HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:
Guest Access | Sign In via User Name/Password

This Article Abstract Full Text (PDF) Free Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Article Archive Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (37) Citing Articles via Google Scholar Google Scholar Articles by Field, S. K. Articles by Cowie, R. L. Search for Related Content PubMed PubMed Citation Articles by Field, S. K. Articles by Cowie, R. L.

Mycobacterium avium complex Pulmonary Disease in Patients Without HIV Infection^*

^* From the Division of Respiratory Medicine, University of Calgary Medical School, Calgary, AB, Canada.

Correspondence to: Stephen K. Field, MD, FCCP, Room 1410, Health Science Centre, 3330 Hospital Dr NW, Calgary, AB, Canada T2N 4N1; e-mail sfield{at}ucalgary.ca

	Abstract

TOP Abstract Introduction Background Effects of Exposure to... Epidemiology of MAC Lung... Clinical Features of MAC... Radiologic Features of MAC... Microbiology of M avium Predisposing Factors for MAC... Conclusion References

 
Mycobacterium avium complex (MAC) is ubiquitous. It is found in various freshwater and saltwater sources around the world, including hot water pipes. Although the organism was identified in the 1890s, its potential to cause human disease was only recognized 50 years later. Only a minority of people exposed to the organism will acquire MAC lung disease, usually those with underlying lung disease or immunosuppression. MAC may, however, cause progressive parenchymal lung disease and bronchiectasis in patients without underlying lung disease, particularly in middle-aged and elderly women. Preliminary data suggest that the interferon- pathways may be deficient in elderly women with MAC lung disease. Other groups of patients who are more likely to harbor MAC in their lungs include patients with a cystic fibrosis or an abnormal ₁-antiproteinase gene and patients with certain chest wall abnormalities. Treatment results continue to be disappointing, and the mortality of patients with MAC lung disease remains high. A PubMed search identified 38 reports of the treatment of MAC lung disease. Apart from the British Thoracic Society study, the only published controlled investigation, the studies published since 1994 have included a macrolide, either clarithromycin or azithromycin, usually in combination with ethambutol and a rifamycin. If success is defined as eradication of the organism without relapse over a period of several years after treatment has been discontinued, the reported treatment success rate with the macrolide containing regimens is approximately 55%. The prolonged treatment period, side effects, and possibly reinfection rather than relapse are responsible for the high failure rate.

Key Words: bronchiectasis • clarithromycin • ethambutol • macrolides • Mycobacterium avium • Mycobacterium intracellulare • nontuberculous mycobacteria • rifabutin rifampin

	Introduction

TOP Abstract Introduction Background Effects of Exposure to... Epidemiology of MAC Lung... Clinical Features of MAC... Radiologic Features of MAC... Microbiology of M avium Predisposing Factors for MAC... Conclusion References

 
The clinical presentation of Mycobacterium avium-intracellulare (Mycobacterium avium complex [MAC]) infection depends on the population studied. MAC can be isolated from the sputum of apparently healthy individuals; it may be present in association with, but not appear to contribute to other lung diseases; or it may cause the progression of underlying lung disease.¹²³⁴ MAC may cause progressive lung disease leading to respiratory failure and even death in previously healthy patients, usually elderly women without a history of lung disease or immunodeficiency.² Disseminated disease may occur in immunocompromised patients, especially in those infected with the HIV.⁵⁶⁷ Paradoxically, patients with disseminated disease rarely have clinically important pulmonary disease.⁸⁹¹⁰ In children, apart from those with cystic fibrosis and those with HIV infection, MAC disease usually presents as cervical lymphadenopathy.¹¹

It appears that MAC lung disease is becoming increasingly prevalent.^1213141516 Some of this apparent increase is due to greater recognition of its role in conditions such as bronchiectasis and cystic fibrosis, but most is likely due to the "graying" of the population.^217181920 It has been suggested that exposure to tuberculosis and bacille Calmette-Guérin vaccination may offer some immunity, and that the declining rates of these conditions may explain some of this increase in infection with nontuberculous mycobacteria (NTM).²¹²²²³ It has been reported that patients with MAC lung disease have a different immune profile than healthy people with skin hypersensitivity to MAC antigen.²⁴ Traditionally, therapy for MAC lung disease has consisted of antituberculosis medications, but treatment outcomes have been disappointing.⁸²⁵ Adverse medication reactions, treatment failures, and frequent relapses often preclude successful therapy. Surgery has been advocated for patients with localized disease, especially for those who fail to eradicate MAC from their sputum with medication.^2627282930 Since the introduction of the newer macrolides, clarithromycin and azithromycin, treatment results have improved.⁸ The purpose of this article is to review the available literature and provide an update on lung disease due to MAC in patients without HIV infection.

	Background

TOP Abstract Introduction Background Effects of Exposure to... Epidemiology of MAC Lung... Clinical Features of MAC... Radiologic Features of MAC... Microbiology of M avium Predisposing Factors for MAC... Conclusion References

 
Mycobacteria that infect other species and mycobacterial saprophytes were identified soon after Koch’s discovery of the tuberculosis bacillus, but were only widely recognized to be human pathogens in the 1950s.¹⁸ Various designations have been used to distinguish these other mycobacterial species from tuberculosis, including: paratubercle or pseudotubercle, anonymous, atypical, environmental, opportunistic, nontuberculous, and mycobacteria other than tuberculosis.³¹ NTM is the currently favored term.^8253132

An avian form of tuberculosis was described in the 1860s, and by 1890 it was recognized that the avian bacillus was different from tuberculosis.³¹ In the 1930s, Pinner³³ reported that NTM could be cultured from water and other environmental sources, and could be isolated from human specimens. He recognized that some mycobacteria were chromogenic and some were rapid growers, but he failed to develop a workable classification for the NTM. The first case of lung infection due to NTM was reported in 1943. The patient was an elderly retired miner with a clinical syndrome consistent with silicotuberculosis. An acid-fast bacillus, which was not Mycobacterium tuberculosis, was isolated repeatedly from his sputum. It was distinct from avian tuberculosis but had many similar features.³⁴

Nearly 20 years after the attempt by Pinner³³ to develop a coherent classification of the NTM, Timpe and Runyon³⁵ analyzed NTM from various parts of the United States and classified them into four groups by colony appearance and growth characteristics. All strains were bacilliform and acid fast. Group 1 organisms are photochromogens; colonies form pigment only after exposure to light, and include Mycobacterium kansasii and Mycobacterium marinum.³⁵ Group 2 organisms are scotochromogens, capable of forming pigment in the dark, and include Mycobacterium gordonae. Organisms in the group 3 belong to the M avium-intracellulare family and are nonchromogenic. Group 4 members are the rapid growers. Colonies can appear within 3 to 7 days, and include the closely related Mycobacterium fortuitum, Mycobacterium chelonae, and Mycobacterium abscessus.

