Pathogenesis of Post Primary Tuberculosis: Immunity and Hypersensitivity in the Development of Cavities

Home , Caseous necrosis

Available online at www.annclinlabsci.org 365 Annals of Clinical & Laboratory Science, vol. 44, no. 4, 2014 Pathogenesis of Post Primary Tuberculosis: Immunity and Hypersensitivity in the Development of Cavities

Robert L. Hunter1, Jeffrey K. Actor1, Shen-An Hwang1, Vadim Karev2, and Chinnaswamy Jagannath1

1Department of Pathology and Laboratory Medicine, University of Texas-Houston Medical School Houston, Texas, USA, and 2Scientific-Research Institute of Childhood Diseases of the Federal Medical and Biological Agency, Saint Petersburg, Russia

Abstract. M. Tuberculosis (MTB) is an obligate human parasite even though humans are more resistant than any of the animals used for study. It is a human parasite because only humans develop post primary tuberculosis (TB) in their lungs that mediates transmission of infection to new hosts. The extreme paucity of human lung tissue with post primary TB has forced scientists to study animal models and human tissues that do not have the disease. Consequently, the unique features of post primary TB remain largely unknown and misconceptions are widely accepted. This manuscript presents a revised pathogenesis of post primary TB based on studies of lung tissues of thousands of patients by multiple authors and related literature. Pri- mary TB stimulates systemic immunity that kills organisms and heals granulomas resulting in both protection from disseminated TB and resistance to new infection. Post primary TB, in contrast, requires systemic immunity that it subverts to produce local susceptibility in the apex of the lung. It begins in the part of lung with the lowest ventilation, perfusion and movement and then proceeds to paralyze alveolar macrophages, block the exits and suppress inflammation to further isolate the area with post obstructive pneumonia. This ovpr ides a safe place for a small number of MTB to drive prolonged accumulation of host lipids and mycobacterial antigens in an otherwise immune person. After many months, the affected lung suddenly undergoes caseation necrosis with vanishingly few MTB. The necrotic tissue fragments to produce a cavity or hardens to develop fibrocaseous disease. Evidence suggests that this is triggered by a hypersensitivity reaction against cord factor and then progresses as the Koch phenomenon against many antigens. MTB grow in perfusion only in dead tissue or on a cavity wall. We anticipate that a more accurate understanding of the pathogenesis of post primary TB will facilitate focusing modern technologies to produce rapid advances in understanding and combating TB.

Keywords: Pulmonary tuberculosis, Mycobacterium tuberculosis, Pathogenesis, Pathology, Cavity and Lung

Introduction Immunity to TB is different from that of any other infection. The enigma of MTB is that it is an obli- Tuberculosis (TB) remains an urgent global health gate human parasite even though humans are more problem with approximately 9 million new cases resistant than any of the animals used for study. It is and 1.4 million deaths each year [1]. A very large a human parasite because only humans develop number of people are latently infected with post primary disease in their lungs [2,3]. Post pri- Mycobacterium tuberculosis (MTB) and at risk of mary TB produces cavities that support prolifera- developing disease in coming decades. The dual tion of vast numbers of MTB in an otherwise im- pandemic of TB and HIV/AIDS and the emer- mune host. Transmission is accomplished by gence of increasingly drug-resistant strains severely coughing MTB into the environment. aggravate the problem and hamper control efforts. While much progress has been made in under- Address correspondence to Robert L Hunter MD, PhD; Professor; standing granulomas produced by animals, little or Department of Pathology and Laboratory Medicine, MSB 2.034, University of Texas-Houston Medical School, 6431 Fannin Street, no progress has been made in understanding the Houston, TX 77030; phone: 713 500 5301; fax: 713 500 0730; e-mail: [email protected] peculiarities of post primary TB that occur only in humans [4]. Unanswered questions include: What

0091-7370/14/0400-365. © 2014 by the Association of Clinical Scientists, Inc. 366 Annals of Clinical & Laboratory Science, vol. 44, no. 4, 2014 is the nature of immunity that protects 95% of followed by discharge of infected necrotic material people from disease? Why does recovery from post into the airways. Recognizing that this is an apt de- primary TB make people more susceptible to new scription of the disease produced by M. bovis, post primary infections, but not to other types of Medlar examined thousands of tubercles looking TB? [5,6]. Why does it occur only in the lung? for evidence of a similar pathogenesis in humans. What is the role of hypersensitivity? He found none [12]. Once formed, all human MTB granulomas remained small. Recent research has served to deepen the mystery. Unlike HIV that modifies its antigens to evade im- In the pre antibiotic era, multiple investigators, mune responses, MTB keeps its antigens remark- each of whom had studied over 1000 cases of adult ably constant with even less variation than essential TB, reported that MTB spread through bronchi to structural genes [7]. Why does TB need exactly produce tuberculous pneumonia that underwent these epitopes? Failure to answer these questions caseation and then cavitation [9-14]. Rich reported has frustrated leaders in the field who increasingly “It has been found by all who have studied early believe that TB can never be eliminated without a pulmonary lesions that they represent areas of case- vaccine that prevents transmission of infection [8]. ous pneumonia rather than nodular tubercles.” Lack of understanding of the immunology and [10]. There was much speculation as to what seeded pathogenesis of late stage TB is widely regarded as the pneumonia [15]. Rich believed that small erod- the major impediment to vaccine development. ed granulomas released MTB into bronchi. One can neither rationally design nor test vaccine However, it was universally recognized that devel- candidates without knowledge of the type of im- oping post primary TB was a pneumonic, not a munity needed for protection. granulomatous process.

Since human lung tissue with developing cavitary Interest in TB declined with the introduction of an- TB has seldom been available since the introduc- tibiotics in the 1950s. It resumed with resurgence tion of antibiotics in the 1940’s, researchers have of the disease in the 1980’s with a new generation been forced to rely on animal models. Unfortunately, of investigators armed with new technologies and nearly all animals eventually die of primary infec- animal models. The rabbit infected with M. bovis tion that most humans clear in weeks. Consequently, was the only common laboratory animal in which research on animals is producing a detailed under- chronic fibrosing TB with cavities could be readily standing of the early phases of TB. However, with- produced [15]. However, the lesions and life cycle out human tissues for reference, modern science of M. bovis are different from those of MTB. M. has not been able to develop models or studies to bovis must be transmitted to new hosts during the approach the late stages of infection that account lifetime of a cow, not the 10-30 years of MTB in for 80% of clinical disease and nearly 100% of people. M. bovis does not produce post primary TB transmission to new hosts. [3,16]. It remains a particularly aggressive form of primary TB with widespread development of case- The pathology of pulmonary TB has a long history. ating granulomas that erode into surfaces of the Two distinct types of pathology, ‘productive’ and pharynx, bronchi, pleura, GI, urinary tracks and ‘exudative’, were recognized by Laennec in 1821 mammary glands to discharge organisms [17]. The and confirmed by virtually all investigators through differences between lesions produced by MTB and the mid 20th century [9-14]. In modern terms ‘pro- M. bovis were forgotten [3]. As a consequence, the ductive’ refers to nodular tubercles or granulomas concept that the caseating granuloma is the charac- and ‘exudative’ refers to pneumonia. Multiple at- teristic lesion of both primary and post primary TB tempts to explain the pathogenesis of these process- became the bed rock paradigm that has guided TB es were published in the late 19th and early 20th research throughout the rise of cellular immunolo- centuries. Some investigators proposed the mecha- gy, molecular microbiology and genetics. This pa- nism that is widely accepted today, that cavities de- per challenges that paradigm. velop by erosion of granulomas into bronchi Pathogenesis of Post Primary Tuberculosis 367

Figure 1. Characteristics of primary and post primary TB [3, 21]. A. X-ray of primary TB as a discrete round lesion in an upper lobe of lung. B. Cluster of caseating granulomas of primary TB that were mistaken for cancer and resected (H&E 3x). C. Caseating granuloma of miliary TB composed primarily of macrophages and lymphocytes with central caseous necrosis (H&E 40x). D. X-ray of acute post primary TB showing diffuse infiltrate of tuberculous pneumonia in an upper lobe. E. Section of lung with caseous pneumonia and developing cavities. No active granulomas were present any place in this persons body (H&E 3x). F. Endogenous lipid pneumonia of early post primary TB. The infection is confined to alveolar spaces and has very few AFB(H&E 100x). Reproduced in part from Hunter et al [21]

Our investigations in this area were initiated by the search for histologic slides of untreated pulmonary chance observation that injections of cord factor TB. Once we located slides so that we could see (trehalose 6, 6’-dimycolate (TDM)) into sensitized what the earlier investigators had described, the mice produced caseating granulomas [18]. This was pieces began falling together. Post primary TB be- a surprise since the literature had stated or decades gins as an endogenous lipid pneumonia, not as a that mice are a poor model of human TB because caseating granuloma, Figure 1 [21]. In continuing they do not produce caseating granulomas. Further studies, we looked for exceptions. This involved studies revealed that variations in the protocols study of histologic sections of infants, people with could produce a spectrum of caseating granulomas immunosuppression, the aged and HIV patients each of which resembled a manifestation of human who had died of pulmonary TB and comparisons disease [18]. The granulomas were found to be de- between disease produced by M. bovis and MTB pendent on CD1d and CD4+ T cell responses [3]. These investigations confirmed and extended against TDM [19,20]. our previous observations as detailed in Figure 2. TB develops as a necrotizing pneumonia and dis- In an effort to determine if these observations had seminates widely in the bodies of infants and im- relevance for humans, we initiated investigation of munosuppressed people. The pulmonary lesions of the histopathology of developing human pulmo- these people are different from those of post pri- nary TB. Very few papers on the pathology of pul- mary TB. They have little lipid, many more MTB monary TB had been published since the 1950s. both in alveoli and in the interstitium, and an acute The older literatureov pr ed exceedingly difficult to inflammatory response. Immunocompetent people understand because the nomenclature had changed infected with MTB for the first time develop case- and there were few pictures. This began a prolonged ating granulomas that contain the organisms and 368 Annals of Clinical & Laboratory Science, vol. 44, no. 4, 2014

