Tuberculosis and Nontuberculous Mycobacterial Infections

Published on 01/06/2015 by admin

Last modified 22/04/2025

Print this page

This article have been viewed 2605 times

Chapter 31 Tuberculosis and Nontuberculous Mycobacterial Infections

The genus Mycobacterium consists of slow-growing organisms that are widely disseminated throughout the world and range from species that cause no human disease to those such as Mycobacterium tuberculosis and Mycobacterium leprae that are responsible for enormous morbidity and mortality. Mycobacteria are aerobic bacilli with high concentrations of lipids in their cell wall, which make them impermeable to most common stains. However, because of their ability to retain carbolfuchsin dye despite decolorization attempts with acid alcohol, they are referred to as acid-fast bacilli (AFB). Although mycobacteria can produce disease in almost any site, two groups of mycobacteria have a propensity for causing pulmonary infections: certain members of the M. tuberculosis complex and the nontuberculous mycobacteria (NTM).

Tuberculosis (TB) is the disease caused by bacteria of the M. tuberculosis complex, which includes the clinically relevant species M. tuberculosis, Mycobacterium bovis, and Mycobacterium africanum. Although M. tuberculosis is the most common cause of TB worldwide, both M. bovis and M. africanum can produce clinically indistinguishable forms of disease. The tubercle bacilli have been around for thousands of years, with evidence of human infection dating back to Neolithic, pre-Columbian, and early Egyptian times. Not until the Industrial Revolution, however, did TB become a major cause of human disease and death. It is estimated that approximately 25% of all adults died from TB in Europe during the 17th and 18th centuries. Throughout this period, the etiology of TB was hotly debated, with some of these early investigators arguing for a hereditary cause and others for a transmissible etiology. Finally, in 1882, Robert Koch presented his momentous discovery: The tubercle bacillus was the cause of TB. Early attempts at therapy, including the sanatorium movement, surgery, and lung collapse therapy, provided little relief from TB, and it was not until the discovery of paraaminosalicylic acid (PAS) and streptomycin in the 1940s that the age of antituberculosis chemotherapy began. Since that time, additional drugs have been developed, and most of the world treats TB with the same four-drug regimen administered for 6 months. More recently, however, co-infection with human immunodeficiency virus (HIV) and the emergence of drug-resistant strains of M. tuberculosis have conspired to complicate clinical management and to create barriers to global TB control.

The NTM group consists of nonlepromatous organisms that are not members of the M. tuberculosis complex. The NTM have been referred to as “mycobacteria other than tuberculosis” (MOTT), atypical mycobacteria, and environmental mycobacteria. The last designation refers to their widespread presence in the environment. The NTM have several features that distinguish them from M. tuberculosis: They have a wide range of pathogenicity, are not always associated with disease, and, unlike M. tuberculosis, are not transmissible from human to human. Of note, however, the incidence of NTM disease is increasing in many areas of the world, and the cause for this increase is unknown. Unfortunately, the pathogenic NTM are relatively drug resistant compared with M. tuberculosis, so NTM infections typically are difficult to treat. Because of the current poor understanding of the transmission and pathogenesis of these infections, little insight into their prevention has emerged, so no public health strategy for the control of disease caused by these ubiquitous organisms has been formulated.

Tuberculosis

Epidemiology, Risk Factors, and Pathogenesis

Epidemiology

The World Health Organization (WHO) estimates that 30% of adults worldwide are infected with organisms in the M. tuberculosis complex. From this large reservoir of infected people, an estimated 9 million new cases of TB occurred in 2009, leading to approximately 1.3 million deaths. In 2008, TB was estimated to be the seventh leading cause of death worldwide, and it is the number one killer of HIV-infected patients.

The burden of TB varies significantly throughout the world, with more than 90% of cases occurring among people residing in developing countries (Figure 31-1). The highest incidence rates for TB are in sub-Saharan Africa, particularly in the southern region of the continent. Not surprisingly, the highest prevalence of HIV co-infection also is in this region. Worldwide, approximately 23% of all persons with TB have underlying HIV co-infection; however, in sub-Saharan Africa, an estimated 50% of persons with TB have HIV/AIDS.

Figure 31-1 Tuberculosis (TB) notification rates, 2009. The legend depicts the number of notified TB cases, which includes both new and relapse cases per 100,000 population.

(From Global tuberculosis control 2010, Geneva, World Health Organization, 2010.)

Recent reports of outbreaks of multidrug-resistant TB (MDR TB) and extensively resistant TB (XDR TB) have highlighted the importance of providing effective antituberculosis therapy to patients. MDR TB refers to disease caused by isolates of M. tuberculosis that are resistant to at least isoniazid (INH) and rifampin, whereas XDR TB refers to that due to MDR TB isolates that also are resistant to fluoroquinolones and at least one second-line injectable agent (amikacin, capreomycin, or kanamycin). Surveys have documented that approximately 4.6% of TB cases worldwide are MDR TB, and 5.4% of these cases are XDR TB. More cases of drug-resistant TB exist today than in all of recorded history, and this trend is likely to continue unless more effective TB control measures are implemented globally.

In the United States, the TB case rate declined by 3% to 5% per year from 1953 to 1984. Between 1986 and 1992, the numbers of TB cases increased by approximately 20%. This increase in the number of cases was the result of at least three major factors: (1) inadequate public health measures, (2) immigration from countries where TB is prevalent, and (3) co-infection with HIV. Fortunately, case numbers have declined since 1992 and are now at a historic low. In 2010, a total of 11,181 cases of TB were reported in the United States, for an incidence of 3.6 per 100,000 population. TB case rates declined an average of 4.5% each year during 2000 to 2010. TB rates were 11 times higher among foreign-born persons than among U.S.-born people. Among U.S.-born persons, blacks were 7 times more likely to have TB than whites. Approximately 1% of new cases in the country had MDR TB, approximately 90% of whom were foreign-born. Of these MDR TB cases, approximately 1% to 2% were XDR TB.

Risk Factors

Certain people are at higher risk for development of TB simply because they are more likely to be exposed to and thus infected with M. tuberculosis (Table 31-1). For example, approximately 60% of reported TB cases in the United States occur among foreign-born people who come from areas where the disease is endemic. Other populations with an increased prevalence of TB infection include certain racial and ethnic groups, low-income populations, the homeless, and injection drug users.

Table 31-1 Criteria for a Positive Reaction on Tuberculin Skin Testing

Induration Size	Risk Groups
≥5 mm	HIV-infected persons Close contact with an infectious tuberculosis case Abnormal-appearing chest radiograph * consistent with previous tuberculosis Immunosuppressed patients receiving the equivalent of ≥15 mg/day of prednisone for at least 1 month

≥10 mm

Foreign-born persons recently arrived (<5 years) from high-prevalence countries

Medical conditions † that increase the risk of tuberculosis

Injection drug users

Medically underserved, low-income populations (e.g., homeless persons)

Residents and staff of long-term care facilities (e.g., nursing homes, correctional institutions, homeless shelters)

Health care workers

Children <4 years of age

Tuberculin skin test converters (increase of ≥10 mm induration within a 2-year period)

≥15 mm

All others; these persons should not be screened in the absence of indication

* The predominant chest radiographic finding consistent with previous tuberculosis is presence of fibrotic lesions; other changes such as pleural thickening or isolated calcified granulomas are not related.

† Medical conditions and factors associated with increased risk for development of active disease in a patient with latent tuberculosis infection include silicosis, end-stage renal disease, malnutrition, diabetes mellitus, carcinoma of the head or neck and lung, immunosuppressive therapy, lymphoma, leukemia, weight loss of more than 10% ideal body weight, gastrectomy, and jejunoileal bypass.

Anyone infected with M. tuberculosis can develop TB disease, but certain groups are at higher-than-normal risk for progression to active disease (see Table 31-1). Patients who have been recently infected with M. tuberculosis and those with medical conditions associated with significant immunosuppression are at particularly high risk for development of TB. HIV co-infection is the strongest known risk factor for the development of TB and is estimated to increase the risk of progression to TB by 50- to 100-fold. Inhibitors of tumor necrosis factor-α (TNF-α) may increase the risk for development of TB by up to 10-fold, and patients taking TNF-α blockers frequently present with disseminated disease. This association appears to be stronger for infliximab and adalimumab than for etanercept. Other medical conditions are associated with a more modest increase in risk for development of disease.