MAC, representative of group 3 of Timpe and Runyon,³⁵ has been found around the world. Reports from Australia, India, Japan, Israel, Thailand, South Africa, various European countries, and North America have documented its presence in different environmental sources, including house dust, soil, birds, and farm animals.^{36373839404142434445} It has even been found in cigarette components, including the tobacco, filter, and paper.⁴⁶

Strain analysis suggests that animal-to-man transmission is not an important source of infection even though MAC is an important cause of disease in swine and poultry.⁴¹⁴² MAC was isolated from 25% of the water samples collected along the East coast of the United States. Most isolates were collected in South Carolina, Georgia, and the Gulf states.⁴⁷ It is found in both freshwater and saltwater, and has even been found in hospital hot water systems.^4748495051 Presumably, aerosolized water serves as the main source of infection.⁸⁵⁰⁵¹ One hypothesis relates the increased prevalence of MAC lung disease to the change in our hygiene habits from bathing to showering, and therefore aerosolizing more water.⁵² It does not appear that person-to-person transmission of MAC is a concern.⁵³

M avium and Mycobacterium intracellulare were initially distinguished by their relative pathogenicity for chickens and rabbits, respectively.⁵⁴ Currently, most laboratories report these as M avium-intracellulare or MAC because the species are so similar. Moreover, demographics, symptoms, radiologic features, and responses to therapy for patients infected with these two species are indistinguishable.⁵⁵

	Effects of Exposure to M avium on Tuberculin Skin Testing Results

TOP Abstract Introduction Background Effects of Exposure to... Epidemiology of MAC Lung... Clinical Features of MAC... Radiologic Features of MAC... Microbiology of M avium Predisposing Factors for MAC... Conclusion References

 
Exposure to NTM can cause cross-reactions with tuberculin. Tuberculin skin testing was performed as part of the large international bacille Calmette-Guérin vaccination programs coordinated by the World Health Organization and the United Nations Children’s Fund in the late 1940s and 1950s.⁵⁶ In Northern countries, reactions to tuberculin were either strong or absent and correlated with the local prevalence of tuberculosis.⁵⁷ In some geographic regions, primarily in warmer climates, a large proportion of the children acquired a low-grade or intermediate-strength reaction, also called a "nonspecific" reaction to tuberculin testing.⁵⁷ The prevalence of nonspecific reactivity varied with geographic location and did not correlate with the local prevalence of tuberculosis or with the prevalence of chest radiograph abnormalities due to tuberculosis infection.⁵⁸ A high proportion of those individuals with nonspecific reactivity to tuberculin responded to protein derivative of M avium.⁵⁸ These and similar reports, and previous veterinary experience that nonspecific reactivity in cattle occurred in the absence of pathologic findings of tuberculosis in animals that were exposed to M avium, suggest that M avium may cause cross-reactions to tuberculin.⁵⁸

Epidemiologic studies⁵⁹ have shown that many healthy people have positive skin test reactions to MAC, but only a small number of patients acquire clinical lung disease. In the late 1950s, screening naval recruits for tuberculosis with routine skin testing and chest radiographs began in the United States of America. Skin test and chest radiograph interpretation were confounded by the presence of abnormal results due to NTM and endemic fungi. The chest radiograph appearances of disease due to NTM and endemic fungi can be confused with tuberculosis.⁶⁰⁶¹⁶² To help clarify the prevalence of tuberculosis infection, skin testing was done with tuberculin, histoplasmin, and antigen derived from killed MAC, or "Battey" antigen.⁶³ MAC has been called the Battey bacillus after the name of the state hospital in Rome, GA, where patients with MAC lung disease were first thoroughly studied.⁶³⁶⁴ Overall, one third of naval recruits had positive skin test reactions to the MAC antigen.⁵⁹ The prevalence of positive skin test reactions to MAC antigen ranged from 10 to 20% in recruits from the north and west United States and increased to > 70% in recruits from the Southeast.⁵⁹ Once it was recognized that MAC was pathogenic, and mycobacterial cultures became routine, MAC was isolated from only a minority of patients with tuberculosis-like illness, even in the southeast United States, where most people have skin hypersensitivity to MAC. Although the skin test antigens have been useful as an epidemiologic tool, surface antigens are shared by different mycobacterial species, and cross-reactions limit their usefulness as diagnostic tools for NTM lung disease.⁶⁵

	Epidemiology of MAC Lung Disease

TOP Abstract Introduction Background Effects of Exposure to... Epidemiology of MAC Lung... Clinical Features of MAC... Radiologic Features of MAC... Microbiology of M avium Predisposing Factors for MAC... Conclusion References

 
As the prevalence of tuberculosis declined sharply after the introduction of effective chemotherapy in the 1950s, reports indicated an increase in the number of patients with tuberculosis-like illness due to NTM. Further instances of NTM disease were detected as more mycobacterial cultures were tested for drug susceptibility. In the 1950s, sputum from approximately 1 to 2% of patients admitted to sanatoria in Florida and Georgia grew NTM.⁶³⁶⁶ Fifteen cases of NTM pulmonary disease were detected among 1,320 patients admitted to a New England hospital with suspected tuberculosis in the early 1960s.⁶⁷ This pathogenic organism belonged to group 3 of Timpe and Runyon,³⁵ or the MAC, in 11 of these 15 patients.⁶⁷ Moreover, 66 of the hospitalized patients (5%) had random isolates of NTM from their sputum on one or more occasion. A Veteran’s Administration hospital survey⁶⁸ between 1963 and 1965 found that the majority of NTM pulmonary disease was due to M kansasii, with approximately one third due to MAC. Patients with MAC lung disease tended to be older than patients with lung disease due to M kansasii, and the majority of patients with MAC had underlying lung disease including COPD, bronchiectasis, healed tuberculosis, or silicosis.⁶⁸