Figure 2. Stages of TB in humans. Disseminated: TB develops as a disseminated infection in the lungs and many other organs in people without competent immune systems such as infants and adults with AIDS (H&E 40x).Primary: Primary TB develops in response to the first infection of immunocompetent individuals. Many animals die of primary TB in months, but most humans heal it within weeks. It produces cell-mediated immunity that effectively mediates life long protection from disease in all parts of the body except the vulnerable parts of the lung (H&E 40x). Post primary TB begins as an endogenous lipid pneumonia in the apices of the lung. Approximately 95% of such lesions resolve spontaneously. The other 5% go on to produce caseous pneumonia that can either soften and fragment to produce a cavity or harden to produce caseating granulomas and fibrocaseous disease. Greater detail of the pathology of post primary TB is presented in the references [3, 21]. Reproduced in part from Hunter et al [3, 21]. systemic immunity sufficient to control the infec- begins after all granulomas have healed. It starts tion in weeks1. MTB, however, remain dormant in with infection of alveolar cells in persons who their bodies. The location and properties of dor- maintain effective immunity to protect the entire mant MTB are controversial [22]. Much of the lit- rest of their bodies. The lesions develop as an en- erature assumes that they reside in granulomas. dogenous lipid pneumonia [21,26]. Most such le- However, multiple studies found that granulomas sions regress spontaneously so that progression to are sterile after 5 years and organisms are more clinical disease is a rare event. If they do not regress, commonly isolated from normal appearing lung the lesions undergo necrosis to produce a caseous and fat cells [10,11,23-25]. In any case, post pri- pneumonia that may fragment and be expelled to mary TB begins by reactivation of dormant MTB leave a cavity or remain as a mass that induces gran- or new infection from the environment. It typically ulomas and fibrocaseous disease.

1The terms granuloma and pneumonia require definition since the term granuloma is frequently used incorrectly to describe any lesion produced by TB. Granuloma: A granuloma is a localized nodular inflammation found in tissues. In TB, a granuloma is a nodular delimited aggregation of mononuclear inflammatory cells within tissue. It is a collection of modified macrophages resembling epithelial cells that are usually surrounded first by a rim of lymphocytes and later by fibroblasts.

Pneumonia is an infection of the alveoli (air sacs) of the lungs. It is typically diffuse within the air spaces of a lobule and does not penetrate into the lung tissue. The alveoli may fill with fluid and/or inflammatory cells in addition to infectious organisms. Pathogenesis of Post Primary Tuberculosis 369 hypothesize that bringing these reactive components together in a lung with endogenous lipid pneumonia triggers the sudden onset of caseation necrosis that further develops into cavities or fibrocaseous TB. The lesions are completely reversible and largely asymptomatic prior the onset of caseation pneumonia, but are irreversible and highly toxic afterwards. The evidence for each of the stages in this process will be presented and discussed.

Early post primary TB: Post primary TB begins when a virulent organism reaches the vulnerable area in the apex of the lung of a person who has immunity sufficient to heal caseating granulomas and prevent new infection in all other parts of the body. For over 100 years, multiple authors have reported that the disease dispro- portionately kills immunocompetent young adults [27-29]. “The higher the resistance of an individual, the greater the tendency to apical localization and death” [10]. It is widely believed that organisms preferentially grow in the apex of the lung because it has increased Figure 3. The pathogenesis of post primary TB. Post primary TB begins in oxygen tension [30]. However, the the most isolated part of the lung and proceeds to isolate it further with bronchial idea that oxygen is the key factor is obstruction and paralysis of macrophages. It accumulates host lipids and myco- contradicted by comparison of the bacterial products before eventually regressing or suddenly undergoing caseous necrosis that then leads to cavitation and/or fibrocaseous disease. distribution of primary and post primary lesions, Figure 4. The lesions of primary TB are distribut- Most of the contemporary literature depicts post ed in the same pattern as dust particles that are randomly primary TB as a war of attrition whereby a battle distributed in proportion to ventilation. One would thus rages between macrophages that must be recruited have to explain why primary and post primary MTB and activated to contain and kill MTB, and MTB have different oxygen requirements and why the apices of the lungs are more susceptible to smoke and other that tries to divide and kill the macrophages. While toxins [30,31]. this is a reasonable explanation of primary TB, post primary TB is different. This manuscript presents In 1983, Goodwin challenged the hypothesis that high studies on the pathogenesis of post primary TB oxygen tension causes TB to localize in the apices of the from both the modern and older literature, Figure lung [30]. He proposed that lymph stasis and impaired 3.A very small number of organisms isolate a sec- clearance of antigenic substances are more important de- tion of lung and suppress inflammation while they terminants of apical localization. The apex is the most quietly accumulate MTB antigens and host lipids isolated part of the lung. It has the lowest ventilation, for a sudden massive necrotizing reaction that de- perfusion and lymphatic drainage. More recent investi- stroys lung tissue to produce a cavity. We gations support this hypothesis. Blood flow in the apex 370 Annals of Clinical & Laboratory Science, vol. 44, no. 4, 2014 occlude perfusion. The oxygen tension in the apices is higher even though ventilation is lower, because there is insufficient blood flow to carry it away.

Several pieces of evidence support the hypothesis that re- duced blood flow is a factor in pathogenesis of early post pri- Figure 4. Distribution of lesions in primary and post primary TB. The localization of lesions of primary and post primary TB of many people were plotted on an X-ray. Each dot mary TB. First, bed marks the location of an individual’s lesion. The distribution of primary TB is consistent with rest was the main- chance distribution of air-borne infection while most post primary lesions were apical. stay of therapy for (Reproduced with permission from Medlar 1948 [87]). TB for many years before introduction of the lung is very low and frequently zero because the to antibiotics [27,32]. Bed rest increases blood flow in pressure within alveoli is lower than the pulmonary ar- the apex of the lung because the right heart does not tery pressure in upright people. The pressure in the pul- have to overcome the effect of gravity ofa recumbent monary artery of ~10-12 mm Hg is insufficient to raise person. Second, there are reports that this effect could be blood from the heart to the apex of the lung of people in further increased by raising the foot of the bed to pro- the upright position. Consequently the pressure within duce still greater blood flow in the apex [33]. In addi- the alveoli of ~2 mm Hg is sufficient to completely tion, pulmonary TB frequently improves when the patients reside in high altitudes where the rarefied air requires more vigorous respiratory efforts. Finally, TB was very rare in people with high pulmonary artery pressure due to mitral valve stenosis and devastating in people with decreased pulmonary artery pressure due to pulmonic valve stenosis [30].

The technologies for measuring the effects of respiration and heart beat motion of individual portions of the lung have become highly sophisticated because of the needs of radiation oncology. When treating lung cancer with radiation, it is important to deliver the maximal dose to the tumor while minimizing radiation of normal tissues. This becomes a difficult feat for a tumor in amoving lung. Four-dimensional CAT scans (4 DCT) visualize the lung in three dimensions plus time to show movement [34]. The vemo ment of the lung is greatest at the base where the diaphragm moves, moderate in the lower chest where the ribcage moves but nonexistent in the apices where the ribs are fixed. The ventilation at the apices is very low in both upright and recumbent positions Figure 5. MTB localization in alveolar macrophages in early post primary TB. AFB are found in alveolar because the lungs don’t expand and contract with respi- macrophages and nowhere else in the bodies of people with ration. Ventilation is driven only by pressure changes in early post primary TB. Similar localization of AFB is the lung [35]. These data suggests that post primary TB found in mice with slowly progressing pulmonary TB or begins in the apices because they do not move, have low reactivated TB. This is a section of mouse with reactivated ventilation and perfusion and, as discussed later, can be TB induced by an i.p. injection of 100 ug of TDM as an oil-in-water emulsion (AFB Stain, 1000x). further isolated by obstruction of bronchi. Pathogenesis of Post Primary Tuberculosis 371

