Pathogenesis of Human Pulmonary Tuberculosis: Insights from the Rabbit Model

This chapter presents an overview of tuberculosis. Tuberculosis is still one of the major diseases of the world, especially in developing countries. It kills over 2 million people each year, more than any other infectious disease. From the primary parenchymal lesion, the bacilli frequently spread via lymphatics and cause caseous lesions of the hilar lymph nodes. The bacilli may also spread via the bloodstream and cause lesions elsewhere in the host. The primary lesion, as well as the metastatic lesions, often progresses until the host succumbs. Adult-type pulmonary tuberculosis is a disease of innately resistant hosts, a category that includes most immunocompetent persons. The active parenchymal lesion frequently forms a cavity, in which the bacilli may multiply extracellularly. If so, the bacilli may spread via the bronchial tree to other parts of the lung. Cavity formation perpetuates tuberculosis in humankind because coughing spreads bacilli from the lungs into the environment, where they may infect other people. Contracting clinical tuberculosis depends on (i) the size and physiological state of the bacillary particle, (ii) its virulence, and (iii) the native and acquired resistance of the host. The chapter discusses factors to protect a personnel against tuberculosis. UV lights (shielded to protect people’s eyes) or high efficiency particle (HEPA)-filtered air purifiers should be used more frequently in hospital areas where tubercle bacilli are likely to be present.

After the inhalation of tubercle bacilli by rabbits and humans, the disease tuberculosis may progress through the following stages. In humans, the disease begins with the establishment of only a single primary pulmonary tubercle. Stage 1: Ingestion and often destruction of bacilli by pulmonary alveolar macrophages. Stage 2: Logarithmic growth of bacilli within nonactivated macrophages that entered the developing tubercle from the bloodstream. Stage 3: Arrest of the logarithmic bacillary growth by delayed-type hypersensitivity, which kills the bacilli-laden macrophages and often forms a solid caseous center in the tubercle. Stage 4a: In hosts with weakly developed cell-mediated immunity, enlargement of the tubercle and its caseous center with hematogenous dissemination of the bacilli. Stage 4b: In hosts with strongly developed cell-mediated immunity, stabilization or regression of the tubercle. Stage 5: Liquefaction of the caseous center, extracellular bacillary growth, cavity formation, and bronchial dissemination of the bacilli. These stages are not distinct but blend into each other. Also, stages 3, 4, and 5 may occur in the same lung and even in different parts of the same lesion, depending on the local concentration of bacilli and their tuberculin-like products. Native and acquired resistance is never absolute, because a large number of tubercle bacilli (which have grown extracellularly in a cavity) can overwhelm even the best-developed host resistance and cause secondary pulmonary lesions.

Human tuberculosis most frequently occurs as a tiny inapparent lesion that stays dormant throughout the life of the host. If clinical disease is produced, it varies from the rapidly progressing, hematogenously spread disease that occurs in infants and immunosuppressed individuals to a chronic, slowly progressing cavitary disease that is commonly found in immunocompetent adults. In this chapter, the gross and histopathological characteristics of the childhood and adult types of pulmonary tuberculosis are described in more detail. Proliferative lesions, sometimes called “hard” tubercles, are more common when small quantities of bacilli and their tuberculin like products are present in a host with high resistance to tuberculosis. Exudative lesions are more common when the host is highly sensitive to tuberculin, especially when large quantities of bacilli and their tuberculin-like products are present at the local site. Miliary tuberculosis occurs when a massive dose of tubercle bacilli is discharged from a caseous or liquefied focus into the bloodstream and the resistance of the host is inadequate, as in early infancy, in old age, and in immunologically depressed persons of any age. Reinfection-type or adult-type tuberculosis consists of a small to large pulmonary lesion that is not accompanied by any marked enlargement of the hilar lymph nodes. In patients with HIV/AIDS, the histologic pattern of tuberculosis generally correlates with the degree of immunosuppression. Tuberculous humans are about 100 times more sensitive than rabbits to tuberculin. Therefore, caseous necrosis and eventual calcification more readily occur in humans.

