Source Case

Transmission of M. tuberculosis is a classic example of an airborne infection. In nearly all instances, tuberculous infection is acquired by inhalation of one or more tubercle bacilli contained in an airborne particle small enough (1 to 5 µm) to reach an alveolus. For a person with tuberculosis to be infectious, the organisms must have access to environmental air and be aerosolized. By and large, this means that only patients with pulmonary tuberculosis can be regarded as infectious. However, respirable particles containing M. tuberculosis may rarely be generated from other sources (e.g., irrigation of a tuberculous abscess). After aerosolized respiratory secretions are expelled from the nose or mouth, their water content evaporates rapidly, leaving only a small residue of solid matter—the droplet nucleus —which may include tubercle bacilli and which may remain suspended in the air for several hours. A single bacillus in a tiny droplet nucleus is more hazardous than several bacilli in larger airborne particles, which when inhaled deposit in airways rather than alveoli and which then are quickly removed by mucociliary clearance or killed.

Coughing is the most effective mechanism for generating aerosols that create droplet nuclei, but it is not the only mechanism. Forced expiratory maneuvers other than coughing—such as sneezing, yelling, singing, and loud talking—all involve, to a greater or lesser extent, the sudden acceleration of air required to disrupt a liquid surface or mucous strands, thereby aerosolizing particles. Thin, watery secretions are more easily fragmented into small respirable droplets than is more viscous mucus. In general, the greater the volume of respiratory secretions, the greater the number of potentially infectious droplets. Riley and coworkers demonstrated marked variability in the infectious potential of tuberculosis patients that, in part, could be related to the severity of coughing. One patient with exceptionally infectious tuberculosis not only had severe pulmonary tuberculosis but tuberculous laryngitis as well. It was calculated that this patient was as contagious as a child with measles is to other susceptible children. Similar studies by Escombe and coworkers, which described the infectiousness of HIV-infected patients with tuberculosis, reported that infectiousness was very heterogeneous. Of 97 patients (118 admissions) hospitalized in an HIV-tuberculosis ward, 10 caused 90% of the infections in guinea pigs exposed to air exhausted from the ward. Of the 10 infectious patients, 6 had MDR tuberculosis that was inadequately treated; in addition to having MDR tuberculosis, being sputum smear positive and being on suboptimal treatment both contributed to the degree of infectiousness. Likewise, in a study of household contacts of tuberculosis patients, the likelihood of transmission was lower if the source case was HIV infected and had a CD4 T-cell count lower than 250/µL. In that study, the effect of HIV and immunodeficiency was larger than could be explained by differences in smear status, delayed treatment, or the presence of cavitary lesions. Even when the bacillary load is taken into account, there is substantial unexplained variability in infectiousness. Jones-Lopez and colleagues used an air-sampling device to capture M. tuberculosis in exhaled (coughed) air and showed that infectiousness was associated with the presence of cultivable organisms in the sampled air. The finding of organisms in the exhaled air was quite variable from patient to patient and did not correlate with sputum smear results.

Simple maneuvers, such as covering the mouth while coughing, can reduce formation of droplet nuclei by deflecting droplets from the air stream. Similarly, a mask worn by the patient is effective because particles are trapped while they are still large, before the water content has evaporated. Masks (disposable particulate respirators) worn by persons exposed to an infectious source are less effective than are masks worn by patients, because most airborne droplet nuclei are much smaller than their parent droplets. However, properly constructed, well-fitting masks are very efficient in removing respirable particles of 1 to 5 µm (see also Chapter 11 ).

A second factor of the source case to be considered in determining infectiousness is the number of organisms contained in the lungs. This can be inferred from the extent and morphology of the disease, as determined by the chest radiograph and more directly estimated by microscopic examination of sputum. Canetti demonstrated that the bacillary population of tuberculous lesions varies greatly, depending on the morphology of the lesion. The number of bacilli in solid nodular lesions ranges from 10 ² to 10 ⁴ organisms, whereas in cavitary lesions, populations are on the order of 10 ⁷ to 10 ⁹ bacilli. Loudon and Spohn, among others, demonstrated that the prevalence of tuberculin reactors among young contacts of patients with newly discovered tuberculosis increased as the radiographic extent of involvement increased. Thus, in tuberculosis control, the contacts of persons with more extensive tuberculosis should be accorded a higher priority for evaluation than the contacts of persons with less severe disease.

The most direct means of estimating bacillary population is microscopic examination of properly stained sputum smears. An average viable bacillary population of 5000 to 10,000 organisms per milliliter of sputum is required for the organisms to be seen in an acid-fast-stained sputum smear. The contacts of patients who have organisms present in sputum smears have a much higher prevalence of infection than do contacts of patients with negative smears and either positive or negative cultures. However, the contacts of sputum smear-negative patients may still acquire tuberculous infection and develop tuberculosis. Transmission from smear-negative patients was estimated to be the cause of approximately 17% of newly diagnosed cases in San Francisco.

A third important factor in determining the infectiousness of a source case is the use of chemotherapy. In studies designed to identify and quantify factors influencing transmissibility of M. tuberculosis, Sultan and associates and Riley and coworkers noted that patients who had positive sputum smears but who were receiving antituberculosis drugs were much less infectious for guinea pigs than were untreated patients. By their calculations, the relative infectiousness of untreated patients in comparison with treated patients was 50 : 1. Escombe and coworkers, as already noted, came to a similar conclusion. Consistent with these experimental observations, a substantial body of clinical data has accumulated that indicates that, once treatment to which the organisms are susceptible is begun, transmission of M. tuberculosis decreases quickly. The quantification of the infectiousness of cough-generated aerosols has shown the same result—that the major factor associated with persistent culture-positive aerosols was lack of effective treatment during the previous week. The most important mechanism by which chemotherapy reduces infectiousness is the direct effect of the drug on the bacillary population in the lungs. Hobby and associates found that, after an average of 15.6 days of multidrug chemotherapy, there was a reduction in the number of tubercle bacilli per milliliter of sputum of at least 2 logs, from approximately 10 ⁶ to approximately 10 ⁴ , or a 99% decrease. These data are similar to those reported by Jindani and coworkers, who demonstrated a reduction in colony counts of nearly 2 logs per milliliter of sputum in the first 2 days of treatment and a further reduction of 1 log during the next 12 days. Thus, in the initial 2 weeks of treatment, there was a decrease from approximately 10 ⁷ to 10 ⁴ organisms per milliliter of sputum, or a reduction of 99.9%. However, even with this profound reduction in the bacillary population, the remaining number of organisms (10,000 per milliliter of sputum) would still be sufficient to produce a positive acid-fast sputum smear. In addition to reducing the number of viable bacilli, chemotherapy also promptly decreases coughing. Loudon and Spohn noted that coughing was reduced by 40% after 1 week of treatment and by 65% after 2 weeks. The sum of these effects is that, once a patient with tuberculosis is placed on effective therapy, transmission of tubercle bacilli ceases to be a concern. The decline in infectiousness is caused mainly by the rapid reduction in bacillary population in the lungs as a result of antituberculosis chemotherapy, particularly isoniazid. Drug regimens that do not include isoniazid probably should not be expected to render the patient noninfectious as rapidly as do those containing isoniazid. Likewise, the prompt reduction in infectiousness cannot be assumed in patients harboring organisms that are resistant to isoniazid.

