70 - 185 Nontuberculous Mycobacterial Infections

185 Nontuberculous Mycobacterial Infections

randomized controlled trial has shown that single-dose rifampin, given once to household contacts, neighbors, and social contacts, reduces the recipients’ risk of leprosy by ~60%. Implementation studies have shown that PEP with single-dose rifampin is feasible and well accepted by patients, contacts, and health workers in a variety of health care settings. Furthermore, modeling studies have indicated the potential impact of PEP on transmission of M. leprae in endemic populations. This intervention was included in the 2018 WHO Guidelines for the Diagnosis, Treatment, and Prevention of Leprosy and is currently being introduced in many countries. Research is ongoing into enhanced PEP regimens for those close contacts who are at increased risk of leprosy (e.g., blood-related household contacts and close contacts of multi­ bacillary leprosy patients).

“Zero Leprosy”  The WHO has formulated its new Global Lep­ rosy Strategy 2021–2030. As in the organization’s previous strategy, a holistic approach to leprosy control is advocated, focusing on zero infection and disease, zero disability, and zero stigma and discrimina­ tion. For 2030, the WHO is setting ambitious targets of achieving 120 countries with zero new autochthonous leprosy cases, reducing the annual number of new cases detected by 70%, reducing the rate of new cases with grade 2 disability per million population (as a proxy for detection delay) by 90%, and reducing the rate of new child cases with leprosy per million children (as a proxy for recent transmission) by 90%. Widespread implementation of PEP with single-dose rifampin is one of the key strategies to achieve these goals. The “Triple Zero Strat­ egy” (zeroleprosy.org) has also been embraced by the partners united in the Global Partnership for Zero Leprosy, the International Federation of Anti-Leprosy Associations, the Novartis Foundation, the Sasakawa Health Foundation, and the International Association for Integration, Dignity, and Economic Advancement. PART 5 Infectious Diseases In July 2023, the WHO launched new “Technical Guidance on Interruption of Transmission and Elimination of Leprosy Disease” (www.who.int/publications/i/item/9789290210467). This contains clear milestones, definitions, and cutoffs for interruption of transmission and elimination of leprosy disease. Using the Leprosy Elimination Framework and the accompanying tools—the Leprosy Elimination Monitoring Tool and Leprosy Programme and Transmission Assessment— countries can track their progress toward these milestones in detail. The outlook for achieving “zero leprosy” is better than ever before, but this goal is admittedly very ambitious. It can be reached only when all leprosy-endemic countries enhance their leprosy control activities to include (1) active case-finding strategies, including improved diag­ nosis; (2) contact screening; (3) implementation of PEP; (4) improved prevention of disability services; and (5) activities to reduce stigma and discrimination and to promote the social inclusion and mental well-being of affected patients and their families. Coincident with these efforts, an important threat must be confronted. With the waning of interest in leprosy and the integration of management of the disease into nonspecialized health systems, the number of medical doctors and health workers at the primary care level who have experience in diagnosing and treating leprosy has decreased substantially all over the world. Once lost, expertise is difficult to regain. Therefore, new energy and resources need to be invested in bolstering technical capacity for all aspects of leprosy services, with a view to strengthening the health system in an integrated way and leaving no one behind. Acknowledgment We thank Dr. Colette L.M. van Hees, dermatologist at Erasmus MC, University Medical Center Rotterdam, for critical review of this chapter. ■ ■FURTHER READING Bratschi MW et al: Current knowledge on Mycobacterium leprae transmission: A systematic literature review. Lepr Rev 86:142, 2015. Collin SM et al: Systematic review of Hansen disease attributed to Mycobacterium lepromatosis. Emerg Infect Dis 29:1376, 2023. Fróes LAR Jr et al: Bacterial, fungal and parasitic co-infections in leprosy: A scoping review. PLoS Negl Trop Dis 17:e0011334, 2023. Kumar B, Kar HK (eds): IAL Textbook of Leprosy, 2nd ed. New Delhi, Jaypee Brothers Medical Publishers (P) Ltd, 2017.