The chronic human-type tuberculosis with liquefaction and cavity formation is easily produced in rabbits, but it is not as easily produced in guinea pigs and cannot be produced in mice. Liquefaction of solid caseous foci in the lung perpetuates tuberculosis in humans because such liquefaction enables the extracellular growth of tubercle bacilli and the formation of cavities that discharge such bacilli into the airways and into the environment. Pulmonary cavities were produced in 2 to 4 weeks by the injection of various materials (emulsified in paraffin oil and lanolin) through the chest wall directly into the lungs of rabbits. Azathioprine, an immunosuppressive drug, also prevented cavities, probably by reducing the amount of delayed-type hypersensitivity (DTH). Hydrolytic enzymes play a major role in liquefaction and cavity formation. Products resulting from the hydrolysis of proteins and other macromolecules increase the osmolarity of the caseum. The tracheobronchial lymph nodes were slightly enlarged and occasionally contained tuberculous granulomas. Humans with well-controlled clinical tuberculosis may not show as great a reduction in tuberculin sensitivity. In clinical tuberculosis, a frequent occurrence is blood-tinged sputum and occasionally frank hemoptysis. The alveolar macrophages frequently had a vacuolated appearance, probably from the ingestion of surfactant. Mycobacterium vaccae is a rapidly growing avirulent acid-fast bacillus that was originally found free in the pastures of cattle and in their milk. Drugs that prevent a composition favorable to extracellular bacillary growth would be a welcome addition to our control of this disease.

Both delayed-type hypersensitivity (DTH) and cell-mediated immunity (CMI) are T-lymphocyte responses to bacillary antigens presented mainly by dendritic cells. In tuberculous lesions, DTH kills (nonactivated) macrophages that contain numerous tubercle bacilli when these bacilli release tissue-damaging local concentrations of tuberculin like products. In the resulting (solid) caseous necrosis, bacillary growth is inhibited and many bacilli die because of low oxygen tension and other factors. Therefore, tissue-damaging DTH has apparently evolved in mammals to stop continuing bacillary growth within the nonactivated macrophages that have permitted such growth. In tuberculous lesions, CMI activates macrophages so that they can inhibit and destroy ingested tubercle bacilli. DTH can also activate macrophages if only low local concentrations of tuberculin-like products are present. Both DTH and CMI exert their control locally. Their main systemic manifestation is to provide an expanded antigen-specific lymphocyte population to infiltrate local sites of bacillary lodgement. Antibodies that aid phagocytosis apparently play little or no role in the destruction of the tubercle bacillus. The bacillus readily enters macrophages without being opsonized by antibodies and evidently can multiply intracellularly within nonactivated macrophages in the presence of antibodies. Antigen-antibody reactions at sites of bacillary lodgement result in the production of chemotactic factors, including the C5a component of complement. In immunized hosts, chemotaxins cause a rapid local accumulation of dendritic cells, macrophages, and antigen-specific T cells—all of which would accumulate more slowly without the local antigen-antibody reaction.

The main types of cells participating in rabbit tuberculous lesions are dendritic cells (DCs), macrophages, natural killer (NK) cells, lymphocytes, and granulocytes. Follicular DCs and antigen-activated helper T cells interact with B cells for antibody production. The primary acquired immune response to the tubercle bacillus is initiated by DCs that activate T cells. Macrophages can inhibit or kill intracellular tubercle bacilli by means of reactive nitrogen and oxygen intermediates (RNIs and ROIs), to which M. tuberculosis is exquisitely susceptible. Tubercle bacilli can live for months in mouse granulomas without multiplying and without dying. Macrophages are the major cells of the mononuclear phagocyte system. This system is composed of promonocytes in the bone marrow, monocytes in the circulation, and macrophages in the tissues. Active collagenase (an enzyme secreted but not stored in macrophages) was only detected in fluids from peak BCG lesions. T lymphocytes have been divided into various subsets by their CD4 and CD8 surface antigens, by their functions (T helper [Th] cells, regulatory T cells, and cytotoxic T cells), and by the cytokines they produce (Th1 and Th2). Perforin is a protein that can polymerize to form a hole (or pore) in the target cell’s membrane. The eosinophils in rabbit BCG lesions seem to show higher ribonuclease activity than any other cell present. At sites of many infections, mast cell cytokines have been found to enhance the recruitment of T cells. They probably play a similar role in lesions produced by the tubercle bacillus.