Pathogenesis

TB is spread from person to person almost exclusively through the air by droplet nuclei, which are particles 1 to 5 µm in diameter that contain viable tubercle bacilli. Droplet nuclei are expelled into the air when patients with infectious TB create an aerosol by talking, coughing, or singing. Three factors determine the likelihood of transmitting TB: the number of bacilli expelled into the air, the concentration of organisms in the air, and the duration of contact with (i.e., breathing of) the infected air. Whether an inhaled tubercle bacillus establishes an infection in the exposed person’s lung depends on both bacterial virulence and host immune defenses.

The tubercle bacillus grows slowly, dividing approximately every 18 to 24 hours. Tubercle bacilli spread through the lymphatics to the hilar lymph nodes or through the bloodstream. Small numbers of bacilli are deposited in other organs, which may then become sites of extrapulmonary disease. An adaptive immune response occurs after 2 to 8 weeks.

Once cell-mediated immunity develops, collections of activated T cells and macrophages form granulomas that wall off the mycobacterial organisms (Figure 31-2). For most persons with normal immune function, infection with M. tuberculosis seems to be arrested once cell-mediated immunity develops, even though small numbers of viable bacilli remain within the granuloma. Although a primary complex can sometimes be seen on chest radiograph, most TB infections are asymptomatic and can be detected only indirectly with a tuberculin skin test (TST) or interferon-γ release assay (IGRA). Persons with “walled-off” TB infection who do not have active disease are not infectious and thus cannot spread the disease to others.

Figure 31-2 Caseating granuloma in lung tissue. This lung biopsy specimen contains a caseating granuloma. The large arrow highlights the caseous center. The smaller arrows point out giant cells, typical of granulomatous inflammation.

If cell-mediated immunity does not contain the tubercle bacilli, the initial infection progresses to active disease. Without treatment, infected persons have approximately a 5% chance of developing TB in the first 1 to 2 years after infection and an additional 5% chance of developing TB during the remainder of their lifetime (Figure 31-3). By contrast, persons who are co-infected with HIV have a 5% to 10% annual risk of active disease developing. When active TB develops soon after infection, the disease is referred to as primary TB. By contrast, when TB develops years or even decades after the initial infection, the disease is referred to as postprimary or reactivation disease. Exogenous reinfection, involving acquisition of a second strain of M. tuberculosis, also can lead to disease and seems to be more common in HIV-infected patients.

Figure 31-3 Pathogenesis of tuberculosis. After exposure to an infectious case of tuberculosis, approximately 30% of close contacts become infected with Mycobacterium tuberculosis. Of the contacts who become infected, approximately 5% will develop active tuberculosis in the first year after infection and an additional 5% will develop active tuberculosis over the course of their lifetime. Therefore, approximately 90% of infected people will not develop tuberculosis. When tuberculosis is introduced into an HIV-infected population, its pathogenesis is altered. Although susceptibility is difficult to establish conclusively, HIV-infected people appear to be more susceptible to initial infection than HIV-negative contacts. In outbreak settings, approximately 40% of HIV-infected contacts develop tuberculosis within the first year of exposure. After tuberculosis infection, HIV-infected people develop active tuberculosis at a rate of 5% to 10% per year. HIV, human immunodeficiency virus.

Genetics

Both susceptibility and resistance to developing TB have long been thought to have a genetic component. Epidemiologic and genetic studies, including those in animal and human models, support this hypothesis. Recent investigations of specific candidate genes and genome-wide scans have identified people at increased risk for TB. Although polymorphisms in more than 10 genes have been associated with active TB, only polymorphisms in the human leukocyte antigen (HLA)-DR molecules and in the genes for the vitamin D₃ receptor, SCL11A-1, IFN-γ promoter, and mannose-binding lectin all have been associated with increased susceptibility to M. tuberculosis. Because each association has been relatively modest at the individual level, the genetic susceptibility is likely to be polygenic in nature. Ultimately, whether an individual with TB infection progresses to disease will depend on the interplay among host, infecting organism, and environment.

Clinical Features

The clinical manifestations of TB are protean. Patterns of disease vary depending on whether the disease is primary or reactivation in nature, the host’s immune status, and possibly the strain of M. tuberculosis. Of note, the clinical features of active TB are the result of a balance between host defenses and bacterial virulence; therefore, a continuum of disease is likely, and the clinical presentation of disease may be altered in severely immunocompromised patients. Most patients initially are seen with pulmonary disease, which is classically divided into primary disease and postprimary disease.

Diagnosis

To diagnose TB, the disease must first be suspected. TB should be suspected in certain high-risk groups reviewed previously (see Table 31-1) and when the clinical and/or radiographic presentation is consistent with TB. The medical history should elicit whether or not the person suspected of having TB has been exposed to M. tuberculosis or has a previous history of TB infection or disease. Symptoms at presentation will vary depending on the sites(s) of involvement and extent of disease as described previously. Guidelines suggest that all persons with an unexplained cough lasting 2 to 3 weeks or more be evaluated for TB. Of note, up to 20% of patients with pulmonary disease are asymptomatic. Findings at physical examination are rather nonspecific and will vary, depending on the site of involvement. Among HIV-infected patients, TB should be considered when any respiratory infection or fever of unknown origin occurs, because the risk for TB in this group is substantially elevated, and signs and symptoms of TB often are atypical.

Tuberculin Skin Test and Interferon-γ Release Assays

The TST (discussed in more detail later on), which uses purified protein derivative (PPD), is the most common way to identify persons with latent tuberculosis infection (LTBI) but should not be used in the diagnosis of active TB. In general, the sensitivity of the TST for detection of active TB ranges from 65% to 94%, but in critically ill patients with disseminated disease, the sensitivity decreases to only 50%. Thus, a negative TST reaction can never exclude a diagnosis of TB.

IGRAs measure the release of IFN-γ in whole blood in response to stimulation by M. tuberculosis antigens. Whole blood is incubated overnight with early secretory antigen target 6 [ESAT-6], culture filtrate protein 10 [CFP10], TB7.7, and control antigens; lymphocytes sensitized by previous exposure to M. tuberculosis release IFN-γ. IGRAs currently available include the QuantiFERON-TB (QFT-TB) Gold and QFT-TB Gold In-Tube (QFT-GIT), which measure IFN-γ in the serum using enzyme-linked immunosorbent assay (ELISA) (Cellestis Limited, Carnegie, Victoria, Australia) and the T-Spot.TB test, which uses enzyme-linked immunospot (ELISPOT) methodology (Oxford Immunotec, Oxford, United Kingdom) to identify IFN-γ–producing cells.

The reported sensitivity of QFT-GIT in patients with active TB has varied, ranging from 62% to 94%, with a pooled sensitivity of 80%, whereas the T-Spot.TB test has a sensitivity of 35% to 100%, with a pooled sensitivity of 81%. By contrast, the TST has a pooled sensitivity of 65% in patients with TB. Although the IGRAs have improved sensitivity compared with the TST, the values are still too low to rule out active TB with confidence, and neither test can differentiate latent from active TB.

Radiographic Examinations

Plain chest radiography is a sensitive but nonspecific test to detect pulmonary TB. Radiographic manifestations of TB vary, depending on whether the patient has primary or postprimary TB and whether co-infection with HIV is present. Patients who have primary pulmonary TB at initial evaluation may demonstrate radiographic opacities in the lower lung zones and an associated pleural effusion (see Figure 31-4). TB caused by reactivation typically involves the apical and posterior segments of the upper lobes or superior segment of the lower lobe (see Figure 31-5). Cavitation and volume loss are common in reactivation disease but unusual in primary disease. Findings on the chest radiograph in patients co-infected with HIV depend on the severity of immunosuppression. Early in the course of HIV disease, the radiograph may show a typical reactivation pattern with cavitation (Figure 31-7), but as the CD4⁺ cell count declines, the radiographic appearance is more like the pattern seen in primary TB (Figure 31-8). Patients co-infected with HIV may sometimes have a normal-appearing chest radiograph despite being sputum AFB smear–positive.

Figure 31-7 Tuberculosis in a human immunodeficiency virus (HIV)-infected patient. The chest radiograph demonstrates a left upper lobe cavitary process (large arrow). In addition, bilateral hilar adenopathy and aortopulmonary window adenopathy (small arrows) are evident.

Figure 31-8 Primary-type presentation of tuberculosis in a human immunodeficiency virus (HIV)-infected patient. The chest radiograph demonstrates right lower lobe and right middle lobe air space consolidation with likely right hilar and paratracheal adenopathy.

Bacteriologic Examination

Sputum Microscopy

Diagnosis of pulmonary TB begins with obtaining two or three spontaneously expectorated sputum samples collected at 8- to 24-hour intervals, with at least one collected in early morning. Two methods are commonly used for acid-fast staining: the carbolfuchsin methods (Ziehl-Neelsen and Kinyoun methods) and a fluorochrome procedure that uses auramine O or auramine-rhodamine dyes (Figure 31-9).