Scollard DM, Gillis TP (eds): International Textbook of Leprosy. Available at https://internationaltextbookofleprosy.org. Accessed February 17, 2024. Smith WC et al: The missing millions: A threat to the elimination of leprosy. PLoS Negl Trop Dis 9:e0003658, 2015. World Health Organization: Guidelines for the diagnosis, treat­ ment and prevention of leprosy. New Delhi, WHO Regional Office for South-East Asia, 2018. Available at https://apps.who.int/iris/ handle/10665/274127.  Accessed February 17, 2024. Steven M. Holland

Nontuberculous

Mycobacterial Infections Several terms—nontuberculous mycobacteria (NTM), atypical myco­ bacteria, mycobacteria other than tuberculosis, and environmental mycobacteria—all refer to mycobacteria other than Mycobacterium tuberculosis, its close relatives (M. bovis, M. caprae, M. africanum,

M. pinnipedii, M. canetti), and M. leprae. The number of identified species of NTM is growing and will continue to do so because of the use of DNA sequence typing for speciation. The number of known spe­ cies currently exceeds 199. NTM are highly adaptable and can inhabit hostile environments, including industrial solvents. ■ ■EPIDEMIOLOGY NTM are ubiquitous in soil and water. Specific organisms have recurring niches, such as M. simiae in certain aquifers, M. fortuitum in pedicure baths, and M. immunogenum in metalworking fluids. Most NTM cause disease in humans only rarely unless some aspect of host defense is impaired, as in bronchiectasis, or breached, as by inoculation (e.g., liposuction, trauma, cardiac surgery). There are few instances of human-to-human transmission of NTM, which occurs almost exclusively in cystic fibrosis. Because infections due to NTM are rarely reported to health agencies and because their identification is sometimes problematic, reliable data on incidence and prevalence are lacking. Disseminated disease denotes significant immune dysfunction (e.g., advanced HIV infection), whereas pulmonary disease, which is much more common, is highly associated with pulmonary epithelial defects but not with systemic immunodeficiency. In the United States, the incidence and prevalence of pulmonary infection with NTM, mostly in association with bronchiectasis (Chap. 301), have for many years been severalfold higher than the corresponding figures for tuberculosis, and rates of the former are increasing among the elderly as rates of tuberculosis continue to fall. Among patients with cystic fibrosis, who often have bronchiectasis, rates of clinical infection with NTM range from 3 to 15%, with even higher rates among older patients. Although NTM may be recovered from the sputa of many individuals, it is critical to differentiate active disease from commensal harboring of the organisms. A scheme to help with the proper diagnosis of pulmonary infection caused by NTM has been developed by the American Thoracic Society and is widely used (https:// doi.org/10.1093/cid/ciaa241). The bulk of nontuberculous mycobacte­ rial disease in North America is due to M. kansasii, organisms of the M. avium complex (MAC), and organisms in the M. abscessus complex. In Europe, Asia, and Australia, the distribution of NTM in clinical specimens is roughly similar to that in North America, with MAC spe­ cies and rapidly growing organisms such as M. abscessus encountered frequently. M. xenopi and M. malmoense are especially prominent in northern Europe. M. ulcerans causes the distinct clinical entity Buruli ulcer, which occurs throughout tropical zones, especially in western Africa.