In humans, rabbits, and guinea pigs, typical tubercles often have a caseous necrotic center. Surrounding this caseous center is tuberculous granulation tissue, which is rich in macrophages and lymphocytes and contains dendritic cells, plasma cells, polymorphonuclear leukocytes, eosinophils, fibroblasts, and many capillaries. This chapter describes the structural components of tuberculous lesions. In humans, tuberculous granulation tissue often contains ill-defined granulomas and areas of caseous necrosis, along with variable numbers of tubercle bacilli. In rabbits, guinea pigs, and humans, caseous necrosis first occurs at the end of the logarithmic (symbiotic) stage of tuberculosis when tuberculin sensitivity develops. When macrophages surrounding the caseous center are poorly activated, the bacillus again grows intracellularly, and again the damaging delayed-type hypersensitivity (DTH) reaction kills such cells. Liquefaction occurs when solid caseous material softens. In contrast to pyogenic abscesses, which liquefy soon after they form, the caseous foci of tuberculosis may not liquefy for months, if ever. In both humans and rabbits, the healing of small endogenous metastatic tubercles (and small tubercles of exogenous reinfection) involves basically the same mechanisms as healing of primary foci, i.e., the local accumulation and activation of macrophages at the site of infection. Different organs of the host have different susceptibilities to tubercle bacilli. Live bovine-type bacilli can apparently survive at a lower oxygen tension than can human type bacilli, which explains some of the differences found in the disease produced by the live strains.

The microvasculature plays an important role in the pathogenesis of tuberculous lesions. Blood vessels bring in the host defense cells, and vascular thrombosis is a major cause of caseous necrosis. This chapter describes the study of microvascular density in tissue sections of developing and healing dermal BCG lesions and in 48-h dermal tuberculin reactions. Rabbits were placed under deep terminal anesthesia, and their entire vasculature was perfused (via the aorta) with a gelatin-colloidal carbon suspension. Then, serial 250-µm thick tissue sections of the dermal BCG lesions were prepared, and the total length of the microvasculature in the whole BCG lesion was calculated from measurements made with a microscope containing an ocular grid. By 3 days, the vascular density in BCG lesions had increased to roughly 1.6 times that found in normal skin. It remained at this level for at least 6 to 7 weeks. The vascular density in tuberculin reactions showed a similar increase. It was concluded that the local microvasculature increases relatively little during the course of this slowly healing infection. A greater blood flow through existing capillaries evidently provides most of the nourishment needed by the infiltrating cells. These studies also demonstrated that microvascular thrombosis is a major cause of the caseous necrosis that occurs during the course of this disease.

In rabbits, 1- to 7-day pulmonary tuberculous lesions produced by aerosols are difficult to find because the inhaled dose of tubercle bacilli cannot be made large enough. A large intravenous dose, however, readily produces many such tubercles. This chapter describes their characteristics and provides information on the activation and multiplication of macrophages within such lesions. To produce these early lesions, the author's group injected rabbits intravenously with 10⁸ to 10⁹ tubercle bacilli (BCG). The blood-borne macrophages that entered the developing tubercles became partly activated during the first day. These entering macrophages retained their ability to divide, i.e., incorporate [³H]thymidine ([³H]TdR), even though they had ingested tubercle bacilli. In contrast, fully activated macrophages within tuberculous lesions lose their ability to divide. Pulmonary alveolar macrophages did not seem to participate in early pulmonary lesions produced by the intravenous route, but accumulated in the surrounding alveolar spaces. However, even though these alveolar macrophages were highly activated, they retained their ability to divide.