Between the 3.5-year periods of January 1963 to July 1966 and October 1968 to April 1972, the number of tuberculosis cases at the Medical College of Wisconsin declined 17%, but the number of cases of infection due to MAC increased from 26 to 123.⁶⁹ Studies³⁸ from other countries suggest that the prevalence of NTM, including MAC, have increased. Between 1967 and 1971, 289 of the 3,150 mycobacterial isolates (9%) from the greater Boston area were NTM, and 117 were identified as MAC.⁷⁰ The prevalence increased in Texas between 1967 and 1976.⁷¹

In 1979 to 1980, one third of the mycobacterial isolates reported to the Centers for Disease Control and Prevention were NTM, and 61% of these were MAC.⁷² Data collected in the United States between 1981 and 1983 demonstrated higher rates of NTM disease in women, African Americans and Hispanics, and urban residents than those in the earlier report from the Centers for Disease Control and Prevention.¹² In the 3.5-year period between April 1982 and October 1985, the Cleveland Clinic reported 129 M tuberculosis and 79 MAC isolates; 92 isolates and 56 isolates, respectively, were from pulmonary specimens.⁵⁴ Prince and colleagues² reported that the number of cases of tuberculosis declined and the number of MAC lung disease cases increased at two Philadelphia hospitals between 1978 and 1987.

Prior to the publication by Prince and colleagues,² most reports of MAC lung disease were from referral institutions or sanatoria, many from within the Veteran’s Administration system.² Not surprisingly, the patients were predominately male with a history of tobacco and/or alcohol abuse and underlying lung disease.⁷³ Most patients had extensive cavitary disease, and only a minority had normal chest radiograph findings or abnormalities with a fibronodular appearance.⁶⁰ Prince and colleagues² recognized that fibronodular bronchiectasis was a common manifestation of MAC in the lung, primarily seen in elderly, often thin, women without apparent immunosuppression or underlying lung disease. Some of the apparent increase in MAC lung disease may be due to the recognition that fibronodular bronchiectasis was also a manifestation of MAC infection, and clinicians began to collect specimens for mycobacterial culture from patients with bronchiectasis and/or chest radiograph abnormalities with a fibronodular appearance after the publication by Prince and coworkers.² Most patients with MAC lung disease in an Oregon health maintenance organization between 1975 and 1986 were elderly women without preexisting lung disease.⁷⁴ They represented 30% of the cases with pleuropulmonary mycobacterial infections. The ratio of M tuberculosis to NTM isolates fell from 3.2 to 1 between 1976 and 1981, to 1 to 1.6 between 1986 and 1991 in a community hospital in South Carolina.⁷⁵

Patients without obvious lung disease may also have MAC isolated from their sputum.¹ The term colonized was previously used to describe this condition.⁶²⁷⁶ Further supporting the concept of transient colonization, MAC was isolated from 11% of patients with pulmonary tuberculosis at Bellevue Hospital, but these isolates did not appear to be clinically significant.⁷⁷ According to the first American Thoracic Society (ATS) statement on the diagnosis and treatment of NTM published in 1990,²⁵ repeated isolation of NTM from the sputum of an individual without fibrocavitary disease does not necessarily represent disease but may be due to colonization that does not require therapy. More recently, the introduction of high-resolution CT (HRCT) and reports⁶⁶ of slow progression of lung disease, sometimes over decades, in patients with fibronodular bronchiectasis, have demonstrated the importance of treating these patients. The repeated isolation of MAC from sputum is now believed to represent infection, and treatment is recommended to prevent the development and progression of lung disease.⁸ If the clinical decision is made to withhold therapy, careful monitoring for disease progression is mandatory.⁸ Unfortunately, there are no prospective, controlled trials allowing estimation of the risk of these patients acquiring progressive lung disease. In the absence of controlled trials, it is unknown how frequently sputum will convert to negative spontaneously, what proportion of patients with M avium in their sputum will acquire progressive lung disease, and whether treatment will alter the course of the disease.

	Clinical Features of MAC Lung Disease

TOP Abstract Introduction Background Effects of Exposure to... Epidemiology of MAC Lung... Clinical Features of MAC... Radiologic Features of MAC... Microbiology of M avium Predisposing Factors for MAC... Conclusion References

 
Lung disease caused by MAC was first described in sanatorium patients with suspected tuberculosis.³⁴⁶⁴ Patients with MAC lung disease tended to be older, were more often white, and were more likely to have underlying lung disease than patients with tuberculosis.³⁶⁶⁹ Unlike tuberculosis, patients with MAC lung disease are not contagious, and family members and other contacts, including health-care workers and other patients in the health-care facilities, were usually not infected.³⁵⁵³

Male patients predominated in the reports of MAC lung disease from the 1950s to the 1980s. Most patients in the treatment trials over the last decade have been elderly women without a history of underlying lung disease. Rather than bilateral upper lobe fibrocavitary disease, the traditional presentation of MAC lung disease, women usually acquire fibronodular disease with bronchiectasis.²

The insidious nature of this illness has been emphasized repeatedly, as symptoms may be present for months or years before the diagnosis is made.⁶⁷⁷⁸ Early cases may be asymptomatic and only discovered by routine screening chest radiographs.⁷⁹ Most patients experience chronic cough, usually productive of purulent sputum but usually without hemoptysis (Table 1 ).²⁸²⁵ Constitutional symptoms including fever, night sweats, weight loss, malaise, lethargy, and fatigue are not common unless patients have extensive lung disease.⁶⁴ Patients with smear-positive sputum are more likely to have fever and hemoptysis.⁸⁰ Assessment of symptoms is complicated in patients with preexisting lung disease.²⁵ Symptoms usually improve with successful therapy.