Figure 31-9 Acid-fast stain in tissue showing Mycobacterium tuberculosis. (Ziehl-Neelsen stain.)

Approximately 5,000 to 10,000 bacilli/mL are necessary to allow detection of these organisms in stained smears. The sensitivity of sputum AFB smears ranges from 50% to 80%, depending on the extent of disease; patients with cavitary disease are more likely than those without cavities to expectorate tubercle bacilli. Light-emitting diode (LED)–based fluorescence microscopy allows for more rapid evaluation of specimens and may increase the sensitivity slightly over that for conventional light microscopy. If patients are unable to produce sputum or have negative sputum smears, additional diagnostic tests may be indicated. In such circumstances, either sputum induction or biopsy using fiberoptic bronchoscopy (FOB) may provide adequate specimens. Studies suggest that sputum induction with hypertonic saline and FOB with bronchoalveolar lavage produce similar yields in smear-negative cases. The primary role of FOB is in smear-negative HIV-infected TB suspects, in whom this technique also can help identify alternative causes of the illness.

Mycobacterial Cultures and Identification

Culture of M. tuberculosis remains the “gold standard” modality for diagnosis of TB, and isolation of the organism is necessary for drug susceptibility testing and genotyping. Thus, all clinical specimens suspected of containing mycobacteria should be inoculated onto culture media. There are three different types of traditional culture media: egg-based (Lowenstein-Jensen or Ogawa), agar-based (Middlebrook 7H10 or 7H11), and liquid (Middlebrook 7H12) media. Growth in the liquid media is faster than that in solid media, and automated commercial broth systems allow for growth detection within 1 to 3 weeks compared with solid media, for which growth takes 3 to 8 weeks. However, solid media allow for observation of colony morphology and the detection of mixed infections. Because only 10 to 100 organisms are required to detect M. tuberculosis, cultures are more sensitive than smears, with reported sensitivity ranging from 80% to 93%. The tubercle bacilli can be identified from cultures and distinguished from NTM species by chemical means or with molecular methods.

Molecular Assays

Nucleic acid amplification assays (NAAs) amplify and detect M. tuberculosis–specific nucleic acid sequences in clinical specimens within 24 to 48 hours. Two U.S. Food and Drug Administration (FDA)-approved NAAs are available in the United States: the AMPLICOR M. tuberculosis (Roche Diagnostic Systems, Inc., Branchburg, New Jersey) and the Amplified Mycobacterium Tuberculosis Direct (MTD) Test (Gen-Probe, Inc., San Diego, California). The enhanced MTD (E-MTD) assay is approved for use with both smear-negative and smear-positive specimens, but the AMPLICOR assay is approved for use with smear-positive specimens only. The assays show sensitivities of at least 80% to 90% in most studies, with specificities of approximately 98% to 99% in smear-positive specimens. The E-MTD assay has been shown to have a sensitivity and specificity close to 100% in smear-positive specimens, and in smear-negative specimens the sensitivity is 90% and specificity is 99%.

NAAs are particularly valuable for rapidly distinguishing TB from NTM infection. When both the NAA and AFB smear are positive, pulmonary TB is almost certain and TB therapy should be initiated. If the NAA result is negative but the sputum smear is positive, testing the sputum for inhibitors and repeating the NAA is advised. If inhibitors are not detected and a second sputum specimen is NAA-negative but smear-positive, the patient probably has an NTM infection. If smears are negative but the clinical suspicion is intermediate to high, an NAA should be ordered, because a positive NAA result in this setting is likely to indicate TB. However, a negative NAA result in a smear-negative patient does not exclude TB. NAA tests should not be performed on sputum specimens from patients at low risk for development of TB. NAAs have been shown to reduce time to diagnosis, to accelerate contact investigations, and to reduce nonindicated TB treatment and are considered a standard component of the TB-diagnostic armamentarium.

Use of NAA for testing extrapulmonary specimens has been systemically reviewed. The sensitivity of commercial assays for detecting M. tuberculosis in CSF and pleural fluid has been approximately 60%, with a specificity of 98%. Therefore, NAAs may be useful in confirming a diagnosis of CNS or pleural TB, but because of the low sensitivity, they cannot be used to rule out disease.

The Xpert MTB/RIF (Cepheid, Sunnyvale, California) is an automated diagnostic system that performs real-time PCR to amplify an MTB-specific sequence of the rpoB gene, which is probed with molecular beacons to identify rifampin resistance. The device, which provides results within two hours, indicates both the presence of M. tuberculosis and the presence of rifampin resistance. In a large multicountry study, the Xpert MTB/RIF correctly identified 98% of patients with smear-positive and culture-positive TB. The sensitivity for smear-negative, culture-positive disease was 72%.

Drug Susceptibility Testing

Drug susceptibility studies should be performed on all initial isolates and only by laboratories that have experience in culturing mycobacteria. Drug susceptibility testing also should be performed on patients whose treatment is failing or who have a recurrence. The agar proportion method and the liquid radiometric or chemoluminescence methods are the ones most commonly used in the Unites States. Automated radiometric procedures for drug susceptibility testing offer more rapid results but often require confirmation with solid media.

More rapid methods of detection of drug resistance are now available. The WHO recommends that rapid methods of drug susceptibility testing for INH and rifampin or rifampin alone be used over conventional testing. The Xpert MTB/RIF test described previously correctly identified 98% of rifampin resistance. Since rifampin resistance rarely occurs in isolation and is generally accompanied with INH resistance, rifampin resistance frequently indicates MDR TB. Thus, Xpert MTB/RIF can provide a presumptive diagnosis of MDR TB within 2 hours. Unlike the Xpert MTB/RIF, other molecular methods are able to identify mutations in the M. tuberculosis genome that confer resistance to multiple agents. Line-probe assays like the Genotype MTBDRplus (Hain Lifesciences, Nehren, Germany) can detect mutations that confer resistance to rifampin and INH whereas the MTBDRsi test also can identify resistance to fluoroquinolones, aminoglycosides, and ethambutol. In the United States, the Molecular Detection of Drug Resistance program at the Centers for Disease Control and Prevention (CDC) uses DNA sequencing to identify mutations that confer resistance to rifampin, INH, fluoroquinolones, kanamycin, amikacin, and capreomycin. With previous authorization, this service accepts and tests isolates from patients at particularly high risk for MDR or XDR TB.

Treatment

Identifying and treating patients with TB is the most effective way of preventing transmission in the community. TB must be treated with at least two drugs to which the organism is susceptible, to prevent the emergence of drug resistance. Dosages of commonly used first-line and second-line drugs are shown in Tables 31-2 and Table 31-3, respectively. In the United States, a regimen consisting of INH and rifampin given for 6 months, plus pyrazinamide for the initial 2 months, is considered standard short-course therapy. Ethambutol should be added to the treatment regimen for the first 2 months of therapy, but once a drug-susceptible isolate has been demonstrated, ethambutol can be stopped (Table 31-4). After the first 2 months of treatment, one of several regimens can be chosen for the continuation phase of treatment. For HIV-negative patients who have a documented negative AFB smear after 2 months of therapy and have no evidence of cavitation on the initial chest radiograph, a once-weekly regimen consisting of rifapentine and INH is effective. The total duration of therapy should be 6 months, but in patients whose 2-month culture remains positive and whose chest radiograph shows evidence of cavitation, the continuation phase should be extended by 3 months, to complete a 9-month treatment course (Figure 31-10).

Table 31-2 First-Line Antituberculosis Drug Regimens for Adults and Children

Table 31-3 Second-Line Antituberculosis Drug Regimens for Adults and Children

Table 31-4 Drug Regimens for Culture-Positive Pulmonary Tuberculosis Caused by Drug-Susceptible Organisms

Figure 31-10 Treatment of pulmonary tuberculosis. Patients suspected to have tuberculosis whose chest radiograph (CXR) shows no evidence of cavitation and who have negative acid-fast bacilli (AFB) sputum smears after 2 months of therapy can be treated in the continuation phase with isoniazid (INH) and rifampin (RIF) to complete a 6-month duration of therapy. Alternately, they may be treated with INH and rifapentine (RPT) administered once weekly. If cultures remain positive after 2 months of therapy, treatment duration should be extended by 3 months. TB suspects who have evidence of cavitation on the chest radiograph or who have a positive AFB smear after 2 months of therapy should be treated with INH and RIF. If the 2-month culture is negative, they can be treated for a total of 6 months. If the 2-month culture is positive and there is no evidence of cavitation on the CXR, they can be treated for 6 months. However, if cavitation is present, the continuation phase should be lengthened by 3 months. EMB, ethambutol; PZA, pyrazinamide.