In rabbit BCG lesions, the turnover of mononuclear cells was most rapid in BCG lesions at 2 to 3 weeks, when the lesion size peaked and tuberculin sensitivity and acquired cellular resistance were well developed. The mononuclear cells were mostly macrophages, with some medium and large lymphocytes and probably some dendritic cells. At this 2- to 3-week peak, more macrophages entered, more died or left, more remained at the site, and more became activated than before or afterward. Before this time, the host had neither delayed-type hypersensitivity nor cell-mediated immunity, so no antigen-specific enhancement of the inflammatory response occurred. After this time, the bacilli and their antigenic products had decreased, so antigen-specific stimuli for cell infiltration and activation were reduced. In “healed” lesions, the mononuclear cell turnover still occurred but was decreased. The continuous entry of live nonactivated macrophages into the viable parts of tuberculous lesions provides fresh intracellular sites where tubercle bacilli can multiply before they are again inhibited by the delayed-type hypersensitivity and cell-mediated immunity of the host. In tuberculosis, bacillary dormancy of long duration can only be present in caseous necrotic tissue where no live host cells exist. In tuberculous lesions produced by BCG in rabbits (and probably those in guinea pigs and humans), macrophages have a continual turnover.

Lurie’s tubercle-count method consists of counting the number of grossly visible primary pulmonary tubercles, present 5 weeks after an aerosol infection of rabbits with virulent human-type tubercle bacilli. It is a quantitative measure of clinically apparent disease. At 5 weeks, the grossly visible primary tubercles are easily recognized, and many microscopic tubercles have regressed. Since human-type tubercle bacilli are not fully virulent for rabbits, the pulmonary-count method has a sensitivity that is not possible with fully virulent strains. The number of grossly visible pulmonary tubercles produced by human-type bacilli decreases (i) when rabbits are infected with bacilli of reduced virulence, (ii) when rabbits of high genetic (innate) resistance are used, and (iii) when rabbits are effectively immunized, so that they can rapidly activate macrophages and stop the development of early tubercles while they are still microscopic in size. Therefore, the pulmonary tubercle-count method can be used to assess (i) bacillary virulence, (ii) the genetic resistance of the host, and (iii) the efficacy of vaccines for tuberculosis.

Using natural airborne infection of virulent bovine-type tubercle bacilli over many months, Lurie showed that resistance to the establishment of tuberculosis and resistance to its progress are separate phenomena: his inbred resistant rabbits converted their tuberculin skin tests an average of 2.7 months sooner than did his inbred susceptible rabbits. The separation of the establishment and progress of tuberculosis is only applicable to experiments in which occasional fully virulent tubercle bacilli are inhaled over many months. Airborne infection of laboratory animals over many months has, however, established other concepts directly applicable to tuberculosis in humans. (i) Only a single grossly visible primary pulmonary lesion will be produced, despite the continuous presence of virulent tubercle bacilli in the air. The immunity developed in response to the primary lesion is evidently sufficient to prevent other occasionally inhaled tubercle bacilli from causing grossly visible lesions. (ii) Some animals (and perhaps a few humans) may convert their dermal tuberculin reactions, and yet show no grossly visible primary lesions in their lungs at necropsy. This occurrence may be due to the early spread of inhaled bacilli out of the lungs to the hilar lymph nodes, where the growth of tubercle bacilli can be more easily controlled. These concepts are consistent with what Riley found when he exposed guinea pigs for months to air from a ward containing sputum-positive tuberculous patients.

The purpose of this chapter is to assemble in one place the characteristics of the disease produced in rabbits by the inhalation of virulent bovine and human types of tubercle bacilli and by the inhalation of BCG. Cavity formation with bronchial spread of the disease followed by death is produced in rabbits only by virulent bovine-type bacilli. Fully virulent bovine-type bacilli usually produced one grossly visible pulmonary tubercle for each inhaled unit of 1 to 3 bacilli that reached the alveolar spaces in both susceptible and resistant inbred Lurie rabbits. In Lurie’s resistant rabbits, virulent bovine-type tubercle bacilli did not multiply to as high a titer as they did in Lurie’s susceptible rabbits. The use of such aerosolized human-type bacilli enabled Lurie to develop his tubercle-count method, which is the most quantitative method for assaying innate and acquired resistance to tuberculosis in rabbits as well as the virulence of the infecting tubercle bacillus. Human-type tubercle bacilli are not fully virulent in rabbits. Most of the inhaled BCG bacilli are apparently destroyed by the alveolar macrophages before they can multiply appreciably. Without such multiplication, the degree of immunization would be negligible. Similar to rabbits, humans should be less immunized by the inhalation of BCG than by parenteral administration, in which higher doses of BCG can be injected and greater bacillary multiplication can occur (because the alveolar macrophages are bypassed).