	Radiologic Features of MAC Lung Disease

TOP Abstract Introduction Background Effects of Exposure to... Epidemiology of MAC Lung... Clinical Features of MAC... Radiologic Features of MAC... Microbiology of M avium Predisposing Factors for MAC... Conclusion References

 
Patients with underlying lung disease and MAC infection often have cavities apparent on the chest radiograph.³⁶⁶⁰⁶⁸ It has been suggested that MAC-related cavities tend to be thinner with less surrounding parenchymal opacification and more contiguous, but less distal bronchogenic, spread of opacification than in tuberculosis.⁸ In older women without underlying lung disease, MAC primarily involves the lingula and right middle lobe. Bronchiectasis is often evident. Often reticular changes, a single nodule, or multiple small nodules are present. If pleural thickening occurs, it usually develops adjacent to parenchymal abnormalities. Pleural effusions are not common in MAC.

HRCT scanning is more sensitive than chest radiographs for detecting the abnormalities associated with MAC lung disease.^79858687 Several studies^{858687888990919293} of the HRCT findings in MAC lung disease have been published. The presence of bronchiectasis and multiple small nodules are predictive of MAC lung disease, especially when they are adjacent to each other.⁷⁹ In a study,⁸⁹ one quarter of patients with typical changes of fibronodular bronchiectasis produced sputum that grew MAC on culture. Infection with MAC was documented in one half of patients with fibronodular bronchiectasis if bronchoscopy with BAL and transbronchial biopsy were performed.⁸⁸⁸⁹ Data from the HRCT studies⁸⁶⁸⁷⁹⁰ indicate that approximately one half of patients have pleural thickening or adhesions, usually adjacent to the pulmonary parenchymal abnormalities. Other reported abnormalities include atelectasis, consolidation, "tree in bud," and "ground-glass" opacities. Gas trapping occurs in most patients with MAC lung disease.^86909192 Studies⁹³⁹⁴ comparing the CT findings with those of tuberculosis found that bronchiectasis is more common in MAC lung disease. Serial CT scanning in MAC lung disease has shown that bronchiectasis tends to become more severe and nodules spread to other segments over time.⁸⁷⁹⁵ The development of consolidation and ground-glass opacities is preceded by the appearance of nodules. Some small nodules will disappear with successful treatment.⁹⁵

Upper lobe fibrocavitary changes have been well described in patients with MAC lung disease.^{606162646869737980} Patients with MAC lung disease who have evidence of cavitary lung disease on CT scan are more likely to have smear-positive sputum than.⁸⁰⁹⁴

	Microbiology of M avium

TOP Abstract Introduction Background Effects of Exposure to... Epidemiology of MAC Lung... Clinical Features of MAC... Radiologic Features of MAC... Microbiology of M avium Predisposing Factors for MAC... Conclusion References

 
Unlike M tuberculosis, the NTM can grow at room temperature.⁶⁶ Runyon and Timpe³⁵ found that the NTM have little, if any, pathogenicity for guinea pigs, but group 1 (M kansasii) and group 3 (MAC) organisms are pathogenic for mice.³⁵ The NTM demonstrate increased catalase activity and niacin negativity (compared to M tuberculosis).⁶⁶ The cell wall of MAC is trilaminar with a basal peptidoglycan layer, an intermediate layer that is transparent as viewed by electron microscopy, and an outer layer that is poorly penetrated by hydrophilic medications.⁹⁶ The relative resistance to antimycobacterials is presumably due to the lipophilicity of the cell wall.⁹⁷

The 1997 ATS statement on the diagnosis and treatment of disease caused by NTM recommended that a minimum of three sputum specimens or other respiratory specimens be stained and cultured in symptomatic patients presenting with infiltrative, nodular, or cavitary lung disease.⁸ The rationale for the requirement for multiple positive culture results is that MAC is present in tap water, even in hospitals, and false-positive culture results may occur.³³⁴⁸⁴⁹

Techniques used for testing for M tuberculosis are satisfactory for staining and culturing for NTM, including MAC.⁸ Patients with bronchiectasis may also have infection with other bacteria, including Pseudomonas or fungi. Care must be taken to prevent the overgrowth of these organisms in culture. MAC can be rapidly identified with commercial DNA probes. Clinical trials have not shown susceptibility testing to antituberculosis medications to be useful. The National Committee for Clinical Laboratory Standards (NCCLS) guidelines for susceptibility testing of mycobacteria, Nocardia, and other aerobic Actinomycetes, published in 2000, recommended that clarithromycin susceptibility testing be done for respiratory MAC isolates from patients with pulmonary disease.⁹⁸ The rationale for this recommendation is that a macrolide (clarithromycin) is the only medication that has demonstrated a correlation between clinical response and the results of in vitro susceptibility testing in the setting of a controlled clinical trial.⁹⁹ The 2000 NCCLS guidelines differ from the 1997 ATS guidelines that recommended clarithromycin susceptibility testing only on isolates from patients who had failed therapy.⁸⁹⁸ The NCCLS guidelines also recommend that clarithromycin susceptibility testing should be done on respiratory isolates from patients receiving prior macrolide therapy and on respiratory isolates from patients who relapse while receiving macrolide therapy.¹⁰⁰ The 1997 ATS guidelines for management of M avium lung disease are summarized in Table 2 .⁸

A single strain of MAC tends to persist in patients with advanced fibrocavitary disease. Patients with fibronodular bronchiectasis may grow different strains over time, suggesting a susceptibility to the organism that allows reinfection after it has been eradicated.¹⁰²

	Predisposing Factors for MAC Lung Disease

TOP Abstract Introduction Background Effects of Exposure to... Epidemiology of MAC Lung... Clinical Features of MAC... Radiologic Features of MAC... Microbiology of M avium Predisposing Factors for MAC... Conclusion References

 
Most patients in the earlier reports were male and had upper lobe cavitary disease. Patients with MAC lung disease tended to be older than patients with M tuberculosis or M kansasii, had a history of tobacco or alcohol consumption, and had underlying lung disease.⁶⁸⁶⁹⁷³ In the more recent reports, ² most patients with fibronodular bronchiectasis were female and elderly. A Japanese study¹⁰³ suggested that being thin was a risk factor for progression of disease in patients with fibronodular bronchiectasis.