(From American Thoracic Society; CDC; Infectious Diseases Society of America: Treatment of tuberculosis, MMWR Recomm Rep 52[RR-11]:1–80, 2003.)

Some patients may be intolerant of a first-line drug or have underlying drug-resistant disease. In such cases, addition of a second-line drug may be necessary (see Table 31-3). These medications include the fluoroquinolones, PAS, ethionamide, cycloserine, clofazimine, and injectables such as kanamycin, capreomycin, and amikacin. The duration of treatment for drug-resistant TB will be determined by the drugs used, the site, and the extent of the disease. Expert consultation should be obtained for treating drug-resistant TB.

To prevent acquired drug resistance, clinicians must prescribe an adequate regimen and ensure that patients adhere to therapy. Directly observed therapy (DOT) should be used whenever possible. If DOT is not available, combination preparations that include INH and rifampin, or INH, rifampin, and pyrazinamide, should be used.

Persons with active untreated pulmonary TB are infectious, particularly those for whom AFB are identified in a sputum specimen. Treatment of TB rapidly renders these patients noninfectious. According to the CDC, patients are not considered infectious if they are on adequate therapy for 2 or more weeks, have a favorable clinical response to therapy, and have three consecutive negative sputum smear results from sputum collected on different days.

Special Circumstances

Human Immunodeficiency Virus Co-infection

Despite being immunocompromised, HIV-infected patients with TB respond well to regimens containing INH and rifampin. Thus, the current recommendations are to begin the same antituberculosis regimens as used in HIV-seronegative cases. A recent study from San Francisco, however, noted that the relapse rate among HIV-infected patients was 9.3 per 100 person-years versus 1.0 in HIV-uninfected patients or those with unknown serostatus. In addition, HIV-infected patients treated with a 6-month regimen were four times as likely to relapse as those treated longer. Studies have demonstrated that intermittent therapy is also associated with a higher rate of relapse and acquired rifampin resistance. Therefore, HIV-infected patients should not be treated with highly intermittent treatment regimens, particularly if they have advanced HIV disease.

The treatment of TB in HIV-infected patients is more complicated because of the potential for drug interactions between the rifamycins and antiretroviral agents, such as the protease inhibitors and non-nucleoside reverse transcriptase inhibitors (NNRTIs), the risk of an immune reconstitution syndrome, and the propensity to develop acquired drug resistance. The rifamycins are inducers of the cytochrome P-450 pathway (rifampin > rifapentine > rifabutin) and thus can increase the metabolism of some antiretroviral drugs. Some combinations of these drugs are contraindicated, and for other combinations, dosages must be adjusted. Therefore, consultation with an expert in the field is necessary to determine the best treatment regimens for HIV-infected patients with TB.

Antiretroviral treatment should be started for all HIV-infected patients with TB, regardless of CD4⁺ count. For patients who are antiretroviral-naive, the timing of antiretroviral initiation has been controversial because of the concern for interactions with TB medications and immune reconstitution syndrome. A recent trial in South Africa demonstrated that “integrated” therapy (in which antiretrovirals are started during TB therapy) was associated with a 56% reduction in mortality relative to “sequential” therapy (in which antiretrovirals are started at the completion of TB therapy). A trial in Cambodia compared initiation of antiretrovirals 2 weeks or 8 weeks after starting TB therapy in patients with CD4⁺ count less than 200 cells/µL. Earlier initiation of antiretrovirals was associated with a 34% reduction in mortality.

Extrapulmonary Tuberculosis

In general, treatment of extrapulmonary TB follows the same principles as those for pulmonary disease. However, both surgery and the use of corticosteroids may be needed more often in extrapulmonary TB. Corticosteroids should be considered in patients with confirmed CNS or pericardial TB. Corticosteroids have been demonstrated to improve the outcomes of tuberculous pericarditis in both acute and later phase disease. In neither setting, however, is there a significant decrease in progression to constriction or need for pericardiectomy. In CNS TB, use of corticosteroids has been shown to decrease the frequency of neurologic sequelae in children. Treatment duration should be prolonged in patients with bone and joint disease to 6 to 9 months and for CNS disease to 9 to 12 months.

Children

Children should be treated with the same regimens as those recommended for adults, but doses should be adjusted appropriately. Although some experts do not recommend the use of ethambutol in children, it seems to be safe and should be used whenever underlying drug resistance is suspected. In addition, some experts recommend increasing the duration of therapy to 9 to 12 months in children with disseminated or meningeal disease. HIV-infected children should receive at least 9 months of therapy.

Pregnancy and Breastfeeding

Pregnant and breastfeeding women with active TB must be treated, because the potential for untreated TB in the fetus or infant is always of greater concern than any small risks associated with therapy. Pyrazinamide currently is not recommended in the United States in pregnant women because of a lack of data regarding teratogenicity, but the drug is recommended by the WHO and the International Union Against Tuberculosis and Lung Disease (IUATLD).

Drug-Resistant Tuberculosis

New guidelines for the programmatic management of drug-resistant TB were recently published by the WHO. The recommendations were formulated on the basis of the results of several large systematic reviews commissioned by the WHO. In patients with MDR TB, four second-line drugs that are likely to be effective, as well as pyrazinamide, should be used in the intensive phase of therapy. Regimens should include pyrazinamide, a later-generation fluoroquinolone, an injectable agent, ethionamide (or prothionamide), and either cycloserine or PAS if cycloserine cannot be used. No second-line parenteral agent was found to be superior to another. Because of its low cost, kanamycin was favored, although amikacin could be used in place of kanamycin. Streptomycin is not recommended. An intensive phase (injectable phase) of at least 8 months’ duration is recommended, with a total duration of at least 20 months. Cure rates for treatment of MDR TB have averaged around 60% to 70% (range of 30% to 95%) in programmatic settings. XDR TB is associated with a high mortality rate and cure rates ranging from 30% to 60%. In view of the complexity of therapy and possible need for surgical resection, experts in the management of drug-resistant TB should be consulted early in the course of therapy.

Clinical Course and Prevention

Monitoring for Adverse Reactions and Response to Therapy

All patients receiving antituberculosis therapy must be educated about possible drug-related adverse reactions. The patients should be warned about insignificant side effects, such as the orange discoloration of urine from rifampin, as well as the symptoms of potentially serious side effects. Baseline measurements of hepatic enzymes, bilirubin, serum creatinine, and blood urea nitrogen, as well as a complete blood cell count including platelets, are obtained before initiation of drug therapy. A serum uric acid level is obtained if pyrazinamide is included in the drug regimen. Visual acuity and red-green color discrimination are monitored in patients receiving ethambutol. Although routine laboratory monitoring for drug toxicity may not be necessary, many centers repeat liver function tests after 1 month of therapy and thereafter if symptoms develop or the liver function tests are elevated significantly.

Sputum examinations at monthly intervals are important to monitor response to therapy. Smears and cultures should be negative after 2 to 3 months of therapy. If the sputum smear remains positive after 3 months, the patient must be reevaluated; special attention is given to monitoring adherence and ruling out acquired drug resistance. Drug susceptibility tests are repeated, and the appropriateness of the drug regimen is reassessed.

Prevention

The best method of preventing TB is to identify active cases and treat them to cure, thus preventing transmission to others. Unfortunately, most transmission has occurred before diagnosis and initiation of therapy, so other methods are needed to prevent the development of TB. Two methods are used to do this. The first is to vaccinate patients with the only available TB vaccine, bacillus Calmette-Guérin (BCG). Although BCG is the most widely used vaccine in the world, serious shortcomings with this approach are recognized (as discussed later on). The second is to diagnose LTBI and treat affected patients with antituberculosis drugs to prevent progression to active disease.

Diagnosis of Latent Tuberculosis Infection

Most people who are infected with M. tuberculosis are able to arrest the development of active disease with adequate cell-mediated immunity. As noted previously, however, approximately 10% of infected people have TB develop during their lifetime. Treatment of LTBI is a pillar of TB prevention in lower incidence countries like the United States because treatment can reduce the risk of progression to TB by up to 92%. Testing for LTBI should be targeted at people with clinical and epidemiologic risks for TB (Figure 31-11). Until recently, the only method to detect LTBI was the TST. Recently, blood-based assays have been developed that provide an alternative to the TST.

Figure 31-11 Flow diagram for screening for latent tuberculosis infection (LTBI). In persons with clinical or epidemiologic risk factors for tuberculosis (TB), a tuberculin skin test (TST) or interferon-γ release assay (IGRA) should be performed. A positive test result warrants clinical evaluation and chest radiograph. If either yields abnormal findings, the patient should be evaluated for the possibility of active TB. If the chest radiograph is normal in appearance, the patient is a candidate for LTBI treatment. If the initial TST reaction or IGRA result is negative but the patient has had contact with infectious cases of TB, treatment may still be indicated.

