Tuberculosis (TB), which is caused by bacteria of the Mycobacterium tuberculosis complex, is one of the oldest diseases known to affect humans and a major cause of death worldwide. Recent population genomic studies suggest that M. tuberculosis may have emerged ~70,000 years ago in Africa and subsequently disseminated along with anatomically modern humans, expanding globally during the Neolithic Age as human density started to increase. Progenitors of M. tuberculosis are likely to have affected prehominids. This disease most often affects the lungs, although other organs are involved in up to one-third of cases. If properly treated, TB caused by drug-susceptible strains is curable in the vast majority of cases. If untreated, the disease may be fatal within 5 years in 50–65% of cases. Transmission usually takes place through the airborne spread of droplet nuclei produced by patients with infectious pulmonary TB.
Mycobacteria belong to the family Mycobacteriaceae and the order Actinomycetales. Of the pathogenic species belonging to the M. tuberculosis complex, which comprises eight distinct subgroups, the most common and important agent of human disease is M. tuberculosis. The complex includes M. bovis (the bovine tubercle bacillus—characteristically resistant to pyrazinamide, once an important cause of TB transmitted by unpasteurized milk, and currently the cause of a small percentage of human cases worldwide), M. caprae (related to M. bovis), M. africanum (isolated from cases in West, Central, and East Africa), M. microti (the “vole” bacillus, a less virulent and rarely encountered organism), M. pinnipedii (a bacillus infecting seals and sea lions in the Southern Hemisphere and recently isolated from humans), M. mungi (isolated from banded mongooses in southern Africa), M. orygis (described recently in oryxes and other Bovidae in Africa and Asia and a potential cause of infection in humans), and M. canetti (a rare isolate from East African cases that produces unusual smooth colonies on solid media and is considered closely related to a supposed progenitor type).
M. tuberculosis is a rod-shaped, non-spore-forming, thin aerobic bacterium measuring 0.5 μm by 3 μm. Mycobacteria, including M. tuberculosis, are often neutral on Gram’s staining. However, once stained, the bacilli cannot be decolorized by acid alcohol; this characteristic justifies their classification as acid-fast bacilli (AFB; Fig. 202-1). Acid fastness is due mainly to the organisms’ high content of mycolic acids, long-chain cross-linked fatty acids, and other cell-wall lipids. Microorganisms other than mycobacteria that display some acid fastness include species of Nocardia and Rhodococcus, Legionella micdadei, and the protozoa Isospora and Cryptosporidium. In the mycobacterial cell wall, lipids (e.g., mycolic acids) are linked to underlying arabinogalactan and peptidoglycan. This structure results in very low permeability of the cell wall, thus reducing the effectiveness of most antibiotics. Another molecule in the mycobacterial cell wall, lipoarabinomannan, is involved in the pathogen–host interaction and facilitates the survival of M. tuberculosis within macrophages. The complete genome sequence of M. tuberculosis comprises 4043 genes encoding 3993 proteins and 50 genes encoding RNAs; its high guanine-plus-cytosine content (65.6%) is indicative of an aerobic “lifestyle.” A large proportion of genes are devoted to the production of enzymes involved in cell wall metabolism.
Acid-fast bacillus smear showing M. tuberculosis bacilli. (Courtesy of the Centers for Disease Control and Prevention, Atlanta.)
More than 5.7 million new cases of TB (all forms, both pulmonary and extrapulmonary) were reported to the World Health Organization (WHO) in 2013; 95% of cases were reported from developing countries. However, because of insufficient case detection and incomplete notification, reported cases may represent only about two-thirds of the total estimated cases. The WHO estimated that 9 million (range, 8.6–9.4 million) new cases of TB occurred worldwide in 2013, 95% of them in developing countries of Asia (5 million), Africa (2.6 million), the Middle East (0.7 million), and Latin America (0.3 million). It is further estimated that 1.49 million (range, 1.32–1.67 million) deaths from TB, including 0.36 million among people living with HIV infection, occurred in 2013, 96% of them in developing countries. Estimates of TB incidence rates (per 100,000 population) and numbers of TB-related deaths in 2013 are depicted in Figs. 202-2 and 202-3, respectively. During the late 1980s and early 1990s, numbers of reported cases of TB increased in industrialized countries. These increases were related largely to immigration from countries with a high incidence of TB; the spread of the HIV epidemic; social problems, such as increased urban poverty, homelessness, and drug abuse; and dismantling of TB services. During the past few years, numbers of reported cases have begun to decline again or have stabilized in most industrialized nations. In the United States, with the re-establishment of stronger control programs, the decline resumed in 1993 and has since been maintained. In 2013, 9582 cases of TB (3.0 cases/100,000 population) were reported to the Centers for Disease Control and Prevention (CDC).
Estimated tuberculosis (TB) incidence rates (per 100,000 population) in 2013. The designations used and the presentation of material on this map do not imply the expression of any opinion whatsoever on the part of the World Health Organization (WHO) concerning the legal status of any country, territory, city, or area or of its authorities or concerning the delimitation of its frontiers or boundaries. Dotted, dashed, and white lines represent approximate border lines for which there may not yet be full agreement. (Courtesy of the Global TB Programme, WHO; with permission.)
Estimated numbers of tuberculosis-related deaths in 2013. (See disclaimer in Fig. 202-2. Courtesy of the Global TB Programme, WHO; with permission.)
In the United States, TB is uncommon among young adults of European descent, who have only rarely been exposed to M. tuberculosis infection during recent decades. In contrast, because of a high risk of transmission in the past, the prevalence of latent M. tuberculosis infection (LTBI) is relatively high among elderly whites. In general, adults ≥65 years of age have the highest incidence rate per capita (4.9 cases/100,000 population in 2013) and children <14 years of age the lowest (0.8 case/100,000 population). Blacks account for the highest proportion of cases (37%; 1257 cases in 2013) among U.S.-born persons. TB in the United States is also a disease of adult members of the HIV-infected population, the foreign-born population (64.6% of all cases in 2013), and disadvantaged/marginalized populations. Of the 6193 cases reported among foreign-born persons in 2013, 37% occurred in persons from the Americas and 32% occurred in persons born in the Western Pacific region. Overall, the highest rates per capita were among Asian Americans (18.7 cases/100,000 population). A total of 536 deaths were caused by TB in the United States in 2011. In Canada in 2013, 1638 TB cases were reported (4.7 cases/100,000 population); 70% (1145) of these cases occurred in foreign-born persons and 19% (309 cases) occurred in members of the Canadian aboriginal peoples, whose per capita rate is disproportionately high (23.4 cases/100,000 population) with a peak in the Nunavut territory of 143 cases/100,000 population—a rate similar to that in many highly endemic countries. Similarly, in Europe, TB has reemerged as an important public health problem, mainly as a result of cases among immigrants from high-incidence countries and among marginalized populations, often in large urban settings like London; in 2013, 41% of all cases reported from the United Kingdom occurred in London, and the rate per capita (36 cases/100,000 population) was similar to that in some middle-income countries. In most Western European countries, there are more cases annually among foreign-born than native populations.
Recent data on global trends indicate that in 2013 the TB incidence was stable or falling in most regions; this trend began in the early 2000s and appears to have continued, with an average annual decline of 2% globally. This global decrease is explained largely by the simultaneous reduction in TB incidence in sub-Saharan Africa, where rates had risen steeply since the 1980s as a result of the HIV epidemic and the lack of capacity of health systems and services to deal with the problem effectively, and in Eastern Europe, where incidence increased rapidly during the 1990s because of a deterioration in socioeconomic conditions and the health care infrastructure (although, after peaking in 2001, incidence in Eastern Europe has since declined slowly).
Of the estimated 9 million new cases of TB in 2013, 13% (1.1 million) were associated with HIV infection, and 78% of these HIV-associated cases occurred in Africa. An estimated 0.36 million persons with HIV-associated TB died in 2013. Furthermore, an estimated 480,000 cases (range, 350,000–610,000) of multidrug-resistant TB (MDR-TB)—a form of the disease caused by bacilli resistant at least to isoniazid and rifampin—occurred in 2013. Only 28% of these cases were diagnosed because of a lack of culture and drug-susceptibility testing capacity in most settings worldwide. The countries of the former Soviet Union have reported the highest proportions of MDR disease among new TB cases (up to 35–40% in some regions of Russia and Belarus). Overall, 60% of all MDR-TB cases occur in China, India, the Russian Federation, Pakistan, and Ukraine. Since 2006, 100 countries, including the United States, have reported cases of extensively drug-resistant TB (XDR-TB), in which MDR-TB is compounded by additional resistance to the most powerful second-line anti-TB drugs (fluoroquinolones and at least one of the injectable drugs amikacin, kanamycin, and capreomycin). Up to 10% of the MDR-TB cases worldwide may actually be XDR-TB, but the vast majority of XDR-TB cases remain undiagnosed because reliable methods for drug susceptibility testing are lacking and laboratory capacity is limited. Lately, cases deemed resistant to all anti-TB drugs have been reported from countries such as India, Italy, and Iran; however, this information must be interpreted with caution because drug susceptibility testing for several second-line drugs is neither accurate nor reproducible.
FROM EXPOSURE TO INFECTION
M. tuberculosis is most commonly transmitted from a person with infectious pulmonary TB by droplet nuclei, which are aerosolized by coughing, sneezing, or speaking. The tiny droplets dry rapidly; the smallest (<5–10 μm in diameter) may remain suspended in the air for several hours and may reach the terminal air passages when inhaled. There may be as many as 3000 infectious nuclei per cough. Other routes of transmission of tubercle bacilli (e.g., through the skin or the placenta) are uncommon and of no epidemiologic significance. The probability of contact with a person who has an infectious form of TB, the intimacy and duration of that contact, the degree of infectiousness of the case, and the shared environment in which the contact takes place are all important determinants of the likelihood of transmission. Several studies of close-contact situations have clearly demonstrated that TB patients whose sputum contains AFB visible by microscopy (sputum smear–positive cases) are the most likely to transmit the infection. The most infectious patients have cavitary pulmonary disease or, much less commonly, laryngeal TB and produce sputum containing as many as 105–107 AFB/mL. Patients with sputum smear–negative/culture-positive TB are less infectious, although they have been responsible for up to 20% of transmission in some studies in the United States. Those with culture-negative pulmonary TB and extrapulmonary TB are essentially noninfectious. Because persons with both HIV infection and TB are less likely to have cavitations, they may be less infectious than persons without HIV co-infection. Crowding in poorly ventilated rooms is one of the most important factors in the transmission of tubercle bacilli because it increases the intensity of contact with a case.
The risk of acquiring M. tuberculosis infection is determined mainly by exogenous factors. Because of delays in seeking care and in making a diagnosis, it is generally estimated that, in high-prevalence settings, up to 20 contacts may be infected by each AFB-positive case before the index case is diagnosed.
FROM INFECTION TO DISEASE
Unlike the risk of acquiring infection with M. tuberculosis, the risk of developing disease after being infected depends largely on endogenous factors, such as the individual’s innate immunologic and nonimmunologic defenses and the level at which the individual’s cell-mediated immunity (CMI) is functioning. Clinical illness directly following infection is classified as primary TB and is common among children in the first few years of life and among immunocompromised persons. Although primary TB may be severe and disseminated, it generally is not associated with high-level transmissibility. When infection is acquired later in life, the chance is greater that the mature immune system will contain it at least temporarily. Bacilli, however, may persist for years before reactivating to produce secondary (or postprimary) TB, which, because of frequent cavitation, is more often infectious than is primary disease. Overall, it is estimated that up to 10% of infected persons will eventually develop active TB in their lifetime—half of them during the first 18 months after infection. The risk is much higher among HIV-infected persons. Reinfection of a previously infected individual, which is common in areas with high rates of TB transmission, may also favor the development of disease. At the height of the TB resurgence in the United States in the early 1990s, molecular typing and comparison of strains of M. tuberculosis suggested that up to one-third of cases of active TB in some inner-city communities were due to recent transmission rather than to reactivation of old latent infection. Age is an important determinant of the risk of disease after infection. Among infected persons, the incidence of TB is highest during late adolescence and early adulthood; the reasons are unclear. The incidence among women peaks at 25–34 years of age. In this age group, rates among women may be higher than those among men, whereas at older ages the opposite is true. The risk increases in the elderly, possibly because of waning immunity and comorbidity.
A variety of diseases and conditions favor the development of active TB (Table 202-1). In absolute terms, the most potent risk factor for TB among infected individuals is clearly HIV co-infection, which suppresses cellular immunity. The risk that LTBI will proceed to active disease is directly related to the patient’s degree of immunosuppression. In a study of HIV-infected, tuberculin skin test (TST)–positive persons, this risk varied from 2.6 to 13.3 cases/100 person-years and increased as the CD4+ T cell count decreased.
TABLE 202-1Risk Factors for Active Tuberculosis in Persons Who Have Been Infected with Tubercle Bacilli ||Download (.pdf) TABLE 202-1Risk Factors for Active Tuberculosis in Persons Who Have Been Infected with Tubercle Bacilli
|Factor ||Relative Risk/Oddsa |
|Recent infection (<1 year) ||12.9 |
|Fibrotic lesions (spontaneously healed) ||2–20 |
|Comorbidities and iatrogenic causes || |
| HIV infection ||21–>30 |
| Silicosis ||30 |
| Chronic renal failure/hemodialysis ||10–25 |
| Diabetes ||2–4 |
| IV drug use ||10–30 |
| Immunosuppressive treatment ||10 |
| Tumor necrosis factor α inhibitors ||4–5 |
| Gastrectomy ||2–5 |
| Jejunoileal bypass ||30–60 |
| Posttransplantation period (renal, cardiac) ||20–70 |
|Tobacco smoking ||2–3 |
|Malnutrition and severe underweight ||2 |
NATURAL HISTORY OF DISEASE
Studies conducted in various countries before the advent of chemotherapy showed that untreated TB is often fatal. About one-third of patients died within 1 year after diagnosis, and more than 50% died within 5 years. The 5-year mortality rate among sputum smear–positive cases was 65%. Of the survivors at 5 years, ~60% had undergone spontaneous remission, while the remainder were still excreting tubercle bacilli. With effective, timely, and proper chemotherapy, patients have a very high chance of being cured. However, improper use of anti-TB drugs, while reducing mortality rates, may also result in large numbers of chronic infectious cases, often with drug-resistant bacilli.