Different strains have different infectiousness. For example, strains of M. tuberculosis resistant to isoniazid may be less pathogenic than fully susceptible organisms. Molecular epidemiologic studies suggest that the mutation conferring resistance may have a role in the pathogenicity of M. tuberculosis, especially mutations associated with isoniazid resistance. Isolates harboring the most common mutation, katG S315T and the inhA promoter 15c-t, were able to cause secondary cases of tuberculosis (defined as genotypic clustering of strains in a population). In contrast, mutations in katG other than the common S315T mutation did not cause secondary cases. Although more information is necessary to confirm this association, this finding may explain the different results regarding the pathogenicity of some drug-resistant strains. It is also clear that the lesser pathogenicity can easily be offset by a prolonged period of infectiousness, as might be expected with ineffective treatment, as suggested by Escombe and associates, or with exposure of an immunocompromised host. Outbreaks of tuberculosis caused by MDR and XDR strains of tubercle bacilli have taken place in hospitals or correctional facilities and disproportionately involve HIV-infected persons, although immunocompetent persons have also been involved.

The operational implications of these assumptions concerning infectiousness should be modified in accordance with the patient’s living and working circumstances. The general principles of infection control entail an assessment of the vulnerability to tuberculous infection of persons who will potentially be exposed and the consequences should those who are exposed become infected.

Environmental Factors

The physical laws that apply to aerosolized particles, described in detail in Chapter 11 dictate that droplet nuclei essentially become part of the environmental air; thus, environmental factors are of extreme importance in influencing transmission of tubercle bacilli. Studies by Loudon and associates showed that, under standard conditions of temperature and humidity indoors, 60% to 71% of aerosolized M. tuberculosis organisms survived for 3 hours, 48% to 56% for 6 hours, and 28% to 32% for 9 hours. Apart from the natural death rate, the only factors influencing the infectiousness of organisms in a droplet nucleus under ordinary circumstances are its removal by venting or filtering and the death of the organisms from exposure to ultraviolet light. Environmental factors may be manipulated to decrease the concentration of tubercle bacilli, mainly removing them by effective filtration, killing them with ultraviolet light, or both. The influence of the concentration of organisms in environmental air in transmission of M. tuberculosis has been well illustrated in several microepidemics in which recirculation of air played an important role. The most dramatic example happened on board a U.S. Navy vessel that had a closed, recirculating ventilation system. The index case had a positive sputum smear and a brisk cough. As a result of this one case, 53 of 60 persons (88%) in his berth compartment acquired tuberculous infections and 6 developed tuberculosis. In a second compartment connected to the same ventilation system, 43 of 81 persons (53%) became infected and 1 developed tuberculosis. Smaller epidemics, one caused by a patient who underwent bronchoscopy in an intensive care unit, have been described.

Since early in the 20th century, it has been known that exposure to ultraviolet radiation kills tubercle bacilli. Riley and coworkers, in their classic studies of infectiousness in a series of patients, demonstrated that ultraviolet irradiation of air passing through an air-conditioning duct completely eliminated transmission of M. tuberculosis to guinea pigs housed beyond the ultraviolet lights. The major usefulness of ultraviolet lights is for providing UV irradiation of room air, using appropriately shielded lamps on the upper walls in hospital areas or clinics where patients with untreated tuberculosis are likely to be encountered. This is especially important in open areas such as waiting rooms, where ventilation may not be adequate to remove infectious particles, and within ventilation systems.

Circumstances of Exposure

The conditions of exposure have a major influence on the number of infectious particles inhaled. If the exposure is of long duration and takes place under conditions that would be associated with a high concentration of droplet nuclei in the air inhaled by the contact, there is obviously a greater likelihood of transmission. This is simply a restatement of what has been known for many years: Crowding and intimacy of contact are important determinants of transmission of M. tuberculosis . This is reflected in data from the United States showing that rates of both clinical tuberculosis and tuberculin reactivity are much higher among close (generally household) than among nonclose (generally out-of-household) contacts. In general, the rate of tuberculosis is in the range of 15 per 1000 close contacts and 3 per 1000 nonclose contacts. Of close contacts, approximately 50% are infected, in comparison with approximately 15% of nonclose contacts. Because the risk of tuberculosis is higher among close contacts, they should also be considered high-priority candidates for isoniazid preventive therapy. However, a number of studies using molecular epidemiology to track transmission dynamics have shown that substantial transmission may take place outside the household, particularly in social gathering places such as bars or informal places where alcohol is consumed.

Host Factors

There is substantial evidence that susceptibility to acquisition of infection with M. tuberculosis is highly variable. Most descriptions of contact investigations report that 40% to 60% of close contacts of an index case become infected, as reflected by conversion of the tuberculin skin test (TST) or interferon (IFN)-γ release assay from negative to positive. Although variations in the intensity of exposure contribute to the likelihood of becoming infected, variations in host susceptibility are also likely to contribute. Strong evidence for variable susceptibility to acquisition of infection with M. tuberculosis was provided by a prospective study of student nurses in a tuberculosis hospital in Philadelphia in the prechemotherapy era. In that study, in which student nurses were assigned to rotations on the same tuberculosis wards, 30% remained uninfected (TST unreactive) after 2 years of nursing school, indicating that despite repeated exposure, some of the student nurses were less susceptible to acquisition of infection than others. However, by the end of the third year of nursing school, 100% of the students had become infected, indicating that resistance to infection is quantitative, not absolute, and can be overcome by sufficient exposure to especially virulent strains of bacteria.

There have been few efforts to identify the determinants of resistance to acquisition of infection, although a recent study in a high-prevalence community in South Africa presented evidence for at least one host genetic determinant that influences the likelihood of acquisition of infection in children. Genomewide linkage analysis of TST-unreactive (zero-mm induration) and TST-reactive (greater than zero; median, 11.2-mm induration) children revealed evidence of a major locus that maps to chromosome region 11p14. The identity of the gene(s) at that locus, together with analysis of the cellular distribution and regulation of expression and the functions of the gene products, may reveal important mechanistic information on human resistance to acquisition of infection with M. tuberculosis .