M. marinum is a common cause of cutaneous and tendon infections in coastal regions and among individuals exposed to fish tanks or swim­ ming pools. The true international epidemiology of infections due to NTM is hard to determine because the isolation of these organisms often is not reported and speciation often is not performed for M. tuberculosis or NTM. The latter issue poses an especially important problem during therapy for tuberculosis when smears positive for acid-fast bacilli are considered evidence of treatment failure. The increasing ease of iden­ tification and speciation of these organisms is already having a major impact on the description of the dynamic international epidemiology of tuberculosis and NTM infections. ■ ■PATHOBIOLOGY Because exposure to NTM is essentially universal and disease is rare, it can be assumed that normal host defenses against these organisms must be strong and that otherwise healthy individuals in whom sig­ nificant disease develops are highly likely to have specific susceptibility factors that permit NTM to become established, multiply, and cause disease. At the advent of HIV infection, CD4+ T lymphocytes were recognized as key effector cells against NTM; the development of dis­ seminated MAC disease was highly correlated with a decline in CD4+ T lymphocyte numbers. Such a decrease has also been implicated in disseminated MAC infection in patients with idiopathic CD4+ T lymphocytopenia. Potent inhibitors of tumor necrosis factor α (TNF-α), such as infliximab, adalimumab, certolizumab, golimumab, and etan­ ercept, neutralize this critical cytokine, with consequent inhibition of granuloma formation. The occasional result is severe mycobacterial or fungal infection; these associations indicate that TNF-α is a crucial element in mycobacterial control. However, in cases without the above risk factors, much of the basis of susceptibility to disseminated infec­ tion with NTM is accounted for by specific mutations in the interferon γ (IFN-γ)/interleukin 12 (IL-12) synthesis and response pathways or autoantibodies to IFN-γ itself. Mycobacteria are typically phagocytosed by macrophages, which respond with the production of IL-12, a heterodimer composed of IL-12p35 and IL-12p40 moieties that together make up IL12p70. IL-12 activates T lymphocytes and natural killer cells through binding to its receptor (composed of IL-12Rβ1 and IL-12Rβ2/IL-23R), with consequent phosphorylation of STAT4. IL-12 stimulation of STAT4 leads to secretion of IFN-γ, which activates neutrophils and macrophages to produce reactive oxidants, to increase expression of the major histocompatibility complex and Fc receptors, and to concen­ trate certain antibiotics intracellularly. Signaling by IFN-γ through its receptor (composed of IFN-γR1 and IFN-γR2) leads to phosphoryla­ tion of STAT1, which in turn regulates IFN-γ-responsive genes, such as those coding for IL-12 and TNF-α. TNF-α signals through its own receptor via a downstream complex containing the nuclear factor-κB (NF-κB) essential modulator (NEMO). Therefore, the positive feed­ back loop between IFN-γ and IL-12/IL-23 drives the immune response to mycobacteria and other intracellular infections. These genes are known to be the critical ones in the pathway of mycobacterial control: specific Mendelian mutations have been identified in IFNG, IFNGR1, IFNGR2, STAT1, GATA2, ISG15, IRF8, IL-12A, IL-12RB1, IL-12RB2, CYBB (which encodes the gp91phox protein of the NADPH oxidase), SPP2A, MCTS1, and IKBKG (which encodes NEMO) (Fig. 185-1). Despite the identification of genes associated with disseminated dis­ ease, only ~70% of cases of disseminated nontuberculous mycobacte­ rial infections that are not associated with HIV infection have a genetic diagnosis; the implication is that more mycobacterial susceptibility genes and pathways remain to be identified. In contrast to the recognized genes and mechanisms associated with disseminated nontuberculous mycobacterial infection, the bestrecognized underlying condition for pulmonary infection with NTM is bronchiectasis (Chap. 301). Most of the well-characterized forms of bronchiectasis, including cystic fibrosis, primary ciliary dyskinesia, STAT3-dominant negative hyper-IgE syndrome (Job’s syndrome), and idiopathic bronchiectasis, have high rates of association with nontu­ berculous mycobacterial infection. The precise mechanism by which