PATHOGENESIS AND IMMUNITY
INFECTION AND MACROPHAGE INVASION
The interaction of M. tuberculosis with the human host begins when droplet nuclei containing viable microorganisms propelled into the air by infectious patients are inhaled by a close bystander. Although the majority of inhaled bacilli are trapped in the upper airways and expelled by ciliated mucosal cells, a fraction (usually <10%) reach the alveoli, a unique immunoregulatory environment. There, alveolar macrophages that have not yet been activated (prototypic alternatively activated macrophages) phagocytose the bacilli. Adhesion of mycobacteria to macrophages results largely from binding of the bacterial cell wall to a variety of macrophage cell-surface molecules, including complement receptors, the mannose receptor, the immunoglobulin GFcγ receptor, and type A scavenger receptors. Phagocytosis is enhanced by complement activation leading to opsonization of bacilli with C3 activation products such as C3b and C3bi. (Bacilli are resistant to complement-mediated lysis.) Binding of certain receptors, such as the mannose receptor, regulates postphagocytic events such as phagosome–lysosome fusion and inflammatory cytokine production. After a phagosome forms, the survival of M. tuberculosis within it seems to depend in part on reduced acidification due to lack of assembly of a complete vesicular proton-adenosine triphosphatase. A complex series of events is generated by the bacterial cell-wall lipoglycan lipoarabinomannan (ManLAM). ManLAM inhibits the intracellular increase of Ca2+. Thus, the Ca2+/calmodulin pathway (leading to phagosome–lysosome fusion) is impaired, and the bacilli survive within the phagosomes. The M. tuberculosis phagosome has been found to inhibit the production of phosphatidylinositol 3-phosphate (PI3P). Normally, PI3P earmarks phagosomes for membrane sorting and maturation, including phagolysosome formation, which would destroy the bacteria. Bacterial factors have also been found to block the host defense of autophagy, in which the cell sequesters the phagosome in a double-membrane vesicle (autophagosome) that is destined to fuse with lysosomes. If the bacilli are successful in arresting phagosome maturation, then replication begins and the macrophage eventually ruptures and releases its bacillary contents. Other uninfected phagocytic cells are then recruited to continue the infection cycle by ingesting dying macrophages and their bacillary content, thus in turn becoming infected themselves and expanding the infection.
VIRULENCE OF TUBERCLE BACILLI
M. tuberculosis must be viewed as a complex formed by a multitude of strains that differ in virulence and are capable of producing a variety of manifestations of disease. Since the elucidation of the M. tuberculosis genome in 1998, large mutant collections have been generated, and many bacterial genes that contribute to M. tuberculosis virulence have been found. Different patterns of virulence defects have been defined in various animal models—predominantly mice but also guinea pigs, rabbits, and nonhuman primates. The katG gene encodes for a catalase/peroxidase enzyme that protects against oxidative stress and is required for isoniazid activation and subsequent bactericidal activity. Region of difference 1 (RD1) is a 9.5-kb locus that encodes two key small protein antigens—early secretory antigen-6 (ESAT-6) and culture filtrate protein-10 (CFP-10)—as well as a putative secretion apparatus that may facilitate their egress; the absence of this locus in the vaccine strain M. bovis bacille Calmette-Guérin (BCG) has been shown to be a key attenuating mutation. The validity of a recent observation in M. marinum needs to be confirmed in M. tuberculosis; in M. marinum, a mutation in the RD1 virulence locus encoding the ESX1 secretion system impairs the capacity of apoptotic macrophages to recruit uninfected cells for further rounds of infection. The results are less replication and fewer new granulomas. Mutants lacking key enzymes of bacterial biosynthesis become auxotrophic for the missing substrate and often are totally unable to proliferate in animals; these include the leuCD and panCD mutants, which require leucine and pantothenic acid, respectively. The isocitrate lyase gene icl1 encodes a key step in the glyoxylate shunt that facilitates bacterial growth on fatty acid substrates; this gene is required for long-term persistence of M. tuberculosis infection in mice with chronic TB. M. tuberculosis mutants in regulatory genes such as sigma factor C and sigma factor H (sigC and sigH) are associated with normal bacterial growth in mice, but they fail to elicit full tissue pathology. Finally, the mycobacterial protein CarD (expressed by the carD gene) seems essential for the control of rRNA transcription that is required for replication and persistence in the host cell. Its loss exposes mycobacteria to oxidative stress, starvation, DNA damage, and ultimately sensitivity to killing by a variety of host mutagens and defensive mechanisms.
INNATE RESISTANCE TO INFECTION
Several observations suggest that genetic factors play a key role in innate nonimmune resistance to infection with M. tuberculosis and the development of disease. The existence of this resistance, which is polygenic in nature, is suggested by the differing degrees of susceptibility to TB in different populations. In mice, a gene called Nramp1 (natural resistance–associated macrophage protein 1) plays a regulatory role in resistance/susceptibility to mycobacteria. The human homologue NRAMP1, which maps to chromosome 2q, may play a role in determining susceptibility to TB, as is suggested by a study among West Africans. Studies of mouse genetics identified a novel host resistance gene, ipr1, which is encoded within the sst1 locus; ipr1 encodes an interferon (IFN)–inducible nuclear protein that interacts with other nuclear proteins in macrophages primed with IFNs or infected by M. tuberculosis. In addition, polymorphisms in multiple genes, such as those encoding for various major histocompatibility complex (MHC) alleles, IFN-γ, T cell growth factor β, interleukin (IL) 10, mannose-binding protein, IFN-γ receptor, Toll-like receptor 2, vitamin D receptor, and IL-1, have been associated with susceptibility to TB.
THE HOST RESPONSE, GRANULOMA FORMATION, AND “LATENCY”
In the initial stage of host–bacterium interaction, prior to the onset of an acquired CMI response, M. tuberculosis disseminates widely through the lymph vessels, spreading to other sites in the lungs and other organs, and undergoes a period of extensive growth within naïve unactivated macrophages; additional naïve macrophages are recruited to the early granuloma. Studies suggest that M. tuberculosis uses specific virulence mechanisms to subvert host cellular signaling and to elicit an early regulated proinflammatory response that promotes granuloma expansion and bacterial growth during this key early phase. A study of M. marinum infection in zebrafish has delineated one molecular mechanism by which mycobacteria induce granuloma formation. The mycobacterial protein ESAT-6 induces secretion of matrix metalloproteinase 9 (MMP9) by nearby epithelial cells that are in contact with infected macrophages. MMP9 in turn stimulates recruitment of naïve macrophages, thus inducing granuloma maturation and bacterial growth. Disruption of MMP9 function results in reduced bacterial growth. Another study has shown that M. tuberculosis–derived cyclic AMP is secreted from the phagosome into host macrophages, subverting the cell’s signal transduction pathways and stimulating an elevation in the secretion of tumor necrosis factor α (TNF-α) as well as further proinflammatory cell recruitment. Ultimately, the chemoattractants and bacterial products released during the repeated rounds of cell lysis and infection of newly arriving macrophages enable dendritic cells to access bacilli; these cells migrate to the draining lymph nodes and present mycobacterial antigens to T lymphocytes. At this point, the development of CMI and humoral immunity begins. These initial stages of infection are usually asymptomatic.
About 2–4 weeks after infection, two host responses to M. tuberculosis develop: a macrophage-activating CMI response and a tissue-damaging response. The macrophage-activating response is a T cell–mediated phenomenon resulting in the activation of macrophages that are capable of killing and digesting tubercle bacilli. The tissue-damaging response is the result of a delayed-type hypersensitivity (DTH) reaction to various bacillary antigens; it destroys unactivated macrophages that contain multiplying bacilli but also causes caseous necrosis of the involved tissues (see below). Although both of these responses can inhibit mycobacterial growth, it is the balance between the two that determines the forms of TB that will develop subsequently. With the development of specific immunity and the accumulation of large numbers of activated macrophages at the site of the primary lesion, granulomatous lesions (tubercles) are formed. These lesions consist of accumulations of lymphocytes and activated macrophages that evolve toward epithelioid and giant cell morphologies. Initially, the tissue-damaging response can limit mycobacterial growth within macrophages. As stated above, this response, mediated by various bacterial products, not only destroys macrophages but also produces early solid necrosis in the center of the tubercle. Although M. tuberculosis can survive, its growth is inhibited within this necrotic environment by low oxygen tension and low pH. At this point, some lesions may heal by fibrosis, with subsequent calcification, whereas inflammation and necrosis occur in other lesions. Some observations have challenged the traditional view that any encounter between mycobacteria and macrophages results in chronic infection. It is possible that an immune response capable of eradicating early infection may sometimes develop as a consequence, for instance, of disabling mutations in mycobacterial genomes rendering their replication ineffective. Individual granulomas that are formed during this phase of infection can vary in size and cell composition; some can contain the spread of mycobacteria, while others cannot. LTBI ensues as a result of this dynamic balance between the microorganism and the host. According to recent developments, latency may not be an accurate term because bacilli may remain active during this “latent” stage, forming biofilms in necrotic areas within which they temporarily hide. Thus, the term persister is probably more accurate to indicate the behavior of the bacilli in this phase. It is important to recognize that latent infection and disease represent not a binary state but rather a continuum along which infection will eventually move in the direction of full containment or disease. The ability to predict, through systemic biomarkers, which affected individuals will progress toward disease would be of immense value in devising prophylactic interventions.
CMI is critical at this early stage. In the majority of infected individuals, local macrophages are activated when bacillary antigens processed by macrophages stimulate T lymphocytes to release a variety of lymphokines. These activated macrophages aggregate around the lesion’s center and effectively neutralize tubercle bacilli without causing further tissue destruction. In the central part of the lesion, the necrotic material resembles soft cheese (caseous necrosis)—a phenomenon that may also be observed in other conditions, such as neoplasms. Even when healing takes place, viable bacilli may remain dormant within macrophages or in the necrotic material for many years. These “healed” lesions in the lung parenchyma and hilar lymph nodes may later undergo calcification.
In a minority of cases, the macrophage-activating response is weak, and mycobacterial growth can be inhibited only by intensified DTH reactions, which lead to lung tissue destruction. The lesion tends to enlarge further, and the surrounding tissue is progressively damaged. At the center of the lesion, the caseous material liquefies. Bronchial walls and blood vessels are invaded and destroyed, and cavities are formed. The liquefied caseous material, containing large numbers of bacilli, is drained through bronchi. Within the cavity, tubercle bacilli multiply, spill into the airways, and are discharged into the environment through expiratory maneuvers such as coughing and talking. In the early stages of infection, bacilli are usually transported by macrophages to regional lymph nodes, from which they gain access to the central venous return; from there they reseed the lungs and may also disseminate beyond the pulmonary vasculature throughout the body via the systemic circulation. The resulting extrapulmonary lesions may undergo the same evolution as those in the lungs, although most tend to heal. In young children with poor natural immunity, hematogenous dissemination may result in highly fatal miliary TB or tuberculous meningitis.
ROLE OF MACROPHAGES AND MONOCYTES
While CMI confers partial protection against M. tuberculosis, humoral immunity plays a less well-defined role in protection (although evidence is accumulating on the existence of antibodies to lipoarabinomannan, which may prevent dissemination of infection in children). In the case of CMI, two types of cells are essential: macrophages, which directly phagocytose tubercle bacilli, and T cells (mainly CD4+ T lymphocytes), which induce protection through the production of cytokines, especially IFN-γ. After infection with M. tuberculosis, alveolar macrophages secrete various cytokines responsible for a number of events (e.g., the formation of granulomas) as well as systemic effects (e.g., fever and weight loss). However, alternatively activated alveolar macrophages may be particularly susceptible to M. tuberculosis growth early on, given their more limited proinflammatory and bactericidal activity, which is related in part to being bathed in surfactant. New monocytes and macrophages attracted to the site are key components of the immune response. Their primary mechanism is probably related to production of oxidants (such as reactive oxygen intermediates or nitric oxide) that have antimycobacterial activity and increase the synthesis of cytokines such as TNF-α and IL-1, which in turn regulate the release of reactive oxygen intermediates and reactive nitrogen intermediates. In addition, macrophages can undergo apoptosis—a defensive mechanism to prevent release of cytokines and bacilli via their sequestration in the apoptotic cell. Recent work also describes the involvement of neutrophils in the host response, although the timing of their appearance and their effectiveness remain uncertain.
Alveolar macrophages, monocytes, and dendritic cells are also critical in processing and presenting antigens to T lymphocytes, primarily CD4+ and CD8+ T cells; the result is the activation and proliferation of CD4+ T lymphocytes, which are crucial to the host’s defense against M. tuberculosis. Qualitative and quantitative defects of CD4+ T cells explain the inability of HIV-infected individuals to contain mycobacterial proliferation. Activated CD4+ T lymphocytes can differentiate into cytokine-producing TH1 or TH2 cells. TH1 cells produce IFN-γ—an activator of macrophages and monocytes—and IL-2. TH2 cells produce IL-4, IL-5, IL-10, and IL-13 and may also promote humoral immunity. The interplay of these various cytokines and their cross-regulation determine the host’s response. The role of cytokines in promoting intracellular killing of mycobacteria, however, has not been entirely elucidated. IFN-γ may induce the generation of reactive nitrogen intermediates and regulate genes involved in bactericidal effects. TNF-α also seems to be important. Observations made originally in transgenic knockout mice and more recently in humans suggest that other T cell subsets, especially CD8+ T cells, may play an important role. CD8+ T cells have been associated with protective activities via cytotoxic responses and lysis of infected cells as well as with production of IFN-γ and TNF-α. Finally, natural killer cells act as co-regulators of CD8+ T cell lytic activities, and γδ T cells are increasingly thought to be involved in protective responses in humans.
MYCOBACTERIAL LIPIDS AND PROTEINS
Lipids have been involved in mycobacterial recognition by the innate immune system, and lipoproteins (such as 19-kDa lipoprotein) have been proven to trigger potent signals through Toll-like receptors present in blood dendritic cells. M. tuberculosis possesses various protein antigens. Some are present in the cytoplasm and cell wall; others are secreted. That the latter are more important in eliciting a T lymphocyte response is suggested by experiments documenting the appearance of protective immunity in animals after immunization with live, protein-secreting mycobacteria. Among the antigens that may play a protective role are the 30-kDa (or 85B) and ESAT-6 antigens. Protective immunity is probably the result of reactivity to many different mycobacterial antigens. These antigens are being incorporated into newly designed vaccines on various platforms.
Coincident with the appearance of immunity, DTH to M. tuberculosis develops. This reactivity is the basis of the TST, which is used primarily for the detection of M. tuberculosis infection in persons without symptoms. The cellular mechanisms responsible for TST reactivity are related mainly to previously sensitized CD4+ T lymphocytes, which are attracted to the skin-test site. There, they proliferate and produce cytokines. Although DTH is associated with protective immunity (TST-positive persons are less susceptible to a new M. tuberculosis infection than TST-negative persons), it by no means guarantees protection against reactivation. In fact, cases of active TB are often accompanied by strongly positive skin-test reactions. There is also evidence of reinfection with a new strain of M. tuberculosis in patients previously treated for active disease. This evidence underscores the fact that previous latent or active TB may not confer fully protective immunity.
TB is classified as pulmonary, extrapulmonary, or both. Depending on several factors linked to different populations and bacterial strains, extrapulmonary TB may occur in 10–40% of patients. Furthermore, up to two-thirds of HIV-infected patients with TB may have both pulmonary and extrapulmonary TB or extrapulmonary TB alone.