Individuals carrying a cystic fibrosis or an abnormal ₁-antiprotease gene appear to be predisposed to the development of MAC lung disease.¹⁰⁴ A recent American multicenter study¹⁰⁵ found that NTM could be cultured from the sputum of 13% of patients with cystic fibrosis > 10 years old. The two most common NTM species were MAC and M abscessus, accounting for 72% and 16% of the NTM isolates, respectively. The patients whose sputum grew NTM tended to be older, had higher FEV₁ values, lower body mass indexes, and were more likely to harbor Staphylococcus aureus and less likely to harbor Pseudomonas aeruginosa than patients with mycobacterial culture-negative sputum. It is not clear whether the association with these bacteria and likelihood of harboring NTM reflects that patients who harbor S aureus have less severe lung disease than those with P aeruginosa, or whether these bacteria modify conditions that influence NTM.¹⁰⁵ Serial HRCT scans indicated that patients with cystic fibrosis and NTM were more likely to experience progressive pulmonary parenchymal changes over the subsequent 15 months.¹⁰⁶ Patients with thoracic disorders including pectus excavatum and kyphoscoliosis are also at greater risk of acquiring MAC lung disease.¹⁰⁷¹⁰⁸

In a murine model, mutations in natural resistance-associated macrophage protein (NRAMP1) lead to overwhelming mycobacterial infection, but a study¹⁰⁹ in elderly women did not find evidence for a genetic NRAMP1 defect. Two Japanese families affected by MAC lung disease were studied to determine if host factors or virulence of the MAC strain were responsible for infection.¹¹⁰ The investigators did not find evidence that an abnormal NRAMP1 gene was causative in either family. Nor did it appear that a particularly virulent strain was responsible, since the infecting strain was different in different family members.¹¹⁰ A recently published retrospective review¹¹¹ found that patients with MAC lung disease were more likely to have sinusitis, had higher IgA levels, and a higher percentage of sputum neutrophils than patients with pulmonary tuberculosis.

An analysis of the major histocompatibility complex (MHC) antigens in patients with MAC lung disease found that the A-26 antigen was more common in patients than in controls.¹¹² In an earlier study, the frequency of haplotype A2-B12 was reported to be significantly higher than expected in patients with MAC lung disease.¹¹³ Other investigative groups have not confirmed the associations between these MHC antigens and MAC lung disease.

Immunology of MAC Lung Disease
Since the organism is ubiquitous, it is fortunate that MAC is not as virulent as M tuberculosis and normal host defenses are effective.⁶⁵ Unlike tuberculosis, most animals are resistant to MAC infection unless host defenses are iatrogenically altered. Certain strains can infect chickens, but these are different than those isolated from humans.⁶⁵ Skin-testing studies indicate that many people are exposed and sensitized to MAC, yet only a minority acquire disease.⁵⁹

Patients with lymphoreticular malignancies and patients with AIDS may acquire disseminated MAC, usually when the CD₄⁺ count is < 50/µL.⁵⁶⁷¹¹⁴ Since patients with AIDS have a constellation of T-cell immune defects, it is unclear which are responsible for the susceptibility to MAC.²⁴ Stimulated CD₄⁺ T-lymphocytes produce interferon (IFN)-, a glycoprotein, and other cytokines. Since the mid 1990s, a number of families have been described with members who have IFN- receptor defects and are predisposed to intracellular pathogens including MAC and other mycobacteria.^{115116117118119120121}

Tumor necrosis factor (TNF)- and IFN- are both important in the defense against mycobacterial disease.¹²² In response to the phagocytosis of mycobacteria, macrophages release interleukin (IL)-12 and IL-18, stimulating production of IFN- by T- and null killer-lymphocytes.^114118121122 IL-12 acts synergistically with rifabutin or clarithromycin to promote the clearance of MAC infection in mice.¹²³ IL-12 and IFN- act on lymphocytes and macrophages to stimulate the production of IL-2, TNF-, other cytokines including granulocyte macrophage-colony stimulating factor, and reactive oxidative intermediates such as superoxide and nitric oxide that participate in the host defense against mycobacteria.¹²⁴ IFN- decreases phagosomal pH and increases the intracellular concentrations of macrolides and fluoroquinolones.¹¹⁸ IFN- increases surface display of the MHC antigens and Fc receptors and activates kinases that, in turn, activate transcription factors.¹²⁵ The specific role for each of these IFN-–stimulated actions on the regulation of mycobacterial growth is not yet clear.¹²¹ At least one of these IFN-–induced transcriptional factors has been shown to play an important role in mycobacterial immunity.¹²⁵ Animal studies¹¹²¹¹⁹ confirm the central role of IFN- in the host defense against mycobacteria. Subcutaneous and aerosolized IFN- has been used experimentally in the treatment of disseminated and pulmonary NTM infection and in patients with multidrug-resistant tuberculosis.^121126127

TNF- also appears to play an important role in the defense against mycobacterial infection. Mice that are TNF- receptor deficient are highly susceptible to mycobacterial infection. Among its contributions to host defense, TNF- concentrates antimycobacterials in macrophages.¹²⁸ In a murine model, it enhanced the antibacterial activity of clofazimine against MAC.¹²⁹ Overproduction of TNF- is believed to have systemic effects including fever production, weight loss, and tissue necrosis.¹²¹

Vankayalapati and colleagues²⁴ compared the effects of MAC antigen on peripheral blood monocytes from patients with MAC lung disease to those from healthy control subjects with a positive delayed hypersensitivity skin response to MAC sensitin. The monocytes from patients with MAC lung disease produced more IL-10, which is immunosuppressive, but less IL-12, IFN-, and TNF- in response to incubation with MAC antigen.²⁴ A subsequent small study¹³⁰ reported lower levels of IFN- in patients with MAC lung disease that did not increase with successful therapy. It is attractive to speculate that abnormal production of cytokines and ILs may contribute to the predisposition to MAC lung disease.