IL-2 IL-2R T/NK IFNf a1 IL-12R a2 IL-18 ? IL-15 IFNfR STAT1 GATA2

IRF8 ISG15 IL-12 NEMO AFB Salm. TNF NRAMP1 MΦ TLR TNFR CD14 LPS FIGURE 185-1  Cytokine interactions of infected macrophages (MΦ) with T and natural killer (NK) lymphocytes. Infection of macrophages by mycobacteria (AFB) leads to the release of heterodimeric interleukin 12 (IL-12). IL-12 acts on its receptor complex (IL-12R), with consequent STAT4 activation and production of homodimeric interferon γ (IFNγ). Through its receptor (IFNγR), IFNγ activates STAT1, stimulating the production of tumor necrosis factor α (TNFα) and leading to the killing of intracellular organisms such as mycobacteria, salmonellae (Salm.), and some fungi. Homotrimeric TNFα acts through its receptor (TNFαR) and requires nuclear factor-κB essential modulator (NEMO) to activate nuclear factor-κB, which also contributes to the killing of intracellular bacteria. Both IFNγ and TNFα lead to upregulation of IL-12. TNFα-blocking antibodies work either by blocking the ligand (infliximab, adalimumab, certolizumab, golimumab) or by providing soluble receptor (etanercept). Mutations in IFNG, IFNGR1, IFNGR2, IL12B, IL12RB1, IL12RB2, STAT1, GATA2, ISG15, IRF8, CYBB, MCTS1, and IKBKG (NEMO) have been associated with predisposition to mycobacterial infections. Other cytokines, such as IL-15 and IL-18, also contribute to IFNγ production. Signaling through the Toll-like receptor (TLR) complex and CD14 also upregulates TNFα production. IRF8, interferon regulatory factor 8; ISG15, interferon-stimulated gene 15; LPS, lipopolysaccharide; NRAMP1, natural resistance-associated macrophage protein 1. CHAPTER 185 Nontuberculous Mycobacterial Infections bronchiectasis predisposes to locally destructive but not systemic involvement is unknown. Unlike disseminated or pulmonary infection, “hot-tub lung” rep­ resents pulmonary hypersensitivity to NTM—most commonly MAC organisms—growing in underchlorinated water, often in indoor hot tubs. ■ ■CLINICAL MANIFESTATIONS Disseminated Disease  Disseminated MAC or M. kansasii infec­ tions in people with advanced HIV infection are now uncommon in North America because of effective antimycobacterial prophylaxis and improved treatment of HIV infection. When such mycobacterial disease was common, the portal of entry was the bowel, with spread to bone marrow and the bloodstream. Surprisingly, disseminated infec­ tions with rapidly growing NTM (e.g., M. abscessus, M. fortuitum) are very rare in people with advanced HIV infection. Because these organ­ isms are of low intrinsic virulence and disseminate only in conjunction with impaired immunity, disseminated disease can be indolent and progressive over weeks to months. Typical manifestations of malaise, fever, and weight loss are often accompanied by organomegaly, lymph­ adenopathy, and anemia. Because special cultures or stains are required to identify the organisms, the most critical step in diagnosis is to sus­ pect infection with NTM. Blood cultures may be negative, but involved organs typically have significant organism burdens, sometimes with a grossly impaired granulomatous response. Disseminated involvement (i.e., involvement of two or more organs) without an underlying iatrogenic cause should prompt a genetic investigation of the IFN-γ/IL-12 pathway. Recessive muta­ tions in IFNGR1 and IFNGR2 typically ablate IFN-γ signaling and lead to severe infection with NTM. In contrast, dominant negative muta­ tions in IFNGR1, which lead to overaccumulation of a defective inter­ fering mutant receptor on the cell surface, inhibit but do not abolish

normal IFN-γ signaling and cause nontuberculous mycobacterial osteomyelitis. Dominant negative mutations in STAT1 and recessive mutations in IL-12RB1 can produce variable phenotypes consistent with their residual capacities for IFN-γ synthesis and response. Male patients who have disseminated nontuberculous mycobacterial infec­ tions along with bacterial or viral infections; conical, peg, or missing teeth; or an abnormal hair pattern should be evaluated for defects in the pathway that activates NF-κB through NEMO (IKBKG). These patients may have associated immune globulin defects as well. Patients with myelodysplasia and mycobacterial disease should be investigated for GATA2 deficiency. A recently recognized group of patients who often develop disseminated infections with both MAC and rapidly growing NTM (predominantly M. abscessus) as well as other opportu­ nistic infections such as Talaromyces have high-titer neutralizing auto­ antibodies to IFN-γ. This syndrome has been reported most frequently in East Asian female patients.