reported that TB pneumonia can develop within one day after surgery for TB [13]. Similar large numbers of foamy macrophages in alveoli have been observed in BCG immunized mice at 24 hours after iv challenge with virulent MTB [43]. Finally, we observed that injections of cord Figure 6. TDM stimulates foamy alveolar macrophages in murine TB. C57Bl/6 mice factor into mice in- were infected i.v. with 106 MTB and injected i.p. one day later with 100ug TDM in an o/w fected with MTB pro- emulsion or the emulsion alone and examined on day 36.A. In the control animal, infection is duces a reduction of confined to small non caseating granulomas in the lung (H&E, 100x).B. In the TDM injected nodular granulomas animal, granulomas are smaller and much of the lung shows tuberculous pneumonia with foamy macrophages in alveoli (H&E, 100x). Most AFB are seen in foamy alveolar macro- and an increased in phages (Insert AFB, 100x). infected alveolar macrophages, Figure 6. Alveolar Macrophages: A characteristic of early post pri- Collectively, these data suggest that TB infects the least mary TB is that MTB are located exclusively in alveolar active part of the lung and then manipulates the macro- macrophages [3,11,21]. They do not enter the intersti- phages in order to prevent formation of granulomas and tium to produce granulomas or infect any other part of to inhibit their normal functions. the body. Alveolar macrophages have dual functions of collecting and eliminating innocuous dust particles and Surfactant: Pulmonary surfactant is a critical regulator recognizing and reacting appropriately to pathogens. It of inflammatory and immune functions within alveoli appears as if MTB are treated as innocuous particles in and very likely has a role in early post primary TB [44]. immune individuals whereas they are treated as patho- Modulation of surfactant function has major effects on gens that are carried into tissue where they induce granu- both macrophage function and fluid balance in alveoli. lomas in immunologically naïve subjects [36]. While Mice with genetic modifications affecting surfactant are much is unknown about this phenomenon, several piec- unable to remove MTB from alveoli to interstitial granu- es of evidence may be relevant. Both innate and adaptive lomas. This results in rapidly progressive tuberculous immune responses have been implicated in this response pneumonia [45]. Peripheral cell wall lipids of MTB are [37]. Killed mycobacterial or purified mycolic acids directly inhibitory to surfactant function [46]. Finally, cause macrophages to become foamy and to accumulate surfactant has been used with chemotherapy in patients lipids [38]. Oxygenated mycolic acids cause differentia- with pulmonary TB. Patients received surfactant inhala- tion of monocyte-derived macrophages into foamy mac- tions for 8 weeks during chemotherapy while the control rophages [39,40]. MTB persisted in these cells in a non- patients received only chemotherapy. The surfactant in- replicative state. Electron microscopy demonstrated creased sputum production, diminished cough, and has- bacilli in the lipid droplets [41]. tened both elimination of MTB from sputum and closure of cavities. Cavernous closure was achieved in The mouse model seems particularly relevant for this 72.9% of the surfactant treated group and in 41.4% of phase of infection [21,42]. In primary TB, alveolar mac- the controls at 4 months [47]. rophages containing MTB migrate into tissue to form small granulomas. In contrast, infected alveolar macro- Bronchial Obstruction: There is much evidence that phages in immune mice remain in alveoli, Figure 5. bronchial obstruction is an essential component of early Acid fast bacilli (AFB) are found only within foamy al- post primary TB [33]. In 1901, Hektoen reported find- veolar macrophages and no place else in the entire body. ing obstruction of bronchi in 100% of lesions of pulmo- MTB may be detected by culture in other organs, but nary TB [48]. Medlar proposed that post primary TB they are too few in number to be detected by AFB stain- should be called bronchogenic TB because it characteris- ing. This preferential localization of AFB in alveolar tically spreads through the bronchi to produce obstruc- macrophages happens only in animals with sufficient tion and TB pneumonia [12]. This is also supported by systemic immunity to prevent infection in all other parts the modern radiologic literature. TB frequently presents of the body. Lesions of foamy alveolar macrophages can radiologically as a wedge shaped lesions in an upper lobe. develop very rapidly in sensitized individuals. Levine The point of the wedge is an obstructed bronchus 372 Annals of Clinical & Laboratory Science, vol. 44, no. 4, 2014

Figure 7. Bronchial obstruction of post primary TB. A. CAT Scan showing tree-in-bud pattern characteristic of TB with obstructed small bronchiole (large arrow) and ground glass in lobular alveoli (small arrow). Reproduced in part from Hunter et al [62]. B,C. Histologic sections of human lung showing bronchial obstruction that is a characteristic of early post primary TB (H&E 40X). [10,49]. The characteristic CT finding of early postprimary pulmonary TB is the tree-in-bud pattern formed by branching linear structures and centrilobular nodules, Figure 7 [50,51]. Histopathologic analysis indicates that these centrilobular ‘tree’ lesions occur as a result of obstruction of the terminal or respiratory bron- chioles and sometimes the alveolar ducts. The ‘buds’ are foci of pneumonia inthe alveoli of the obstructed ducts. Centrilobular nodules may grow and coalesce into lobular consolidation that con- Figure 8. Abnormal cilia in post primary TB. Electron micrographs sists of centrally located lesions that con- show lesions in cilia. Reproduced with permission from Ahonen and tain caseation necrosis and surrounding Valavirta [55]. A. Cross section of a normal cilia of a control patient (x 52 000). B. Cross section of one cilia with 10+0 microtubular arrangement. perifocal inflammation. Cavitation, when Dynein arms are partially lost. The other cilia is enlarged, deformed and it occurs, is centrilobular. Several centri- dark stained (x104 000). lobular cavities may progressively coalesce into a larger cavity.

Epituberculosis: A particular type of TB, known as epituberculosis, provides addi- tional evidence of the role of bronchial obstruction in developing post primary TB [52]. Epituberculosis usually presents as a wedge shaped lesion associated with bronchial obstruction that almost always heals completely [10]. It is more common in children with a bronchus obstructed by a node with TB [52]. Some physicians were unsure if such lesions were TB. However, needle biopsy produced MTB and Langhans type giant cells, with no caseation [53]. Removal of bronchial obstruction causes epituberculosis to heal. In one case, an epituberculosis shadow had Figure 9. Endogenous lipid pneumonia produced by cancer and remained stationary for a month. TB. The lesions both have lipid rich foamy alveolar macrophages with Bronchoscopy demonstrated an ulcerated apoptosis and interstitial lymphocytic inflammation. Those caused by cancer are similar but do not have Langhans giant cells and contain fewer and plugged bronchus leading to the area. lymphocytes. A. Cancer, B. TB (H&E 200x). The obstruction was removed and found to be tuberculous. The shadow cleared in Pathogenesis of Post Primary Tuberculosis 373

Figure 10. Bronchial obstruction by cancer and chemotherapy produced a TB like cavity. A. Anterior view CAT scan showing a large mass in the left side of the mediastinum measuring approximately 8.2 x 9 x 9.5 cm (arrow). It encases the left upper lobe bronchi, pulmonary artery and pulmonary vessels. Although it does not show well in this reproduction, the pulmonary parenchyma of the left upper lobe had a ground glass appearance characteristic of obstructive pneumonia. (Reproduced with permission in part from Hunter et al [62]). B. Repeat CAT scan 5 weeks later showing a large cavity in the left upper lobe measuring 12 x 8 cm where the ground glass area had been (arrow). The radiologist’s report stated “Given its rapid development and the presence of numerous tree-in-bud opacities … the patient needs to be assessed for possible TB.” C. Posterior photograph of both lungs of the patient at autopsy showing a large thin walled cavity in the left upper lobe (arrow). This cavity, like those caused by TB, extended to the pleural surface, but did not penetrate. D. Section of the wall of the cavity of our cancer case showing a layer of necrotic lung overlying granulation tissue and lipid pneumonia. The detached, necrotic fragment of lung (arrow) resembles the fragments of lung that are coughed up by people with developing cavitary TB (H&E Stain 40). E. Giant cell resembling a Langhans Giant cell. (H&E stain 400x). F. Area on necrosis resembling caseation necrosis with cholesterol crystals.(H&E stain 100x).

six days by x-ray. There are multiple other reports in the distortion and swelling of other cilia, Figure 8. These pre antibiotic literature where surgical relief of bronchial unique changes suggest that MTB has a specific effect on obstruction produced resolution of pulmonary TB [10]. cilia. Finally, the role of bronchial obstruction in developing post primary TB is illustrated by the observation that The effects of tobacco smoke support an important role extensive unilateral pulmonary TB has been reported for cilia in the pathogenesis of developing TB [31]. In with segmental atresia of a bronchus [54]. epidemiologic studies, cigarette smoking has been con- sistently reported to markedly increase the risk of devel- Cilia: Obstruction of small bronchi by developing TB oping pulmonary TB. It is well established that cigarette may be viewed as a failure of ciliary clearance. Most par- smoke inhibits cilia action. Thus it is likely that paralysis ticles that reach alveoli are ingested by alveolar macro- of cilia by the smoke contributes to the pathogenesis of phages that migrate to small bronchi where they contact post primary TB [56,57]. cilia. The cilia then carry the macrophages and particles into the trachea and up to the mouth where they are Post obstructive pneumonia: We reported that post pri- swallowed. In patients with pulmonary TB, the bron- mary TB begins as an endogenous lipid pneumonia chial mucosal is generally intact although there may be [3,21]. We now know that bronchial obstruction from areas of ulcerated infected tissue extending into the sub- any cause produces post-obstructive pneumonia charac- mucosal glands. Ahonen and Valavirta studied cilia of terized by gradual accumulation lipid rich foamy macro- TB patients by electron microscopy [55]. They demon- phages in alveoli behind the obstruction [58]. The lipid strated loss of some ciliary components and major has been identified as neutral lipids and cholesterol that 374 Annals of Clinical & Laboratory Science, vol. 44, no. 4, 2014