Treatment Trials in MAC Lung Disease
Trials of antituberculosis medications were begun once MAC was recognized to be a pathogen. A PubMed search was undertaken in July 2002 to identify published reports of MAC lung disease therapy. Key terms included M avium, M intracellulare, MAC, M avium-intracellulare, and NTM. The terms were combined with therapy, treatment, clarithromycin, azithromycin, rifabutin, rifampin, ethambutol, isoniazid, streptomycin, amikacin, kanamycin, capreomycin, viomycin, clofazimine, pyrazinamide, ethionamide, cycloserine, para-aminosalicylic acid (PAS), ciprofloxacin, fluoroquinolones, and surgery. The bibliographies of the identified articles were also searched.

A total of 38 treatment reports of MAC lung disease were identified with this search strategy. There were seven reports^{6364666768131132} published between 1957 and 1971 (Table 3 ). The drugs primarily used in these reports were isoniazid, PAS, and streptomycin. Rifampin and ethambutol were used in most of the 19 treatment studies published between 1974 and 1993 and in the British Thoracic Society (BTS) trial (Table 4 ).^{2697374133134135136137138139140141142143144145146147} Clarithromycin or azithromycin were included in the 12 studies published after 1993 (Table 5 ).^{148149150151152153154155156157158159} In comparison to tuberculosis, MAC is relatively resistant to chemotherapy. It is difficult to compare the different published reports since patients were treated over a period of 5 decades and patient populations differed between the reports. Criteria for diagnosis, age and gender mix, extent of lung disease, type and severity of comorbidity, and the proportion of retreated patients differed between the studies.¹⁴¹ Some were treated as inpatients, allowing for strict adherence to prescribed therapy and minimizing the number of patients abandoning therapy before the study was completed, and others were treated out of hospital. Criteria for treatment success also differed between the reports; some included patients who were clinically and/or radiologically stable as successfully treated cases, some included cases that required adjunctive surgery as treatment successes, whereas the more recently reported studies used sputum conversion to negative with medication therapy as the criterion of success. From the 1950s until the mid-1960s, the available medications included isoniazid, PAS, ethionamide, cycloserine, streptomycin, kanamycin, viomycin, and sulfisoxasole. Crow and colleagues⁶⁴ reported their experience with 64 patients with MAC lung disease hospitalized between 1950 and 1955; 56 patients received a regimen that included isoniazid and streptomycin for periods ranging from 3 to 57 months but averaged 17 months. Sputum converted to negative in 17 patients (30%) with chemotherapy alone, 7 patients required surgery, and 32 patients did not convert to negative.⁶⁴ The results of the other six reports before the availability of ethambutol and rifampin were similar (Table 3). ^{63666768131132}

Results with antituberculosis medications improved after the introduction of ethambutol and rifampin in the mid 1960s and early 1970s, respectively. The reports of regimens that included ethambutol and/or rifampin are summarized in Table 4. Most of the reports were from the United States, but some were published elsewhere, including Japan,^133143145 Canada,⁷³ Scandinavia,^138139144 and the United Kingdom.¹³⁷ Success rates varied between studies, but outcomes in the different countries were similar.

The reported rates of treatment success have ranged from 20 to 90% in individual studies⁶⁷¹⁴⁰ (Tables 3, 4). Much of this variability depended on whether an intention-to-treat strategy was used and whether relapses were included in the calculation of treatment success. If patients who were unable or unwilling to complete treatment, those requiring surgery, those who died, and those with relapse during an adequate follow-up period are included, the cure rate was approximately 40%. One group of investigators¹⁴⁰ reported a conversion-to-negative rate of 91% (42 of 46 patients) with a regimen of rifampin, isoniazid, ethambutol, and 6 months of streptomycin; but if the patients who dropped out of the treatment protocol and relapses are included, the success rate was only 42%. Mortality was high in many of the reported series since a large number of the patients were elderly and had serious comorbid conditions.

Results were better with regimens that contained more medications but only for the patients who were able to tolerate the multidrug regimens. Serious toxicity limited the number of medications, dose, and duration of therapy in these regimens. The study by Horsburgh and colleagues¹⁴² was exceptional; all 75 inpatients completed therapy with an average of more than five medications, and the sputum converted to negative in 50 of them. They were able to prevent patients dropping out of the study despite 46% discontinuing one or more drugs because of toxicity.¹⁴²

Some have suggested that adjusting therapy to sensitivity testing results improves outcome in MAC lung disease, but others have not been able to show a benefit.¹⁴²¹⁴⁴ Its relative resistance to individual medications but the possible existence of a synergistic effect of combined therapy on MAC may explain the lack of a correlation between sensitivity testing and clinical outcome results.¹⁶⁰¹⁶¹ Despite the lack of any controlled trials of antituberculous therapy for MAC lung disease at that time, Tsukumura¹⁴³ argued that the increased "minimum inhibitory concentrations" after treatment and the response of the more sensitive organisms to antimicrobial therapy provided evidence of a beneficial effect from treatment.

Apart from the recently published BTS trial, none of the other trials were controlled.¹⁴⁶¹⁴⁷ The BTS trial was started in 1987 and compared 24 months of treatment with rifampin and ethambutol to rifampin, ethambutol, and isoniazid. It found that the patients treated with three drugs had a slightly better conversion-to-negative rate but a higher mortality rate. Five years after treatment, only 23 of the 75 patients (31%) with MAC lung disease were alive and in remission.¹⁴⁷ Unfortunately, the data are no longer relevant since the introduction of macrolide-based regimens.