Tuberculin Skin Test

The traditional test to diagnose LTBI for more than 100 years has been the TST. The reaction to intradermally injected tuberculin is a classic example of a delayed-type hypersensitivity reaction, characterized by a peak reaction at 48 to 72 hours marked by induration and, rarely, vesiculation and necrosis. The standard tuberculin test consists of intradermal administration of 0.1 mL (5 tuberculin units) of PPD, usually given in the volar surface of the forearm. Sensitization is induced by infection with M. tuberculosis or other cross-reacting mycobacteria antigens. The TST reaction should be read by trained readers 48 to 72 hours after injection. The basis of the reading is the degree of induration present, not erythema.

Over time, delayed-type hypersensitivity resulting from mycobacterial infection may wane in some people, resulting in a nonreactive TST despite the fact that they are truly infected. The stimulus of this initial negative TST reaction in these persons may “boost” or increase the size of the reaction to a second test administered later, resulting in a positive TST reaction and incorrectly suggesting tuberculin conversion. This ability of the TST to recall the waned reactivity is known as the booster phenomenon.

Because of the difficulty in distinguishing boosting (indicating previous infection many years ago) from tuberculin conversion (indicating recent infection), two-step testing is recommended in persons who require annual TSTs and persons older than 55 years of age.

TST is not entirely specific for the diagnosis of infection with M. tuberculosis and false-positive tests can occur from either infection with NTM or BCG vaccination. In a patient who has received BCG, the likelihood that a positive TST reaction represents a false-positive result due to BCG rather than LTBI depends on the age at vaccination. In 24 studies that involved almost 250,000 subjects vaccinated in infancy with BCG, only 1% had a positive TST attributable after 10 years. BCG vaccination after infancy had a more significant and lasting impact on the TST reaction. In practice, most clinicians ignore previous BCG vaccination when interpreting the results of TSTs if at least several years have elapsed since the time of vaccination.

Three different thresholds—5, 10, and 15 mm—have been set for defining a positive tuberculin reaction, depending on the individual or population being tested (see Table 31-1). For persons at highest risk for development of TB, a cutoff of 5 mm or more is recommended. This group includes persons known or suspected of being HIV-infected, close contacts of active TB cases, persons with an abnormal-appearing chest radiograph showing fibrosis consistent with previous TB (Figure 31-12), and other immunosuppressed patients. A cutoff of 10 mm of induration or greater is classified as positive for people at intermediate risk for TB. The remaining group consists of persons at low risk for TB who have no risk factors. In these persons, TST reactions are classified as positive if the induration is 15 mm or more across; in general, such patients should not be screened.

Figure 31-12 Chest radiograph obtained in a patient with previous tuberculosis. Changes include right upper lobe linear and nodular opacities with pleural thickening. In addition, note the elevation of the right hemidiaphragm and superior retraction of the right hilum secondary to volume loss.

Recent TST converters also are at high risk for development of TB and have therefore been identified as a high-priority group for treatment of LTBI. In the United States, conversion is defined as an increase in induration of at least 10 mm within a 2-year period.

Interferon-γ Release Assays

As noted previously, there are three new T cell–based tests for the diagnosis of LTBI: QFT-TB Gold, QFT-GIT, and T-Spot.TB test. The QFT-TB Gold assays use an ELISA method to measure antigen-specific production of IFN-γ, whereas the T-Spot.TB test uses ELISPOT to measure the number of cells that produce IFN-γ. The CDC has recommended that these assays be used in any situation in which the TST is used, whereas the U.K. National Institute for Clinical Excellence has suggested that the IGRAs be used as confirmatory tests to the TST.

Numerous studies have assessed the test characteristics of these assays by use of different versions of the tests, in different populations, and under different laboratory conditions. Overall, the specificity rates for QFT-GIT, T-Spot.TB, and TST are approximately 99.8%, 97.8%, and 88.7%, respectively. Of note, however, a “gold standard” test for LTBI is lacking, so the true test characteristics of these assays are unknown. In studies assessing the correlation with degree of exposure, both IGRAs correlated with exposure better than the TST, and neither was affected by previous BCG vaccination.

The advantages of IGRAs compared with the standard TST are several: These blood-based tests have less cross-reactivity from vaccination with BCG and NTM, are less susceptible to reader variability that occurs with TST interpretation, require only one patient visit to obtain results, and may be more specific for identifying M. tuberculosis infection. Additionally, IGRAs may be more predictive of the development of active disease than the TST. The rate of progression to active disease among patients who tested positive for LTBI and who refused preventive therapy has ranged from 2.3% to 3.3% for TST, 2.8% to 14.3% for QFT-GIT, and 3.3% to 10.0% for T-Spot.TB.

Treatment of Latent Tuberculosis Infection

Treatment is not recommended for all persons with LTBI. Instead, therapy should be provided to those persons at higher risk for TB infection and/or TB. For persons who are at increased risk of progressing to disease, treatment of latent infection is indicated, regardless of age. It is critical that patients being considered for LTBI therapy receive a clinical and radiographic evaluation to exclude the possibility of active disease. The two most commonly used drugs for the treatment of LTBI are INH and rifampin (Table 31-5).

Table 31-5 Recommended Drug Regimens for Treatment of Latent Tuberculosis Infection in Adults

INH was evaluated in randomized controlled trials conducted by the United States Public Health Service that included more than 70,000 participants encompassing a variety of populations. In these studies, the effectiveness of the drug compared with placebo in reducing the incidence of active TB averaged approximately 60%, with a range of 25% to 92%, the higher values being associated with better adherence to the drug. On the basis of these studies, the American Thoracic Society (ATS) and the CDC recommend that INH be administered as a single daily dose or twice weekly. Completion of treatment is based on the total number of doses administered and not on duration of therapy alone.

Hepatitis is the most important adverse reaction related to INH. Although liver enzyme abnormalities are relatively common in persons taking INH, symptomatic hepatitis is uncommon. The rate of INH-related hepatitis is estimated to be 1 per 1000 persons and increases with age. The most important cofactor for the development of INH hepatitis is alcohol consumption, so patients should be advised not to drink alcohol when they are taking INH. In addition, all persons taking INH should be educated about the signs and symptoms of hepatitis, including nausea, vomiting, extreme fatigue, abdominal pain, dark urine, and jaundice, so that they can be evaluated before the hepatitis becomes severe.

The other potential side effect of INH is peripheral neuropathy that is caused by interference with the metabolism of pyridoxine. In persons predisposed to neuropathy (such as patients with diabetes, uremia, malnutrition, and HIV infection), in pregnant women, and persons with seizure disorders, pyridoxine (at a dose of 25 or 50 mg/day) should be given concurrently with INH.

Rifampin alone for 4 months is an alternative option for treatment of LTBI although only one large randomized trial has evaluated the efficacy of rifampin monotherapy for LTBI. This trial, conducted among patients with silicosis at extremely high risk for LTBI, found that 3 months of daily rifampin was as efficacious as 6 months of daily INH. Nonetheless, approximately 10% of the patients in each treatment group progressed to active TB. Subsequent nonefficacy trials indicate that 4 months of daily rifampin monotherapy is associated with fewer serious adverse effects and better rates of treatment completion than 9 months of daily INH. Rifampin alone has a very low rate of hepatotoxicity. Rifampin should be used for patients who are intolerant of INH or who are presumed to have infection with INH-resistant strains of M. tuberculosis.

Weekly rifapentine plus INH for 3 months is a third option for treatment of LTBI in low-incidence settings. In 2011, the CDC released results of the PREVENT TB trial, a 10-year multinational study in which more than 8000 patients with LTBI at high risk for progression to active disease were randomized to receive once-weekly, directly observed rifapentine plus INH for 3 months or standard daily self-administered INH for 9 months. The once-weekly rifapentine plus INH group had equivalent efficacy (7 TB cases versus 15 in the INH-only group), was safe and tolerable, and was associated with better adherence (82% completion versus 69% completion in the INH group). A critical point to emphasize is that the PREVENT-TB trial evaluated directly observed rifapentine plus INH, and the effectiveness of this regimen in practice if self-administered is unknown.

Clinical Monitoring

Baseline laboratory testing of liver enzymes is not routinely indicated but is recommended in the following patient groups: HIV-infected persons, pregnant women and those within 3 months post partum, patients with a history of liver disease, and persons who use alcohol regularly. Follow-up laboratory testing of liver enzymes is indicated only if baseline liver enzyme tests are abnormal or when symptoms of hepatitis occur. Drugs should be withheld if a patient’s serum transaminase level is greater than three times normal if associated with symptoms and five times normal if asymptomatic.