It is currently unclear whether HIV infection, which has a major effect on progression to active tuberculosis, also affects susceptibility to acquisition of infection.

Intracellular Trafficking of Mycobacterium Tuberculosis

Seminal studies by Armstrong and D’Arcy-Hart and colleagues revealed that pathogenic mycobacteria survive and replicate in host phagocytes, including macrophages, by perturbing the normal pathway of phagosome maturation that eventuates in fusion with lysosomes and killing and digestion of other pathogens. Subsequent studies in basic cell biology have revealed that phagolysosome formation follows an ordered series of interactions between phagosome membrane proteins and phospholipids, relying on proteins that regulate intracellular traffic. One of these regulatory proteins, a low-molecular-weight GTPase termed Rab7, is essential for the late step of phagosome acquisition of lysosomal constituents; M. tuberculosis disrupts recruitment of Rab7 to the phagosome membrane. Normal trafficking also requires conversion of phagosome membrane phosphatidylinositol to phosphatidylinositol-3-phosphate; virulent mycobacteria also interfere with this essential step by mechanisms that remain incompletely defined. Together, these mechanisms allow M. tuberculosis to survive and replicate intracellularly. Recent studies have revealed an additional requirement for the ESCRT (Endosomal Sorting Complex Required for Transport) pathway of intracellular vesicle trafficking in phagosome maturation, and there is strong evidence that EsxH, a protein secreted by M. tuberculosis, disrupts the ESCRT pathway. In addition, the autophagy system, originally characterized for its role in degradation of endogenous cellular proteins and organelles, has received considerable attention for its role in restricting growth of intracellular mycobacteria, especially after stimulation of cells by the cytokine, IFN-γ. However, M. tuberculosis employs mechanisms to evade autophagy, including disruption of the phagosome membrane, giving the bacteria access to the host cell cytoplasm. Thus, M. tuberculosis can occupy several distinct intracellular niches: immature phagosomes, autophagosomes, and the cytoplasm.

ESX-1 Protein Secretion System

The early secreted antigen 6 kilodaltons (ESAT-6) secretion system 1 (ESX-1) was the first bacterial type VII secretion system to be discovered and is essential for virulence of M. tuberculosis . This was first indicated by the discovery that all strains of BCG, the attenuated strain of M. bovis used as a vaccine for tuberculosis, lack a functional ESX-1 system. Of the proteins secreted by the M. tuberculosis ESX-1 system, ESAT-6 and culture filtrate protein of 10 kD (CFP-10) are the best characterized. These proteins are not present in BCG and, thus, confer specificity to IFN-γ release assays in the diagnosis of latent tuberculous infection.

Although multiple properties have been attributed to ESX-1–secreted proteins, the best characterized are their effects on host membrane integrity. ESX-1–deficient bacteria do not escape the phagosome and do not stimulate host signaling pathways whose sensing mechanisms are in the cytoplasm, indicating that one target of the ESX-1 proteins is the phagosome membrane. ESX-1–deficient mycobacteria are also defective in cell-to-cell spread, most likely because they are less able to cause host cell necrosis; with cell necrosis, bacteria are released to the extracellular space, where they can access adjacent cells that support subsequent rounds of bacterial replication. Studies of the purified protein have directly implicated ESAT-6 in membrane perturbation. ESAT-6 has also been implicated in stimulating epithelial cells to express matrix metalloproteinase-9, which promotes migration of uninfected macrophages to the region adjacent to dying infected macrophages when they release viable bacteria, thus sustaining the infection cycle.

M. tuberculosis possesses five type VII secretion systems, of which ESX-1 is the best characterized. Of the remainder, ESX-3 is essential for survival of M. tuberculosis and contributes to pathogenesis; it is responsible for secretion of the protein, EsxH (also known as Tb10.4), which interferes with the host ESCRT pathway and is also a frequent target for recognition by T cells of people infected with M. tuberculosis .

Induction of Type I Interferons

An important aspect of tuberculosis pathogenesis is the induction of type I IFN (IFN-α and/or IFN-β) secretion. Whole blood transcriptome analysis to discover human genes that are differentially expressed in tuberculosis revealed a transcriptional signature dominated by IFN-responsive genes, and this finding has been replicated in additional patient cohorts. The strength of the IFN transcriptional signature correlates with the extent of disease, and the transcriptional signature is rapidly reversed with effective chemotherapy. Whether type I IFN plays a pathogenic role or is a secondary effect of active tuberculosis in humans remains to be defined. However, in mice, there is evidence that type I IFNs play a pathogenic role in M. tuberculosis infection, at least partially by enhancing recruitment of mononuclear cells that support intracellular bacterial replication. In addition, type I IFNs limit expression of interleukin (IL)-1-β, a cytokine that is essential for control of M. tuberculosis . A link between virulence and induction of a host-detrimental type I IFN response has been established by the observation that intact ESX-1 secretion and phagosome permeabilization are required for M. tuberculosis induction of type I IFN expression. Despite the growing evidence for a detrimental role of type I IFNs in tuberculosis, therapeutic administration of type I IFN for multiple sclerosis or hepatitis C has not been accompanied by a striking increase in the frequency of active tuberculosis.

Biologically Active Mycobacterial Lipids

M. tuberculosis has long been known to be rich in lipids, including mycolic acids that possess acyl chains containing up to 90 carbons. However, rather than being a mere “waxy coat” that acts as a barrier to drugs and other polar molecules, mycobacterial lipids interact directly with the host to contribute to pathogenesis.

Trehalose dimycolate (TDM), an abundant cell wall lipid traditionally known as mycobacterial “cord factor,” modulates innate immune and inflammatory responses to M. tuberculosis and is sufficient to induce transient granuloma formation when injected into experimental animals. Deficiency of TDM decreases pathogenicity of a mutant strain of M. tuberculosis experimental animals, indicating that the responses induced by TDM favor the bacteria. TDM is recognized by two related C-type lectin receptors, termed Mincle (also termed Clec4e) and MCL (also termed Clec4d), expressed on macrophages and dendritic cells, which transduce signals that result in proinflammatory cytokine production.