Pulmonary TB is conventionally categorized as primary or postprimary (adult-type, secondary). This distinction has been challenged by molecular evidence from TB-endemic areas indicating that a large percentage of cases of adult pulmonary TB result from recent infection (either primary infection or reinfection) and not from reactivation.
Primary pulmonary TB occurs soon after the initial infection with tubercle bacilli. It may be asymptomatic or may present with fever and occasionally pleuritic chest pain. In areas of high TB transmission, this form of disease is often seen in children. Because most inspired air is distributed to the middle and lower lung zones, these areas are most commonly involved in primary TB. The lesion forming after initial infection (Ghon focus) is usually peripheral and accompanied by transient hilar or paratracheal lymphadenopathy, which may or may not be visible on standard chest radiography (Fig. 202-4). Some patients develop erythema nodosum on the legs (see Fig. 25e-40) or phlyctenular conjunctivitis. In the majority of cases, the lesion heals spontaneously and becomes evident only as a small calcified nodule. Pleural reaction overlying a subpleural focus is also common. The Ghon focus, with or without overlying pleural reaction, thickening, and regional lymphadenopathy, is referred to as the Ghon complex.
Chest radiograph showing right hilar lymph node enlargement with infiltration into the surrounding lung tissue in a child with primary tuberculosis. (Courtesy of Prof. Robert Gie, Department of Paediatrics and Child Health, Stellenbosch University, South Africa; with permission.)
In young children with immature CMI and in persons with impaired immunity (e.g., those with malnutrition or HIV infection), primary pulmonary TB may progress rapidly to clinical illness. The initial lesion increases in size and can evolve in different ways. Pleural effusion, which is found in up to two-thirds of cases, results from the penetration of bacilli into the pleural space from an adjacent subpleural focus. In severe cases, the primary site rapidly enlarges, its central portion undergoes necrosis, and cavitation develops (progressive primary TB). TB in young children is almost invariably accompanied by hilar or paratracheal lymphadenopathy due to the spread of bacilli from the lung parenchyma through lymphatic vessels. Enlarged lymph nodes may compress bronchi, causing total obstruction with distal collapse, partial obstruction with large-airway wheezing, or a ball-valve effect with segmental/lobar hyperinflation. Lymph nodes may also rupture into the airway with development of pneumonia, often including areas of necrosis and cavitation, distal to the obstruction. Bronchiectasis (Chap. 312) may develop in any segment/lobe damaged by progressive caseating pneumonia. Occult hematogenous dissemination commonly follows primary infection. However, in the absence of a sufficient acquired immune response, which usually contains the infection, disseminated or miliary disease may result (Fig. 202-5). Small granulomatous lesions develop in multiple organs and may cause locally progressive disease or result in tuberculous meningitis; this is a particular concern in very young children and immunocompromised persons (e.g., patients with HIV infection).
Chest radiograph showing bilateral miliary (millet-sized) infiltrates in a child. (Courtesy of Prof. Robert Gie, Department of Paediatrics and Child Health, Stellenbosch University, South Africa; with permission.)
Postprimary (Adult-Type) Disease
Also referred to as reactivation or secondary TB, postprimary TB is probably most accurately termed adult-type TB because it may result from endogenous reactivation of distant LTBI or recent infection (primary infection or reinfection). It is usually localized to the apical and posterior segments of the upper lobes, where the substantially higher mean oxygen tension (compared with that in the lower zones) favors mycobacterial growth. The superior segments of the lower lobes are also more frequently involved. The extent of lung parenchymal involvement varies greatly, from small infiltrates to extensive cavitary disease. With cavity formation, liquefied necrotic contents are ultimately discharged into the airways and may undergo bronchogenic spread, resulting in satellite lesions within the lungs that may in turn undergo cavitation (Figs. 202-6 and 202-7). Massive involvement of pulmonary segments or lobes, with coalescence of lesions, produces caseating pneumonia. While up to one-third of untreated patients reportedly succumb to severe pulmonary TB within a few months after onset (the classic “galloping consumption” of the past), others may undergo a process of spontaneous remission or proceed along a chronic, progressively debilitating course (“consumption” or phthisis). Under these circumstances, some pulmonary lesions become fibrotic and may later calcify, but cavities persist in other parts of the lungs. Individuals with such chronic disease continue to discharge tubercle bacilli into the environment. Most patients respond to treatment, with defervescence, decreasing cough, weight gain, and a general improvement in well-being within several weeks.
Chest radiograph showing a right-upper-lobe infiltrate and a cavity with an air-fluid level in a patient with active tuberculosis. (Courtesy of Dr. Andrea Gori, Department of Infectious Diseases, S. Paolo University Hospital, Milan, Italy; with permission.)
CT scan showing a large cavity in the right lung of a patient with active tuberculosis. (Courtesy of Dr. Elisa Busi Rizzi, National Institute for Infectious Diseases, Spallanzani Hospital, Rome, Italy; with permission.)
Early in the course of disease, symptoms and signs are often nonspecific and insidious, consisting mainly of diurnal fever and night sweats due to defervescence, weight loss, anorexia, general malaise, and weakness. However, in up to 90% of cases, cough eventually develops—often initially nonproductive and limited to the morning and subsequently accompanied by the production of purulent sputum, sometimes with blood streaking. Hemoptysis develops in 20–30% of cases, and massive hemoptysis may ensue as a consequence of the erosion of a blood vessel in the wall of a cavity. Hemoptysis, however, may also result from rupture of a dilated vessel in a cavity (Rasmussen’s aneurysm) or from aspergilloma formation in an old cavity. Pleuritic chest pain sometimes develops in patients with subpleural parenchymal lesions or pleural disease. Extensive disease may produce dyspnea and, in rare instances, adult respiratory distress syndrome. Physical findings are of limited use in pulmonary TB. Many patients have no abnormalities detectable by chest examination, whereas others have detectable rales in the involved areas during inspiration, especially after coughing. Occasionally, rhonchi due to partial bronchial obstruction and classic amphoric breath sounds in areas with large cavities may be heard. Systemic features include fever (often low-grade and intermittent) in up to 80% of cases and wasting. Absence of fever, however, does not exclude TB. In some cases, pallor and finger clubbing develop. The most common hematologic findings are mild anemia, leukocytosis, and thrombocytosis with a slightly elevated erythrocyte sedimentation rate and/or C-reactive protein level. None of these findings is consistent or sufficiently accurate for diagnostic purposes. Hyponatremia due to the syndrome of inappropriate secretion of antidiuretic hormone has also been reported.
In order of frequency, the extrapulmonary sites most commonly involved in TB are the lymph nodes, pleura, genitourinary tract, bones and joints, meninges, peritoneum, and pericardium. However, virtually all organ systems may be affected. As a result of hematogenous dissemination in HIV-infected individuals, extrapulmonary TB is seen more commonly today than in the past in settings of high HIV prevalence.
Lymph Node TB (Tuberculous Lymphadenitis)
The most common presentation of extrapulmonary TB in both HIV-seronegative and HIV-infected patients (35% of cases worldwide and more than 40% of cases in the United States in recent series), lymph node disease is particularly frequent among HIV-infected patients and among children (Fig. 202-8). In the United States, besides children, women (particularly non-Caucasians) seem to be especially susceptible. Once caused mainly by M. bovis, tuberculous lymphadenitis today is due largely to M. tuberculosis. Lymph node TB presents as painless swelling of the lymph nodes, most commonly at posterior cervical and supraclavicular sites (a condition historically referred to as scrofula). Lymph nodes are usually discrete in early disease but develop into a matted nontender mass over time and may result in a fistulous tract draining caseous material. Associated pulmonary disease is present in fewer than 50% of cases, and systemic symptoms are uncommon except in HIV-infected patients. The diagnosis is established by fine-needle aspiration biopsy (with a yield of up to 80%) or surgical excision biopsy. Bacteriologic confirmation is achieved in the vast majority of cases, granulomatous lesions with or without visible AFBs are typically seen, and cultures are positive in 70–80% of cases. Among HIV-infected patients, granulomas are less well organized and are frequently absent entirely, but bacterial loads are heavier than in HIV-seronegative patients, with higher yields from microscopy and culture. Differential diagnosis includes a variety of infectious conditions, neoplastic diseases such as lymphomas or metastatic carcinomas, and rare disorders like Kikuchi’s disease (necrotizing histiocytic lymphadenitis), Kimura’s disease, and Castleman’s disease.
Tuberculous lymphadenitis affecting the cervical lymph nodes in a 2-year-old child from Malawi. (Courtesy of Prof. S. Graham, Centre for International Child Health, University of Melbourne, Australia; with permission.)
Involvement of the pleura accounts for ~20% of extrapulmonary cases in the United States and elsewhere. Isolated pleural effusion usually reflects recent primary infection, and the collection of fluid in the pleural space represents a hypersensitivity response to mycobacterial antigens. Pleural disease may also result from contiguous parenchymal spread, as in many cases of pleurisy accompanying postprimary disease. Depending on the extent of reactivity, the effusion may be small, remain unnoticed, and resolve spontaneously or may be sufficiently large to cause symptoms such as fever, pleuritic chest pain, and dyspnea. Physical findings are those of pleural effusion: dullness to percussion and absence of breath sounds. A chest radiograph reveals the effusion and, in up to one-third of cases, also shows a parenchymal lesion. Thoracentesis is required to ascertain the nature of the effusion and to differentiate it from manifestations of other etiologies. The fluid is straw colored and at times hemorrhagic; it is an exudate with a protein concentration >50% of that in serum (usually ~4–6 g/dL), a normal to low glucose concentration, a pH of ~7.3 (occasionally <7.2), and detectable white blood cells (usually 500–6000/μL). Neutrophils may predominate in the early stage, but lymphocyte predominance is the typical finding later. Mesothelial cells are generally rare or absent. AFB are rarely seen on direct smear, and cultures often may be falsely negative for M. tuberculosis; positive cultures are more common among postprimary cases. Determination of the pleural concentration of adenosine deaminase (ADA) may be a useful screening test, and TB may be excluded if the value is very low. Lysozyme is also present in the pleural effusion. Measurement of IFN-γ, either directly or through stimulation of sensitized T cells with mycobacterial antigens, can be helpful. Needle biopsy of the pleura is often required for diagnosis and is recommended over pleural fluid; it reveals granulomas and/or yields a positive culture in up to 80% of cases. Pleural biopsy can yield a positive result in ~75% of cases when real-time automated nucleic acid amplification is used (the Xpert MTB/RIF assay [Cepheid, Sunnyvale, CA]; see “Nucleic Acid Amplification Technology,” below), although pleural fluid testing with this assay is not recommended because of low sensitivity. This form of pleural TB responds rapidly to chemotherapy and may resolve spontaneously. Concurrent glucocorticoid administration may reduce the duration of fever and/or chest pain but is not of proven benefit.
Tuberculous empyema is a less common complication of pulmonary TB. It is usually the result of the rupture of a cavity, with spillage of a large number of organisms into the pleural space. This process may create a bronchopleural fistula with evident air in the pleural space. A chest radiograph shows hydropneumothorax with an air-fluid level. The pleural fluid is purulent and thick and contains large numbers of lymphocytes. Acid-fast smears and mycobacterial cultures are often positive. Surgical drainage is usually required as an adjunct to chemotherapy. Tuberculous empyema may result in severe pleural fibrosis and restrictive lung disease. Removal of the thickened visceral pleura (decortication) is occasionally necessary to improve lung function.
Nearly always a complication of advanced cavitary pulmonary TB, TB of the upper airways may involve the larynx, pharynx, and epiglottis. Symptoms include hoarseness, dysphonia, and dysphagia in addition to chronic productive cough. Findings depend on the site of involvement, and ulcerations may be seen on laryngoscopy. Acid-fast smear of the sputum is often positive, but biopsy may be necessary in some cases to establish the diagnosis. Carcinoma of the larynx may have similar features but is usually painless.
Genitourinary TB, which accounts for ~10–15% of all extrapulmonary cases in the United States and elsewhere, may involve any portion of the genitourinary tract. Local symptoms predominate, and up to 75% of patients have chest radiographic abnormalities suggesting previous or concomitant pulmonary disease. Urinary frequency, dysuria, nocturia, hematuria, and flank or abdominal pain are common presentations. However, patients may be asymptomatic and their disease discovered only after severe destructive lesions of the kidneys have developed. Urinalysis gives abnormal results in 90% of cases, revealing pyuria and hematuria. The documentation of culture-negative pyuria in acidic urine should raise the suspicion of TB. IV pyelography, abdominal computed tomography (CT), or magnetic resonance imaging (MRI) (Fig. 202-9) may show deformities and obstructions; calcifications and ureteral strictures are suggestive findings. Culture of three morning urine specimens yields a definitive diagnosis in nearly 90% of cases. Severe ureteral strictures may lead to hydronephrosis and renal damage. Genital TB is diagnosed more commonly in female than in male patients. In female patients, it affects the fallopian tubes and the endometrium and may cause infertility, pelvic pain, and menstrual abnormalities. Diagnosis requires biopsy or culture of specimens obtained by dilation and curettage. In male patients, genital TB preferentially affects the epididymis, producing a slightly tender mass that may drain externally through a fistulous tract; orchitis and prostatitis may also develop. In almost half of cases of genitourinary TB, urinary tract disease is also present. Genitourinary TB responds well to chemotherapy.
MRI of culture-confirmed renal tuberculosis. T2-weighted coronary plane: coronal sections showing several renal lesions in both the cortical and the medullary tissues of the right kidney. (Courtesy of Dr. Alberto Matteelli, Department of Infectious Diseases, University of Brescia, Italy; with permission.)
In the United States, TB of the bones and joints is responsible for ~10% of extrapulmonary cases. In bone and joint disease, pathogenesis is related to reactivation of hematogenous foci or to spread from adjacent paravertebral lymph nodes. Weight-bearing joints (the spine in 40% of cases, the hips in 13%, and the knees in 10%) are most commonly affected. Spinal TB (Pott’s disease or tuberculous spondylitis; Fig. 202-10) often involves two or more adjacent vertebral bodies. Whereas the upper thoracic spine is the most common site of spinal TB in children, the lower thoracic and upper lumbar vertebrae are usually affected in adults. From the anterior superior or inferior angle of the vertebral body, the lesion slowly reaches the adjacent body, later affecting the intervertebral disk. With advanced disease, collapse of vertebral bodies results in kyphosis (gibbus). A paravertebral “cold” abscess may also form. In the upper spine, this abscess may track to and penetrate the chest wall, presenting as a soft tissue mass; in the lower spine, it may reach the inguinal ligaments or present as a psoas abscess. CT or MRI reveals the characteristic lesion and suggests its etiology. The differential diagnosis includes tumors and other infections. Pyogenic bacterial osteomyelitis, in particular, involves the disk very early and produces rapid sclerosis. Aspiration of the abscess or bone biopsy confirms the tuberculous etiology, as cultures are usually positive and histologic findings highly typical. A catastrophic complication of Pott’s disease is paraplegia, which is usually due to an abscess or a lesion compressing the spinal cord. Paraparesis due to a large abscess is a medical emergency and requires rapid drainage. TB of the hip joints, usually involving the head of the femur, causes pain; TB of the knee produces pain and swelling. If the disease goes unrecognized, the joints may be destroyed. Diagnosis requires examination of the synovial fluid, which is thick in appearance, with a high protein concentration and a variable cell count. Although synovial fluid culture is positive in a high percentage of cases, synovial biopsy and tissue culture may be necessary to establish the diagnosis. Skeletal TB responds to chemotherapy, but severe cases may require surgery.