Special Circumstances

Co-infection with Human Immunodeficiency Virus

All HIV-infected persons should be screened for TB and LTBI. HIV-infected persons with LTBI should receive 9 months of INH once active TB has been ruled out. Rifampin should be used with caution in HIV-infected persons taking PIs or NNRTIs because of drug interactions described previously. In HIV-infected persons taking these antiretroviral drugs who have LTBI, rifabutin (often at an adjusted dose) can be substituted for rifampin in some circumstances.

Abnormal-Appearing Chest Radiographs

In patients with evidence of LTBI who have an abnormal-appearing chest radiograph with parenchymal fibrotic lesions (see Figure 31-12) who have not been previously treated, sputum should be collected to exclude active TB. Once active TB has been excluded, treatment options for LTBI include INH for 9 months and rifampin for 4 months.

Pregnancy and Lactating Women

Although INH can be given safely during pregnancy, most clinicians wait until after delivery to begin treatment for LTBI, unless the woman has HIV infection or has been in known contact with an infectious case.

Infants and Children

Once infected, infants and children younger than 5 years of age are at high risk for progression of the initial infection to active TB and should receive INH for 9 months.

Contacts with Drug-Resistant Cases

Contacts of patients with INH-resistant, rifampin-susceptible TB should be treated with rifampin for 4 months. The treatment of persons in recent contact with an MDR TB case is challenging, however, and needs to be individualized on the basis of the susceptibility pattern of the source patient’s organism, the probability that infection has occurred, and risk factors for progression to active TB. Treatment of contacts of patients with MDR TB often entails the administration of two drugs (to which the source case’s isolate is susceptible) for 9 to 12 months.

Vaccination with Bacillus Calmette-Guérin

BCG is a live attenuated vaccine derived from a strain of M. bovis. BCG is used to vaccinate children and, in some cases, adolescents and adults throughout much of the world. Although the vaccine seems to be protective in children, it confers little protection in adults, who are the ones who usually transmit disease. Data suggest that BCG vaccination decreases the risk of disseminated disease in young children—the primary reason for its use today. In countries in which BCG vaccination was suspended, TB rates did not increase significantly.

As noted previously, BCG vaccination can influence the results of tuberculin skin testing because of cross-reaction with PPD. The effects of BCG vaccination on tuberculin reactivity vary, depending on the specific vaccine used, age at vaccination, and interval between vaccination and skin testing. Despite widespread use of BCG vaccination, it has been difficult to demonstrate significant impact on TB control globally. New vaccines are urgently needed.

Diseases Caused by Nontuberculous Mycobacteria

NTM comprise more than 135 different species that are widely distributed throughout the environment. The list of mycobacterial species has grown dramatically over recent years because of the availability of DNA sequencing, and because organisms that differ in sequence by 1% or more from all known species are defined as new species.

The NTM were not widely recognized as a cause of human disease until the late 1950s. In 1959, Runyon reported a classification system that organized NTM organisms into four groups on the basis of microbiologic characteristics, including the formation of pigment and the speed of growth. This system has become less useful with the development of rapid molecular methods of diagnosis, but differentiating NTM species on the basis of their rate of growth is still used today. NTM typically are divided into rapidly and slowly growing species (Table 31-6).

Table 31-6 Clinically Significant Nontuberculous Examples of Mycobacteria by Rate of Growth

Slowly Growing (>7 days of incubation for mature growth)	Rapidly Growing (≤7 days of incubation for mature growth)
Mycobacterium avium Mycobacterium intracellulare Mycobacterium kansasii Mycobacterium xenopi Mycobacterium simiae Mycobacterium szulgai Mycobacterium scrofulaceum Mycobacterium malmoense Mycobacterium haemophilum Mycobacterium genavense Mycobacterium marinum*	Mycobacterium fortuitum Mycobacterium peregrinum Mycobacterium abscessus Mycobacterium massiliense Mycobacterium bolletii Mycobacterium chelonae Mycobacterium mucogenicum

* This organism often grows within 7 to 10 days.

Epidemiology, Risk Factors, and Pathogenesis

Epidemiology

NTM have been isolated from soil and water, including both natural and treated water sources, from throughout the world. Because these organisms do not seem to be transmitted from human to human, the source of human infection is thought to be through environmental exposures. Skin test surveys of U.S. Navy recruits in the 1960s demonstrated higher rates of reactivity in those from the southeastern United States than from the northern states, suggesting higher rates of infection. Studies evaluating antibody to lipoarabinomannin (LAM) have demonstrated anti-LAM antibodies beginning early in life and rapidly rising through age 12.

Epidemiologic data on NTM infections are largely lacking, because unlike TB, NTM infections are not routinely considered reportable to public health authorities. The median rates of pulmonary NTM isolation have been estimated to be approximately 6.2 per 100,000 population in North America, 8.3 per 100,000 in Europe, 15 per 100,000 in Asia, and 7.2 per 100,000 in Australia. Studies from Oregon have reported that the prevalence of NTM infections was higher in females than males (6.4 per 100,000 versus 4.7 per 100,000) and was highest in persons aged more than 50 years (15.5 per 100,000). Among four integrated health care systems in the United States, the annualized prevalence ranged from 1.4 to 6.6 per 100,000 and among persons aged 60 years or more, the annualized prevalence was 26.7 per 100,000.

Recent data support the view that the incidence of NTM infections is increasing. Studies from Canada, Australia, Taiwan, the Netherlands, and the United States have reported increases in the incidence or prevalence of NTM. In a retrospective cohort review from 1997 to 2003 in Ontario, Canada, 222,247 pulmonary isolates from 10,231 patients were identified. The prevalence was 9.1 per 100,000 in 1997 and increased to 14.1 per 100,000 by 2003 (P < .0001), with an average annual increase of 8.4%. Increases were noted among Mycobacterium avium complex (MAC), Mycobacterium xenopi, rapidly growing mycobacteria, and Mycobacterium kansasii. Of note, the rate of TB declined 4.0% over the study period. Of 200 patients who were evaluated in more detail, 33% fulfilled the clinical, radiologic, and bacteriologic ATS criteria. In the United States, pulmonary hospitalizations for NTM infections increased significantly among both males and females between 1998 and 2005.

The most common pathogenic NTM vary geographically. MAC is the most common cause of NTM-related pulmonary disease in almost all studies. The next most common NTM species depends on the origin of reported data. For example, in the United States, M. kansasii is the second most common cause of pulmonary disease, followed by Mycobacterium abscessus. In Canada, and some parts of Europe, M. xenopi is the second most common, whereas in northern Europe and Scandinavia, Mycobacterium malmoense is second.

Risk Factors

Skin test reactivity to mycobacterial antigens has ranged in prevalence from 11% to 33% in the United States. Predictors of skin test reactivity have included residence in the southeastern region of the country, male gender, black race, birth outside of the United States, and degree of soil exposure. Of note, however, many of the patients who develop pulmonary disease related to NTM do not have these characteristics. For example, patients with the nodular-bronchiectatic form of pulmonary disease (see “Pulmonary Disease” further on) frequently are white, postmenopausal women who were born in the United States.

Patients who have pulmonary infections caused by NTM often have structural lung disease such as chronic obstructive pulmonary disease, bronchiectasis, cystic fibrosis, pneumoconiosis, previous TB, alveolar proteinosis, and chronic aspiration. A study from France noted that more than 50% of the patients who met ATS disease criteria (except for those with M. kansasii) had underlying predisposing factors such as preexisting pulmonary or immune deficiency. HIV-infected patients and those with abnormalities in the IFN-γ or interleukin-12 (IL-12) pathways are predisposed to severe NTM infections.

Pathogenesis

Although little is known about the pathogenesis of NTM infections, several observations have provided some insight into this disease. First, in HIV-infected patients, disseminated NTM infections typically occur only after the CD4⁺ lymphocyte count falls below 50 cells/µL, suggesting that specific T cell products or activities are required for mycobacterial resistance. Second, in HIV-uninfected patients, certain genetic syndromes have been identified that are associated with disseminated NTM infections. These syndromes have been traced to mutations in IFN-γ and IL-12 synthesis and response pathways. Third, a striking association has been noted among bronchiectasis, nodular pulmonary NTM infections, and a particular body habitus, predominantly in postmenopausal women. In the last instance, it remains to be seen whether or not these women have some sort of subtle immune deficiency that predisposes them to NTM pulmonary infections, or whether their predisposition is related to ineffective mucociliary clearance or poor tracheobronchial secretion drainage. Investigators have hypothesized that decreased leptin, increased adiponectin, and/or decreased estrogen may account for the increased susceptibility in these latter patients. In addition, in those with the distinctive morphotype, abnormalities in fibrillin have been hypothesized to further increase the susceptibility to NTM infections through expression of the immunosuppressive cytokine tumor growth factor-β.