The newer macrolides, clarithromycin and azithromycin, are concentrated in alveolar macrophages and have good activity against MAC, an intracellular organism.^{162163164165166} Clarithromycin when administered alone and in combination with a rifamycin, especially rifabutin, was effective both for prophylaxis and treatment of disseminated MAC in patients with AIDS.^167168169 Trials of clarithromycin, and more recently azithromycin, administered alone and in combination with other medications, were also undertaken in patients with MAC lung disease without HIV infection (Table 5). ^{148149150151152153154155156157158159} Clarithromycin demonstrated good activity against MAC, but resistance developed when it was administered alone.¹⁴⁸ Emergence of resistance to both clarithromycin and azithromycin may be prevented by the addition of ethambutol or a rifamycin.¹⁷⁰ Most published trials of macrolide-based treatment regimens have included a rifamycin and at least one other antibiotic, usually ethambutol. Streptomycin has also been administered during the first 2 months in many of the macrolide-based trials.⁸

Unfortunately, elderly patients with MAC lung disease frequently experience side effects from the long-term use of clarithromycin.¹⁷¹ Taste perversion, GI upset, hearing problems, allergic reactions, and hepatitis may prevent patients from completing therapy.¹⁷² Azithromycin appears to be as effective as clarithromycin and may be better tolerated.¹⁵⁸ It may also have less pharmacokinetic interaction with rifabutin than clarithromycin.¹⁷³ Rifabutin is more active than rifampin against MAC but causes troubling immune side effects, including polyarthralgias, anterior uveitis, and leukopenia.^174175176 Both rifabutin and rifampin lower serum levels of clarithromycin.¹⁷⁷ Rifampin interferes with the metabolism of a number of drugs metabolized through the hepatic P 450 3A4 cytochrome system. Since most patients with MAC lung disease are elderly, many have other health problems and require medications that may be affected by rifampin. The published macrolide-based trials are summarized in Table 5. More than 80% of the patients in two French trials¹⁵⁴¹⁵⁵ who could tolerate relatively high doses of clarithromycin converted to negative. Unfortunately, nearly 20% of the patients were unable to tolerate treatment. Similar dropout and conversion-to-negative rates were seen in the other trials^149152156158 with clarithromycin. The risk of side effects is likely related to serum drug levels.¹⁷⁸ Early results suggest that azithromycin may be better tolerated than clarithromycin and that it is as effective against MAC.^151153158 Another advantage of azithromycin is that a thrice-weekly regimen is feasible.¹⁵¹¹⁵³ Unfortunately, none of the macrolide-based trials were controlled because of the small numbers of patients with MAC lung disease in any one center. If one includes treatment dropouts and relapses, the cure rate with the macrolide-containing regimens is approximately 56% (Table 5), comparing favorably with the earlier antituberculosis regimens (Tables 3, 4), but a treatment failure rate approaching 45% is unsatisfactory. The treatment success rate can probably be improved by 15 to 20% by substituting drugs that are better tolerated, eg, azithromycin for clarithromycin. However, some patients will remain resistant to antimicrobial therapy, and alternate treatments will require consideration.

Some of the newer fluoroquinolones and linezolid demonstrate activity against MAC and other mycobacteria in in vitro testing. Their role in treatment remains to be determined since none of these drugs have undergone clinical testing.¹⁷⁹

The recognition that these patients have a failure of the IFN- pathways has led to some preliminary studies of cytokine therapy in MAC lung disease, and preliminary results¹²⁷¹⁸⁰ suggest that IFN- may have a role in the treatment of MAC. Animal trials and in vitro studies¹⁸¹ with human monocyte-derived macrophages indicate that IFN- activates macrophage antimicrobial activity. Trials¹²⁶¹²⁷ in patients with MAC lung disease who are refractory to therapy and in patients without HIV infection with disseminated MAC have been undertaken. Although the work on the immune response to MAC is preliminary, the findings raise the possibility that cytokine therapy may have a role in treatment in the future.

	Conclusion

TOP Abstract Introduction Background Effects of Exposure to... Epidemiology of MAC Lung... Clinical Features of MAC... Radiologic Features of MAC... Microbiology of M avium Predisposing Factors for MAC... Conclusion References

 
MAC is an important pulmonary pathogen. It can complicate preexisting lung diseases and can cause lung disease in previously healthy people, usually thin, elderly women. Preliminary data suggest an important role for the IFN- pathways in the host defense against MAC and other mycobacteria. The newer macrolides, clarithromycin and azithromycin, have improved the results of therapy for MAC lung disease, but the failure rate remains high and work is needed to improve the outcome in MAC lung disease. Investigators should collaborate to develop a multicenter trial group to facilitate studies of MAC lung disease and its treatment that are properly powered and controlled.

	Acknowledgements

 
The authors thank Ms. L.G. Field for editorial review.

	Footnotes

 
Abbreviations: ATS = American Thoracic Society; BTS = British Thoracic Society; HRCT = high-resolution CT; IFN = interferon; IL = interleukin; MAC = Mycobacterium avium complex; MHC = major histocompatability complex; NCCLS = National Committee for Clinical Laboratory Standards; NRAMP1 = natural resistance-associated macrophage protein; NTM = nontuberculous mycobacteria; PAS = para-aminosalicylic acid; TNF = tumor necrosis factor

Received for publication October 18, 2002. Accepted for publication November 18, 2003.

	References

TOP Abstract Introduction Background Effects of Exposure to... Epidemiology of MAC Lung... Clinical Features of MAC... Radiologic Features of MAC... Microbiology of M avium Predisposing Factors for MAC... Conclusion References

 