CT scan demonstrating destruction of the right pedicle of T10 due to Pott’s disease. The patient, a 70-year-old Asian woman, presented with back pain and weight loss and had biopsy-proven tuberculosis. (Courtesy of Charles L. Daley, MD, University of California, San Francisco; with permission.)
Tuberculous Meningitis and Tuberculoma
TB of the central nervous system accounts for ~5% of extrapulmonary cases in the United States. It is seen most often in young children but also develops in adults, especially those infected with HIV. Tuberculous meningitis results from the hematogenous spread of primary or postprimary pulmonary TB or from the rupture of a subependymal tubercle into the subarachnoid space. In more than half of cases, evidence of old pulmonary lesions or a miliary pattern is found on chest radiography. The disease often presents subtly as headache and slight mental changes after a prodrome of weeks of low-grade fever, malaise, anorexia, and irritability. If not recognized, tuberculous meningitis may evolve acutely with severe headache, confusion, lethargy, altered sensorium, and neck rigidity. Typically, the disease evolves over 1–2 weeks, a course longer than that of bacterial meningitis. Because meningeal involvement is pronounced at the base of the brain, paresis of cranial nerves (ocular nerves in particular) is a frequent finding, and the involvement of cerebral arteries may produce focal ischemia. The ultimate evolution is toward coma, with hydrocephalus and intracranial hypertension.
Lumbar puncture is the cornerstone of diagnosis. In general, examination of cerebrospinal fluid (CSF) reveals a high leukocyte count (up to 1000/μL), usually with a predominance of lymphocytes but sometimes with a predominance of neutrophils in the early stage; a protein content of 1–8 g/L (100–800 mg/dL); and a low glucose concentration. However, any of these three parameters can be within the normal range. AFBs are infrequently seen on direct smear of CSF sediment, and repeated lumbar punctures increase the yield. Culture of CSF is diagnostic in up to 80% of cases and remains the gold standard. Real-time automated nucleic acid amplification (the Xpert MTB/RIF assay; see “Nucleic Acid Amplification Technology,” below) has a sensitivity of up to 80% and is the preferred initial diagnostic option. Treatment should be initiated immediately upon a positive Xpert MTB/RIF result. A negative result does not exclude a diagnosis of TB and requires further diagnostic workup. Imaging studies (CT and MRI) may show hydrocephalus and abnormal enhancement of basal cisterns or ependyma. If unrecognized, tuberculous meningitis is uniformly fatal. This disease responds to chemotherapy; however, neurologic sequelae are documented in 25% of treated cases, in most of which the diagnosis has been delayed. Clinical trials have demonstrated that patients given adjunctive glucocorticoids may experience faster resolution of CSF abnormalities and elevated CSF pressure. In one study, adjunctive dexamethasone significantly enhanced the chances of survival among persons >14 years of age but did not reduce the frequency of neurologic sequelae. The dexamethasone schedule was (1) 0.4 mg/kg per day given IV with tapering by 0.1 mg/kg per week until the fourth week, when 0.1 mg/kg per day was administered; followed by (2) 4 mg/d given by mouth with tapering by 1 mg per week until the fourth week, when 1 mg/d was administered.
Tuberculoma, an uncommon manifestation of central nervous system TB, presents as one or more space-occupying lesions and usually causes seizures and focal signs. CT or MRI reveals contrast-enhanced ring lesions, but biopsy is necessary to establish the diagnosis.
Gastrointestinal TB is uncommon, making up 3.5% of extrapulmonary cases in the United States. Various pathogenetic mechanisms are involved: swallowing of sputum with direct seeding, hematogenous spread, or (largely in developing areas) ingestion of milk from cows affected by bovine TB. Although any portion of the gastrointestinal tract may be affected, the terminal ileum and the cecum are the sites most commonly involved. Abdominal pain (at times similar to that associated with appendicitis) and swelling, obstruction, hematochezia, and a palpable mass in the abdomen are common findings at presentation. Fever, weight loss, anorexia, and night sweats are also common. With intestinal-wall involvement, ulcerations and fistulae may simulate Crohn’s disease; the differential diagnosis of this entity is always difficult. Anal fistulae should prompt an evaluation for rectal TB. Because surgery is required in most cases, the diagnosis can be established by histologic examination and culture of specimens obtained intraoperatively.
Tuberculous peritonitis follows either the direct spread of tubercle bacilli from ruptured lymph nodes and intraabdominal organs (e.g., genital TB in women) or hematogenous seeding. Nonspecific abdominal pain, fever, and ascites should raise the suspicion of tuberculous peritonitis. The coexistence of cirrhosis (Chap. 363) in patients with tuberculous peritonitis complicates the diagnosis. In tuberculous peritonitis, paracentesis reveals an exudative fluid with a high protein content and leukocytosis that is usually lymphocytic (although neutrophils occasionally predominate). The yield of direct smear and culture is relatively low; culture of a large volume of ascitic fluid can increase the yield, but peritoneal biopsy (with a specimen best obtained by laparoscopy) is often needed to establish the diagnosis.
Pericardial TB (Tuberculous Pericarditis)
Due either to direct extension from adjacent mediastinal or hilar lymph nodes or to hematogenous spread, pericardial TB has often been a disease of the elderly in countries with low TB prevalence. However, it also develops frequently in HIV-infected patients. Case–fatality rates are as high as 40% in some series. The onset may be subacute, although an acute presentation, with dyspnea, fever, dull retrosternal pain, and a pericardial friction rub, is possible. An effusion eventually develops in many cases; cardiovascular symptoms and signs of cardiac tamponade may ultimately appear (Chap. 288). In the presence of effusion, TB must be suspected if the patient belongs to a high-risk population (HIV-infected, originating in a high-prevalence country); if there is evidence of previous TB in other organs; or if echocardiography, CT, or MRI shows effusion and thickness across the pericardial space. A definitive diagnosis can be obtained by pericardiocentesis under echocardiographic guidance. The pericardial fluid must be submitted for biochemical, cytologic, and microbiologic evaluation. The effusion is exudative in nature, with a high count of lymphocytes and monocytes. Hemorrhagic effusion is common. Direct smear examination is very rarely positive. Culture of pericardial fluid reveals M. tuberculosis in up to two-thirds of cases, whereas pericardial biopsy has a higher yield. High levels of ADA, lysozyme, and IFN-γ may suggest a tuberculous etiology.
Without treatment, pericardial TB is usually fatal. Even with treatment, complications may develop, including chronic constrictive pericarditis with thickening of the pericardium, fibrosis, and sometimes calcification, which may be visible on a chest radiograph. Systematic reviews and meta-analyses show that adjunctive glucocorticoid treatment remains controversial, with no conclusive evidence of benefits for all principal outcomes of pericarditis—i.e., no significant impact on resolution of effusion, no significant difference in functional status after treatment, and no significant reduction in the frequency of development of constriction or death. However, in HIV-infected patients, glucocorticoids do improve functional status after treatment.
Caused by direct extension from the pericardium or by retrograde lymphatic extension from affected mediastinal lymph nodes, tuberculous myocarditis is an extremely rare disease. Usually it is fatal and is diagnosed postmortem.
Miliary or Disseminated TB
Miliary TB is due to hematogenous spread of tubercle bacilli. Although in children it is often the consequence of primary infection, in adults it may be due to either recent infection or reactivation of old disseminated foci. The lesions are usually yellowish granulomas 1–2 mm in diameter that resemble millet seeds (thus the term miliary, coined by nineteenth-century pathologists). Clinical manifestations are nonspecific and protean, depending on the predominant site of involvement. Fever, night sweats, anorexia, weakness, and weight loss are presenting symptoms in the majority of cases. At times, patients have a cough and other respiratory symptoms due to pulmonary involvement as well as abdominal symptoms. Physical findings include hepatomegaly, splenomegaly, and lymphadenopathy. Eye examination may reveal choroidal tubercles, which are pathognomonic of miliary TB, in up to 30% of cases. Meningismus occurs in fewer than 10% of cases.
A high index of suspicion is required for the diagnosis of miliary TB. Frequently, chest radiography (Fig. 202-5) reveals a miliary reticulonodular pattern (more easily seen on underpenetrated film), although no radiographic abnormality may be evident early in the course and among HIV-infected patients. Other radiologic findings include large infiltrates, interstitial infiltrates (especially in HIV-infected patients), and pleural effusion. Sputum-smear microscopy is negative in most cases. Various hematologic abnormalities may be seen, including anemia with leukopenia, lymphopenia, neutrophilic leukocytosis and leukemoid reactions, and polycythemia. Disseminated intravascular coagulation has been reported. Elevation of alkaline phosphatase levels and other abnormal values in liver function tests are detected in patients with severe hepatic involvement. The TST may be negative in up to half of cases, but reactivity may be restored during chemotherapy. Bronchoalveolar lavage and transbronchial biopsy are more likely to provide bacteriologic confirmation, and granulomas are evident in liver or bone-marrow biopsy specimens from many patients. If it goes unrecognized, miliary TB is lethal; with proper early treatment, however, it is amenable to cure. Glucocorticoid therapy has not proved beneficial.
A rare presentation seen in the elderly, cryptic miliary TB has a chronic course characterized by mild intermittent fever, anemia, and—ultimately—meningeal involvement preceding death. An acute septicemic form, nonreactive miliary TB, occurs very rarely and is due to massive hematogenous dissemination of tubercle bacilli. Pancytopenia is common in this form of disease, which is rapidly fatal. At postmortem examination, multiple necrotic but nongranulomatous (“nonreactive”) lesions are detected.
Less Common Extrapulmonary Forms
TB may cause chorioretinitis, uveitis, panophthalmitis, and painful hypersensitivity-related phlyctenular conjunctivitis. Tuberculous otitis is rare and presents as hearing loss, otorrhea, and tympanic membrane perforation. In the nasopharynx, TB may simulate granulomatosis with polyangiitis. Cutaneous manifestations of TB include primary infection due to direct inoculation, abscesses and chronic ulcers, scrofuloderma, lupus vulgaris (a smoldering disease with nodules, plaques, and fissures), miliary lesions, and erythema nodosum. Tuberculous mastitis results from retrograde lymphatic spread, often from the axillary lymph nodes. Adrenal TB is a manifestation of disseminated disease presenting rarely as adrenal insufficiency. Finally, congenital TB results from transplacental spread of tubercle bacilli to the fetus or from ingestion of contaminated amniotic fluid. This rare disease affects the liver, spleen, lymph nodes, and various other organs.
TB may cause persisting pulmonary damage in patients whose infection has been considered cured on clinical grounds. Chronic impairment of lung functions, bronchiectasis, aspergillomas, and chronic pulmonary aspergillosis (CPA) have been associated with TB. CPA may manifest as simple aspergilloma (fungal ball) or chronic cavitary aspergillosis. Early studies revealed that, especially in the presence of large residual cavities, Aspergillus fumigatus may colonize the lesion and produce symptoms such as respiratory impairment, hemoptysis, persistent fatigue, and weight loss, often resulting in the erroneous diagnosis of TB recurrence. The detection of Aspergillus precipitins (IgG) in the blood suggests CPA, as do radiographic abnormalities such as thickening of the cavitary walls or the presence of a fungal ball inside the cavity. Treatment is difficult. Recent preliminary studies on the use of itraconazole for 6 months suggest that treatment with this agent may be superior to conservative treatment in improving radiologic and clinical manifestations of CPA. Surgical removal of lesions is risky.
(See also Chap. 226) TB is one of the most common diseases among HIV-infected persons worldwide and a major cause of death in this population; more specifically, it is responsible for an estimated 24% of all HIV-related mortality. In certain urban settings in some African countries, the rate of HIV infection among TB patients reaches 70–80%. A person with a positive TST who acquires HIV infection has a 3–13% annual risk of developing active TB. A new TB infection acquired by an HIV-infected individual may evolve to active disease in a matter of weeks rather than months or years. TB can appear at any stage of HIV infection, and its presentation varies with the stage. When CMI is only partially compromised, pulmonary TB presents in a typical manner (Figs. 202-6 and 202-7), with upper-lobe infiltrates and cavitation and without significant lymphadenopathy or pleural effusion. In late stages of HIV infection, when the CD4+ T cell count is <200/μL, a primary TB–like pattern, with diffuse interstitial and subtle infiltrates, little or no cavitation, pleural effusion, and intrathoracic lymphadenopathy, is more common. However, these forms are becoming less common because of the expanded use of antiretroviral treatment (ART). Overall, sputum smears are less frequently positive among TB patients with HIV infection than among those without; thus, the diagnosis of TB may be difficult, especially in view of the variety of HIV-related pulmonary conditions mimicking TB. Extrapulmonary TB is common among HIV-infected patients. In various series, extrapulmonary TB—alone or in association with pulmonary disease—has been documented in 40–60% of all cases in HIV-co-infected individuals. The most common forms are lymphatic, disseminated, pleural, and pericardial. Mycobacteremia and meningitis are also common, particularly in advanced HIV disease. The diagnosis of TB in HIV-infected patients may be complicated not only by the increased frequency of sputum-smear negativity (up to 40% in culture-proven pulmonary cases) but also by atypical radiographic findings, a lack of classic granuloma formation in the late stages, and a negative TST. The Xpert MTB/RIF assay (see “Nucleic Acid Amplification Technology,” below) is the preferred initial diagnostic option, and therapy should be started on the basis of a positive result because treatment delays may be fatal. A negative Xpert MTB/RIF result does not exclude a diagnosis of TB, and culture remains the gold standard.