Lurie’s rabbits were inbred for either susceptibility or resistance to the progress of tuberculosis. When infected with virulent bovine-type tubercle bacilli, the susceptible rabbits developed a rapidly progressing, hematogenously spreading “childhood type” of tuberculosis, and the resistant rabbits developed a slowly progressing, cavitary, bronchial-spreading “adult type” of tuberculosis. In lesions produced by virulent bovine-type bacilli, by human-type bacilli, and by BCG, the same manifestations of genetic resistance to tuberculosis were evident histologically: mature epithelioid cells (now known as highly activated macrophages) were always more numerous in the lesions of resistant rabbits than in the lesions of susceptible rabbits, irrespective of the differences in virulence of the infecting bacillary strain. The genetic resistance of these rabbits resides in their ability to activate macrophages to control the growth of tubercle bacilli, both nonspecifically and immune-specifically. Crossbreeding showed that the genetic resistance to tuberculosis is multifactorial, with genes associated with resistance being dominant over susceptibility genes. The commercially available outbred New Zealand White rabbits seem almost as resistant as Lurie’s inbred resistant strain III rabbits. Thorbecke inbred rabbits were distinctly more susceptible than commercial outbred rabbits, but apparently not as susceptible as Lurie’s inbred C and FC rabbits. van Zutphen’s inbred rabbits (which are hypo- and hyper responsive to dietary cholesterol, respectively) have not been adequately studied for resistance to tuberculosis.

In recent times, mice have been by far the most frequently used animal for the study of tuberculosis. Guinea pigs and rabbits are used less often, and monkeys are used only occasionally. Rabbits die after an infection with virulent bovine-type tubercle bacilli, but eventually heal an infection with virulent human-type tubercle bacilli. Mice develop slowly progressing pulmonary tubercles with both bovine and human strains of tubercle bacilli, but the disease progresses more rapidly with the bovine strain. With virulent human-type bacilli, the tubercles of mice contain a larger number of viable bacilli than do the tubercles of rabbits and guinea pigs. Apparently, the low levels of delayed-type hypersensitivity (DTH) in mice and the rarity of caseous necrosis allow the bacillus to grow to higher titers in the logarithmic stage. After these titers are reached, the good cell-mediated immunity (CMI) developed by mice reduces the intracellular multiplication of the majority of the bacilli to almost a dormant state. However, because of the extensive lung destruction caused by tissue-damaging DTH, guinea pigs often die in less time than do rabbits and mice. Rhesus monkeys are very susceptible to tuberculosis, but cynomolgus monkeys are more resistant. Some cynomolgus monkeys can even stop the progression of the disease. Due to the differences in the pathogenesis of tuberculosis in rabbits, mice, and guinea pigs, all three species should be used for testing new vaccines prior to their trials in humans.

Pharmacological amounts of glucocorticoids are frequently given as therapy for a variety of allergic, autoimmune, and inflammatory conditions, such as asthma and rheumatoid arthritis. When such drugs are continued for long periods of time, latent tuberculosis may reactivate. In tuberculous rabbits infected with human-type tubercle bacilli, pharmacological amounts of glucocorticoids decreased cell-mediated immunity and delayed-type hypersensitivity. Macrophages were poorly activated, and tubercle bacilli grew to large numbers within these phagocytes. After glucocorticoid administration was stopped, the tuberculin sensitivity returned, and (because of the large numbers of bacilli) liquefaction, cavity formation, tuberculous bronchopneumonia, and hematogenous dissemination occurred in some of the rabbits. One of Lurie’s inbred rabbit strains was evidently deficient in glucocorticoid production. In these rabbits, the administration of physiological doses of adrenocorticotropic hormone increased their resistance to tuberculosis. This chapter would not be complete without mention of dehydroepiandrosterone (DHEA) and 3β,17β-androstenediol (AED). These corticosteroids counteract the adverse effects of glucocorticoids on tuberculosis. The adrenals of different animal species secrete different proportions of glucocorticoids: e.g., rabbits and rats have low ratios of 17- hydroxycorticosterone (hydrocortisone) to corticosterone, whereas monkeys and humans have high ratios. Five weeks after infection with human-type tubercle bacilli, Lurie’s resistant strain III rabbits apparently produced higher levels of both hormones than did his susceptible strain FC rabbits.