Figure 11. MTB antigen in immunocompetent and immune compromised people. A. Immunohistochemical stain for MTB (IHC) in alveolar macrophages of an immunocompetent person with developing post primary TB (IHC Stain 200x). B. Higher power of same section showing that the antigen is entirely intracellular. (IHC 1000x). These sections were negative by AFB stain demonstrating a paucibacillary infection. C. Many AFB in a patient with HIV and disseminated TB (AFB stain, 1000x). D. IHC stain for MTB of section C reflects AFB in number and distribution of MTB. (IHC Stain 1000x).

are probably derived from action of macrophages on sur- Accumulation of mycobacterial antigens: factant. In 1925, Pagel used fat stains to access the Immunohistochemical staining tissues with polyclonal amount of lipid in alveoli during the course of develop- antiserum against whole killed MTB has been developed ing post primary TB. He found that stainable lipid in- as a diagnostic test for TB [64]. The amount and distri- creased progressively with time [59]. bution of MTB antigen staining in tissues varies with the type of infection. In tissues from people with rapidly Today, cancer is the most common cause of post obstruc- progressive TB associated with HIV, the staining of tive pneumonia [60]. It is an endogenous lipid pneumo- MTB antigen is similar to that of acid fast staining. The nia with many similarities to early post primary TB. The staining of early post primary TB is different. It typically disease produced by cancer has fewer lymphocytes and demonstrates far larger amounts mycobacterial antigen no Langhans type giant cells, but has a similar pattern of than can be accounted for by organisms observed with lipid rich foamy macrophages filling alveoli, Figure 9. acid fast stains. In fact, there is frequently intense stain- The lipid is derived from surfactant and is similar tothat ing for mycobacterial products in tissues with no detect- found in tuberculous pneumonia. Post obstructive pneu- able AFB, Figure 11. The finding of MTB antigens in monia due to cancer can undergo necrosis resembling tissues with no AFB is the justification for use of immu- the caseous necrosis of TB and cavitation. One study nohistochemistry for diagnosis [64-66]. The staining for found cavities in 13% of patients with post obstructive MTB antigens in immunocompetent patients is typical- pneumonia caused by squamous cell carcinomas [61]. In ly diffuse within the cytoplasm of foamy alveolaro- macr one case, chest x-ray demonstrated ground glass appear- phages. It is not found extracellularly except in necrosis. ance of post obstructive pneumonia immediately before These studies demonstrate that mycobacterial antigens initiation of chemotherapy. A similar x-ray one week accumulate with the lipid in foci of endogenous lipid later demonstrated a very large cavity, Figure 10 [62]. pneumonia of post primary TB. The patient subsequently died. An autopsy revealed lipid pneumonia with multinucleated giant cells and necrosis Our earlier studies demonstrated that slowly progressive resembling caseation in the wall of the cavity with strik- pulmonary TB in the mouse has many similarities to ing similarities to that of recently formed human tuber- early post primary TB in humans [3,21]. This includes culous cavities. Others have reported similar cases [63]. obstruction of bronchi an accumulation of foamy mac- These results demonstrate that post obstructive pneumo- rophages containing lipid, few AFB and mycobacterial nia establishes conditions for the development of caseous antigens in portions of the lung distal to the obstruction. necrosis and cavitation [60]. In this model, ManLAM, MPT64 and polyclonal Pathogenesis of Post Primary Tuberculosis 375 mycobacterial antigens were all shown to progressively accumulate in alveolar macrophages over several months [80]. ManLAM has anti-inflammatory properties that inhibit production of TNF-α and IL-12 by human dendritic cells and macrophages and modulates MTB-induced apoptosis via binding to macrophage mannose receptors. This is particularly important in deactivating host macro- Figure 12. Perifocal inflammation and pulmonary alveolar proteinosis. A. Perifocal inflammation of TB is a lipid rich edematous inflammatory response that sur- phages to allow the bacteria to rounds active TB sites of infection. While it typically has few or no AFB, perifocal inflam- survive within them [67,68]. mation contributes significantly to caseation, cavitation and morbidity of TB (H&E Despite the low numbers of 100X). B. Pulmonary alveolar proteinosis (PAP) has many similarities to perifocal inflam- AFB, the MTB were metaboli- mation. TB (H&E 100X). cally active during chronic TB in murine lungs and they have a unique transcriptome signature [69,70].

Perifocal inflammation is a toxic edema or desquamating pneumonia near tuberculous lesions that is thought to be a manifestation of hypersensitivity because it has few or no MTB detectable by acid fast staining [11,12]. It can occur with both primary and post Figure 13. Spontaneous regression of post primary TB. Only about 5% of people develop symptomatic post primary TB. However, many have apical scars and other evi- primary TB and has been dence of asymptomatic infection suggesting that most lesions resolve spontaneously. These identified as a major contribu- sections are from a resolving lesion were found incidentally at autopsy. A. Low magnifica- tor to mortality in miliary TB tion showing encapsulated lesion measuring 3/8 inch in diameter (H&E 4x). B. Higher [71]. The morphology of a magnification of the same lesion showing alveoli filled with fibrin and macrophages (H&E, perifocal reaction is essentially 20x). that of an exudative tuberculous tissue reaction [72]. It is composed mainly of blood plasma, fibrin, red blood corpuscles, relatively few poly- morphonuclear leukocytes, many lymphocytes, and desquamated foamy alveolar macrophages. The histological character of a perifocal reaction, however, will vary from purely hemorrhagic perifocal zones to zones of leukocyte in- Figure 14. Histopathology formation of cavities. A typical case showed tuberculous pneumonia and perifocal inflammation undergoing caseation necrosis and fragmentation filtrations with marked des- to form cavities. No granulomas were present any place in her body. All of the lesions ap- quamation and proliferation peared consistent with the 10-day history. MTB was identified in tissue by PCR. AFB were of alveolar cells with lympho- present, but extremely rare. Only three were found by acid fast staining in all of her lung cytes and plasma. Or there sections. There were also signs of acute vasculitis and intravascular blood coagulation, may be merely an exudation of sludging of erythrocytes in capillaries, and microthrombi in small vessels, in larger branches of the pulmonary artery and in veins resulting in the lung tissue infarction. A. Section plasma with a little fibrin and a showing necrotic lung (top) surrounded by granulation tissue and caseous tuberculous few lymphocytes. Erokhin pneumonia (H&E 1x).B. A nearby area where the necrotic tissue has been coughed out to used electron microscopy to leave a cavity traversed by a blood vessel (arrow) (H&E 1x). Annals of Clinical & Laboratory Science, vol. 44, no. 4, 2014 376