Exacerbations in systemic (lymphadenopathy) or respiratory symptoms, signs, and laboratory or radiographic manifestations of TB—termed the immune reconstitution inflammatory syndrome (IRIS) or TB immune reconstitution disease (TB-IRD)—have been associated with the administration of ART and occur in ~10% of HIV-infected TB patients. Usually developing 1–3 months after initiation of ART, IRIS is more common among patients with advanced immunosuppression and extrapulmonary TB. “Unmasking IRIS” may also develop after the initiation of ART in patients with undiagnosed subclinical TB. The earlier ART is started and the lower the baseline CD4+ T cell count, the greater the risk of IRIS. Death due to IRIS is relatively infrequent and occurs mainly among patients who have a high preexisting mortality risk. The presumed pathogenesis of IRIS consists of an immune response that is elicited by antigens released as bacilli are killed during effective chemotherapy and that is temporally associated with improving immune function. There is no diagnostic test for IRIS, and its confirmation relies heavily upon case definitions incorporating clinical and laboratory data; a variety of case definitions have been suggested. The first priority in the management of a possible case of IRIS is to ensure that the clinical syndrome does not represent a failure of TB treatment or the development of another infection. Mild paradoxical reactions can be managed with symptom-based treatment. Glucocorticoids have been used for more severe reactions, and prednisolone given for 4 weeks at a low dosage (1.5 mg/kg per day for 2 weeks and half that dose for the remaining 2 weeks) has reduced the need for hospitalization and therapeutic procedures and hastened alleviation of symptoms, as reflected by Karnofsky performance scores, quality-of-life assessments, radiographic response, and C-reactive protein levels. The effectiveness of glucocorticoids in alleviating the symptoms of IRIS is probably linked to suppression of proinflammatory cytokine concentrations, as these medications reduce serum concentrations of IL-6, IL-10, IL-12p40, TNF-α, IFN-γ, and IFN-γ-inducible protein 10 (IP-10). Recommendations for the prevention and treatment of TB in HIV-infected individuals are provided below.
The key to the diagnosis of TB remains a high index of suspicion. Diagnosis is not difficult in persons belonging to high-risk populations who present with typical symptoms and a classic chest radiograph showing upper-lobe infiltrates with cavities (Fig. 202-6). On the other hand, the diagnosis can easily be missed in an elderly nursing-home resident or a teenager with a focal infiltrate. Often, the diagnosis is first entertained when the chest radiograph of a patient being evaluated for respiratory symptoms is abnormal. If the patient has no complicating medical conditions that cause immunosuppression, the chest radiograph may show typical upper-lobe infiltrates with cavitation (Fig. 202-6). The longer the delay between the onset of symptoms and the diagnosis, the more likely is the finding of cavitary disease. In contrast, immunosuppressed patients, including those with HIV infection, may have “atypical” findings on chest radiography—e.g., lower-zone infiltrates without cavity formation.
The several approaches to the diagnosis of TB require, above all, a well-organized laboratory network with an appropriate distribution of tasks at different levels of the health care system. At the peripheral and community levels, screening and referral are the principal tasks—besides clinical assessment and radiography—that can be accomplished through AFB microscopy and/or real-time automated nucleic acid amplification technology (the Xpert MTB/RIF assay; see below). At a secondary level (e.g., a traditional district hospital in a high-incidence setting), additional technology can be adopted, including rapid culture and drug susceptibility testing.
A presumptive diagnosis is commonly based on the finding of AFB on microscopic examination of a diagnostic specimen, such as a smear of expectorated sputum or of tissue (e.g., a lymph node biopsy). Although inexpensive, AFB microscopy has relatively low sensitivity (40–60%) in culture-confirmed cases of pulmonary TB. The traditional method—light microscopy of specimens stained with Ziehl-Neelsen basic fuchsin dyes—is nevertheless satisfactory, although time-consuming. Most modern laboratories processing large numbers of diagnostic specimens use auramine–rhodamine staining and fluorescence microscopy; this approach is more sensitive than the Ziehl-Neelsen method. However, it is expensive because it requires high-cost mercury vapor light sources and a dark room. Less expensive light-emitting diode (LED) fluorescence microscopes are now available. They are as sensitive as—or more sensitive than—traditional fluorescence microscopes. As a result, conventional light and fluorescence microscopes are being replaced with this more recent technology, especially in developing countries. For patients with suspected pulmonary TB, it has been recommended that two or three sputum specimens, preferably collected early in the morning, should be submitted to the laboratory for AFB smear and mycobacterial culture. Two specimens collected on the same visit may be as effective as three. If tissue is obtained, it is critical that the portion of the specimen intended for culture not be put in formaldehyde. The use of AFB microscopy in examining urine or gastric lavage fluid is limited by the presence of commensal mycobacteria that can cause false-positive results.
NUCLEIC ACID AMPLIFICATION TECHNOLOGY
Several test systems based on amplification of mycobacterial nucleic acid have become available in the past few years. These tests are most useful for the rapid confirmation of TB in persons with AFB-positive specimens, but some also have utility for the diagnosis of AFB-negative pulmonary and extrapulmonary TB. One system that permits rapid diagnosis of TB with high specificity and sensitivity (approaching that of culture) is the fully automated, real-time nucleic acid amplification technology known as the Xpert MTB/RIF assay. Xpert MTB/RIF can simultaneously detect TB and rifampin resistance in <2 h and has minimal biosafety and training requirements. Therefore, it can be housed in nonconventional laboratory settings. The WHO recommends its use worldwide as the initial diagnostic test in adults and children presumed to have MDR-TB or HIV-associated TB. Taking into account the availability of resources, the test may also be used in any adult or child presumed to have TB or as a follow-up test after microscopy in adults presumed to have TB but not at risk of MDR-TB or HIV-associated TB. Xpert MTB/RIF should be the initial test applied to CSF from patients in whom TB meningitis is suspected as well as a replacement test (over conventional microscopy, culture, and histopathology) for selected nonrespiratory specimens—obtained by gastric lavage, fine-needle aspiration, or pleural or other biopsies—from patients in whom extrapulmonary TB is suspected. This test has a sensitivity of 98% among AFB-positive cases and ~70% among AFB-negative specimens. Other tests, such as those based on manual amplification platforms, have not yet been deemed satisfactory for introduction into clinical practice as replacements for existing tests.
Definitive diagnosis depends on the isolation and identification of M. tuberculosis from a clinical specimen or the identification of specific DNA sequences in a nucleic acid amplification test. Specimens may be inoculated onto egg- or agar-based medium (e.g., Löwenstein-Jensen or Middlebrook 7H10) and incubated at 37°C (under 5% CO2 for Middlebrook medium). Because most species of mycobacteria, including M. tuberculosis, grow slowly, 4–8 weeks may be required before growth is detected. Although M. tuberculosis may be identified presumptively on the basis of growth time and colony pigmentation and morphology, a variety of biochemical tests have traditionally been used to speciate mycobacterial isolates. In modern, well-equipped laboratories, liquid culture for isolation and species identification by molecular methods or high-pressure liquid chromatography of mycolic acids has replaced isolation on solid media and identification by biochemical tests. A widely used technology is the mycobacterial growth indicator tube (BBL™ MGIT™; BD, Franklin Lakes, NJ), which uses a fluorescent compound sensitive to the presence of oxygen dissolved in the liquid medium. The appearance of fluorescence detected by fluorometric technology indicates active growth of mycobacteria. A low-cost, rapid immunochromatographic lateral-flow assay based on detection of MTP64 antigen may also be used for species identification of the M. tuberculosis complex in culture isolates. These new methods, which are also being introduced in low-income countries, have decreased the time required for bacteriologic confirmation of TB to 2–3 weeks.
DRUG SUSCEPTIBILITY TESTING
Any initial isolate of M. tuberculosis should be tested for susceptibility to isoniazid and rifampin in order to detect drug resistance and/or MDR-TB, particularly if one or more risk factors for drug resistance are identified or if the patient either fails to respond to initial therapy or has a relapse after the completion of treatment (see “Treatment Failure and Relapse,” below). In addition, expanded susceptibility testing for second-line anti-TB drugs (especially the fluoroquinolones and the injectable drugs) is mandatory when MDR-TB is found. Susceptibility testing may be conducted directly (with the clinical specimen) or indirectly (with mycobacterial cultures) on solid or liquid medium. Results are obtained rapidly by direct susceptibility testing on liquid medium, with an average reporting time of 3 weeks. With indirect testing on solid medium, results may be unavailable for ≥8 weeks. Highly reliable genotypic methods for the rapid identification of genetic mutations in gene regions known to be associated with resistance to rifampin (such as those in rpoB) and isoniazid (such as those in katG and inhA) have been developed and are being widely implemented for screening of patients at increased risk of drug-resistant TB. Apart from the Xpert MTB/RIF assay, which, as mentioned above, detects rifampin resistance, the most widely used are molecular line probe assays. After extraction of DNA from M. tuberculosis isolates or from clinical specimens, the resistance gene regions are amplified by polymerase chain reaction (PCR), and labeled and probe-hybridized PCR products are detected by colorimetric development. This assay reveals the presence of M. tuberculosis as well as mutations in target resistance gene regions. A similar approach has been developed for second-line anti-TB drugs such as the fluoroquinolones, the aminoglycosides kanamycin and amikacin, and capreomycin, but the diagnostic accuracy of the current technology is not yet sufficient to recommend its use in clinical practice. Finally, a few noncommercial, inexpensive culture and drug-susceptibility testing methods (e.g., microscopically observed drug susceptibility, or MODS; nitrate reductase assays; and colorimetric redox indicator assays) may be useful in resource-limited settings. Their use is restricted to national reference laboratories with proven proficiency and adequate external quality control as an interim solution while genotypic or automated liquid culture technology is introduced.
As noted above, the initial suspicion of pulmonary TB is often based on abnormal chest radiographic findings in a patient with respiratory symptoms. Although the “classic” picture is that of upper-lobe disease with infiltrates and cavities (Fig. 202-6), virtually any radiographic pattern—from a normal film or a solitary pulmonary nodule to diffuse alveolar infiltrates in a patient with adult respiratory distress syndrome—may be seen. In the era of AIDS, no radiographic pattern can be considered pathognomonic. CT (Fig. 202-7) may be useful in interpreting questionable findings on plain chest radiography and may be helpful in diagnosing some forms of extrapulmonary TB (e.g., Pott’s disease; Fig. 202-10). MRI is useful in the diagnosis of intracranial TB.
ADDITIONAL DIAGNOSTIC PROCEDURES
Other diagnostic tests may be used when pulmonary TB is suspected. Sputum induction by ultrasonic nebulization of hypertonic saline may be useful for patients who cannot produce a sputum specimen spontaneously. Frequently, patients with radiographic abnormalities that are consistent with other diagnoses (e.g., bronchogenic carcinoma) undergo fiberoptic bronchoscopy with bronchial brushings and endobronchial or transbronchial biopsy of the lesion. Bronchoalveolar lavage of a lung segment containing an abnormality may also be performed. In all cases, it is essential that specimens be submitted for AFB smear, mycobacterial culture, and molecular testing with the Xpert MTB/RIF assay. For the diagnosis of primary pulmonary TB in children, who often do not expectorate sputum, induced sputum specimens and specimens from early-morning gastric lavage may yield positive cultures and/or Xpert MTB/RIF assay results.
Invasive diagnostic procedures are indicated for patients with suspected extrapulmonary TB. In addition to testing of specimens from involved sites (e.g., CSF for tuberculous meningitis, pleural fluid and biopsy samples for pleural disease), biopsy and culture of bone marrow and liver tissue have a good diagnostic yield in disseminated (miliary) TB, particularly in HIV-infected patients, who also have a high frequency of positive blood cultures. In some cases, culture or Xpert MTB/RIF assay results are negative but a clinical diagnosis of TB is supported by consistent epidemiologic evidence (e.g., a history of close contact with an infectious patient) and a compatible clinical and radiographic response to treatment. In the United States and other industrialized countries with low rates of TB, some patients with limited abnormalities on chest radiographs and sputum positive for AFB are infected with nontuberculous mycobacteria, most commonly organisms of the M. avium complex or M. kansasii (Chap. 204). Factors favoring the diagnosis of nontuberculous mycobacterial disease over TB include an absence of risk factors for TB and the presence of underlying chronic pulmonary disease.
Patients with HIV-associated TB pose several diagnostic problems (see “HIV-Associated TB,” above). Moreover, HIV-infected patients with sputum culture–positive, AFB-positive TB may present with a normal chest radiograph. The Xpert MTB/RIF assay is the preferred rapid diagnostic test in this population of patients because of simplicity of use and a sensitivity of ~60% among AFB-negative culture-positive cases and of 97% among AFB-positive cases. With the advent of ART, the occurrence of disseminated M. avium complex disease that can be confused with TB has become much less common.
SEROLOGIC AND OTHER DIAGNOSTIC TESTS FOR ACTIVE TB
A number of serologic tests based on detection of antibodies to a variety of mycobacterial antigens are marketed in developing countries but not in the United States. Careful independent assessments of these tests suggest that they are not useful as diagnostic aids—especially in persons with a low probability of TB—because of their low sensitivity and specificity and their poor reproducibility. After a rigorous evaluation of the tests, the WHO issued a “negative” recommendation in 2011 in order to prevent their abuse in the private sector of many resource-limited countries. Various methods aimed at detection of mycobacterial antigens in diagnostic specimens are being investigated but are limited at present by low sensitivity. Determinations of ADA and IFN-γ levels in pleural fluid may be useful adjunctive tests in the diagnosis of pleural TB; their utility in the diagnosis of other forms of extrapulmonary TB (e.g., pericardial, peritoneal, and meningeal) is less clear.
DIAGNOSIS OF LATENT M. TUBERCULOSIS INFECTION
In 1891, Robert Koch discovered that components of M. tuberculosis in a concentrated liquid culture medium, subsequently named “old tuberculin,” were capable of eliciting a skin reaction when injected subcutaneously into patients with TB. In 1932, Seibert and Munday purified this product by ammonium sulfate precipitation to produce an active protein fraction known as tuberculin purified protein derivative (PPD). In 1941, PPD-S, developed by Seibert and Glenn, was chosen as the international standard. Later, the WHO and UNICEF sponsored large-scale production of a master batch of PPD (RT23) and made it available for general use. The greatest limitation of PPD is its lack of mycobacterial species specificity, a property due to the large number of proteins in this product that are highly conserved in the various species. In addition, subjectivity of the skin-reaction interpretation, deterioration of the product, and batch-to-batch variations limit the usefulness of PPD.
The skin test with tuberculin PPD (TST) is most widely used in screening for LTBI. It probably measures the response to antigenic stimulation by T cells that reside in the skin rather than the response of recirculating memory T cells. The test is of limited value in the diagnosis of active TB because of its relatively low sensitivity and specificity and its inability to discriminate between LTBI and active disease. False-negative reactions are common in immunosuppressed patients and in those with overwhelming TB. False-positive reactions may be caused by infections with nontuberculous mycobacteria (Chap. 204) and by BCG vaccination. Repeated TST can produce larger reaction sizes due to either boosting or true conversion. The “boosting phenomenon” is a spurious TST conversion resulting from boosting of reactivity on subsequent TST 1–5 weeks after the initial test. Distinguishing boosting from true conversion is difficult yet important and can be based on clinical and epidemiologic considerations. For instance, true conversions are likely after BCG vaccination in a previously TST-negative person or in a close contact of an infectious patient.