Since estrogen increased the hyaluronic acid and water content of the skin, it decreased the spread of intradermally injected virulent bovine-type tubercle bacilli in rabbits. Chorionic gonadotropin had the reverse effect. These two sex hormones had no appreciable effect on the innate or acquired ability of the host to control the progression of tuberculosis, because neither hormone appreciably changed the number of primary pulmonary tubercles generated by aerosols of virulent human-type tubercle bacilli. However, estrogen markedly suppressed the development of amyloid in the spleens of rabbits dying of the more chronic form of tuberculosis caused by bovine-type bacilli. In rabbits, triiodothyronine or thyroxine increased host resistance in that they decreased the number of grossly visible primary pulmonary tubercles produced by the inhalation of virulent human-type tubercle bacilli, whereas thyroidectomy or propylthiouracil treatment increased the number of such primary tubercles. Thyroid hormones were most beneficial in inbred rabbits of intermediate resistance. The resistance of the most susceptible inbred C rabbits and that of the most resistant III(r) rabbits were not appreciably increased by thyroid hormones.

In an experiment to study the effects of sublethal wholebody irradiation, commercial rabbits were irradiated with 400 rads of whole-body X-irradiation—a sublethal dose. At 2 or 10 days thereafter, they were injected intradermally with BCG. Between 2 and 4 weeks after irradiation, the BCG lesions and 48-h tuberculin reactions in the irradiated group were smaller than those of the nonirradiated controls. The BCG lesions in the irradiated group also contained more bacilli. Pulmonary alveolar macrophages (AM) recovered by bronchoalveolar lavage (BAL) from irradiated rabbits contained higher levels of hydrolytic enzymes than did AM from nonirradiated controls. The AM from the irradiated group were apparently an older (more activated) cell population, because they had ingested inhaled particles for a longer period of time. The irradiation evidently had reduced the young macrophage population that replenishes the AM population. Rabbits were infected by aerosol with virulent human-type tubercle bacilli (H37Rv) at 12 or 30 days after irradiation. In each case, 5 weeks after infection, the number of primary pulmonary tubercles in their lungs was the same in both the irradiated and the nonirradiated groups. The number of viable bacilli in these tubercles was the same. Therefore, this sublethal dose of irradiation had no appreciable effect on the development and progress of primary pulmonary tubercles in rabbits. In brief, X-irradiation reduces the bone marrow’s capacity to provide defense cells to protect the host against infection. When the host is challenged by inhaled virulent human-type tubercle bacilli, an adequate supply of defense cells is available.

A sequential histochemical study of cytokines in developing and healing rabbit tuberculous (BCG) lesions is described in this chapter. In tissue sections, interleukin-1β (IL-1β), tumor necrosis factor alpha (TNF- α), macrophage chemoattractant (activating) protein 1 (MCP-1), and IL-8 are evaluated for cytokine mRNA by in situ hybridization techniques and for cytokine protein by immunohistochemical techniques. In tissue homogenates, gamma interferon (IFN-γ) mRNA was evaluated by reverse transcription-PCR. In the BCG lesions, the percentage of mononuclear cells that contained the mRNAs of these cytokines showed a biphasic pattern. Mononuclear cells containing IL-1β and IL-8 mRNAs were more numerous surrounding the caseous center. These cytokines evidently recruited the polymorphonuclear leukocytes that were common in this location. Mononuclear cells containing MCP-1 mRNA were more numerous in the outer third of the lesion where new macrophages and lymphocytes were being recruited. Both the nonspecific and antigen-specific cytokine responses of BCG vaccines are evidently synergistic. The early nonspecific cytokine (chemokine) response causes a local accumulation of antigen-presenting cells and lymphocytes, which explains, at least in part, why tubercle bacilli are good immunological adjuvants. This adjuvant effect should be considered in developing improved vaccines for the prevention of tuberculosis, because vaccines producing a strong early nonspecific cytokine (chemokine) response should be more immunogenic than vaccines with similar antigens producing a weak response.