Other complex mycobacterial lipids contribute to pathogenesis by mechanisms that remain under investigation. Phthiocerol dimycocerosate (PDIM) is present in all clinical isolates examined, despite its frequent loss during serial passage of M. tuberculosis in liquid culture; the persistence of PDIM in clinical isolates implies that it is essential for pathogenesis but dispensable in culture. PDIM-deficient mutants are less pathogenic than PDIM-replete bacteria in mice ; PDIM masks the molecules recognized by Toll-like receptors and dampens initial inflammatory responses. The structurally related lipoglycans, phosphatidylinositol mannans (PIMs) and lipoarabinomannan (LAM), are implicated in specific interactions with the host. PIMs can stimulate innate immune and inflammatory responses by serving as agonists for Toll-like receptor-2 (TLR2), and studies with LAM-coated beads have indicated a potential role for LAM in altering phagosome maturation.

Granulomas

Granulomas, consisting of aggregates of macrophages, often including multinucleated giant cells and “epithelioid” macrophages together with variable numbers of lymphocytes, are a pathologic hallmark of tuberculosis. Granulomas can also contain variable numbers of necrotic cells and microscopic and macroscopic necrotic centers; some also exhibit caseating necrosis (characterized by complete loss of tissue structure and a texture resembling soft cheese) and may become calcified. Granulomas have been traditionally considered to be host-protective structures, thought to “wall off” bacteria and keep them from disseminating. Whereas this view may apply in later stages of fibrotic and calcified granulomas, early granuloma formation actually promotes infection by facilitating cell-to-cell spread within the macrophage aggregates, thus optimizing expansion of the bacterial population. In addition, intravital microscopy has revealed that macrophages and lymphocytes in granulomas are dynamic, with lymphocytes freely migrating between the apparently closely apposed macrophages. Together, these studies and others indicate that granulomas are dynamic structures that may benefit either the pathogen or the host, depending on the stage of infection.

Modulation of Apoptosis

As a facultative intracellular pathogen, M. tuberculosis shapes its environment to optimize its survival and growth. One mechanism is to inhibit apoptosis (programmed cell death) and prolong the life of infected cells, allowing the bacteria to grow to a larger population size in each infected cell before spreading to adjacent cells. There are other benefits to M. tuberculosis from inhibiting apoptosis. Because uptake of apoptotic cell fragments by naive macrophages results in killing of the bacteria formerly residing in an apoptotic cell, inhibition of apoptosis allows M. tuberculosis to evade this mode of death. Because bacterial antigens associated with apoptotic cell fragments are subject to uptake and “cross-presentation” to CD8 T cells, inhibition of apoptosis of the infected cells minimizes the frequency of CD8 T-cell activation. Evidence that inhibition of apoptosis is a virulence mechanism in animal models of tuberculosis is provided by the findings that proapoptotic mutants of M. tuberculosis are less pathogenic in vivo.

Phylogenetic Lineage- and Strain-Dependent Variation in Pathogenesis

As noted earlier, use of high-resolution genotyping methods has recently revealed genetic diversity in the M. tuberculosis complex. This has provided a basis for comparative studies to understand the phenotypic consequences of genetic diversity. For example, M. africanum, which represents a distinct genetic lineage of the M. tuberculosis complex and is endemic in West Africa, compared with strains of other lineages, is less pathogenic at specific steps in the infection cycle. It is transmitted as readily as members of other lineages, but the progression to disease is less frequent than other lineages. Yet once active tuberculosis develops with M. africanum, it is clinically indistinguishable from tuberculosis due to other lineages. Evidence that these findings are not strictly attributable to host population differences is provided by finding that M. africanum is also less pathogenic in inbred mice. As another example, a subfamily of the Beijing family has been found to be more readily transmitted than other strains in the same community and causes more severe disease in experimentally infected guinea pigs. These examples provide evidence for intrinsic differences in strains of M. tuberculosis and indicate that studies linking epidemiology and pathogenesis (using animal models) can provide complementary information.

Latency/Dormancy and Reactivation

One of the most important characteristics of tuberculosis is the state of latent infection, which develops in the majority of infected humans, with the ability to reactivate and cause active, transmissible disease. Although host factors (described later) contribute to establishing and maintaining the latent state, the bacteria also possess highly evolved mechanisms that are involved in latency and reactivation. A considerable amount of recent work has focused on certain M. tuberculosis genes in which expression is induced by hypoxia, and which are believed to be involved in latency (modeled as bacterial dormancy, wherein most of the bacterial population is not actively dividing). One group of genes whose expression is controlled by the transcription factor, dosR, is induced transiently by hypoxia, while other genes, collectively termed the Enduring Hypoxic Response, are under alternative regulation. Together with evidence that reoxygenation reverses the changes in gene expression, these findings provide a paradigm for understanding how M. tuberculosis can reversibly adapt to its environment and alter its metabolic and growth state. Evidence that such mechanisms operate during latent tuberculosis infection in humans is provided by the finding that T-cell responses to proteins encoded by these “dormancy” genes are more common in individuals with latent infection than in those with active tuberculosis.

The contributions of bacterial determinants to reactivation of latent infection are less well understood than are those associated with latency. However, the M. tuberculosis genome encodes 5 proteins with homology to a family of “ resuscitation promoting factors ” (RPFs) characterized in other bacteria for their ability to stimulate growth of bacteria from stagnant cultures. Although RPFs in other bacteria may act in interbacterial communication, the M. tuberculosis RPFs characterized to date are peptidoglycan glycosidases and appear to be involved in remodeling of the cell wall. Mutant strains of M. tuberculosis in which the genes encoding RPFs have been deleted are defective for reactivation in animal models, suggesting that RPFs may contribute to progression from latent to active tuberculosis in humans.

Innate Immunity to Mycobacterium Tuberculosis

Innate immunity includes cellular and humoral defenses that do not depend on clonal rearrangement of antigen receptor genes, the defining feature of acquired immunity as carried out by B and T lymphocytes. Although innate immune responses play an important role in tuberculosis, evidence from studies in individuals with HIV and in animal models have established that innate immunity is insufficient for control of M. tuberculosis infection.

Adaptive Immunity to Mycobacterium Tuberculosis

Early studies of the mechanisms of immunity to mycobacteria found that adoptive transfer of cells, but not serum, conferred cutaneous hypersensitivity to tuberculin on the recipients. This initial observation was followed by studies that revealed that cell transfer also improved the ability to control mycobacteria and that the responsible cells were T lymphocytes. Thus, the focus of immunology studies in tuberculosis has been on T cells rather than antibodies.

Contributions of Immune Responses to Tuberculosis Pathology

Tuberculosis is a prime example of an infection in which immune responses clearly contribute to the pathology of the disease, suggesting that the balance between protection and pathology requires tight regulation. For example, treatment with immunosuppressive corticosteroids predisposes to tuberculosis, can mask the symptoms and reduce the radiographic manifestations of pulmonary tuberculosis, and prevents mortality in some patients with tuberculous meningitis.