Two in vitro assays that measure T cell release of IFN-γ in response to stimulation with the highly TB-specific antigens ESAT-6 and CFP-10 are available. The T-SPOT¯.TB test (Oxford Immunotec, Oxford, United Kingdom) is an enzyme-linked immunospot (ELISpot) assay, and the QuantiFERON¯-TB Gold test (Qiagen GmbH, Hilden, Germany) is a whole-blood enzyme-linked immunosorbent assay (ELISA) for measurement of IFN-γ. The QuantiFERON¯-TB Gold In-Tube assay, which facilitates blood collection and initial incubation, also contains another specific antigen, TB7.7. These tests likely measure the response of recirculating memory T cells—normally part of a reservoir in the spleen, bone marrow, and lymph nodes—to persisting bacilli producing antigenic signals.
In settings or population groups with low TB and HIV burdens, IFN-γ release assays (IGRAs) have previously been reported to be more specific than the TST as a result of less cross-reactivity due to BCG vaccination and sensitization by nontuberculous mycobacteria. Recent studies, however, suggest that IGRAs may not perform well in serial testing (e.g., among health care workers) and that interpretation of test results is dependent on cutoff values used to define positivity. Potential advantages of IGRAs include logistical convenience, the need for fewer patient visits to complete testing, and the avoidance of somewhat subjective measurements such as skin induration. However, IGRAs require that blood be drawn from the individual and then delivered to the laboratory in a timely fashion. IGRAs also require that testing be performed in a laboratory setting. These requirements pose challenges similar to those faced with the TST, including cold-chain requirements and batch-to-batch variations. Because of higher specificity and other potential advantages, IGRAs have usually replaced the TST for LTBI diagnosis in low-incidence, high-income settings. However, in high-incidence TB and HIV settings and population groups, there is limited and inconclusive evidence about the performance and usefulness of IGRAs. In view of higher costs and increased technical requirements, the WHO does not recommend the replacement of the TST by IGRAs in low- and middle-income countries.
A number of national guidelines on the use of IGRAs for LTBI testing have been issued. In the United States, an IGRA is preferred to the TST for most persons over the age of 5 years who are being screened for LTBI. However, for those at high risk of progression to active TB (e.g., HIV-infected persons), either test—or, to optimize sensitivity, both tests—may be used. Because of the paucity of data on the use of IGRAs in children, the TST is preferred for LTBI testing of children under age 5. In Canada and some European countries, a two-step approach for those with positive TSTs—i.e., initial TST followed by an IGRA—is recommended. However, a TST may boost an IGRA response if the interval between the two tests exceeds 3 days. Similar to the TST, current IGRAs have only modest predictive value for incident active TB, are not useful in identifying patients with the highest risk of progression toward disease, and cannot be used for diagnosis of active TB.
The two aims of TB treatment are (1) to prevent morbidity and death by curing TB while preventing the emergence of drug resistance and (2) to interrupt transmission by rendering patients noninfectious. Chemotherapy for TB became possible with the discovery of streptomycin in 1943. Randomized clinical trials clearly indicated that the administration of streptomycin to patients with chronic TB reduced mortality rates and led to cure in the majority of cases. However, monotherapy with streptomycin eventually was associated with the development of resistance to this drug and the resulting failure of treatment. With the introduction into clinical practice of para-aminosalicylic acid (PAS) and isoniazid, it became axiomatic in the early 1950s that cure of TB required the concomitant administration of at least two agents to which the organism was susceptible. Furthermore, early clinical trials demonstrated that a long period of treatment—i.e., 12–24 months—was required to prevent recurrence. The introduction of rifampin (rifampicin) in the early 1970s heralded the era of effective short-course chemotherapy, with a treatment duration of <12 months. The discovery that pyrazinamide, which was first used in the 1950s, augmented the potency of isoniazid/rifampin regimens led to the use of a 6-month course of this triple-drug regimen as standard therapy. DRUGS
Four major drugs are considered first-line agents for the treatment of TB: isoniazid, rifampin, pyrazinamide, and ethambutol (Table 202-2). These drugs are well absorbed after oral administration, with peak serum levels at 2–4 h and nearly complete elimination within 24 h. These agents are recommended on the basis of their bactericidal activity (i.e., their ability to rapidly reduce the number of viable organisms and render patients noninfectious), their sterilizing activity (i.e., their ability to kill all bacilli and thus sterilize the affected tissues, measured in terms of the ability to prevent relapses), and their low rate of induction of drug resistance by selection of mutant bacilli. Two additional rifamycins, rifapentine and rifabutin, are also available in the United States; however, the level of cross-resistance with rifampin is high. For a detailed discussion of the drugs used for the treatment of TB, see Chap. 205e.
Because of a lower degree of efficacy and a higher degree of intolerability and toxicity, six classes of second-line drugs are generally used only for the treatment of patients with TB resistant to first-line drugs: (1) the fluoroquinolone antibiotics; (2) the injectable aminoglycosides kanamycin, amikacin, and streptomycin; (3) the injectable polypeptide capreomycin; and the oral agents (4) ethionamide and prothionamide, (5) cycloserine and terizidone (therizidone), and (6) PAS. Streptomycin, formerly a first-line agent, is now rarely used for drug-resistant TB because resistance levels worldwide are high and it is more toxic than the other drugs in the same class; however, the level of cross-resistance with the other injectables is low. Of the quinolones, later-generation agents such as levofloxacin and moxifloxacin are preferred. Gatifloxacin (no longer marketed in several countries, including the United States, because of previously observed dysglycemia) has recently been tested in a 4-month regimen that produced no detectable major side effects; thus, this drug could be reconsidered as a good alternative. Other drugs (referred to by the WHO as “group 5”) whose efficacy is not clearly defined are used in the treatment of patients with TB resistant to most of the first- and second-line agents; these drugs include clofazimine, linezolid, amoxicillin/clavulanic acid, clarithromycin, and carbapenems such as imipenem/cilastatin and meropenem. Today amithiozone (thiacetazone) is used very rarely because it has been associated with severe and at times fatal skin reactions among HIV-infected patients. Two novel drugs belonging to two new antibiotic classes—the diarylquinoline bedaquiline and the nitroimidazole delamanid—have recently been approved for use in severe cases of MDR-TB by stringent regulatory authorities (the U.S. Food and Drug Administration [FDA] and the European Medicine Agency [EMA] in the case of bedaquiline; the EMA and the Pharmaceuticals and Medical Devices Agency of Japan in the case of delamanid). REGIMENS
Standard short-course regimens are divided into an initial, or bactericidal, phase and a continuation, or sterilizing, phase. During the initial phase, the majority of the tubercle bacilli are killed, symptoms resolve, and usually the patient becomes noninfectious. The continuation phase is required to eliminate persisting mycobacteria and prevent relapse. The treatment regimen of choice for virtually all forms of drug-susceptible TB in adults consists of a 2-month initial (or intensive) phase of isoniazid, rifampin, pyrazinamide, and ethambutol followed by a 4-month continuation phase of isoniazid and rifampin (Table 202-3). This regimen can cure TB in more than 90% of patients. In children, most forms of TB in the absence of HIV infection or suspected isoniazid resistance can be safely treated without ethambutol in the intensive phase. Treatment should be given daily throughout the course. However, daily treatment during the intensive phase and intermittently (three times weekly) during the continuation phase is an alternative for patients who can be directly supervised and properly supported. A fully supervised, three-times-weekly regimen throughout the course also can be offered in the absence of HIV infection, although the risk of acquired drug resistance is higher than that among patients treated daily for the full course. In addition, if the infecting strain is resistant to isoniazid, the risks of both acquired resistance and treatment failure are higher with three-times-weekly intensive therapy than with daily treatment in the intensive phase. HIV-infected patients should always receive their initial-phase regimen daily (see below). A continuation phase of once-weekly rifapentine and isoniazid has been shown to be equally effective for HIV-seronegative patients with noncavitary pulmonary TB who have negative sputum cultures at 2 months. Patients with cavitary pulmonary TB and delayed sputum-culture conversion (i.e., those who remain culture-positive at 2 months) should be tested immediately for drug-resistant TB, and a change of regimen should be considered. To prevent isoniazid-related neuropathy, pyridoxine (10–25 mg/d) should be added to the regimen given to persons at high risk of vitamin B6 deficiency (e.g., alcoholics; malnourished persons; pregnant and lactating women; and patients with conditions such as chronic renal failure, diabetes, and HIV infection, which are also associated with neuropathy). A full course of therapy (completion of treatment) is defined more accurately by the total number of doses taken than by the duration of treatment, although the course should not include interruptions of longer than 4 weeks. Specific recommendations on the required number of doses for each of the various treatment regimens have been published jointly by the American Thoracic Society, the Infectious Diseases Society of America, and the CDC. In some developing countries where the ability to ensure adherence to treatment is limited, a continuation-phase regimen of daily isoniazid and ethambutol for 6 months was used in the past. However, this regimen is associated with a higher rate of relapse, failure, and death, especially among HIV-infected patients, and is no longer recommended by the WHO.
Lack of adherence to treatment is recognized worldwide as the most important impediment to cure. Moreover, the tubercle bacilli infecting patients who do not fully adhere to the prescribed regimen are likely to become drug resistant. Both patient- and provider-related factors may affect adherence. Patient-related factors include a lack of belief that the illness is significant and/or that treatment will have a beneficial effect; the existence of concomitant medical conditions (notably alcohol or substance abuse); lack of social support; fear of stigma and discrimination associated with TB; and poverty, with attendant joblessness and homelessness. Provider-related factors that may promote adherence include the support, education, and encouragement of patients and the offering of convenient clinic hours. In addition to specific measures promoting adherence, two other strategic approaches are used: direct supervision of treatment with support to the patient, consisting of incentives and enablers such as meals, travel vouchers, cash transfers, and grants to replace income loss; and provision of fixed-drug-combination products that reduce the number of tablets the patient needs to swallow. Because it is difficult to predict which patients will adhere to the recommended treatment for a disease that has important public as well as individual health implications, all patients should have their therapy directly supervised, especially during the initial phase, with proper social support including education, psychosocial counseling, and material sustainment. In an increasing number of countries, personnel to supervise therapy are usually available through TB control programs of local public health departments and from members of the community who are accepted by the patient to undertake that role and who have been properly educated by health workers. Direct supervision with patient support usually increases the proportion of patients completing treatment in all settings and greatly lessens the chances of failure, relapse, and acquired drug resistance. Fixed-drug-combination products (e.g., isoniazid/rifampin, isoniazid/rifampin/pyrazinamide, and isoniazid/rifampin/pyrazinamide/ethambutol) are available and are strongly recommended as a means of minimizing the likelihood of prescription error and of the development of drug resistance as the result of monotherapy. In some formulations of these combination products, the bioavailability of rifampin has been found to be substandard. Stringent regulatory authorities ensure that combination products are of good quality; however, this type of quality assurance is not always operative in low-income countries. Alternative regimens for patients who exhibit drug intolerance or adverse reactions are listed in Table 202-3. However, severe side effects prompting discontinuation of any of the first-line drugs and use of these alternative regimens are uncommon. The fluoroquinolones moxifloxacin and gatifloxacin have been tested as 4-month treatment-shortening regimens for drug-susceptible TB. Recently published results from these clinical trials failed to show that a 4-month regimen substituting gatifloxacin for ethambutol or moxifloxacin for either ethambutol or isoniazid is noninferior to the standard 6-month regimen. Thus, currently there is no 4-month regimen available for TB treatment. MONITORING TREATMENT RESPONSE AND DRUG TOXICITY
Bacteriologic evaluation through culture and/or smear microscopy is essential in monitoring the response to treatment for TB. In addition, the patient's weight should be monitored regularly and the drug dosage adjusted with any significant weight change. Patients with pulmonary disease should have their sputum examined monthly until cultures become negative to allow early detection of treatment failure. With the recommended regimen, more than 80% of patients will have negative sputum cultures at the end of the second month of treatment. By the end of the third month, the sputum of virtually all patients should be culture negative. In some patients, especially those with extensive cavitary disease and large numbers of organisms, AFB smear conversion may lag behind culture conversion. This phenomenon is presumably due to the expectoration and microscopic visualization of dead bacilli. As noted above, patients with cavitary disease in whom sputum culture conversion does not occur by 2 months require immediate testing for drug resistance. When a patient's sputum cultures remain positive at ≥3 months, treatment failure and drug resistance or poor adherence to the regimen are likely, and testing of drug resistance should guide the choice of the best treatment option (see below). A sputum specimen should be collected by the end of treatment to document cure. If mycobacterial cultures are not practical, then monitoring by AFB smear examination should be undertaken at 2, 5, and 6 months. Smears that are positive after 3 months of treatment when the patient is known to be adherent are indicative of treatment failure and possible drug resistance. Therefore, if not done at the start of treatment, drug susceptibility testing is mandatory at this stage. Bacteriologic monitoring of patients with extrapulmonary TB is more difficult and often is not feasible. In these cases, the response to treatment must be assessed clinically and radiographically.
Monitoring of the response during chemotherapy by nucleic acid amplification technology has not been shown to be suitable. Thus Xpert MTB/RIF should not be used to monitor treatment. Likewise, serial chest radiographs are not recommended because radiographic changes may lag behind bacteriologic response and are not highly sensitive. After the completion of treatment, neither sputum examination nor chest radiography is recommended for routine follow-up purposes. However, a chest radiograph obtained at the end of treatment may be useful for comparative purposes should the patient develop symptoms of recurrent TB months or years later. Patients should be instructed to report promptly for medical assessment if they develop any such symptoms. In addition, an end-of-treatment chest radiograph may reveal earlier the post-TB complications described above.
During treatment, patients should be monitored for drug toxicity. The most common adverse reaction of significance is hepatitis. Patients should be carefully educated about the signs and symptoms of drug-induced hepatitis (e.g., dark urine, loss of appetite) and should be instructed to discontinue treatment promptly and see their health care provider should these symptoms occur. Although biochemical monitoring is not routinely recommended, all adult patients should undergo baseline assessment of liver function (e.g., measurement of serum levels of hepatic aminotransferases and bilirubin). Older patients, those with concomitant diseases, those with a history of hepatic disease (especially hepatitis C), and those using alcohol daily should be monitored especially closely (i.e., monthly), with repeated measurements of aminotransferases, during the initial phase of treatment. Up to 20% of patients have small increases in aspartate aminotransferase (up to three times the upper limit of normal) that are not accompanied by symptoms and are of no consequence. For patients with symptomatic hepatitis and those with marked (five- to sixfold) elevations in serum levels of aspartate aminotransferase, treatment should be stopped and drugs reintroduced one at a time after liver function has returned to normal. Hypersensitivity reactions usually require the discontinuation of all drugs and rechallenge to determine which agent is the culprit. Because of the variety of regimens available, it usually is not necessary—although it is possible—to desensitize patients. Hyperuricemia and arthralgia caused by pyrazinamide can usually be managed by the administration of acetylsalicylic acid; however, pyrazinamide treatment should be stopped if the patient develops gouty arthritis. Individuals who develop autoimmune thrombocytopenia secondary to rifampin therapy should not receive the drug thereafter. Similarly, the occurrence of optic neuritis with ethambutol is an indication for permanent discontinuation of this drug. Other common manifestations of drug intolerance, such as pruritus and gastrointestinal upset, can generally be managed without the interruption of therapy. TREATMENT FAILURE AND RELAPSE
As stated above, treatment failure should be suspected when a patient’s sputum smears and/or cultures remain positive after 3 months of treatment. In the management of such patients, it is imperative that the current isolate be urgently tested for susceptibility to first- and second-line agents. Initial molecular testing for rifampin resistance should be done if the technology is available. When the results of susceptibility testing are based on molecular methods and are expected to become available within a few days, changes in the regimen can be postponed until that time. However, if the patient’s clinical condition is deteriorating, an earlier change in regimen may be indicated. A cardinal rule in the latter situation is always to add more than one drug at a time to a failing regimen: at least two and preferably three drugs that have never been used and to which the bacilli are likely to be susceptible should be added. The patient may continue to take isoniazid and rifampin along with these new agents pending the results of susceptibility tests.