Reinfection BCG lesions provide a simple model of how tuberculosis vaccines would affect an exogenous infection with virulent tubercle bacilli. Tissue sections of the lesions were prepared, and the types of cells and their cytokine mRNAs and proteins were analyzed by histochemical methods. The cytokines studied were interleukin-1β, macrophage chemoattractant (activating) protein (MCP-1), interleukin-8, and tumor necrosis factor alpha. The most informative findings were with MCP-1, one of the main chemokines attracting mononuclear cells (MN). At 3 h, both the reinfection lesions and the primary lesions contained the same percentage of MN labeled for MCP-1 mRNA. However, the reinfection lesions were 400 to 500 times larger and therefore contained many more of these MN. This high cell number alone would cause the total chemokine production to exceed by far that occurring in the primary lesions. By 1 day, the percentage of MN containing MCP-1 mRNA (and protein) had markedly decreased in the reinfection lesions, but remained high for at least 2 days in the primary lesions, which were beginning to increase in size. The rapid local accumulation of MN (macrophages, dendritic cells, and antigen specific lymphocytes) in the early reinfection BCG lesions seemed to be due to the presence of antibodies that developed during the first BCG infection. The antigen-antibody complexes formed at the site of reinfection evidently produced chemotactic factors that markedly hastened the cell infiltration.

Vascular adhesion molecules enable host defense cells to leave the bloodstream and enter tuberculous lesions. After the inhalation of tubercle bacilli, Lurie’s resistant rabbits had a larger number of mononuclear cells within developing lesions than did his susceptible rabbits. Therefore, the rapid local accumulation of mononuclear cells seems to be one of the factors associated with resistance to the progress of this disease. Vascular adhesion molecules enable such an accumulation to occur. With immunohistochemical techniques, the author's group evaluated the rise and fall of three major vascular adhesion molecules as rabbit dermal BCG lesions developed and healed. Intercellular adhesion molecule 1 (ICAM- 1) is important for the adherence of polymorphonuclear leukocytes (PMN), monocytes, and lymphocytes to activated vascular endothelium before they emigrate from the bloodstream into sites of inflammation and infection. Vascular cell adhesion molecule 1 (VCAM-1) is a major factor in monocyte, lymphocyte, and eosinophil emigration. Endothelial-leukocyte adhesion molecule 1 (ELAM-1, now called E-selectin) aids the emigration of granulocytes (and some monocytes and T lymphocytes). In tuberculosis, epithelioid cells are macrophages that adhere to one another in an epithelial-like pattern. This adherence seems to be due in part to the ICAM-1 of one macrophage’s binding to its ligand LFA-1 (lymphocyte function-associated antigen 1) (CD11a/CD18) on a neighboring macrophage. ICAM, VCAM, and ELAM are markers for activated vascular endothelial cells. In tuberculous lesions, such activated endothelial cells can capture and present local mycobacterial antigens and therefore may be killed by antigen-specific cytotoxic T lymphocytes.

Tuberculosis vaccines have little or no effect on the establishment of a microscopic pulmonary lesion produced by the inhalation of a virulent tubercle bacillus. Such a lesion is established only when the pulmonary alveolar macrophages fail to destroy the inhaled bacillus. Alveolar macrophages do not expand their population in response to specific antigens. Therefore, the establishment of a microscopic pulmonary tubercle is not affected by vaccination. Effective tuberculosis vaccines may, however, stop the progression of a tiny established lesion, because the vaccination has expanded antigen-specific lymphocyte populations. These lymphocytes enter the early lesion, where they cause a rapid local delayed-type hypersensitivity (DTH) and cell-mediated immunity (CMI) response that often prevents progression of the disease. When comparing their relative efficacies, two or more live vaccines should be standardized for equal numbers of live and dead bacilli, equal numbers of log-phase and dormant bacilli, and equal numbers of clumps and isolated bacilli. If vaccines more effective than BCG are ever developed, they would probably produce in the host a higher CMI/DTH ratio, i.e., an expanded antigen-specific lymphocyte population capable of producing increased numbers of activated macrophages and decreased amounts of tissue necrosis. To do this, the improved vaccine would probably contain increased bacillary glycolipid-protein components and decreased tissue-damaging tuberculin-like protein components. The vaccine should also contain components that increase the Th1/Th2 ratio. Dendritic cell-vaccine carriers may find their best use as immunotherapy in immunocompetent patients with multidrugresistant tuberculosis.