IV catheters can become infected with NTM, usually as a conse­ quence of contaminated water. M. abscessus and M. fortuitum some­ times infect deep indwelling lines as well as fluids used in eye surgery, subcutaneous injections, and local anesthetics. Infected catheters should be removed. Pulmonary Disease  Lung disease is by far the most common form of nontuberculous mycobacterial infection in North America and the rest of the industrialized world. In North America, rates of NTM lung disease far exceed rates of tuberculosis. The clinical presentation typi­ cally consists of months or years of throat clearing, nagging cough, and slowly progressive fatigue. Patients will often have seen physicians mul­ tiple times and received symptom-based or transient therapy before the diagnosis is entertained and samples are sent for mycobacterial stains and cultures. Because not all patients can produce sputum, bronchos­ copy may be required for diagnosis. The typical lag between onset of symptoms and diagnosis is ~5 years in older women. Predisposing fac­ tors include underlying lung diseases such as bronchiectasis (Chap. 301), pneumoconiosis (Chap. 300), chronic obstructive pulmonary disease (Chap. 303), primary ciliary dyskinesia (Chap. 301), α1 antitrypsin deficiency (Chap. 303), and cystic fibrosis (Chap. 302). Bronchiectasis and nontuberculous mycobacterial infection often coexist and progress in tandem. This situation makes causality difficult to determine in a given index case, but bronchiectasis is certainly among the most critical predisposing factors that are exacerbated by infection. PART 5 Infectious Diseases MAC are the most common cause of pulmonary nontuberculous mycobacterial infection in North America, but rates vary somewhat by region. MAC infection most commonly develops during the sixth or seventh decade of life in women who have had months or years of nagging intermittent cough and fatigue, with or without sputum pro­ duction or chest pain. The constellation of pulmonary disease due to NTM in a tall and thin woman who may have chest wall abnormalities is often referred to as “Lady Windermere syndrome,” after an Oscar Wilde character of the same name. In fact, pulmonary MAC infec­ tion does afflict older nonsmoking white women more than men, with onset at ~60 years. Patients tend to be taller and thinner than the general population, with high rates of scoliosis, mitral valve prolapse, and pectus anomalies. Whereas male smokers with upper-lobe cavi­ tary disease tend to carry the same single strain of MAC indefinitely, nonsmoking females with nodular bronchiectasis tend to carry several strains of MAC simultaneously, with changes over the course of their disease. M. kansasii can cause a clinical syndrome that strongly resembles tuberculosis, consisting of hemoptysis, chest pain, and cavitary lung disease. The rapidly growing NTM, such as M. abscessus, have been associated with esophageal motility disorders such as achalasia. Patients with pulmonary alveolar proteinosis are prone to pulmonary nontuberculous mycobacterial and Nocardia infections; the underlying mechanism may be inhibition of alveolar macrophage function due to the autoantibodies to granulocyte-macrophage colony-stimulating fac­ tor found in many of these patients. Cervical Lymph Nodes  The most common form of nontubercu­ lous mycobacterial infection among young children in North America

is isolated cervical lymphadenopathy, caused most frequently by MAC organisms but also by other NTM. The cervical swelling is typically firm and relatively painless, with a paucity of systemic signs. Because the differential diagnosis of painless adenopathy includes malignancy, many children have infection with NTM diagnosed inadvertently at biopsy; cultures and special stains may not have been requested because mycobacterial disease was not ranked high in the differential. Local fistulae usually resolve completely with resection and/or anti­ biotic therapy. Likewise, the entity of isolated pediatric intrathoracic nontuberculous mycobacterial infection, which is probably related to cervical lymph node infection, is usually mistaken for cancer. In neither isolated cervical nor isolated intrathoracic infections with NTM have children with underlying immune defects been commonly identified, nor do the affected children usually go on to develop other opportunistic infections. Skin and Soft Tissue Disease  Cutaneous involvement with NTM usually requires a break in the skin for introduction of the bacteria. Pedicure bath–associated infection with M. fortuitum is more likely if skin abrasion (e.g., during leg shaving) has occurred just before the pedicure. Outbreaks of skin infection are often caused by rapidly growing NTM (especially M. abscessus, M. fortuitum, and M. chelonae) acquired via skin contamination from surgical instruments (especially in cosmetic surgery), injections, and other procedures. These infec­ tions are typically accompanied by painful, erythematous, draining subcutaneous nodules, usually without associated fever or systemic symptoms. M. marinum lives in many water sources and can be acquired from fish tanks, swimming pools, barnacles, and fish scales. This organ­ ism typically causes papules or ulcers (“fish-tank granuloma”), but the infection can progress to tendinitis with significant impairment of manual dexterity. Lesions appear days to weeks after inoculation of organisms by a typically minor trauma (e.g., incurred during the cleaning of boats or the handling of fish). Tender nodules due to