Analogous to autoimmune diseases such as type 1 diabetes mellitus and multiple sclerosis, T-cell responses can contribute to tissue destruction in tuberculosis. In mice selectively lacking PD-1, a molecule that transmits inhibitory signals in T cells, M. tuberculosis infection is rapidly lethal, due to excessive CD4 ⁺ T-cell activation and very high levels of proinflammatory cytokines. In humans coinfected with HIV, the frequency of cavitary lesions on chest radiographs is directly correlated with the CD4 ⁺ T-cell count at the time of diagnosis of tuberculosis. Since cavitary tuberculosis causes more secondary cases than noncavitary tuberculosis, mycobacterial induction of detrimental T-cell responses may benefit the bacteria by promoting transmission. Examination of the human T-cell epitopes in diverse strains of M. tuberculosis revealed a high level of sequence conservation (rather than variation, as in the antigenic targets in other pathogens), implying an evolutionary benefit to the bacteria from T-cell recognition. These results suggest that T-cell responses to M. tuberculosis, which provide a benefit to infected people, can also benefit the bacteria by promoting pathology in a fraction of infected people, resulting in enhanced transmission. This finding may guide tuberculosis vaccine antigen selection and emphasizes the need to evaluate the potential adverse effects of new tuberculosis vaccines.

Exogenous Versus Endogenous Infection

One of the historical controversies in tuberculosis has been the extent to which tuberculosis can be attributed to new infection by recently inhaled exogenous organisms (i.e., from the environment) as opposed to a reactivation of viable bacilli that have been maintained for many years in a dormant or growth-restricted state within the body. This concept is important in that current tuberculosis control efforts are based largely on the idea that most tuberculosis in low-incidence areas is the result of endogenous reactivation. Thus, prevention entails identification of infected persons and giving them preventive therapy with isoniazid.

Since the early 1990s, genotyping of M. tuberculosis has been successfully used to determine if tuberculosis is due to exogenous or endogenous infection. M. tuberculosis DNA fingerprinting derived from different genotyping markers and methods has been used to track specific isolates of M. tuberculosis in a community. For example, isolates of tubercle bacilli from HIV-positive individuals, in health care facilities or correctional institutions, frequently show identical genotypes ( Fig. 35-4 ), indicating that these individuals were probably infected by the same source within the facility. In such situations, tuberculosis can be assumed to be due to a recent exogenous infection with rapid progression to active tuberculosis. In contrast, patients with a unique DNA fingerprint are considered to have tuberculosis due to reactivation of latent infection acquired previously. Genotyping methods can similarly be used to differentiate relapse from reinfection in a patient with recurrent tuberculosis: a patient with relapse will have a similar M. tuberculosis DNA fingerprint in both episodes while a patient with reinfection will have different strains. Since the early 1990s, these markers have also enabled tuberculosis control programs, mainly in high-income settings, to track specific isolates of M. tuberculosis in a community to determine population-level risk factors for transmission, establish tailored public health strategies, and gauge the success of control measures.

Gel electrophoresis of DNA extracted from organisms isolated from patients in an outbreak of tuberculosis in San Francisco.

Cases 3 through 14 were linked epidemiologically, whereas cases 1 and 2 were not. This is an example of the use of restriction fragment length polymorphism analysis to track strains of Mycobacterium tuberculosis.

(From Daley CL, Small PM, Schecter GF, et al: An outbreak of tuberculosis with accelerated progression among persons infected with the human immunodeficiency virus: an analysis using restriction fragment length polymorphisms. N Engl J Med 326:231–235, 1992.)

The genetic markers used to track strains in the community are sufficiently polymorphic to distinguish among unrelated isolates yet stable enough to recognize isolates that are part of the chain of transmission. These markers include (1) the insertion element (IS) 6110 , (2) the polymorphic GC-rich repetitive sequences (PGRS), (3) polymorphisms in the clustered regularly interspaced short palindromic repeats (CRISPRs), and (4) the mycobacterial interspersed repetitive unit (MIRU)– variable number tandem repeats (VNTRs). The genotypes using CRISPR (also known as spoligotypes) and MIRU-VNTR (MIRU-type) are currently widely used to track a strain in the community. For both methods, there are online databases in which researchers can compare (and submit) their genotypes. Most recently, the availability of high-throughput technology and the dramatic decrease in costs have allowed for the use of whole genome sequencing to identify mutations as markers to study transmission of M. tuberculosis in a community . Compared with the other genetic markers, whole genome sequencing allows the identification of microevolutionary events (i.e., single nucleotide polymorphisms) within a chain of transmission, identifies epidemiologic links, and determines the directionality of the transmission events.

Tuberculin Skin Test

Tuberculin was first prepared by Robert Koch in 1890 and was touted by him as being therapeutic for tuberculosis. Shortly thereafter, the diagnostic capabilities of the material were recognized through its use in animals. In 1934, Seibert and Glenn prepared the first batch of a much more purified preparation, which they termed purified protein derivative (PPD). The most prevalent antigenic proteins in PPD are now known to be the bacterial heat shock proteins (or chaperonins) GroES, GroEL2, HspX, and DnaK. The antigen is prepared in liquid form containing the detergent Tween 80 to decrease adsorption of protein to the glass of the vial. The standard intermediate tuberculin test consists of the intracutaneous injection (Mantoux method) of 0.1 mL of material, which contains 5 tuberculin units (TU). The site usually chosen is the volar surface of the forearm, but any accessible area can be used. A short-beveled 26- or 27-gauge needle should be used with a 1-mL graduated syringe. A properly placed intracutaneous injection should cause a well-demarcated wheal 6 to 10 mm in diameter in which the hair follicles form dimples. Conventionally, the reading is done 48 to 72 hours after the injection, but it may be delayed for up to 1 week.

In persons such as hospital employees, who are likely to be tested repeatedly—if tuberculin negative at the outset—a two-step testing procedure is recommended to avoid confusing a boosted reaction with a true conversion. The phenomenon of boosting can happen in a person who is infected but loses skin test reactivity after several years. In this situation, a single tuberculin test can be falsely negative, but the test itself can recall (i.e., boost) the waned reactivity. A subsequent test can then elicit a positive reaction. A positive reaction following a negative one may cause the person tested to be classified as a tuberculin converter. To elicit any potential boosted response and to categorize the person more accurately as infected or not infected, a second 5-TU tuberculin test is applied within 1 to 2 weeks of the first test. If the second test shows a positive reaction, it is interpreted as a boosted response indicative of prior infection; if the reaction remains negative, it is assumed to be truly negative.