Patients who experience a recurrence after apparently successful treatment (relapse) are less likely to harbor drug-resistant strains (see below) than are patients in whom treatment has failed. Acquired resistance is uncommon among strains from patients in whom relapse follows the completion of a standard short-course regimen. However, pending the results of susceptibility testing, it is prudent to begin the treatment of all patients whose infections have relapsed with a standard regimen containing all four first-line drugs plus streptomycin. In less affluent countries and other settings where facilities for culture and drug susceptibility testing are not yet routinely available and where the prevalence of MDR-TB is low, the WHO recommends that a standard regimen with all four first-line drugs plus streptomycin be used in all instances of relapse and treatment default. Patients with treatment failure and those relapsing or defaulting with a high likelihood of MDR-TB should receive a regimen that includes second-line agents and is based on their history of anti-TB treatment and the drug resistance patterns in the population (Table 202-3). Once drug susceptibility test results are available, the regimen can be adjusted accordingly. DRUG-RESISTANT TB
Strains of M. tuberculosis resistant to individual drugs arise by spontaneous point mutations in the mycobacterial genome that occur at low but predictable rates (10–7–10–10 for the key drugs). Resistance to rifampin is associated with mutations in the rpoB gene in 95% of cases; that to isoniazid with mutations mainly in the katG (50–95% of cases) and inhA (up to 45%) genes; that to pyrazinamide in the pncA gene (up to 98%); that to ethambutol in the embB gene (50–65%); that to the fluoroquinolones in the gyrA–gyrB genes (75–95%); and that to the aminoglycosides mainly in the rrs gene (up to 80%). Because there is no cross-resistance among the commonly used drugs, the probability that a strain will be resistant to two drugs is the product of the probabilities of resistance to each drug and thus is low. The development of drug-resistant TB is almost invariably the result of monotherapy—i.e., the failure of the health care provider to prescribe at least two drugs to which tubercle bacilli are susceptible or of the patient to take properly prescribed therapy. In addition, the use of drugs of substandard quality may cause the emergence of drug resistance. Drug-resistant TB may be either primary or acquired. Primary drug resistance is that which develops in a patient infected from the start by a drug-resistant strain. Acquired resistance is that which develops during treatment with an inappropriate regimen. In North America, Western Europe, most of Latin America, and the Persian Gulf States, rates of primary resistance are generally low and isoniazid resistance is most common. In the United States, although rates of primary isoniazid resistance have been stable at ~7–8%, the rate of primary MDR-TB has declined from 2.5% in 1993 to 1% since 2000. As described above, MDR-TB is an increasingly serious problem in some regions, especially in the states of the former Soviet Union and some countries of Asia (Fig. 202-11). Even more serious is the recently described occurrence of XDR-TB due to MDR strains that are also resistant to any fluoroquinolones and to any of three second-line injectable agents (amikacin, kanamycin, and capreomycin). Creation of drug-resistant TB can be prevented by adherence to the principles of sound treatment: inclusion of at least two qualityassured, bactericidal drugs to which the organism is susceptible; use of fixed-drug-combination products; supervision of treatment with patient support; and verification that patients complete the prescribed course. Transmission of drug-resistant strains can be prevented by implementation of respiratory infection-control measures (see below).
Although the 6-month regimen described in Table 202-3 is generally effective for patients with initial isoniazid-resistant disease, it is prudent to include at least ethambutol and possibly pyrazinamide for the full 6 months and to consider extending the treatment course to 9 months. In such cases, isoniazid probably does not contribute to a successful outcome and could be omitted. In case of documented resistance to both isoniazid and ethambutol, a 9- to 12-month regimen of rifampin, pyrazinamide, and a fluoroquinolone can be used. Any patients whose isolates exhibit resistance to rifampin should be managed as if they had MDR-TB (see below), with the addition of isoniazid if susceptibility to this agent is confirmed via rapid testing or is presumed. MDR-TB, in which bacilli are resistant to (at least) isoniazid and rifampin, is more difficult to manage than is disease caused by drug-susceptible organisms because these two bactericidal drugs are the most potent agents available and because associated resistance to other first-line drugs as well (e.g., ethambutol) is not uncommon. For treatment of MDR-TB, the WHO recommends that in most patients five drugs be used in the initial phase of at least 8 months: a later-generation fluoroquinolone, an injectable agent (the aminoglycosides amikacin or kanamycin or the polypeptide capreomycin), ethionamide (or prothionamide), either cycloserine or PAS, and pyrazinamide. Ethambutol can be added (Table 202-3). Although the optimal duration of treatment is not known, a course of at least 20 months is recommended for previously untreated patients, including the initial phase with an injectable agent, which is usually discontinued at 4 months after culture conversion.
In late 2012, the FDA granted accelerated approval of bedaquiline, a diarylquinoline antibiotic. This new drug, when given for the first 24 weeks (400 mg daily for 2 weeks followed by 200 mg thrice weekly for 22 weeks), has been shown to increase the efficacy of the WHO standard regimen for MDR-TB with faster sputum conversion. Bedaquiline should be used with caution in people >65 years of age and in HIV-infected patients; its use is not advised in children and pregnant women. In early 2014, the European Medical Agency granted accelerated approval of another new agent, the nitroimidazole compound delamanid. Data from a phase 2B clinical trial in which delamanid was added to the WHO-recommended standard MDR-TB regimen have shown increased culture conversion at 2 months. Pending phase 3 trial results and in view of potential side effects of both new drugs (including QT interval prolongation in both cases and hepatotoxicity in the case of bedaquiline), the WHO recommends limiting the use of bedaquiline and delamanid to cases of MDR-TB when an effective WHO-recommended standard MDR-TB regimen cannot be designed because of known resistance, intolerance, or nonavailability of any second-line drugs in the regimen. Patients treated with bedaquiline or delamanid should be counseled, should give informed consent, and should be closely monitored during treatment. In particular, patients with cardiac anomalies such as prolonged QT interval or a history of ventricular arrhythmias should not be given these drugs. Currently, there is no information about simultaneous use of these two agents; therefore, combining them is not recommended.
Finally, a shorter (9-month) regimen consisting of gatifloxacin or moxifloxacin, clofazimine, ethambutol, and pyrazinamide given throughout the treatment period and supplemented by prothionamide, kanamycin, and high-dose isoniazid during an intensive phase of at least 4 months is reportedly effective for MDR-TB in certain settings. Further investigations are necessary to elucidate the role of this shorter regimen in MDR-TB treatment.
Patients with XDR-TB have fewer treatment options and a much poorer prognosis. However, observational studies have shown that aggressive management of cases comprising early drug-susceptibility testing, rational combination of at least five drugs, readjustment of the regimen, strict directly observed therapy, monthly bacteriologic monitoring, and intensive patient support may result in cure and avert death. Table 202-4 summarizes the management of patients with XDR-TB. Some recently published studies regarding the use of linezolid in patients with XDR-TB suggest that, although it carries a high level of toxicity, this drug increases culture conversion.
For patients with localized disease and sufficient pulmonary reserve, lobectomy or pneumonectomy may be considered. Because the management of patients with MDR- and XDR-TB is complicated by both social and medical factors, care of these patients is ideally provided in specialized centers or, in their absence, in the context of programs with adequate resources and capacity, including community support. HIV-ASSOCIATED TB
Several observational studies and randomized controlled trials have shown that treatment of HIV-associated TB with anti-TB drugs and simultaneous use of ART are associated with significant reductions in mortality risk and AIDS-related events. Evidence from randomized controlled trials shows that early initiation of ART during anti-TB treatment is associated with a 34–68% reduction in mortality rates, with especially good results in patients with CD4+ T cell counts of <50/μL. Therefore, the main aim in the management of HIV-associated TB is to initiate anti-TB treatment and to immediately consider initiating or continuing ART. All HIV-infected TB patients, regardless of CD4+ T cell count, are candidates for ART, which optimally is initiated as soon as possible after the diagnosis of TB and within the first 8 weeks of anti-TB therapy. However, ART should be started within the first 2 weeks of TB treatment for patients with CD4+ T cell counts of <50/μL. In general, the standard 6-month daily regimen is equally efficacious in HIV-negative and HIV-positive patients for treatment of drug-susceptible TB. As for any other adult living with HIV (Chap. 226), first-line ART for TB patients should consist of two nucleoside reverse transcriptase inhibitors (NRTIs) plus a nonnucleoside reverse transcriptase inhibitor (NNRTI). Although TB treatment modalities are similar to those in HIV-negative patients, adverse drug effects may be more pronounced in HIV-infected patients. In this regard, three important considerations are relevant: an increased frequency of paradoxical reactions, interactions between ART components and rifamycins, and development of rifampin monoresistance with intermittent treatment. IRIS—i.e., the exacerbation of symptoms and signs of TB—has been described above. Rifampin, a potent inducer of enzymes of the cytochrome P450 system, lowers serum levels of many HIV protease inhibitors and some NNRTIs—essential drugs used in ART. In such cases, rifabutin, which has much less enzyme-inducing activity, has been used in place of rifampin. However, dosage adjustments for rifabutin and protease inhibitors are still being assessed. Several clinical trials have found that patients with HIV-associated TB whose degree of immunosuppression is advanced (e.g., CD4+ T cell counts of <100/μL) are prone to treatment failure and relapse with rifampin-resistant organisms when treated with “highly intermittent” (i.e., once- or twice-weekly) rifamycin-containing regimens. Consequently, it is recommended that all TB patients who are infected with HIV receive a rifampin-containing regimen on a daily basis. Because recommendations are frequently updated, consultation of the following websites is advised: www.who.int/hiv, www.who.int/tb, www.cdc.gov/hiv, and www.cdc.gov/tb. SPECIAL CLINICAL SITUATIONS
Although comparative clinical trials of treatment for extrapulmonary TB are limited, the available evidence indicates that most forms of disease can be treated with the 6-month regimen recommended for patients with pulmonary disease. The WHO and the American Academy of Pediatrics recommend that children with bone and joint TB, tuberculous meningitis, or miliary TB receive up to 12 months of treatment. Treatment for TB may be complicated by underlying medical problems that require special consideration. As a rule, patients with chronic renal failure should not receive aminoglycosides and should receive ethambutol only if serum drug levels can be monitored. Isoniazid, rifampin, and pyrazinamide may be given in the usual doses in cases of mild to moderate renal failure, but the dosages of isoniazid and pyrazinamide should be reduced for all patients with severe renal failure except those undergoing hemodialysis. Patients with hepatic disease pose a special problem because of the hepatotoxicity of isoniazid, rifampin, and pyrazinamide. Patients with severe hepatic disease may be treated with ethambutol, streptomycin, and possibly another drug (e.g., a fluoroquinolone); if required, isoniazid and rifampin may be administered under close supervision. The use of pyrazinamide by patients with liver failure should be avoided. Silicotuberculosis necessitates the extension of therapy by at least 2 months.
The regimen of choice for pregnant women (Table 202-3) is 9 months of treatment with isoniazid and rifampin supplemented by ethambutol for the first 2 months. Although the WHO has recommended routine use of pyrazinamide for pregnant women, this drug has not been recommended in the United States because of insufficient data documenting its safety in pregnancy. Streptomycin is contraindicated because it is known to cause eighth-cranial-nerve damage in the fetus. Treatment for TB is not a contraindication to breast-feeding; most of the drugs administered will be present in small quantities in breast milk, albeit at concentrations far too low to provide any therapeutic or prophylactic benefit to the child.
Medical consultation on difficult-to-manage cases is provided by the U.S. CDC Regional Training and Medical Consultation Centers (www.cdc.gov/tb/education/rtmc/).
TABLE 202-2Recommended Dosagea for Initial Treatment of Tuberculosis in Adultsb ||Download (.pdf) TABLE 202-2Recommended Dosagea for Initial Treatment of Tuberculosis in Adultsb
| ||Dosage |
|Drug ||Daily Dose ||Thrice-Weekly Dose |
|Isoniazid ||5 mg/kg, max 300 mg ||10 mg/kg, max 900 mg |
|Rifampin ||10 mg/kg, max 600 mg ||10 mg/kg, max 600 mg |
|Pyrazinamide ||25 mg/kg, max 2 g ||35 mg/kg, max 3 g |
|Ethambutolc ||15 mg/kg ||30 mg/kg |
TABLE 202-3Recommended Antituberculosis Treatment Regimens ||Download (.pdf) TABLE 202-3Recommended Antituberculosis Treatment Regimens
TABLE 202-4Management Guidelines for Patients with Documented or Strongly Suspected Extensively Drug-Resistant Tuberculosis (XDR-TB) ||Download (.pdf) TABLE 202-4Management Guidelines for Patients with Documented or Strongly Suspected Extensively Drug-Resistant Tuberculosis (XDR-TB)
Use pyrazinamide and any first-line oral agents that may be effective.
Use an injectable agent to which the strain is susceptible, and consider an extended duration of use (12 months or possibly the whole treatment period). If the strain is resistant to all injectable agents, use of one that the patient has not previously received is recommended.a
Use a later-generation fluoroquinolone, such as moxifloxacin, high-dose levofloxacin, or possibly gatifloxacin.b
Use all second-line oral bacteriostatic agents (para-aminosalicylic acid, cycloserine, and ethionamide or prothionamide) that have not been used extensively in a previous regimen or any such agents that are likely to be effective.
Add bedaquiline or delamanid and one or more of the following drugsc: clofazimine, linezolid, amoxicillin/clavulanic acid, clarithromycin, and carbapenems such as imipenem/cilastatin and meropenem.
The simultaneous use of bedaquiline and delamanid is not recommended at the moment in view of the current lack of information on the potential of adverse reactions when these drugs are administered together.
Consider treatment with high-dose isoniazid if low-level resistance to this drug is documented.
Consider adjuvant surgery if there is localized disease.
Enforce strong infection-control measures.
Implement strict directly observed therapy and full adherence support as well as comprehensive bacteriologic and clinical monitoring.
Percentage of new tuberculosis cases with multidrug resistance in all countries surveyed by the World Health Organization (WHO) Global Drug Resistance Surveillance Project during 1994–2013. (See disclaimer in Fig. 202-2. Courtesy of the Global TB Programme, WHO; with permission.)
