probe, San Diego, CA) is currently recommended for identification of M. tuberculosis complex, M. avium complex (as well as M. avium and M. intracellulare separately), M. kansasii, and M. gordonae.
The clinical usefulness of drug susceptibility testing in the management of patients with NTM disease remains controversial because in vitro results do not correlate well with clinical outcomes for some mycobacterial species. For slowly growing mycobacteria, no single susceptibility method is recommended for all species. For MAC, a broth-based culture method—with both microdilution and macrodilution methods—is considered acceptable. Initial isolates should be tested for response to clarithromycin, as should isolates from treatment failures and relapses, from patients who have taken macrolides previously, and from AIDS patients who develop bacteremia on macrolide prophylaxis. For M. kansasii, isolates should be tested for response to rifampin (rifampicin) because resistance to rifampin is associated with treatment failure or relapse. If rifampin resistance is documented, additional drugs should be tested. For RGM, broth microdilution minimal inhibitory concentration determination for susceptibility testing is recommended.
Slowly Growing Mycobacteria
Although there are fewer species of slowly growing NTM than rapidly growing species, the slow growers are more common causes of lung disease (see Table 36-1 ). The slow growers include more than 70 species with a wide range of pathogenicity, from organisms such as M. kansasii that are probably second only to M. tuberculosis in terms of disease-producing capability to M. gordonae, which rarely causes lung disease.
Mycobacterium avium Complex
Mycobacterium avium complex includes the NTM species M. avium, of which there are several subspecies, M. intracellulare, M. chimaera, M. colombiense, and M. vulneris, as well as some less well-described species. As noted previously, M. avium and M. intracellulare are likely acquired from different environmental sources. Differentiation of the separate MAC species is not possible with routine laboratory techniques, although DNA probes are available for M. avium and M. intracellulare. Currently, in most circumstances, the differentiation of M. avium and M. intracellulare does not make a significant difference clinically or therapeutically (although M. avium is more often seen with disseminated disease and, in the United States, M. intracellulare is more often isolated as a respiratory pathogen). However, recent data from South Korea suggest that M. intracellulare may be more virulent than M. avium: compared to those with M. avium , patients with M. intracellulare presented with more severe disease and had a worse prognosis.
MAC lung disease typically presents as apical fibrocavitary lung disease similar to tuberculosis, sometimes with large cavities, usually in males but also in females in their fifth and sixth decades who have a history of cigarette smoking and alcohol abuse ( Fig. 36-3 ). If left untreated, this form of disease is generally progressive, can result in extensive cavitary lung destruction, and is associated with increased mortality compared with noncavitary MAC lung disease. The necessity for starting therapy with cavitary MAC lung disease is clearly much more pressing than with noncavitary MAC disease.
MAC lung disease also presents with nodular and interstitial nodular opacities, which frequently involve the right middle lobe or lingula, predominantly in postmenopausal nonsmoking white females ( Fig. 36-4 ; eFig. 36-1 ). This form of disease, termed nodular/bronchiectatic disease, tends to have a much slower progression than cavitary disease, so that long-term follow-up may be necessary to demonstrate clinical or radiographic changes. Even with this more indolent form of disease, however, death may result from progressive disease. Nodular/bronchiectatic MAC lung disease is radiographically characterized by HRCT findings that include multiple small peripheral bronchiolocentric pulmonary nodules with branching configurations, and cylindrical bronchiectasis (see Fig. 36-4 ; see eFig. 36-1 ). The HRCT pattern of these predominantly peripheral small nodular branching opacities has been described as “tree-in-bud,” and reflects inflammatory changes including bronchiolitis.
Patients with nodular/bronchiectatic MAC lung disease often have additional microbiologic findings associated with bronchiectasis, including respiratory cultures positive for Pseudomonas aeruginosa, Nocardia species, and occasionally other NTM such as M. abscessus. Because nonmycobacterial exacerbations of bronchiectasis often complicate the assessment and management of MAC disease, strategies aimed at treatment of the bronchiectasis may improve patients’ symptoms.
For many patients, perhaps the majority of patients with nodular/bronchiectatic disease, it is unknown if bronchiectasis is the result of the mycobacterial infection or if bronchiectasis due to some other process then predisposes to subsequent mycobacterial infection. In some patients with NTM lung disease, bronchiectasis is clearly the result of diseases such as CF, primary ciliary dyskinesia, or alpha 1 -antitrypsin deficiency and thus antedates the MAC disease. The routine evaluation of underlying causes of bronchiectasis, such as CF and alpha 1 -antitrypsin deficiency, in patients with nodular/bronchiectatic MAC disease is currently controversial and no consensus exists.
Therapy of MAC Lung Disease.
Several aspects of treatment for MAC lung disease are difficult to explain, and even counterintuitive. The greatest misunderstanding about treatment regimens for NTM pathogens results from the expectation that all NTM infections should respond in a predictable manner to antimicrobial therapy, similar to infections due to M. tuberculosis. However, even when treatment regimens are based on in vitro susceptibility testing, the NTM pathogen very often does not respond to antimicrobial agents as predicted from the susceptibility results. The most difficult and frustrating aspect of NTM therapy for many clinicians is the lack of a clear association between the in vitro susceptibility results and the in vivo, clinical response. For many NTM, including MAC, laboratory cutoffs for “susceptible” and “resistant” do not have a demonstrable clinical correlate and have not been confirmed to be clinically meaningful. Response to treatment of disease caused by MAC correlates primarily with in vitro susceptibility to macrolides (clarithromycin and azithromycin), and not to most other agents. However, evidence is emerging that in vitro susceptibility or resistance to amikacin does impact the outcomes of treatment regimens that include amikacin. Several other NTM species (e.g., M. abscessus, M. simiae, M. malmoense, M. xenopi ) lack established correlation between in vitro susceptibilities and in vivo responsiveness to any antimicrobial agent. The reason for the discordance between laboratory susceptibility testing results and clinical benefit for many NTM is not known. Accordingly, clinicians must use in vitro susceptibility data for many NTM with the awareness that the results are an imperfect guide to treatment outcome. Two recent reviews summarize the multiple and complicated factors that likely impact this troubling aspect of NTM disease therapy.
Because the macrolides—clarithromycin and azithromycin—are the principal and most important antimicrobial agents for which there is a demonstrated correlation between in vitro susceptibility and in vivo response for MAC lung disease, these agents, partnered with ethambutol, serve as the cornerstones of MAC therapy ( Table 36-4 ). As also shown, this regimen must be reinforced with a rifamycin and, possibly, an injectable aminoglycoside (amikacin or streptomycin), because companion drugs are necessary to prevent the emergence of macrolide-resistant MAC isolates. Macrolides should never be used as monotherapy for treatment of MAC disease, either pulmonary or disseminated.