1. Edwards, LB, Palmer, CE (1959) Isolation of "atypical" mycobacteria from healthy persons. Am Rev Respir Dis 80,747-749[ISI][Medline]
2. Prince, DS, Peterson, DD, Steiner, RM, et al Infection with Mycobacterium avium complex in patients without predisposing conditions. N Engl J Med 1989;321,863-868[Abstract]
3. Tsukumura, M Background factors for casual isolation of Mycobacterium intracellulare from sputum of patients with tuberculosis. Am Rev Respir Dis 1973;108,679-683[ISI][Medline]
4. Tsukumura, M Clinical significance of casual isolation of acid-fast organisms from sputum of tuberculous patients. Am Rev Respir Dis 1973;108,1429-1430[ISI][Medline]
5. Horsburgh, CJ, Jr, Selik, RM The epidemiology of disseminated nontuberculous mycobacterial in the acquired immunodeficiency syndrome (AIDS). Am Rev Respir Dis 1989;139,4-7[ISI][Medline]
6. Hawkins, CC, Gold, WM, Whimbey, E, et al Mycobacterium avium-complex infections in patients with the acquired immunodeficiency syndrome. Ann Intern Med 1986;105,184-188[ISI][Medline]
7. Nightingale, SD, Byrd, LT, Southern, PM, et al Incidence of Mycobacterium avium-intracellulare in human immunodeficiency virus-positive patients. J Infect Dis 1992;165,1082-1085[ISI][Medline]
8. American Thoracic Society. Diagnosis and treatment of disease caused by nontuberculous mycobacteria. Am J Respir Crit Care Med 1997;156,S1-S25[ISI][Medline]
9. Griffith, DE Mycobacteria as pathogens of respiratory infection. Infect Dis North Am 1998;12,593-611[ISI][Medline]
10. Hocqueloux, L, Lesprit, P, Herrmann, JL, et al Pulmonary Mycobacterium avium complex disease without dissemination in HIV-infected patients. Chest 1998;113,542-548[Medline]
11. Lincoln, EM, Gilbert, LA Disease in children due to mycobacteria other than Mycobacterium tuberculosis. Am Rev Respir Dis 1972;105,683-714[ISI][Medline]
12. O’Brien, RJ, Geiter, LJ, Snider, DE, Jr The epidemiology of nontuberculous mycobacterial diseases in the United States. Am Rev Respir Dis 1987;135,1007-1014[ISI][Medline]
13. Horsburgh, CR Epidemiology of human disease caused by Mycobacterium avium complex. Can J Infect Dis 1994;5,5B-9B
14. Elwood, RK, Opazo, AM, Lentini, V, et al Incidence of pulmonary disease caused by mycobacteria other than tuberculosis in British Columbia. Can Respir J 2002;9,319-323[Medline]
15. Kennedy, MP, O’Connor, TM, Ryan, C, et al Nontuberculous mycobacteria: incidence in Southwest Ireland from 1987 to 2000. Respir Med 2003;97,257-263[CrossRef][ISI][Medline]
16. Marras, TK, Daley, CL Epidemiology of human pulmonary infection with nontuberculous mycobacteria nontuberculous. Clin Chest Med 2002;23,553-567[CrossRef][ISI][Medline]
17. Barker, AF Bronchiectasis. N Engl J Med 2002;346,1383-1393[Free Full Text]
18. Angrill, J, Agusti, C, Torres, A Bronchiectasis. Curr Opin Infect Dis 2001;14,193-197[CrossRef][ISI][Medline]
19. Kilby, JM, Gilligan, PH, Yankaskas, JR, et al Nontuberculous mycobacteria in adult patients with cystic fibrosis. Chest 1992;102,70-75[CrossRef][ISI][Medline]
20. Oliver, A, Maiz, L, Canton, R, et al Nontuberculous mycobacteria in patients with cystic fibrosis. Clin Infect Dis 2001;32,1298-1303[CrossRef][Medline]
21. Ahn, CH, Lowell, JR, Onstad, GD, et al A demographic study of disease due to Mycobacterium kansasii or M intracellulare-avium in Texas. Chest 1979;75,120-125[Medline]
22. Trnka, L, Dankova, D, Svandova, E Six years’ experience with the discontinuation of BCG vaccination: 4. Protective effect of BCG vaccination against the Mycobacterium avium intracellulare complex. Tuberc Lung Dis 1994;75,348-352[CrossRef][ISI][Medline]
23. Romanus, V, Hallander, HO, Wahlen, P, et al Atypical mycobacteria in extrapulmonary disease among children: incidence in Sweden from 1969 to 1990, related to changing BCG-vaccination coverage. Tuber Lung Dis 1995;76,300-310[CrossRef][ISI][Medline]
24. Vankayalapati, R, Wizel, B, Samten, B, et al Cytokine profiles in immunocompetent persons infected with Mycobacterium avium complex. J Infect Dis 2001;183,478-484[CrossRef][ISI][Medline]
25. American Thoracic Society. Diagnosis and treatment of disease caused by nontuberculous mycobacteria. Am Rev Respir Dis 1990;142,940-953[ISI][Medline]
26. Corpe, RF Surgical management of pulmonary disease due to Mycobacterium avium-intracellulare. Rev Infect Dis 1981;3,1064-1067[ISI][Medline]
27. Moran, JF, Alexander, LG, Staub, EW, et al Long-term results of pulmonary resection for atypical mycobacterial disease. Ann Thorac Surg 1983;35,597-604[Abstract]
28. Nelson, KG, Griffith, DE, Brown, BA, et al Results of operation in Mycobacterium avium-intracellulare lung disease. Ann Thorac Surg 1998;66,325-330[Abstract/Free Full Text]
29. Pomerantz, M, Madsen, L, Goble, M, et al Surgical management of resistant Mycobacterial tuberculosis and other mycobacterial pulmonary infections. Ann Thorac Surg 1991;52,1108-1112[Abstract]
30. Shiriaishi, Y, Nakajima, Y, Takasuna, K, et al Surgery for Mycobacterium avium complex lung disease in the clarithromycin era. Eur J Cardiothorac Surg 2002;21,314-318[Abstract/Free Full Text]
31. Wolinsky, E Nontuberculous mycobacteria and associated diseases. Am Rev Respir Dis 1979;119,107-159[ISI][Medline]
32. Jenkins, PA Nontuberculous mycobacteria and disease. Eur J Respir Dis 1981;62,69-71[ISI][Medline]
33. Pinner, M Atypical acid-fast microorganisms. Am Rev Tuberc 1935;32,424-439
34. Feldman, WH, Davies, R, Moses, HE, et al An unusual mycobacterium isolated from sputum of a man suffering from pulmonary disease of long duration. Am Rev Tuberc 1943;48,82-93
35. Timpe, A, Runyon, EH The relationship of "atypical" acid-fast bacteria to human disease: a preliminary report. J Lab Clin Med 1954;44,202-209[Medline]
36. Lewis, AG, Jr, Dunbar, FP, Lasche, EM, et al Chronic pulmonary disease due to atypical mycobacterial infections. Am Rev Respir Dis 1959;80,188-199[ISI][Medline]
37. Reznikov, M, Leggo, JH Investigation by seroagglutination of strains of the Mycobacterium intracellulare-M scrofulaceum group from house dusts and sputum in southeastern Queensland. Am Rev Respir Dis 1971;104,951-953[ISI][Medline]
38. Bhatia, SVN A study of mycobacterial species isolated from patients with chronic respiratory disease. Am Rev Respir Dis 1973;108,378-381