The best way to prevent TB is to diagnose and isolate infectious cases rapidly and to administer appropriate treatment until patients are rendered noninfectious (usually 2–4 weeks after the start of proper treatment) and the disease is cured. Additional strategies include BCG vaccination and treatment of persons with LTBI who are at high risk of developing active disease.
BCG was derived from an attenuated strain of M. bovis and was first administered to humans in 1921. Many BCG vaccines are available worldwide; all are derived from the original strain, but the vaccines vary in efficacy, ranging from 80% to nil in randomized, placebo-controlled trials. A similar range of efficacy was found in recent observational studies (case–control, historic cohort, and cross-sectional) in areas where infants are vaccinated at birth. These studies and a meta-analysis also found higher rates of efficacy in the protection of infants and young children from serious disseminated forms of childhood TB, such as tuberculous meningitis and miliary TB. BCG vaccine is safe and rarely causes serious complications. The local tissue response begins 2–3 weeks after vaccination, with scar formation and healing within 3 months. Side effects—most commonly, ulceration at the vaccination site and regional lymphadenitis—occur in 1–10% of vaccinated persons. Some vaccine strains have caused osteomyelitis in ~1 case per million doses administered. Disseminated BCG infection (“BCGitis”) and death have occurred in 1–10 cases per 10 million doses administered, although this problem is restricted almost exclusively to persons with impaired immunity, such as children with severe combined immunodeficiency syndrome or adults with HIV infection. BCG vaccination induces TST reactivity, which tends to wane with time. The presence or size of TST reactions after vaccination does not predict the degree of protection afforded.
BCG vaccine is recommended for routine use at birth in countries with high TB prevalence. However, because of the low risk of transmission of TB in the United States and other high-income countries, the unreliable protection afforded by BCG, and its impact on the TST, the vaccine is not recommended for general use. HIV-infected adults and children should not receive BCG vaccine. Moreover, infants whose HIV status is unknown but who have signs and symptoms consistent with HIV infection or who are born to HIV-infected mothers should not receive BCG.
Over the past decade, renewed research and development efforts have been made toward a new TB vaccine. In mid-2014, 16 candidates were in clinical trials and 12 were being field tested. The first new vaccine, for which results of a clinical trial became available in early 2013, is MVA85A/AERAS-485; unfortunately, this viral-vectored vaccine did not show clinical benefit as a booster to BCG.
TREATMENT Latent Tuberculosis Infection
It is estimated that about 2 billion people, or nearly one-third of the human population, have been infected with M. tuberculosis. Although only a small fraction of these infections will progress toward active disease, new active cases will continue to emerge from this pool of “latently” infected individuals. Unfortunately, there is no diagnostic test at present that can predict which individuals with LTBI will develop active TB. Treatment of selected persons with LTBI aims at preventing active disease. This intervention (also called preventive chemotherapy or chemoprophylaxis) is based on the results of a large number of randomized, placebo-controlled clinical trials demonstrating that a 6- to 9-month course of isoniazid reduces the risk of active TB in infected people by up to 90%. Analysis of available data indicates that the optimal duration of treatment is ~9 months. In the absence of reinfection, the protective effect is believed to be lifelong. Clinical trials have shown that isoniazid reduces rates of TB among TST-positive persons with HIV infection. Studies in HIV-infected patients have also demonstrated the effectiveness of shorter courses of rifampin-based treatment.
Candidates for treatment of LTBI are listed in Table 202-5. They can be identified by TST or IGRA of persons in defined high-risk groups. For skin testing, 5 tuberculin units of polysorbate-stabilized PPD should be injected intradermally into the volar surface of the forearm (i.e., the Mantoux method). Multipuncture tests are not recommended. Reactions are read at 48–72 h as the transverse diameter (in millimeters) of induration; the diameter of erythema is not considered. In some persons, TST reactivity wanes with time but can be recalled by a second skin test administered ≥1 week after the first (i.e., two-step testing). For persons periodically undergoing the TST, such as health care workers and individuals admitted to long-term-care institutions, initial two-step testing may preclude subsequent misclassification of those who have boosted reactions as TST converters. The cutoff for a positive TST (and thus for treatment) is related both to the probability that the reaction represents true infection and to the likelihood that the individual, if truly infected, will develop TB. Table 202-5 suggests possible cutoff by risk group. Thus, positive reactions for persons with HIV infection, recent close contacts of infectious cases, organ transplant recipients, previously untreated persons whose chest radiograph shows fibrotic lesions consistent with old TB, and persons receiving drugs that suppress the immune system are defined as an area of induration ≥5 mm in diameter. A 10-mm cutoff is used to define positive reactions in most other at-risk persons. For persons with a very low risk of developing TB if infected, a cutoff of 15 mm is used. (Except for employment purposes where longitudinal screening is anticipated, the TST is not indicated for these low-risk persons.) A positive IGRA is based on the manufacturers’ recommendations; however, good clinical practice requires that epidemiologic and clinical factors also guide the decision to implement treatment for LTBI and that active TB be definitively excluded before the initiation of chemoprophylaxis.
Some TST- and IGRA-negative individuals are also candidates for treatment. Once an appropriate clinical evaluation has excluded active TB, infants and children who have come into contact with infectious cases should be treated for presumed LTBI. HIV-infected persons who have been exposed to an infectious TB patient should receive treatment regardless of the TST result. Any HIV-infected candidate for LTBI treatment must be screened carefully to exclude active TB, which would necessitate full treatment. The use of a clinical algorithm based on four symptoms (current cough, fever, weight loss, and night sweats) helps to define which HIV-infected person is a candidate for LTBI treatment. The absence of all four symptoms tends to exclude active TB. The presence of one of the four symptoms, on the other hand, warrants further investigation for active TB before treatment of LTBI is started. Although administering a TST is prudent, this test is not an absolute requirement—given the logistical challenges—among people living with HIV in high-TB-incidence and low-resource settings.
Before treatment of LTBI begins, it is mandatory to carefully exclude active TB. Several regimens can be used to treat LTBI. The most widely used is that based on isoniazid alone at a daily dose of 5 mg/kg (up to 300 mg/d) for 9 months. On the basis of cost-benefit analyses and concerns about feasibility, a 6-month period of treatment is currently recommended by the WHO, especially in highly TB-endemic countries. Isoniazid can be administered intermittently (twice weekly) at a dose of 15 mg/kg (up to 900 mg) but only as directly observed therapy. An alternative regimen for adults is 4 months of daily rifampin. A 3-month regimen of daily isoniazid and rifampin is used in some countries (e.g., the United Kingdom) for both adults and children who are known not to have HIV infection. A previously recommended regimen of 2 months of rifampin and pyrazinamide has been associated with serious or even fatal hepatotoxicity and now is generally not recommended. The rifampin-containing regimens should be considered for persons who are likely to have been infected with an isoniazid-resistant strain. A recent clinical trial showed that a regimen of isoniazid (900 mg) and rifapentine (900 mg) given once weekly for 12 weeks is as effective as the standard 9-month isoniazid regimen. This regimen was associated with higher treatment completion (82% vs 69%) and less hepatotoxicity (0.4% vs 2.7%) than isoniazid alone, although the rate of permanent discontinuation due to an adverse event was higher (4.9% vs 3.7%).
Currently, the isoniazid–rifapentine regimen is not recommended for children <2 years of age, people living with HIV infection who are receiving ART, or pregnant women. Rifampin and rifapentine are contraindicated in HIV-infected individuals receiving protease inhibitors and most NNRTIs. (Efavirenz is the safest agent in this class of antiretrovirals for simultaneous administration with a rifamycin.) Clinical trials to assess the efficacy of long-term isoniazid administration (i.e., for at least 3 years) among people living with HIV in high-TB-transmission settings have shown that this regimen can be more effective than 9 months of isoniazid and is therefore recommended under those circumstances. Isoniazid should not be given to persons with active liver disease. All persons at increased risk of hepatotoxicity (e.g., those abusing alcohol daily and those with a history of liver disease) should undergo baseline and then monthly assessment of liver function. All patients should be carefully educated about hepatitis and instructed to discontinue use of the drug immediately should any symptoms develop. Moreover, patients should be seen and questioned monthly during therapy about adverse reactions and should be given no more than a 1-month supply of drug at each visit. Treatment of LTBI among persons likely to have been infected by a multidrug-resistant strain is a challenge because no regimens have yet been tested in clinical trials. Close observation for early signs of disease is one option; consultation with a TB expert is advised.
It may be more difficult to ensure compliance when treating persons with LTBI than when treating those with active TB. If family members of active cases are being treated, compliance and monitoring may be easier. When feasible, supervised therapy may increase the likelihood of completion. As in active cases, the provision of incentives may also be helpful.
TABLE 202-5Tuberculin Reaction Size and Treatment of Latent Mycobacterium tuberculosis Infection ||Download (.pdf) TABLE 202-5Tuberculin Reaction Size and Treatment of Latent Mycobacterium tuberculosis Infection
|Risk Group ||Tuberculin Reaction Size, mm |
|HIV-infected persons ||≥5 |
|Recent contacts of a patient with TB ||≥5a |
|Organ transplant recipients ||≥5 |
|Persons with fibrotic lesions consistent with old TB on chest radiography ||≥5 |
|Persons who are immunosuppressed, e.g., due to the use of glucocorticoids or tumor necrosis factor α inhibitors ||≥5 |
|Persons with high-risk medical conditionsb ||≥5 |
|Recent immigrants (≤5 years) from high-prevalence countries ||≥10 |
|Injection drug users ||≥10 |
|Mycobacteriology laboratory personnel; residents and employees of high-risk congregate settingsc ||≥10 |
|Children <5 years of age; children and adolescents exposed to adults in high-risk categories ||≥10 |
|Low-risk personsd ||≥15 |
The highest priority in any TB control program is the prompt detection of cases and the provision of short-course chemotherapy to all TB patients under proper case-management conditions, including directly observed therapy. In addition, screening of high-risk groups, including immigrants from high-prevalence countries, migratory workers, prisoners, homeless individuals, substance abusers, and HIV-seropositive persons, is recommended. TST-positive high-risk persons should be treated for LTBI as described above. Contact investigation is an important component of efficient TB control. In the United States and other countries worldwide, a great deal of attention has been given to the transmission of TB (particularly in association with HIV infection) in institutional settings such as hospitals, homeless shelters, and prisons. Measures to limit such transmission include respiratory isolation of persons with suspected TB until they are proven to be noninfectious (at least by sputum AFB smear negativity), proper ventilation in rooms of patients with infectious TB, use of ultraviolet irradiation in areas of increased risk of TB transmission, and periodic screening of personnel who may come into contact with known or unsuspected cases of TB. In the past, radiographic surveys, especially those conducted with portable equipment and miniature films, were advocated for case finding. Today, however, the prevalence of TB in industrialized countries is sufficiently low that “mass miniature radiography” is not cost-effective.
In high-prevalence countries, most TB control programs have made remarkable progress in reducing morbidity and mortality since the mid-1990s by adopting and implementing the strategy promoted by the WHO. Between 2000 and 2013, 37 million lives were saved, and since 1995, 61 million TB cases have been successfully treated. The essential elements of good TB care and control (the DOTS strategy) are (1) political commitment with increased and sustained financing; (2) case detection through quality-assured bacteriology (starting with examination of sputum from patients with cough of >2–3 weeks’ duration); (3) administration of standardized short-course chemotherapy, with direct supervision and patient support; (4) an effective drug supply and management system; and (5) a monitoring and evaluation system, with impact measurement (including assessment of treatment outcomes—e.g., cure, completion of treatment without bacteriologic proof of cure, death, treatment failure, and default—in all cases registered and notified). In 2006, the WHO indicated that, although these essential elements remain the fundamental components of any control strategy, additional steps must be undertaken to reach the 2015 international TB control targets set within the United Nations Millennium Development Goals. Thus, a new “Stop TB Strategy” with six components has been promoted since 2006: (1) Pursue high-quality DOTS expansion and enhancement. (2) Address HIV-associated TB, MDR-TB, and the needs of poor and vulnerable populations. (3) Contribute to health system strengthening. (4) Engage all care providers. (5) Empower people with TB and their communities. (6) Enable and promote research. As part of the fourth component, evidence-based International Standards for Tuberculosis Care—focused on diagnosis, treatment, and public health responsibilities—have been introduced for wide adoption by medical and professional societies, academic institutions, and all practitioners worldwide (http://www.who.int/tb/publications/ISTC_3rdEd.pdf?ua=1).
Care and control of HIV-associated TB are particularly challenging in developing countries because existing interventions require collaboration between HIV/AIDS and TB programs as well as standard services. While TB programs must test every patient for HIV in order to provide access to trimethoprim-sulfamethoxazole prophylaxis against common infections and ART, HIV/AIDS programs must regularly screen persons living with HIV/AIDS for active TB, provide treatment for LTBI, and ensure infection control in settings where people living with HIV congregate.
Early and active case detection is considered an important intervention not only among persons living with HIV/AIDS but also among other vulnerable populations, as it reduces transmission in a community and provides early effective care. For TB control efforts to succeed and for elimination to become a realistic target, programs must optimize their performance and include additional interventions as described. Moreover, the approach to TB control and care needs to become holistic and engage beyond dedicated programs. Therefore, the WHO’s “End TB” strategy has been designed and builds on three pillars for the post-2015 era of increased efforts by governments and national programs worldwide: (1) integrated, patient-centered care and prevention; (2) bold policies and supportive systems; and (3) intensified research and innovation. The first pillar incorporates all technological innovations, such as early diagnostic approaches (including universal drug-susceptibility testing and systematic screening of identified, setting-specific, high-risk groups); well-designed treatment regimens for all forms of TB; proper management of HIV-associated TB and other comorbidities; and preventive treatment of persons at high risk. The second pillar is fundamental and is normally beyond the control of dedicated programs, relying on policies forged by the highest-level health and governmental authorities: availability of adequate and well-identified human and financial resources; engagement of civil society organizations and all relevant public and private providers to facilitate care and prevention of all patients; a policy of universal health coverage (which implies avoidance of catastrophic expenditures caused by TB among the poorest); regulatory frameworks for case notifications, vital registration, quality and rational use of medicines, and infection control; social protection mechanisms; poverty alleviation strategies; and interventions on the broader determinants of TB. Finally, the third pillar of the new strategy emphasizes intensification of engagement in research and development of new tools and interventions as well as optimization of implementation and rapid adoption of new tools in endemic countries. In the end, besides specific clinical care and control interventions as described in this chapter, elimination of TB ultimately will require control and attenuation of the multitude of risk factors (e.g., HIV, smoking, and diabetes) and socioeconomic determinants (e.g., extreme poverty, inadequate living conditions and bad housing, alcoholism, malnutrition, and indoor air pollution) with clearly implemented policies within the health sector and other sectors linked to human development and welfare.
The contributions of Richard J. O’Brien to this chapter in previous editions are gratefully acknowledged.