The reaction to the test should be read by inspecting and palpating the area where the tuberculin was injected. The reaction size is determined by measuring the diameter of any induration with a ruler. The amount of erythema should not be taken into account; only the extent of induration is important. Readings must be recorded accurately in millimeters.

The interpretation of tuberculin tests requires clinical judgment, as well as an understanding of the test. In a population in which the only mycobacterial species causing infection is M. tuberculosis, the curve describing the distribution of reaction sizes in otherwise healthy infected persons given 5-TU PPD would be bell shaped, having a mode of 17 to 18 mm, with very few reactions less than 10 mm. Thus, defining the minimum reaction size indicative of tuberculous infection would be simple. However, in many parts of the world, a portion of the population is infected with nontuberculous mycobacteria, which induce some degree of sensitization to tuberculin; inoculation with BCG, for many years the world’s most commonly used vaccination, has the same enhancing effect on tuberculin reactivity. Although these reactions are on the whole smaller than those caused by M. tuberculosis, they blur the distinction between reactions in persons infected with M. tuberculosis and those not infected. On the basis of a large amount of epidemiologic data and skin testing with antigens prepared from nontuberculous mycobacteria, the best compromise between false-positive and false-negative readings to 5-TU PPD tuberculin tests is 10 mm. Thus, under most circumstances, a reading of 10 mm or more is considered indicative of infection with M. tuberculosis. However, in some situations, smaller reactions should be taken to indicate tuberculous infection. For example, a reaction of 5 mm in a child who is a contact of a person with smear-positive tuberculosis probably indicates tuberculous infection and is considered positive. Likewise, a 5-mm reaction in a person with known HIV infection should be considered positive. In the United States, the history of BCG administration is generally ignored in interpreting the results of the TST.

There are several reasons why the tuberculin reaction may be interpreted as negative in the presence of tuberculous infection. These include errors in application or reading of the test result, usually related to the inexperience of the tester/reader, and should be easily correctable with proper training. Problems with the antigen are infrequent unless it has been improperly handled. Many disease states, especially HIV infection, interfere with cell-mediated immune responses. Lymphoreticular malignancies such as Hodgkin disease are potent suppressors of cell-mediated immunity. Corticosteroids and immunosuppressive drugs decrease tuberculin reactivity if the patient is on a sufficient dose for a sufficient period of time; for corticosteroids, the minimum dose is 15 to 20 mg of prednisone or the equivalent of another preparation, given daily for 2 to 3 weeks. As stated previously, advancing age is associated with loss of tuberculin reactivity, although it may be recalled. Malnutrition also may cause defects in cell-mediated immunity with consequent diminished tuberculin reactivity. Finally, pleural tuberculosis and overwhelming tuberculosis itself may cause diminished or absent tuberculin responsiveness.

Even when the test is applied and the result is read with particular care in patients with proven tuberculosis and no apparent immunosuppression at the time they are admitted to a hospital, only 80% to 85% have reactions of 10 mm or more to 5-TU PPD. Thus, a negative tuberculin test result cannot be used to exclude tuberculosis as a diagnostic possibility.

The interpretation of the TST in persons with HIV infection is a particular problem because of the progressive immunosuppression in HIV disease. Individuals infected with HIV are less likely to have a positive tuberculin reaction. In a cross-sectional study of persons with HIV infection and a wide range of CD4 ⁺ lymphocyte counts, anergy—defined as lack of any reaction to tuberculin (5 TU) plus mumps and Candida antigens—was more common when the CD4 ⁺ count was less than 400 cells/µL.

Interferon-γ Release Assays

IFN-γ release assays (IGRAs) are used for the diagnosis of latent tuberculous infection (LTBI); these assays cannot distinguish LTBI from active tuberculosis. Two IGRAs are currently approved in the United States, the QuantiFERON-TB test and the T-SPOT.TB test. The QuantiFERON-TB (Qiagen, Alameda, CA) tests, specifically QuantiFERON-TB Gold and QuantiFERON-TB Gold In-tube (QFT-GIT), measure the amount of IFN-γ released from sensitized lymphocytes in whole blood incubated overnight with mixtures of M. tuberculosis antigens, ESAT-6 and CFP-10. A newer-generation test includes an additional antigen, Tb7.7. The other approved test, the T-SPOT.TB, utilizes an ELISPOT format to quantify the number of cells in peripheral blood that secrete IFN-γ when stimulated with ESAT-6 and CFP-10 (Oxford Immunotec, Abingdon, United Kingdom).

Neither the TST nor IGRAs have value for the diagnosis of active tuberculosis in adults. A systematic review and meta-analysis showed that the sensitivity to diagnose tuberculosis in low- and middle-income countries was 76% for T-SPOT.TB and 60% for QFT-GIT. The specificity was 61% and 52%, respectively.

Compared with the TST, the IGRAs have several advantages: The tests can be performed in one patient visit, they are more specific in the presence of BCG vaccination or infection with nontuberculous mycobacteria, they are not subject to reader variability, and they do not stimulate waned immunity (the booster reaction, described earlier). A systematic review showed that IGRAs have excellent specificity and are not affected by BCG vaccination (due to the absence of ESAT-6 and CFP-10 from all strains of BCG). The sensitivity to diagnose infection (using culture-positive M. tuberculosis cases as the reference) was variable in the different studies, but T-SPOT.TB appeared to be more sensitive than QuantiFERON-TB Gold and QFT-GIT and the TST. In patients with HIV infection, the responses to TST and QFT-GIT correlate with the degree of immunodeficiency, whereas results of the T-SPOT.TB are independent of the level of CD4 T-cell depletion, which may explain the variable results of the different IFN-γ-release assays and TST among patients with HIV infection.

IGRAs have a few disadvantages: The tests require drawing blood and processing it within a specific time, and they are less well known than the TST and, therefore, have much less evidence to characterize their performance for diagnosing latent tuberculous infection and for epidemiologic studies. For example, risk factors for conversion (from negative to positive) and reversion (from positive to negative) were different for T-SPOT.TB compared with the TST. The authors attributed these findings to a lack of understanding of the dynamics of the IFN-γ response to ESAT-6 and CFP-10. Also, because there is no knowledge about the time required for the IFN-γ release assays to become positive after the onset of infection, the interpretation of a negative IFN-γ release assay requires caution .