NTM do not seem to live in a state of dormancy like that for M. tuberculosis. Moreover, unlike with TB, simply isolating an NTM from a respiratory specimen does not mean that the patient has NTM-related disease. For years the term colonization has been used to describe the infective status of patients who have a single or small number of positive cultures over time and in whom progressive disease cannot be demonstrated. With longer follow-up, however, many of these patients do, in fact, demonstrate clinical and/or radiographic progression of disease, so it may be more appropriate to think of these patients as having indolent infection.

Genetics

The genetics of NTM infection are not well understood. Studies from Japan and South Korea have not found an association of such infections with mutations in genes encoding IFN-γ, IL-12, or HLA. However, associations with NTM pulmonary disease and mutations in the major histocompatibility complex class I chain–related A (MICA) gene and natural resistance-associated macrophage protein 1 gene (Nramp1) have been reported. The role of Toll-like receptor-2 remains controversial.

Clinical Features

The most common clinical manifestation of NTM infection is chronic lung disease. However, lymphatic, skin or soft tissue, and bone or joint involvement, as well as disseminated disease, also is important. The propensity for a specific manifestation varies with the NTM species and certain host factors. For example, HIV-infected patients typically are seen with disseminated disease caused by Mycobacterium avium, whereas elderly white women often have pulmonary disease caused by Mycobacterium intracellulare.

Pulmonary Disease

Chronic pulmonary disease is the most common clinical presentation of NTM disease, and patients usually present with chronic cough, fatigue, malaise, dyspnea, fever, hemoptysis, chest pain, and weight loss. Patients should be evaluated for possible gastroesophageal disorders that lead to aspiration, lipoid pneumonia, cystic fibrosis, and α₁-antitrypsin anomalies. Physical examination may identify certain morphologic characteristics in postmenopausal women, and occasionally men, that include thin body habitus, scoliosis, pectus excavatum, and mitral valve prolapse.

NTM pulmonary infections manifest with two prototypical radiographic patterns: fibrocavitary disease, consisting of upper lobe opacities with cavities and volume loss, and nodular bronchiectatic disease, consisting of nodules and bronchiectasis. Classically, fibrocavitary disease was recognized in older men with underlying lung disease (Figure 31-13). This radiographic pattern mimics that of TB, and NTM infection frequently is identified in the course of evaluation for TB. By contrast, nodular bronchiectatic disease typically occurs among women without known structural lung disease. Chest radiographs reveal opacities in the middle and lower lung fields. High-resolution CT scans demonstrate bronchiectasis often in the middle lobe and lingula, with evidence of small noncavitating nodules, centrilobular in location (Figure 31-14). It is important to recognize that there is a great deal of overlap between the two classic radiographic presentations and between the patterns of disease produced by the various NTM species.

Figure 31-13 A, Chest radiograph obtained in a patient with upper lobe emphysema and severe lung disease caused by Mycobacterium avium complex. B, Chest computed tomography from the same patient demonstrates severe emphysema with cavitation posteriorly (arrows) in the right lower lobe.

Figure 31-14 A, Chest radiograph showing right middle lobe and lingular bronchiectasis in a patient with Mycobacterium avium complex and Mycobacterium simiae infection. B, Chest computed tomography from the same patient demonstrating bronchiectasis in the right middle lobe and lingula. Note also the nodular opacities and tree-in-bud opacities (arrows).

Patients with M. kansasii infection usually present with upper lobe cavitary opacities, and the cavities typically are thin-walled (Figure 31-15). The chest radiograph in patients with rapidly growing mycobacterial infections usually shows multilobar, patchy, reticulonodular or mixed interstitial alveolar opacities, with an upper lobe predominance (Figure 31-16). Cavitation is reported to occur in 15% to 40%. High-resolution CT scans will show bronchiectasis and small nodules, similar to those in MAC infection.

Figure 31-15 A, Chest radiograph showing a thin-walled cavity (arrow) in the left upper lobe caused by infection with Mycobacterium kansasii. B, Chest computed tomography from the same patient demonstrating a thin-walled cavity in the left upper lobe.

Figure 31-16 Chest radiograph obtained in a patient with infection caused by Mycobacterium abscessus. Note the multiple cavities (arrows), diffuse nodular opacities, and kyphoscoliosis.

Lymphadenitis

In children, the most common form of NTM disease is cervical lymphadenitis. M. avium is the most common etiologic agent, accounting for 80% of culture-proven cases. Mycobacterium scrofulaceum is the second most common cause of lymphadenitis in the United States and Australia, whereas M. malmoense and Mycobacterium haemophilum are in Scandinavia, the United Kingdom, and other areas of northern Europe. M. tuberculosis is isolated in only 10% of cases of culture-proven mycobacterial cervical lymphadenitis in the United States, but in adults, 90% are due to TB.

Infection usually involves the submandibular, submaxillary, cervical, and preauricular lymph nodes in children between 1 and 5 years of age. The disease is of insidious onset and is rarely associated with systemic symptoms. Involvement of the lymph nodes usually is unilateral, and affected tissue is nontender. The lymph nodes may enlarge and eventually rupture, producing sinus tracts just as in tuberculous lymphadenitis. Diagnosis usually is made by fine needle aspiration or surgical excision of the involved lymph nodes. Only 50% to 82% of excised nodes will be culture positive. Treatment for most cases of NTM-related cervical lymphadenitis is surgical excision.

Soft Tissue, Skin, and Bone Infections

Although virtually any NTM can cause skin, soft tissue, and bone infections, the most common species to do so are the rapid growers Mycobacterium marinum and Mycobacterium ulcerans. Rapid growers often produce infections at the site of punctures or surgery. Joint and bone infections have occurred after surgery and traumatic injuries. Tissue biopsy is the most sensitive way to diagnose these infections.

Disseminated Infections

Disseminated infections are most commonly associated with HIV infection and other forms of severe immunosuppression. More than 90% of reported cases of disseminated infection in HIV-infected patients are due to MAC, and almost all of these are due to M. avium. The next most common cause of disseminated NTM disease in HIV-infected patients is M. kansasii, but a number of other species also have been implicated. Most patients have advanced HIV disease at clinical presentation and complain of fever, night sweats, and weight loss. Abdominal pain and diarrhea also may be reported. In non–HIV-infected patients, fever of unknown origin is a common presentation. Diagnosis usually is through detection of the causative organism in the blood. In HIV-infected patients, M. avium can be isolated in more than 90% of cases.

Diagnosis

Integration of Clinical, Microbiologic, and Radiographic Data

Patients suspected of having a NTM infection should be evaluated with a chest radiograph or high-resolution CT, or both, particularly in the absence of evidence of cavitation on the radiograph. If radiographic abnormalities are consistent with an NTM infection, at least three sputum specimens should be obtained for AFB examination and mycobacterial culture. TB, as well as other disorders, should be excluded.

To diagnose NTM infection, the clinician must weigh clinical, bacteriologic, and radiographic information (Box 31-1). NTM pulmonary infections should be suspected when a patient presents with a compatible clinical picture and nodular or cavitary opacities on the chest radiograph or multifocal bronchiectasis with multiple small nodules on a high-resolution CT scan. In addition to these clinical criteria, the patient should have at least two positive cultures from separate sputum specimens or a positive culture from at least one bronchial wash or lavage procedure. Additional diagnostic criteria include transbronchial or other lung biopsy specimens with mycobacterial histopathologic features and culture-positive for NTM, or a biopsy specimen showing mycobacterial histopathologic features and one or more sputum or bronchial washings that are culture-positive. In patients who do not meet the preceding definition for disease, close follow-up is indicated, because many will demonstrate progression over time.

Box 31-1

Clinical and Microbiologic Criteria for Diagnosis of Lung Disease Due to Nontuberculous Mycobacteria

Clinical *

1. Pulmonary symptoms, nodular or cavitary opacities on the chest radiograph, or multifocal bronchiectasis with multiple small nodules on a high-resolution computed tomography scan

AND

2. Appropriate exclusion of other diagnoses

Microbiologic

1. Positive culture results from at least two separate expectorated sputum samples. If results are nondiagnostic, consider repeat sputum acid-fast bacilli (AFB) smears and cultures.