demonstrate destruction of alveolar epithelium, particu- Onset of clinical disease: Patients with post primary TB who larly Type 2 cells [29]. Tissue with perifocal inflamma- develop disease typically do so within one or two years follow- tion is seldom seen in surgical resections since it clears ing infection [42,80]. The onset of clinical pulmonary TBis rapidly with antibiotic therapy [10,12]. Reversal of peri- extraordinarily diverse causing much difficulty with diagnosis focal inflammation may be observed in X-rays within 72 [27]. It can present as acute pneumonia, chronic dyspeptic and hours of the initiation of therapy [73]. anemic symptoms or bronchitic with so little symptoms or fever that TB is discovered only accidentally by sputum. The Perifocal inflammation, at first glance, appears tobean most instructive cases, however, are those who undergo a rapid epiphenomenon: something that takes place as a result onset of caseous pneumonia, Figure 14. Investigators from of TB, not a primary contributor to it. However, several Laennec in the early 19th century through Erokhin in the early pieces of information suggest that it may be central to he 21st century described this mode of presentation [29,81]. pathogenesis of TB. First, perifocal inflammation under- Available evidence suggests that the disease process is similar, goes caseous necrosis to contribute a major portion of but less extensive and less well synchronized, in patients with lung afflicted with caseous pneumonia. Next, it can be less severe onset. induced in tuberculous people by injections of tuberculin and is associated with clinical exacerbation of disease. Caseous pneumonia typically begins with sudden acute onset This is the Koch phenomenon to be discussed later. of chill, fever, and rapidly advancing symptoms of severe in- Finally, the fact that it resolves quickly with therapy in- toxication [27,29]. Patients often complain of pain in the dicates that it requires live MTB. chest, dyspnea and coughing with expectoration of sometimes rust-colored sputum. Physical examination in the first days of Histologically, perifocal inflammation of TB resembles disease reveals intensive deadening of percussion sounds, bron- pulmonary alveolar proteinosis (PAP), Figure 12. This is chial breathing with crackling rales of high pitch and sonority. a condition in which alveoli are filled with lipid rich The disease at this early stage is typically indistinguishable from fluid and foamy macrophages and lymphocytes butfew community acquired bacterial pneumonia [29,82]. The X-rays other inflammatory cells. It can be caused by a deficiency show only diffuse density of lung lobes or lobules, which be- in surfactant protein B or by an autoimmune disease comes intensive over a few days. Diagnosis is especially difficult with altered GM-CSF signaling [74,75]. PAP can be because of the lack of AFB in sputum [83]. In a recent study, caused by a number of infections including TB. We sus- only 12.8% of caseous pneumonia cases had AFB in their spu- pect that some of these were cases of perifocal inflamma- tum in first two weeks of the disease [29]. This number in- tion of TB rather than cases of alveolar proteinosis com- creased to 39.2% by the end of the first month, and to 67% by plicated by TB [76,77]. The disease can frequently be the end of the second month. treated simply by lavage that washes out the lipid rich material from alveoli. Patients with caseous pneumonia did have two characteristic laboratory findings [29]. The first was profound hyper coagula- PAP frequently occurs in association with endogenous tion with development of disseminated intravascular coagula- lipid pneumonia [78]. Most cases analyzed showed mor- tion (DIC). This is a well known complication of tuberculosis phological features of both PAP and endogenous lipid [84]. The second was lymphopenia of CD4+ cells. Patients pneumonia. Type II pneumocytes, macrophages, and with caseous pneumonia had significant relative and absolute neutrophils play a significant part in the pathogenesis of decrease in CD4+-lymphocytes versus patients with chronic both. Historically, endogenous lipid pneumonia has pulmonary TB and age matched controls. At the same time, been reported to be caused by bronchial obstruction by the numbers of CD8+ cells in caseous pneumonia patients did cancer, but it also can occur with infection and other not differ from those found in patients with chronic pulmo- diseases that are not associated with bronchial obstruc- nary TB. The CD4+/CD8+ ratio was 0.7±0.03 for patients tion. Additionally, in PAP, the alveoli are usually filled with caseous pneumonia versus 1.4±0.09 for patients with with protein and lipid material resembling surfactant chronic pulmonary TB and 1.4±0.07 for normal controls. [78,79]. The Koch Phenomenon: The clinical presentation of caseous Regression of post primary TB: The large majority of pneumonia has striking resemblance to the Koch phenomenon early post primary TB cases regress prior to development [85]. Koch proposed that tuberculin could be a specific cure of symptoms. Only about 5% of people develop disease for TB. In early attempts at therapy, large doses of tuberculin [42]. Since some people resolve infections repeatedly, it were injected subcutaneously. The treatment caused severe lo- is likely that over 99% of early post primary infections cal and systemic reactions in patients with relatively mild dis- resolve spontaneously. Consequently, it is a rare event for ease. It induced perifocal reactions that progressed rapidly to the process to progress to clinical disease. Scars found in caseation and cavitation. Of the 230 patients with advanced the apices of the lungs of people born in the preantibi- cavitary disease who received this treatment, 30 died [86]. otic era have been attributed to healed TB that regressed Koch, himself, had a severe constitutional reaction following in the very early stages, Figure 13. the injection of old tuberculin [72]. Pathogenesis of Post Primary Tuberculosis 377 cavities will grow in culture [11,12,89]. It has been as- sumed that MTB that do not grow in culture are dead. Recent investigations demonstrate that many such organisms can be resuscitated and that proliferation of MTB is tightly controlled by both host and parasite [135]. Organisms that proliferate massively in any part of the body except a cavity endanger their host and thereby themselves. Increasing data indicates that MTB controls its proliferation to insure our survival and to develop conditions for its transmission to new hosts.

Production of cavities: Once a lesion develops caseous Figure 15. AFB in viable and necrotic lung. Section of pneumonia, it may further develop in one of two direc- lung of a person with necrotizing tuberculous pneumonia. tions. First, it may soften and fragment to the point that In immunocompetent people, very few and frequently no pieces of lung can be coughed out to produce a cavity. AFB are ever found in viable tissue (Right). Large numbers The fragmentation and removal of necrotic lung may be of AFB are found only in newly necrotic tissue (Arrow nearly complete in some patients. In these cases, the re- Left). However, AFB in necrotic tissue seldom grow in culture [11, 12]. Their numbers decline as the nuclear debris sultant cavities begin with a shaggy wall of debris. Over is removed (AFB Stain 100x). time this resolves and they develop a thin fibrous wall with little or no inflammation and bacteriaow gr only on More recent studies have confirmed and extended these the inner surface of the cavity. Laennec noticed that such results. Perifocal reactions occur only around existing ac- cavities are always centered on bronchi and that they are tive tuberculous lesions. The systemic symptoms include lined with a smooth waxy material that pushes its way fever often up to 40°C or higher with toxemia [85]. into the bronchi. Histologically, the inner surface is a Injections of tuberculin have the potential to increase the mass of organisms that grow as a pellicle [3,90]. MTB severity of TB markedly [72]. The morphology of an in- grown in liquid media without surfactants grows prefer- flammatory reaction induced by tuberculin around a tu- entially as a surface pellicle which forces its way across berculous focus is essentially that of a perifocal reaction the surface and climbs the sides of the vessel, Figure 16. of active TB. It is composed mainly of blood plasma, fi- This is due to rigidity of the monolayer of TDM[90]. brin, red blood corpuscles, relatively few polymorpho- nuclear leucocytes, many lymphocytes, desquamated al- Fibrocaseous TB: The caseous pneumonic material that veolar epithelial cells and alveolar macrophages. is not coughed out will dehydrate and become surrounded by granulomatous inflammation [11]. However, un- Growth and cultivability of MTB in tissue: The litera- like granulomas of primary TB in which caseation ne- ture frequently states that massive numbers of MTB crosis develops within granulomas, the granulomas in overwhelm host defenses to produce caseation necrosis post primary TB develop around preexisting foci of case- [15]. This is at odds with the observations that the onset ous pneumonia. If a caseous pneumonic lesion is small, of caseous pneumonia is a pauci bacillary process with it forms the nidus of a caseating granuloma superficially very few MTB. Many papers report that some early le- similar to those of primary TB, Figure 17. However, the sions are paucibacillary while others have many AFB underlying alveolar structure can be observed with con- [11,12]. A likely explanation is that most of these inves- nective tissue stains. If a caseous pneumonic is large, the tigators are reporting on chronic pulmonary TB that has formation of granuloma around it results in fibrocaseous multiple stages simultaneously. In addition, MTB do TB that is the most common type of chronic disease. grow in massive numbers producing necrosis in immune compromised people such as those with HIV. This is not TB is the largest cause of pulmonary fibrosis worldwide the case in immunocompetent people. Canetti and [91,92]. It has been reported that the rabbit is the only Medlar who independently examined thousands of cases animal model that will produce fibrocaseous TB. report that AFB are found in large numbers only in re- However, such lesions can be produced in mice by injec- cently necrotic tissue or on cavity walls [11,12,87]. They tions of mycobacteria in an oil emulsion of TDM [18]. are never found in large numbers in viable tissue of im- In addition, it appears that fibrocaseous TB is a reaction munocompetent people. This is our observation as well, to dehydrated caseous necrosis; a lipid matrix containing Figure 15. An unexplained, but consistent finding, is mycobacterial antigens. Freund’s complete adjuvant that the vast majority of AFB in closed necrotic tissues consisting of mycobacteria in an oil-in-water emulsion will not grow in culture nor will they infect guinea pigs reproduces this rather well [93]. [11,88]. In contrast, nearly 100% of AFB present in 378 Annals of Clinical & Laboratory Science, vol. 44, no. 4, 2014

Figure 16. Maturation of Cavities. A. A cavity in the upper lobe of the lung with connection to a bronchus. The disease was largely asymptomatic and the patient died of unrelated illness. The wall of the cavity was composed of fibrous tissue and a layer of necrotic debris [10]. Reproduced in part from Rich & Hunter [3]. B. The surface of this cavity demonstrated massive numbers of acid-fast organisms. No organisms or evidence of active TB were present anywhere else in the body (AFB Stain, 200x). C. Other areas of the lung demonstrated scars of healed TB with no evidence of recent activity (H & E Stain 100 x). D. MTB growing as a pellicle in a petri dish. The layer of MTB growing on a cavity wall has the characteristics of a pellicle that spreads and climbs the walls due to the pressure of the ex- panding layer of TDM [9, 90].

Discussion yield clean numbers, they frequently throw off the results, sometimes dramatically. Post primary TB is an For generations, the metaphor of host resistance to extreme example of the streetlight effect. It is a disease TB has been a war of attrition. Can the host recruit that occurs only in human lungs. Unfortunately, lung and activate macrophages fast enough to contain tissue suitable for study has not been available to re- and kill MTB or can the MTB divide and kill the searchers since the dawn of the antibiotic age in the macrophages to produce an enlarging lesion. 1950’s before the development of cellular immunolo- Evidence presented herein demonstrates that this gy, molecular microbiology or genetics. Researchers metaphor is inappropriate. A better one is the ex- have been forced to use animal models and human plosive reaction produced by dropping pure sodi- tissues that do not have post primary TB. As a conse- um into water. Two reactive components, host T quence, generations of scientists have been led to cells and MTB antigens, are produced and stored adopt an erroneous conception of the pathogenesis of separately in developing post primary TB. When post primary TB. The new sciences of cellular immu- brought together, they react violently to produce nology and molecular genetics were unable to focus caseous pneumonia that evolves to cavities and/or light on phenomena they could not see. fibrocaseous TB. Live MTB are not necessary for this process, but are standing by to move in and Stimulated by inconsistencies in the literature, our in- populate the cavity as the violence subsides. vestigations began with searching literature from the preantibiotic era. This ovpr ed to be of little value be- The Streetlight effect. “Many, and possibly most, cause the nomenclature and concepts were so differ- scientists spend their careers looking for answers ent from what we had been taught and there were few where the light is better rather than where the truth pictures. Progress was made only when we gained ac- is more likely to lie. They don’t always have much cess to slides from autopsies of adults who died of un- choice. It is often extremely difficult or even impos- treated acute pulmonary TB. Their lungs wesho d ca- sible to cleanly measure what is really important, so seous pneumonia with very few organisms and no scientists instead cleanly measure what they can, granulomas. With this, the pieces began to fall into hoping it turns out to be relevant.” David H. place suggesting plausible answers to long standing Freedman [94]. While surrogate measurements questions. Pathogenesis of Post Primary Tuberculosis 379 speculation as to what causes the softening and fragmentation of caseous material [15]. Granulomas in post primary TB occur only in response to caseous pneumonia that fails to fragment [11,12].