M. marinum can advance up the arm in a pattern also seen with Sporothrix schenckii (sporotricoid spread). The typical carpal tendon involvement may be the first presenting manifestation and may lead to surgical exploration or steroid injection. The index of suspicion for M. marinum infections must be high to ensure that proper specimens obtained during procedures are sent for culture. M. ulcerans, another waterborne skin pathogen, is found mainly in the tropics, especially in tropical areas of Africa. Infection follows skin trauma or insect bites that allow admission to contaminated water. The skin lesions are typically painless, clean ulcers that slough and can cause osteomyelitis. The toxin mycolactone accounts for the dimin­ ished host inflammatory response and the painless ulcerations. ■ ■DIAGNOSIS NTM can be detected on acid-fast or fluorochrome smears of sputum or other body fluids. When the organism burden is high, the organ­ isms may appear as gram-positive beaded rods, but this finding is unreliable. (In contrast, nocardiae may appear as gram-positive and beaded but filamentous bacteria.) Again, the requisite and most sensi­ tive step in the diagnosis of any mycobacterial disease is to think of including it in the differential. In almost all laboratories, mycobacte­ rial sample processing, staining, and culture are conducted separately from routine bacteriologic tests; thus, many infections go undiagnosed because of physician failure to request the appropriate test. In addition, mycobacteria usually require separate blood culture media. NTM are broadly differentiated into rapidly growing (<7 days) and slowly growing (≥7 days) forms. Because M. tuberculosis typically takes ≥2 weeks to grow, many laboratories refuse to consider culture results final until 6 weeks have elapsed. Newer techniques using liquid culture media permit more rapid isolation of mycobacteria from specimens than is possible with traditional media. Species more readily detected with incubation at 30°C include M. marinum, M. haemophilum, and

M. ulcerans. M. haemophilum prefers iron supplementation or blood, whereas M. genavense requires supplemented medium with the addi­ tive mycobactin J. Bacterial formation of pigment in light conditions

(photochromogenicity) or dark conditions (scotochromogenicity) or a lack of bacterial pigment formation (nonchromogenicity) was his­ torically used to help categorize NTM. In contrast to NTM colonies,

M. tuberculosis colonies are beige, rough, dry, and flat. Current iden­ tification schemes reliably use biochemical, nucleic acid, or cell wall composition, as assessed by high-performance liquid chromatography or mass spectrometry, for speciation. With the remarkable decline in U.S. cases of tuberculosis over recent decades, NTM have become the mycobacteria most commonly isolated from humans in North America. However, not all isolations of NTM, especially from the lung, reflect pathology and require treatment. Whereas identification of an organism in a blood or organ biopsy specimen in a compatible clinical setting is considered diagnostic, the American Thoracic Society rec­ ommends that pulmonary infection due to NTM be diagnosed only when disease is clearly demonstrable—i.e., in an appropriate clinical and radiographic setting (nodules, bronchiectasis, cavities) and with repeated isolation of NTM from expectorated sputum or recovery of NTM from bronchoscopy or biopsy specimens. Given the large num­ ber of species of NTM and the importance of accurate diagnosis for the implementation of proper therapy, identification of these organisms is ideally taken to the species level. The purified protein derivative (PPD) of tuberculin is delivered intradermally to evoke a memory T-cell response to mycobacterial antigens. This test is variously referred to as the PPD test, the tubercu­ lin skin test, and the Mantoux test, among other designations. Unfor­ tunately, the cutaneous immune response to these tuberculosis-derived filtrate proteins does not differentiate well between infection with some NTM and that with M. tuberculosis. Because intermediate reactions (~10 mm) to PPD in latent tuberculosis and nontuberculous myco­ bacterial infections can overlap significantly, the progressive decline in active tuberculosis in the United States means that NTM probably account for increasing proportions of PPD reactivity. In addition, bac­ ille Calmette-Guérin (BCG) can cause some degree of cross-reactivity in PPD testing, posing problems of interpretation for patients who have received BCG vaccine. Assays to measure the elaboration of IFN-γ in response to the relatively tuberculosis-specific proteins ESAT6 and CFP10 form the basis for IFN-γ-release assays (IGRAs). These assays can be performed with whole blood or on membranes. It is important to note that M. marinum, M. kansasii, and M. szulgai also have ESAT6 and CFP10 and may cause false-positive reactions in IGRAs. Despite cross-reactivity with NTM, large PPD reactions (>15 mm) most com­ monly signify tuberculosis. Conversely, in the setting of anti-IFNγ autoantibodies, the IGRA test is indeterminate (failure of IFNγ detec­ tion in response to specific antigens and mitogens, due to neutralizing anti-IFNγ autoantibodies). Isolation of NTM from blood specimens is clear evidence of disease. Whereas rapidly growing mycobacteria may proliferate in routine blood culture media, slow-growing NTM typically do not; therefore, it is imperative to suspect the diagnosis and to use the correct bottles for cultures. Isolation of NTM from a biopsy specimen constitutes strong evidence for infection, but cases of laboratory contamination do occur. Identification of organisms on stained sections of biopsy material confirms the authenticity of the culture. Certain NTM require lower incubation temperatures (M. genavense) or special additives (M. haemophilum) for growth. Some NTM (e.g., M. tilburgii) remain noncultivable but can be identified molecularly in clinical samples. The radiographic appearance of nontuberculous mycobacterial dis­ ease in the lung depends on the underlying disease, the severity of the infection, and the imaging modality used. The advent and increase in the use of computed tomography (CT) scanning has allowed the iden­ tification of characteristic changes that are highly consistent with non­ tuberculous mycobacterial infection, such as the “tree-in-bud” pattern of bronchiolar inflammation (Fig. 185-2). Involvement of the lingual and right-middle lobes is commonly seen on chest CT but is difficult to appreciate on plain film. Severe bronchiectasis and cavity formation are common in more advanced disease. Isolation of NTM from respiratory samples can be confusing.