In rabbits (and humans), dermal BCG lesions have the same components as lesions caused by virulent tubercle bacilli, i.e., a caseous liquefied center surrounded by tuberculous granulation tissue. Therefore, BCG lesions can be used as a model in which to study the host response to this disease. In immunocompetent hosts, BCG lesions do not progress and always heal. The rate of healing of dermal BCG lesions reflected the native and acquired resistance of Lurie’s susceptible and resistant inbred rabbit strains. Similarly, the rate of healing of BCG lesions has been shown to reflect the resistance to tuberculosis of human populations. However, the healing of BCG lesions in individual rabbits and humans may vary considerably from the mean. In Lurie’s natively resistant rabbits, immunization with BCG was quite effective in preventing many grossly visible primary pulmonary tubercles. In his natively susceptible rabbits, BCG was hardly effective at all. In other words, the rabbits that needed it the least were helped by BCG vaccination the most, and the rabbits that needed it the most were helped the least. This principle also applies to human populations: persons with poor resistance to clinical tuberculosis would develop less immunity from BCG administration than persons with strong resistance. The efficacies of various BCG and Mycobacterium microti (the vole bacillus) vaccines were evaluated in commercial outbred rabbits by the tubercle-count method. Both of these vaccine types were effective. Recombinant BCG vaccines that contain additional antigenic components may be more promising than currently used BCG vaccines.

The pathogenesis of tuberculosis can be considered a series of battles between the host and the tubercle bacillus, each of which has its own weapons that can be used against the other. In addition, both the host and the bacillus have sites of vulnerability that can be exploited by the adversary. The weapons of the host are: cell-mediated immunity (CMI), which activates macrophages so that they can kill or inhibit tubercle bacilli that they ingest; and delayed-type hypersensitivity (DTH), which stops the intracellular growth of bacilli in nonactivated macrophages by killing these macrophages. DTH transforms an environment that is favorable for the bacillus into an environment that is inhibitory, i.e., solid caseous tissue. In general, each stage of tuberculosis is won by the host with increasing difficulty. Furthermore, in the same lung, some lesions may progress while other lesions may regress. Tuberculosis is a locally controlled disease that depends on the growth of bacilli in nonactivated macrophages or in liquefied caseum, or on the inhibition of bacilli in activated macrophages or in solid caseum. To control bacillary multiplication, both CMI and tissue-damaging DTH are required. This statement was proved by correlating bacillary growth curves with the observed gross pathology and histopathology. The type of tuberculosis described in the susceptible rabbits resembles that found in infants and immunocompromised adults. Acute and chronic bacterial infections show a spectrum of host responses. At one end of the spectrum is typical pneumococcal pneumonia, a rapidly progressing acute disease somewhat like anthrax and plague.

This section contains five parts. Part I talks about the host response and its implications; part II suggests research on the bacillus; part III suggests research on prophylactic vaccines and immunotherapy; part IV deals with development of new drugs for the treatment and prevention of tuberculosis; and part V provides sources of additional information. The cells participating in tuberculous lesions show a heterogeneity of functions, even among cells of identical appearance. The cytokines and cytokine receptors that are involved in the immune response within different organs should be identified by molecular biology techniques. Both DTH and CMI are mediated by Th1 lymphocytes and their cytokines. Tissue necrosis in clinical tuberculosis could be reduced by vaccines that cause little or no sensitivity to tuberculin. In rabbits, guinea pigs, and humans, microvascular thrombosis causes much of the tissue damage found in tuberculosis. Polymorphonuclear leukocytes (PMN) (neutrophils in humans, mice, and guinea pigs, but eosinophilic heterophils in rabbits) are common near solid and liquefied caseum. The prevention of clinical tuberculosis by a more effective vaccine than currently available BCG vaccines would be the most efficient way to reduce the incidence of this disease in the world. Immunotherapy should be pursued more extensively, because it would be of great benefit in treating multidrug-resistant tuberculosis, and because it could shorten the treatment of drug-susceptible tuberculosis.