This conception of the pathogenesis of TB is based upon a mixture of facts, correlations and specula- tions. While some of the ideas presented in this paper will surely change with advancing knowledge, the facts will remain and must be considered. The following statements are supported by multiple investigations each of which involves hundreds or thousands of patients and should be considered to Figure 17. Chronic Fibrocaseous TB. Aging caseous be facts. pneumonia that fails to fragment becomes surrounded by granulomatous tissue and fibrosis. The central necrotic core 1. In post primary TB, granulomas develop in re- retains the structure of alveoli. It became surrounded by an active granulomatous process with epitheloid cells, giant cells sponse to caseation necrosis and are not the cause of and lymphocytes, but no foamy macrophages. Fibrosis devel- it [10-12]. ops gradually. This is a characteristic lesion of late post primary TB (Trichrome Stain, 40 x magnification). 2. Cavities develop from fragmentation of caseous pneumonia, not from eroding granulomas [10-13]. Primary TB stimulates systemic immunity. This ac- 3. Caseous pneumonia with developing cavities is tivates macrophages to ingest and kill MTB and to typically paucibacillary (contains very few MTB) contain them within granulomas that eventually [11,12,27,29]. heal. The result is protection from disseminated TB 4. Developing post primary TB is characterized and effective resistance to new infection. by bronchial obstruction that produces post ob- Granulomas don’t occur in developing post prima- structive pneumonia that can cavitate [12,13,48, ry TB because, there is sufficient immunity from 51,95,96]. primary TB to prevent growth of MTB in viable 5. TB antigens accumulate within cells in areas tissue. This protects the entire body except for the with very few AFB [21,64-66,97]. vulnerable region of the lung that has special mechanisms to facilitate transmission of MTB to new The value of scientific theory is often judged by hosts. how well it explains natural phenomena. The o-pr posed pathogenesis of post primary TB suggests Post primary TB does not result from a lack of sys- answers to long standing questions. temic immunity, but rather from local susceptibility in the apex of the lung. It begins in the part of • How can MTB be an obligate human para- lung with the lowest ventilation, perfusion and site when people are more resistant than any of movement and proceeds to paralyze alveolar micro- the animals studied? MTB is an obligate human phages, blockade the exits and suppress inflamma- parasite because it can only be transmitted to new tion to further isolate the area in order to provide a hosts by people. Humans are more resistant to TB place for prolonged accumulation of host lipids and because most develop effective immunity against mycobacterial products. At some point, there is a primary TB in weeks whereas most animals die sudden onset of a necrotizing hypersensitivity reac- within months of progressive disease. This benefits tion to the accumulated products. This appears to the organism because it maintains the health of the be a combination of perifocal inflammation due to host who can then spread infection frequently for protein antigens and caseous necrosis due to cord decades [27,98]. Only humans develop post pri- factor. Once induced, the caseous pneumonia may mary TB that progresses to pulmonary cavities either fragment to produce cavities or harden to from which infection can be transmitted to new produce fibrocaseous TB. There has been much hosts. 380 Annals of Clinical & Laboratory Science, vol. 44, no. 4, 2014 • What is the nature of the immunity that pro- Damage to bronchi probably facilitates mainte- tects most people from post primary TB? People nance of bronchial obstruction by TB. Modified do not develop post primary TB unless they have immune responses may contribute, but there is lit- sufficient immunity to heal primary TBe- andpr tle data. vent new infection in every part of their bodies except the vulnerable area in the apex of the lung. The • What is the role of immunosuppression in appropriate question, therefore, is: How do MTB post primary TB? Suppression of T cell function manipulate an otherwise effective host responses in by steroid drugs, HIV or old age reactivates latent this particular part of the lung? MTB selects the infection and increases susceptibility to disseminat- least active part of lung and uses an otherwise effec- ed TB. This is frequently extrapolated to imply that tive immune response to isolate it further and mod- all reactivation of TB is initiated by immunosup- ulate macrophages to accumulate host and myco- pression. However, there is little evidence that peo- bacterial products quietly in preparation for an ple with pulmonary TB are immunosuppressed. abrupt and massive necrotizing reaction. Since pro- They are not more susceptible to other viral or bac- gression to disease is a rare event, resistance can be terial infections. They are not even more susceptible due to any of the many factors that could prevent to other intracellular bacterial infections [100]. In developing conditions for the necrotizing reaction. addition, people with suppression of T cell function are less likely to develop cavities and spread the • How can multiple pulmonary lesions in a disease to others than immunocompetent young single lung act independently as if the others did adults [101-103]. In fact, young adults with the not exist? Patients with chronic TB frequently have strongest immune responses are the most likely to multiple stages of TB lesions in their lungs that de- have apical localization of infection, to produce velop independently of one another [10,11,15,27]. cavities and to die of acute disease [10,27]. A strong Many stages of lesions in post primary TB may co- immune response is necessary to produce a cavity. exist in a single lung because they are controlled Once a cavity is produced, a strong immune re- locally, particularly by bronchial obstruction that sponse is needed to keep the massive number of produces stasis and facilitates storage of lipids and MTB produced in the cavity from spreading infec- antigens. Obstruction is a local process that can be- tion to the rest of the body and killing its host. gin in different parts of the lung at different times. Immunosuppression may be necessary to reactivate In contrast, the lesions in persons with primary latent MTB and to get organisms to the apex of the (miliary) or disseminated TB all look similar to one lung. From then on its role is unclear. another because they are controlled by systemic immunity. • What are the roles of hypersensitivity and immunity? It has long been known that much of • Why doesn’t recovery from post primary TB the tissue damage in TB cannot be attributed di- produce immunity? Recovery from most infections rectly to the organisms, but is a hypersensitivity results in immunity. TB is different. “To have that resembles the Koch phenomenon [10,72,86, passed through a period of high mortality risk con- 104]. Possible roles of hypersensitivity in immunity fers not protection, but added hazard in later life.” have been widely discussed [15,37]. The studies (The White Plague, Dubos [28]). Recent studies presented herein suggest that hypersensitivity and report that people who have recovered from TB are immunity are separate and distinct phenomena. highly susceptible to new infections rather than re- Primary TB stimulates immunity, the ability to reactivation of previous infections [5,99]. Such peo- cruit and activate macrophages sufficiently to con- ple are only susceptible to new post primary TB, trol and kill MTB and prevent new infection. This and not to primary or disseminated TB. Immunity immunity may be weakened by immunosuppres- typically remains sufficient to protect the rest of sion such as HIV to allow disseminated infection. their bodies. Susceptibility in such persons is prob- However, it typically persists and protects the body ably due to production of more vulnerable areas of from new infection in every place except for the lung due to disruption of bronchial structures and vulnerable region of the lung where MTB has es- clearance mechanisms by the previous lesions. tablished a unique environment for production of Pathogenesis of Post Primary Tuberculosis 381 cavities that facilitate transmission to new hosts. fatal [72]. Tuberculous pneumonia has developed Immunity protects the host from systemic infec- overnight following surgery for TB [13]. Perifocal tion. Hypersensitivity reactions destroy lung tissue reactions in the form of a sudden increase in the to produce a cavity. The organism needs both in size of pulmonary lesions were probably the result order to survive. If there is no cavity, MTB can of dissemination of tuberculoprotein. [72]. Cord never escape to a new host. If here is no systemic factor produced by MTB is far more toxic that that immunity, MTB will produce disseminated infec- produced by other mycobacteria [108]. MTB cord tion and kill the host thereby killing itself. factor also has hypersensitivity epitopes that are not present on BCG [109]. Furthermore, mixing the MTB is an obligate human parasite because it can MTB with oil in Freund’s adjuvant induces a far only be transmitted from host to host by humans. more severe and long lasting toxicity in tuberculous Consequently, all of its genes have been selected to humans than that the MTB materials alone. This is enhance survival and transmission in people. probably relevant since caseation necrosis consists Everything that it makes or does has been selected of mycobacterial antigens in a lipid matrix. for that one purpose. MTB is an ancient human pathogen that can persist indefinitely in very small Given the nature of natural selection, epitopes are populations of as little as 200 people [105]. The hyper conserved because