M. gordonae is often recovered from respiratory samples but is not usually seen on smear and is almost never a pathogen. Patients with

FIGURE 185-2  Chest computed tomography of a patient with pulmonary Mycobacterium avium complex infection. Arrows indicate the “tree-in-bud” pattern of bronchiolar inflammation (peripheral right lung) and bronchiectasis (central right and left lungs). bronchiectasis occasionally have NTM recovered from sputum culture with a negative smear. The American Thoracic Society has developed guidelines for the diagnosis of infection with MAC, M. abscessus, and M. kansasii. A positive diagnosis requires the growth of NTM from two of three sputum samples, regardless of smear findings; a positive bronchoscopic alveolar sample, regardless of smear findings; or a pul­ monary parenchyma biopsy sample with granulomatous inflammation or mycobacteria found on section and NTM found on culture. These guidelines probably apply to other NTM as well. CHAPTER 185 Although many laboratories use DNA probes to identify M. tuber­ culosis, MAC, M. gordonae, and M. kansasii, speciation of NTM helps determine the antimycobacterial therapy to be used. Only testing of MAC organisms for susceptibility to clarithromycin and of M. kansasii for susceptibility to rifampin is indicated; few data support other in vitro susceptibility tests, attractive though they appear. MAC isolates that have not been exposed to macrolides are almost always suscepti­ ble. NTM that have persisted beyond a course of antimicrobial therapy are often tested for antibiotic susceptibility, but the value and meaning of these tests are undetermined. Nontuberculous Mycobacterial Infections ■ ■PREVENTION Prophylaxis of MAC disease in patients infected with HIV is started when the CD4+ T lymphocyte count falls to <50/μL. Azithromycin (1200 mg weekly), clarithromycin (1000 mg daily), and rifabutin (300 mg daily) are effective. Macrolide prophylaxis in immunodefi­ cient patients who are susceptible to NTM (e.g., those with defects in the IFN-γ/IL-12 axis) has not been prospectively validated but seems prudent. TREATMENT Nontuberculous Mycobacteria NTM cause chronic infections that evolve relatively slowly over a period of weeks to years. Therefore, it is rarely necessary to initi­ ate treatment on an emergent basis before the diagnosis is clear and the infecting species is known. Treatment of NTM is complex, often poorly tolerated, and potentially toxic. Just as in tuberculosis, inadequate single-drug therapy is almost always associated with the emergence of antimicrobial resistance and relapse. MAC infection often requires multidrug therapy, the founda­ tion of which is a macrolide (clarithromycin or azithromycin), ethambutol, and a rifamycin (rifampin or rifabutin). For dis­ seminated nontuberculous mycobacterial disease in HIV-infected patients, the use of rifamycins poses special problems—i.e., rifa­ mycin interactions with protease inhibitors. For pulmonary MAC disease, thrice-weekly administration of a macrolide, a rifamycin,