Other authors, who also found a substantial proportion of reversions, attributed their findings to the decrease of the bacterial load due to treatment and to the natural resolution of the infection. In a study in the United Kingdom, Wilkinson and colleagues, using T-SPOT.TB, showed an early rise in spot-forming counts in patients given isoniazid and rifampin, then a reduction at month 3. No changes were observed in those without treatment. In India, Pai and colleagues followed 216 nursing and medical students and observed that 9 (24%) of the 38 individuals with an initial positive QFT-GIT test had reversion, which was associated with a negative TST result. Interestingly, those with skin test conversion also had an increase in levels of IFN-γ. Taken together, these data underscore the uncertainty regarding the interpretation of the IFN-γ release assays in providing information about the outcome of infection with M. tuberculosis, both treated and untreated.

Currently, the U.S. Centers for Disease Control and Prevention (CDC) recommends using the QuantiFERON-TB Gold test for the same indications as the TST: for evaluating persons suspected of having tuberculosis and for screening, including contacts of an infectious case of tuberculosis, children younger than 17 years of age, pregnant women, and persons at increased risk of tuberculosis, particularly those with HIV infection, recent immigrants who have had BCG vaccination, and health care workers. The QuantiFERON-TB Gold test usually can be used in place of—and not in addition to—the TST.

Diagnostic Evaluation

Clinicians must recognize that, in evaluating persons who may have tuberculosis, they are assuming an essential public health function and providing care to an individual patient. Early and accurate diagnosis is critical to tuberculosis care and control. Despite dramatically improved access to high-quality tuberculosis services worldwide during the past 2 decades, there is substantial evidence that failure to identify cases early is a major weakness in efforts to ensure optimal outcomes for the patient and to control the disease. Diagnostic delays result in ongoing transmission in the community and more severe, progressive disease in the affected person.

Globally there are three main reasons for delays in diagnosing tuberculosis: the affected person not seeking or not having access to care; the provider not suspecting the disease; and the lack of sensitivity of the most commonly available diagnostic test, sputum (or other specimen) smear microscopy. Approaches to reducing these delays are, obviously, quite different. Reducing delays on the part of the affected person entails providing accessible health care facilities, enhancing community and individual awareness, and active case finding in high-risk populations. Reducing provider delay is best approached by increasing provider awareness of the risks for and symptoms of tuberculosis and of the appropriate and available diagnostic tests. Rapid molecular tests that increase both the speed and the sensitivity for identifying Mycobacterium tuberculosis are increasingly available and, in some situations, are recommended by the WHO as the initial diagnostic test.

Patient History

There must be a clinical suspicion of tuberculosis before proper diagnostic tests are ordered. Clinical suspicion is prompted largely by the presence of symptoms and by awareness of comorbidities and epidemiologic circumstances that increase the risk of tuberculosis in an individual patient. These risks are summarized in the WHO guidelines for screening for tuberculosis. The most commonly reported symptom of pulmonary tuberculosis is persistent cough that generally, but not always, is productive of mucus and sometimes blood. In persons with tuberculosis, the cough is often accompanied by systemic symptoms, such as fever, night sweats, and weight loss. In addition, findings such as lymphadenopathy, consistent with concurrent extrapulmonary tuberculosis, may be noted, especially in patients with HIV infection. However, chronic cough with sputum production is not always present, even among persons with sputum smears showing acid-fast bacilli. Data from several tuberculosis prevalence surveys in countries with a high incidence of the disease show that an important proportion of persons with active tuberculosis do not have cough of 2 or more weeks, a criterion that, conventionally, has been used to define suspected tuberculosis. In these studies, 10% to 25% of patients with bacteriologically confirmed tuberculosis do not report cough. These data suggest that evaluation for tuberculosis, using a symptom review that includes cough of any duration, fever, night sweats or weight loss, may be indicated in selected risk groups, especially in areas where there is a high prevalence of the disease and in high-risk populations and individuals with increased susceptibility, such as persons with HIV infection. Use of this broadened set of questions in a population of persons with HIV infection was found to have a negative predictive value of 97.7% for tuberculosis. The presence of any one of the symptoms should be viewed as an indication for an evaluation for tuberculosis in high-risk groups or in high-incidence areas.

Hemoptysis is usually seen with more extensive involvement but does not necessarily indicate an active tuberculous process. Hemoptysis may also result from bronchiectasis left as a residual of healed tuberculosis; from rupture of a dilated vessel in the wall of an old cavity ( Rasmussen aneurysm ); from bacterial or fungal infection (especially in the form of a fungus ball [ aspergilloma or mycetoma]) in an old residual cavity ( Fig. 35-5 ); or from erosion of calcified lesions into the lumen of an airway ( broncholithiasis ).

Aspergilloma developing in an old tuberculosis cavity.

A, Front chest radiograph in a patient with tuberculosis shows bilateral upper lobe fibronodular changes with a right apical cavity ( arrow ). B, Frontal chest radiograph several years after ( A ) when the patient complained of hemoptysis shows development of an opacity within the right apical cavity ( arrowheads ) representing aspergilloma. C-E, Focused axial chest CT confirms the presence of aspergilloma.

(Courtesy Michael Gotway, MD.)

The systemic features of tuberculosis include fever in approximately 35% to 80%, malaise, and weight loss; there may be a variety of hematologic abnormalities, especially leukocytosis and anemia.

Physical Examination

In most cases, physical findings are not particularly helpful. Crackles may be heard in the area of involvement, along with bronchial breath sounds, when lung consolidation is close to the chest wall. Amphoric breath sounds (like the low sound of blowing across the top of an open bottle) may be indicative of a cavity. Findings such as lymph node enlargement, suggestive of extrapulmonary tuberculosis, may also indicate concurrent pulmonary involvement.

Radiographic Features

Radiographic examination of the chest is commonly the first diagnostic study undertaken, after the history and physical examination. However, in resource-limited settings, a chest radiograph is not necessarily included as part of the routine evaluation because of cost, complexity, and nonspecificity of the findings.

Pulmonary tuberculosis nearly always causes detectable abnormalities on the chest radiograph, although in patients with HIV infection, a chest radiograph may be normal in up to 11% of patients with positive sputum cultures. In primary tuberculosis, resulting from recent infection, the process is generally seen as a middle or lower lung zone opacity, often associated with ipsilateral hilar adenopathy ( Fig. 35-6 ). Atelectasis may result from compression of airways by enlarged lymph nodes. If the primary process persists beyond the time when specific cell-mediated immunity develops, cavities may form (so-called progressive primary tuberculosis).

Primary tuberculosis.