they help the organism lifecycle of MTB is to infect a person, frequently a survive in their only natural host, people. To do child, and then hide for 10-30 years before reacti- this, they must have two properties. The first is to vating to produce a cavity in the lung from which produce particularly necrotizing tissue reactions. organisms are coughed up over a period of decades The second is that they must remain hidden until to infect new individuals. needed. Premature release of MTB antigens would not stimulate a large enough reaction to produce a Highly conserved immunologic epitopes. Many cavity from which MTB could escape to infect new microbial antigens are highly variable in order to hosts. There is evidence that hiding of antigens is evade immune responses. MTB is different. Musser accomplished via multiple mechanisms. First, the reported remarkably little variation in antibody epi- isolation and blockade of sections of the lung pro- topes of MTB [106]. In 2012, Comas et al reported vides a protected environment for MTB. Second, that the T epitopes of MTB are hyper conserved. antigens detectable by immunohistochemistry are [7]. More than 95% of the 491 individual epitopes located exclusively in live cells in alveoli. Third, the analyzed had identical amino acid sequences. The mycobacteria produce anti-inflammatory agents researchers speculated that MTB needs these epit- like man-LAM and sulfolipid [80, 110]. Finally, as opes because they induce particularly toxic reac- will be discussed later, cord factor is non-toxic until tions that are necessary for tissue destruction and its toxicity is activated by interaction with appro- the formation of cavities that mediate transmission priate lipids. It seems reasonable to hypothesize, of infection. therefore, that the ability to cause sudden onset of caseous pneumonia depends both on the toxicity of Several pieces of evidence support this hypothesis. the hypersensitivity reactions and the ability to hide First, infected humans are much more sensitive to the antigens until sufficient quantities are available tuberculin than the guinea pig or other animals to produce a cavity able to mediate transmission of [85]. Tuberculin test sites in humans with TB fre- infection to new hosts. quently become necrotic. Necrosis does not occur in animals or in people with positive skin-tests in- Cord factor as the trigger for caseous pneumonia: duced by BCG or tuberculoid leprosy [107]. MTB In the 1950’s, Middlebrook identified two proper- shares epitopes with some saprophytic mycobacte- ties of virulent MTB that distinguished them from ria. These shared mycobacterial epitopes evoke mild avirulent mycobacteria [111]. The first was forma- skin-test reactivity in TB patients [107]. Cases have tion of elongated aggregates of organisms, called been reported in which tuberculin skin tests pro- serpentine cords. The second was staining with duced pulmonary perifocal inflammation that was neutral red. Cord formation is due to cord factor or 382 Annals of Clinical & Laboratory Science, vol. 44, no. 4, 2014 trehalose 6’6 dimycolate (TDM) [112,113]. a hydrophobic surface. When virulent MTB con- Neutral red staining is largely due to sulfolipid that tact lipid droplets, they instantly enter the droplet is an inhibitor of the toxicity of cord factor shedding the TDM that assumes the monolayer [110,114,115]. It is probably more than coinci- configuration on the water-lipid interface. The dence that two long recognized characteristics of monolayer of TDM is stable with a half life of 4 virulent MTB, cord formation and neutral red days as a single molecular monolayer in mice staining relate to TDM and its sulfolipid inhibitor. [123,124]. It is highly toxic with an LD50 of 30 µg and produces foreign body granulomas in naïve TDM is the leading candidate for triggering sud- mice and hypersensitive granulomas in immunized den onset of caseous necrosis because its potent mice [19,126,127]. The hypersensitivity reactions ability to stimulate release of TNF from macro- are dependent on CD1d and CD4+ T cells [19]. phages depends on its physical conformation. Injections of 10 µg of TDM emulsion every other While it is found in all mycobacteria, the quantities day induce hemorrhagic pneumonitis, cachexia and excreted, chemical structure, toxicity and antigenic- death of C57BL1/6 mice [117]. Much larger doses, ity of that produced by MTB are unique. TDM is 50-100 µg are rarely lethal. the only chemically defined substance ever to be shown to produce caseation necrosis [18]. It is the The key property of TDM that makes it a candidate most abundant and by orders of magnitude most trigger for caseous pneumonia is its ability to the toxic lipid of MTB [116]. While MTB contains change configuration to suddenly stimulate release many lipids, TDM is unique in being free on the of large amounts of TNF. Geisel et al demonstrated bacterial surface in large quantity. Extraction of lip- that the ability of TDM to induce release of TNF ids from the surface of viable virulent MTB yields from macrophages increases with particle size nearly pure TDM [116,117]. Removal of TDM [128]. TDM on the surface of 90 micron diameter from the surface of MTB reduces virulence for beads induced release over 2600 pg/ml TNF while mice. Adding purified TDM back restores the viru- the same amount on the same surface area of 1 mi- lence [118]. TDM generated controversy immedi- cron beads induces less than 100 pg/ml. In retro- ately after its discovery because its toxic effects were spect, similar observations had been made earlier. only demonstrable when it was injected with oil Bloch demonstrated that injection of TDM in oil is that was considered ‘unphysiologic’ [117,119]. far more toxic than injection of the same materials More recent studies have suggested that rather than as an oil-in -water emulsion [112,116]. Yarkoni being unphysiologic, interactions of TDM with demonstrated that oil emulsions of TDM with lipids are central to the pathogenesis of TB [90]. large particle size were far more effective as cancer immunotherapeutic agents than those with the The most striking feature of TDM is that ithas same qualities in small droplets [129]. We demon- multiple distinct sets of biologic activities that de- strated that small beads coated with TDM are in- pend upon its physical configuration. TDM is so gested by macrophages and induced non-allergic large and insoluble that it remains as a thick coat on granulomas. TDM on a plate (very large bead), in the organisms. As such, it prevents phagosome lyso- contrast, causes macrophages to adhere, spread and some fusion, prevents acidification of phagosomes, lyse within minutes [130]. protects organisms from killing by macrophages and impedes antigen presentation [120,121]. In sa- The hypothesis is that post primary TB isolates a line, TDM is nontoxic; the LD50 for mice is great- section of lung to facilitate the buildup of host lip- er than 50,000 µg, the highest dose ever injected. ids and mycobacterial products to generate condi- TDM removed from MTB and placed as a single tions for a hypersensitive reaction that produces molecular monolayer on the surface of lipid drop- caseation pneumonia and cavitation [107,131]. lets has a completely different physical structure Lipid progressively accumulates with increasing and set of biologic activities [122,123]. It spontane- droplet size in post obstructive pneumonia. The ously forms a two dimensional crystalline mono- trigger for initiating caseous pneumonia may be an layer on oil-water interfaces that is the most rigid increase in the size of lipid droplets coated with lipid monolayer known [122-125]. The monolayer TDM to the point that they activate release of large spontaneously forms on any interface of water with amounts of TNF that drives a positive feedback for Pathogenesis of Post Primary Tuberculosis 383 the rest of the lesions. In support of this, we have many models could yield valuable insights into par- observed that injection of a TDM emulsion into ticular aspects the disease. The challenge is to de- mice with slowly progressive pulmonary TB induc- fine models that address relevant aspects of the hu- es production of larger lipid droplets, greater in- man disease. flammation, necrosis and proliferation of organisms. Vaccine candidates: Many investigators believe that a vaccine that acts against pulmonary TB is the Animal Experimental Models: Most studies using only hope of eventually eradicating MTB [8]. The animal models of TB have focused on understand- current vaccine, BCG, provides a degree of protec- ing the early stages of granuloma formation. A mation against disseminated TB, saving the lives of ap- jor gap is a failure to address later stages leading to proximately 50 thousand children every year. transmission [132]. This gap cannot be addressed However, it has no measurable impact on post pri- without a better understanding of pathogenesis. mary TB or transmission of infection. The funda- Our observations suggest that most, if not all, ani- mental problem is that one can’t rationally develop mals develop components of both primary and post an effective vaccine for post primary TB without primary TB. For example, the late stages of TB in understanding protective immunity [134]. rabbits, mice and guinea pigs are all similar to human TB in that disease is not driven by increasing Vaccine development projects typically focus on re- numbers of MTB, but by a host response to myco- ducing the numbers of MTB in tissue and prolong- bacterial products [104]. Humans, however, are ing the life of animals. This is appropriate for pri- the only species in which the lesions progress to mary TB, but not for post primary disease because produce cavities that mediate transmission of infec- the disease has very few MTB during developmen- tion to new hosts. Most current literature assumes tal stages and needs