2. Positive culture result from at least one bronchial wash or lavage.

3. Transbronchial or other lung biopsy with mycobacterial histopathologic features (granulomatous inflammation or AFB) and positive culture for nontuberculous mycobacteria (NTM) or biopsy showing mycobacterial histopathologic features (granulomatous inflammation or AFB) and one or more sputum or bronchial washings that are culture-positive for NTM.

Modified from Griffith DE, Aksamit T, Brown-Elliott BA, et al; ATS Mycobacterial Diseases Subcommittee, American Thoracic Society; Infectious Diseases Society of America: An official ATS/IDSA statement: diagnosis, treatment, and prevention of nontuberculous mycobacterial diseases, Am J Respir Crit Care Med 175:367–416, 2007.

^* Presence of both clinical criteria is required.

Laboratory Diagnosis

The same methods that are used to stain and grow M. tuberculosis are used for NTM. Both solid and liquid culture media support growth of the NTM, and both are recommended for use in the clinical laboratory. Although cultures in broth media have a higher yield and provide a more rapid result, they do not allow for observation of colony morphology, growth rates, and recognition of mixed cultures, as do solid media. For most mycobacteria, the optimum temperature for growth is 28° to 37° C, and most clinically significant NTM organisms grow at 35° to 37° C. However, some NTM species—M. marinum, Mycobacterium chelonae, M. ulcerans, and M. haemophilum—require lower temperatures for optimum growth. Others, such as M. xenopi, grow best at higher temperatures. Some species require special media, such as iron or heme for M. haemophilum and mycobactin J for Mycobacterium genavense. Most NTM organisms grow within 2 to 3 weeks on subculture, but it may take up to 8 to 12 weeks to grow M. ulcerans and M. genavense. Rapidly growing mycobacteria usually grow within 7 days on subculture.

Species identification can be performed biochemically or, more commonly, with high-performance liquid chromatography (HPLC), genetic probes, and/or 16S ribosomal DNA sequencing. Genetic probes are commercially available only for M. tuberculosis, M. kansasii, M. avium, M. intracellulare, and Mycobacterium gordonae. These probes have a sensitivity of between 85% and 100%, with a specificity of 100%. HPLC is a practical and rapid way to detect differences in mycolic acid content between NTM species, although this method cannot differentiate some species of NTM. DNA sequence analysis is able to differentiate strains on the basis of two hypervariable sequences.

An alternative approach to the clinical problems of rapidly distinguishing pulmonary MAC infection from TB and distinguishing MAC disease from MAC colonization has been proposed by Japanese investigators who have developed an enzyme immunoassay (EIA) for a MAC-specific glycopeptidolipid. This assay identified MAC pulmonary disease with 84% sensitivity and 100% sensitivity. Although promising, this assay is not widely available and awaits validation in additional clinical cohorts.

Drug Susceptibility Testing

The role of in vitro susceptibility in management of patients with NTM disease is controversial because, with the few specific exceptions mentioned further on, in vitro susceptibility results are not demonstrated to correlate with clinical response to therapy. This is in contrast with M. tuberculosis, for which susceptibility results have clear clinical implications.

Macrolide resistance in MAC infection is an important instance in which susceptibility results have prognostic and therapeutic implications. Development of macrolide resistance is associated with azithromycin or clarithromycin monotherapy and portends a poor prognosis. Macrolide resistance is an indication for aggressive therapy including the use of an aminoglycoside and surgical resection if feasible. Current ATS guidelines recommend clarithromycin testing for new, previously untreated MAC isolates, those who fail macrolide-based treatment regimens, or prophylaxis regimens.

Rifampin resistance among previously untreated M. kansasii isolates represents a second instance in which in vitro susceptibility informs management, because rifampin resistance has been correlated with poor outcomes. Isolates resistant to rifampin also should be tested against rifabutin, ethambutol, isoniazid, clarithromycin, fluoroquinolones, amikacin, and sulfonamides.

The role of testing for macrolide susceptibility with M. abscessus is unclear. M. abscessus isolates routinely appear to be macrolide-susceptible on in vitro testing but patients with M. abscessus infections are noted to have poor clinical responses to macrolides. This discrepancy may be explained with discovery of the M. abscessus erm(41) gene, which confers inducible macrolide resistance. A subset of isolates classified as M. abscessus on routine testing are shown with molecular methods to be Mycobacterium massiliense. M. massiliense has a deletion in the erm(41) gene that makes it nonfunctional, and a better clinical response to macrolide-based therapy was demonstrated in patients infected with M. massiliense than in those infected with M. abscessus.

In summary, because NTM susceptibility results do not always correlate with clinical response and interpretation differs for different organisms, results should be interpreted with caution and, if possible, in consultation with clinicians expert in the care of NTM infections.

Treatment

When the patient’s culture specimen grows M. tuberculosis, treatment is always indicated, assuming that the isolate was not due to laboratory cross-contamination. With NTM infections, however, isolation should not always lead to treatment. The decision to treat is based on the potential risks and benefits for the individual patient. Thus, management of patients with NTM infections is complicated and requires a great deal of individualization of therapy. In addition, because in vitro susceptibility results for many NTM species do not correlate well with clinical response to antimicrobial therapy, clinicians should use such data with a clear understanding of the limitations.

Pulmonary Infections

Mycobacterium avium Complex

Before the availability of the newer macrolides, the long-term cure rate for patients treated with antituberculosis regimens was approximately 50%. Small, noncomparative studies of azithromycin- and clarithromycin-containing regimens suggest higher bacteriologic response rates, but long-term follow-up data often are lacking. The ATS currently recommends that the treatment regimen be based on the presence or absence of cavitary disease and whether or not the patient has been treated previously (Table 31-7). For patients with noncavitary nodular bronchiectasis disease, a three-times-a-week regimen may be considered. For patients with cavitary or advanced disease and in those who have been treated previously, daily therapy is recommended. An aminoglycoside should be considered in patients, at least for the first 2 to 3 months of therapy, if they have cavitary disease or if previous treatment has failed. Surgery should be considered in patients in whom the infection is due to a macrolide-resistant strain of MAC, in whom treatment has failed, or in whom cavitary disease is localized and potentially resectable.

Table 31-7 Nontuberculous Mycobacteria and Recommended Therapy

Common Etiologic Agents	Recommended Antimicrobial Therapy	Other Etiologic Mycobacterial Species
M. avium complex		M. simiae M. szulgai M. celatum M. asiaticum M. haemophilum M. smegmatis M. chelonae M. fortuitum M. scrofulaceum M. shimodei

Nodular bronchiectasis Clarithromycin 1000 mg 3×/wk OR azithromycin 500-600 mg 3×/wk; rifampin 600 mg 3×/wk; ethambutol 25 mg/kg 3×/wk Cavitary disease Clarithromycin 1000 mg daily OR azithromycin 250 mg daily; rifampin 450-600 mg daily; ethambutol 15 mg/kg daily
Also consider amikacin 15 mg/kg (for the first 2-3 mo) 3×/wk Previously treated Clarithromycin 1000 mg daily OR azithromycin 250 mg daily; rifampin 450-600 mg daily OR rifabutin 150-300 mg daily; amikacin 15 mg/kg (for the first 2-3 months) 3×/wk M. kansasii Isoniazid 300 mg/day; rifampin 600 mg/day; ethambutol 15 mg/kg/day
Can add aminoglycoside in severe disease M. abscessus Clarithromycin 1000 mg/daily OR azithromycin 250 mg/daily; cefoxitin max. 12 g/d divided in 3-4 doses/day OR imipenem 500-1000 mg q8-12 h PLUS amikacin 15 mg/kg 3×/wk M. xenopi Clarithromycin 1000 mg daily OR azithromycin 250 mg daily; rifampin 450-600 mg daily; ethambutol 15 mg/kg daily
Also consider isoniazid and/or amikacin 15 mg/kg (for the first 2-3 mo) 3×/wk M. malmoense Clarithromycin 1000 mg daily OR azithromycin 250 mg daily; rifampin 450-600 mg daily; ethambutol 15 mg/kg daily

Mycobacterium kansasii

Patients with lung disease caused by M. kansasii should be treated with INH, rifampin, and ethambutol (see Table 31-7). Substitution of clarithromycin for INH has been associated with good short-term outcomes, and in one study, no relapses were seen at follow-up evaluation after 46 months. As with other NTM infections, the treatment duration should encompass 12 months of negative sputum cultures. For patients whose isolate is resistant to the rifamycins, a three-drug regimen is recommended on the basis of data for in vitro susceptibility to clarithromycin or azithromycin, moxifloxacin, ethambutol, sulfamethoxazole, or streptomycin. Surgical resection is almost never necessary in patients infected with M. kansasii.