Frontal chest radiograph in a young adult shows superior segment right lower lobe consolidation associated with right hilar lymphadenopathy ( arrow ) due to primary Mycobacterium tuberculosis infection. Mild right paratracheal lymph node enlargement ( arrowhead ) is also visible.

(Courtesy Michael Gotway, MD.)

Tuberculosis that develops at a time remote from the original infection (endogenous reactivation) usually involves the upper lobes of one or both lungs. Cavitation is common in this form of tuberculosis. The most frequent sites are the apical and posterior segments of the right upper lobe ( Fig. 35-7 ) and the apical-posterior segment of the left upper lobe. Healing of the tuberculous lesions usually results in development of a fibrotic scar with shrinkage of the lung parenchyma and, often, calcification. Involvement of the anterior segments alone is unusual. In the immunocompetent adult with tuberculosis, intrathoracic adenopathy is uncommon. When the disease progresses, infected material may be spread via the airways (i.e., “bronchogenic” spread) into the lower portions of the involved lung or to the other lung. Erosion of a parenchymal focus of tuberculosis into a blood or lymph vessel may result in dissemination of the organism and a miliary pattern on the chest imaging ( Fig. 35-8 , see Fig. 18-25 ). Radiographic findings in HIV-infected patients are affected by the degree of immunosuppression. As further explained and illustrated in Chapter 90 , tuberculosis relatively early in the course of HIV infection tends to produce typical radiographic findings with predominantly upper lobe infiltration and cavitation. With more advanced HIV disease, the radiographic findings become more “atypical”: cavitation is uncommon, and lower lung zone or diffuse opacities and intrathoracic adenopathy are frequent ( Fig. 35-9 ). Surprisingly, a substantial number of HIV-infected patients with pulmonary tuberculosis had normal radiographs at the end of their course of treatment.

Cavitary tuberculosis.

Frontal chest radiograph in a patient with tuberculosis shows extensive right upper lobe cavitation.

(Courtesy Michael Gotway, MD.)

Disseminated tuberculosis.

Frontal chest radiograph in a patient with disseminated tuberculosis shows numerous small, randomly distributed nodules bilaterally, representing the miliary pattern.

(Courtesy Michael Gotway, MD.)

Tuberculosis in an HIV patient.

Frontal chest radiograph in a patient with tuberculosis shows bilateral hilar lymph node enlargement ( arrows ) associated with poorly defined parenchymal nodular opacities ( arrowheads ) and nodular areas of consolidation ( double arrowheads ).

(Courtesy Michael Gotway, MD.)

The activity of a presumed tuberculous process cannot be determined simply from a single radiographic examination of the chest. A cavity might be a sterile residual of an old infection, whereas a fibrotic-appearing lesion may be active. Conversely, not all radiographic worsening of the residua of prior tuberculous lesions can be ascribed to reactivation of the disease, although such worsening should always be of concern. Superimposed infections with other organisms or bleeding from bronchiectasis or from residual cavities may cause new infiltrations to appear. In addition, carcinomas may arise from within the area of scarring (so-called scar carcinomas) and be the cause of radiographic changes.

From this discussion, it should be obvious that the chest radiograph, although extremely valuable, cannot provide a definitive diagnosis of tuberculosis. Because of the radiographic similarities among the other disorders in the differential diagnosis, and because of the uncertainties in assessing disease activity and in determining the reasons for progressive radiographic changes, careful microbiologic evaluation is always indicated. A nondiagnostic microbiologic evaluation should prompt a careful assessment for other causes of the radiographic abnormality.

Bacteriologic Evaluation

As noted previously, a definitive diagnosis of tuberculosis can be established only by isolation of tubercle bacilli in culture or by identification of specific nucleic acid sequences. When the lung is involved, sputum is the initial specimen of choice. Two sputum specimens should be collected, which can be obtained the same day because the sensitivity of tests using same-day specimens is similar to tests using specimens collected on different days. The collection of the sputum in 1 day allows results to be available the same day, thereby increasing the efficiency of sputum smear microscopy. The strategies for same-day microscopy include the preparation of 2 or 3 slides from sputum samples obtained the first day the patient is assessed. Collecting more than two sputum specimens increases the yield only slightly.

There are several options for obtaining specimens from patients who are not producing sputum. The first and most useful in terms of yield and avoidance of patient discomfort is inducing sputum production by the inhalation of a hypertonic (3% to 5%) saline mist generated by an ultrasonic nebulizer. Sputum induced by this technique is clear and resembles saliva; thus, it must be properly labeled or it may be discarded by the laboratory. This is a benign and well-tolerated procedure, although bronchospasm may be precipitated in asthmatic patients.

Sampling of gastric contents via a nasogastric tube has a lower yield than sputum induction and is more complicated and uncomfortable for the patient. However, in children and some adults, gastric contents may be the only specimen that can be obtained. Gastric lavage should be performed early in the morning before the patient has gotten out of bed, eaten, or performed dental hygiene. Once the specimen is obtained, the specimen should be sent to the laboratory and processed the same day. To prolong the viability of the bacteria, neutralization of gastric acid with an equal volume of sterile 1% sodium bicarbonate is recommended when the specimen is not processed inmediately.

Depending on the clinical circumstances, if the sputum is negative or cannot be obtained, the next diagnostic step is usually fiberoptic bronchoscopy with bronchoalveolar lavage, and in some instances transbronchial lung biopsy. The yield of bronchoscopy is high in miliary tuberculosis and in focal disease as well. Bronchoscopic procedures have been especially helpful in the diagnostic evaluations of patients with HIV infection with negative sputum smear microscopy. Needle aspiration biopsy may also provide specimens from which mycobacteria are isolated, but the technique is especially suited to the evaluation of peripheral nodular lesions for which there is a suspicion of malignancy.

In some situations, a therapeutic trial of antituberculosis chemotherapy may be indicated before more invasive studies are undertaken. For example, a radiographic abnormality consistent with tuberculosis in a person who is younger than 40, is a nonsmoker, and comes from a country where there is a high prevalence of tuberculosis, either current or past tuberculosis is much more likely than a neoplasm, even in the presence of negative smears and cultures of sputum. In such a patient, improvement in the chest radiograph concomitant with antituberculosis treatment would be sufficient reason for making a diagnosis of tuberculosis and continuing with a full course of therapy. A response should be seen within 2 months of starting treatment. If no improvement is noted, the abnormality must be the result of either old tuberculosis or another process. An algorithm illustrating this approach is shown in Figure 35-10 .

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

1. Innate Immunity
2. Parasitic Infections
3. Obstructive Sleep Apnea
4. Dyspnea

Stay updated, free articles. Join our Telegram channel

Join