none to produce disease. There that caseating granulomas are the characteristic le- has long been realization that the immune response sion of both primary and post primary TB. is responsible for tissue damage as well as control- ling proliferation of MTB. Targeting latent MTB A central point of this paper is that granulomas are or those active in early post primary disease could characteristic of primary TB, but are not involved be effective. However, current, animal models of in developing post primary disease. The concept protection from primary TB are not helpful. They that the TB in guinea pigs is more human like than only serve to try to induce animals to produce the that in mice because it develops large caseating type of response that most humans develop sponta- granulomas is incorrect. Large caseating granulo- neously in weeks. The challenge for preventing post mas are characteristic only of progressive primary primary TB is to find ways to prevent MTB from TB in infants. Mice are regularly criticized because co-opting and using our strongest response for its they fail to produce caseating granulomas when, in advantage. This may be possible since progression fact, slowly progressive TB in mice is a rather good of developing post primary TB is a rare event. At model of parts of the early stages of post primary least 95% of early infections regress spontaneously. TB. In addition, many reviewers and funding agen- The people who develop disease do not have weaker cies insist that animals be infected by low dose aero- or less effective immune responses. MTB uses and sol infection to reproduce the conditions of human subverts immunity to produce an isolated area of infection. This reproduces only the first few weeks lung in which it can slowly accumulate host lipids of infection and has little or no relevance for post and mycobacterial antigens necessary to produce a primary TB that occurs decades later. As exempli- necrotizing hypersensitivity reaction. A vaccine that fied by our studies with caseating granulomas in blocks any one of multiple factors required for this mice, one can develop protocols to reproduce in process could produce regression. animals the conditions and lesions that exist in humans at different stages of infection [18]. A recent The Streetlight Effect states that one can only ef- publication used this approach to produce cavitary fectively study things that can be seen [94]. The lesions in rabbits [133]. While there is no complete capabilities of modern science have increased phe- animal model of post primary TB, we believe that nomenally since tissues with post primary TB 384 Annals of Clinical & Laboratory Science, vol. 44, no. 4, 2014 ceased to be readily obtainable due to introduction 13. Levine, E. R. Chapter 7, Classification of reinfection pulmonary tuberculosis. In The Fundamentals of Pulmonary of antibiotics and declining autopsy rates. This pa- Tuberculosis and its Complications for Students, Teachers and per presents a synthesis of old and new investiga- Practicing Physicians. E. Hayes, ed. Charles C Thomas, tions to produce a revised conception of the patho- Springfield. 1949. 97-113. 14. Pagel, W., F. Simmonds, N. MacDonald, and E. Nassau. genesis of post primary TB. Progression of infection Chapter 3. The Morbid Anatomy and Histologyof Tuberculosis, from implantation of MTB in a lung to clinical An Introduction in Simple Terms. In In Pulmonary disease is a rare event because most infections re- Tuberculosis, Bacteriology, Pathology, Management, Epidemiology and Prevention. Oxford University Press, solve spontaneously. Examination of the process London. 1964. 36-63. suggests many places where it might be interrupted 15. Dannenberg, A. M., Jr. Pathogenisis of human pulmonary tuberculosis. Insights from the rabbit model. 2006 ASM press, thereby preventing clinical disease and transmission Washington, DC. of infection to new hosts. We anticipate that this 16. Creighton, C. Bovine Tubercuosis in Man: An account of the will move the streetlight to facilitate use of modern Pathology of Suspected Cases. 1881 MacMillan and Co., London. technologies to rapidly address previously refracto- 17. Liebana, E., L. Johnson, J. Gough, P. Durr, K. Jahans, R. ry questions of the world’s most lethal bacterial Clifton-Hadley, Y. Spencer, R. G. Hewinson, and S. H. pathogen. Downs. Pathology of naturally occurring bovine tuberculosis in England and Wales. Vet J 2008. 176: 354-360. 18. Hunter, R. L., M. Olsen, C. Jagannath, and J. K. Actor. Acknowledgements Trehalose 6,6’-dimycolate and lipid in the pathogenesis of caseating granulomas of tuberculosis in mice. Am J Pathol 2006. Supported in part by USPHS grants HL 68537, AI78420, 168: 1249-1261. AI49534, HL080313. 19. Guidry, T. V., R. L. Hunter, Jr., and J. K. Actor. Mycobacterial glycolipid trehalose 6,6’-dimycolate-induced hypersensitive References granulomas: contribution of CD4+ lymphocytes. Microbiology 2007. 153: 3360-3369. 1. World_Health_Organization. Global tuberculosis report 20. Guidry, T. V., M. Olsen, K. S. Kil, R. L. Hunter, Jr., Y. J. Geng, 2013. 2013. and J. K. Actor. Failure of CD1D-/- mice to elicit hypersensi- 2. Russell, D. G., C. E. Barry, 3rd, and J. L. Flynn. Tuberculosis: tive granulomas to mycobacterial cord factor trehalose 6,6’-di- what we don’t know can, and does, hurt us. Science 2010. 328: mycolate.J Interferon Cytokine Res 2004. 24: 362-371. 852-856. 21. Hunter, R. L., C. Jagannath, and J. K. Actor. Pathology of 3. Hunter, R. L. Pathology of post primary tuberculosis of the postprimary tuberculosis in humans and mice: contradiction lung: An illustrated critical review. Tuberculosis (Edinb)E of long-held beliefs. Tuberculosis (Edinb) 2007. 87: 267-278. publication ahead of print July 4, 2011. 22. Ehlers, S. Lazy, dynamic or minimally recrudescent? On the 4. Russell, D. G. Who puts the tubercle in tuberculosis? Nat Rev elusive nature and location of the mycobacterium responsible Microbiol 2007. 5: 39-47. for latent tuberculosis. Infection 2009. 37: 87-95. 5. den Boon, S., S. W. van Lill, M. W. Borgdorff, D. A. Enarson, 23. Lin, P. L., C. B. Ford, M. T. Coleman, A. J. Myers, R. S. Verver, E. D. Bateman, E. Irusen, C. J. Lombard, N. W. Gawande, T. Ioerger, J. Sacchettini, S. M. Fortune, and J. L. White, C. de Villiers, and N. Beyers. High prevalence of tuber- Flynn. Sterilization of granulomas is common in active and la- culosis in previously treated patients, Cape Town, South tent tuberculosis despite within-host variability in bacterial Africa. Emerg Infect Dis 2007. 13: 1189-1194. killing. Nat Med 2014. 20: 75-79. 6. van Rie, A., T. C. Victor, M. Richardson, R. Johnson, G. D. 24. Neyrolles, O., R. Hernandez-Pando, F. Pietri-Rouxel, P. van der Spuy, E. J. Murray, N. Beyers, N. C. Gey van Pittius, Fornes, L. Tailleux, J. A. Barrios Payan, E. Pivert, Y. Bordat, D. P. D. van Helden, and R. M. Warren. Reinfection and Mixed Aguilar, M. C. Prevost, C. Petit, and B. Gicquel. Is adipose Infection Cause Changing Mycobacterium tuberculosis Drug- tissue a place for Mycobacterium tuberculosis persistence? PloS resistance Patterns. Am J Respir Crit Care Med 2005. one 2006. 1: e43. 7. Comas, I., J. Chakravartti, P. M. Small, J. Galagan, S. 25. Agarwal, P., S. R. Khan, S. C. Verma, M. Beg, K. Singh, K. Niemann, K. Kremer, J. D. Ernst, and S. Gagneux. Human T Mitra, A. N. Gaikwad, M. S. Akhtar, and M. Y. Krishnan. cell epitopes of Mycobacterium tuberculosis are evolutionarily Mycobacterium tuberculosis persistence in various adipose de- hyperconserved. Nature genetics 2010. 42: 498-503. pots of infected mice and the effect of anti-tubercular therapy. 8. Brennan, M. J., and J. Thole. Tuberculosis vaccines: a strategic Microbes Infect 2014. 16: 571-580. blueprint for the next decade. Tuberculosis (Edinb) 2012. 92 26. Hadda, V., and G. C. Khilnani. Lipoid pneumonia: an over- Suppl 1: S6-13. view. Expert review of respiratory medicine 2010. 4: 799-807. 9. Laennec, R. A treatise on diseases of the chest in which they are 27. Osler, W. Chapter 26, Tuberculosis. In The principles and described according to their anatomical characters, and their practice of medicine. D. Appleton and Company, New York. diagnosis established on a new principle by means of acoustick 1892. 184-255. instruments. 1821 T&G Underwood, London (reprinted 1979 28. Dubos, R., and G. Dubos. The White Plague. Tuberculosis, by The Classics of Medicine Library. Birmingham AL). Man, and Society. 1987 Rutgers Univesity Press, Brunswick, 10. Rich, A. The Pathogenesis of Tuberculosis, Second Edition. NJ. 1951 Charles C Thomas, Springfield, Illinois. 29. Erokhin, V. V., V. Y. Mishin, V. I. Chukanov, and D. V. 11. Canetti, G. The tubercle bacillus in the pulmonary lesion of Guiller. Caseous pneumonia, Pathologic anatomy, pathogene- man. Histobacteriology and its bearing on the therapy of pul- sis, diagnosis, clinical course and treatment, A manual for prac- monary tuberculosis. 1955 Springer Publishing Compani Inc., titioners 2008 Meditsina Publishers, Moscow. New Yoyk. 30. Goodwin, R. A., and R. M. Des Prez. Apical localization of 12. Medlar, E. M. The behavior of pulmonary tuberculous lesions; pulmonary tuberculosis, chronic pulmonary histoplasmosis, a pathological study. Am Rev Tuberc 1955. 71: 1-244. and progressive massive fibrosis of the lung. Chest 1983. 83: 801-805. Pathogenesis of Post Primary Tuberculosis 385

31. Pai, M., A. Mohan, K. Dheda, C. C. Leung, W. W. Yew, D. J. 48. Hektoen, L., and D. Reisman. A Text-Book of Pathology for Christopher, and S. K. Sharma. Lethal interaction: the collid- the use of students and practitoners of medicine and surgery. ing epidemics of tobacco and tuberculosis. Expert review of 1901 W.B. Saunders & Company, London and Philadelphia. anti-infective therapy 2007. 5: 385-391. 49. Coats, J., and L. K. Sutherland. Manual Of Pathology, Chapter 32. Murray, J. F. Bill Dock and the location of pulmonary tubercu- 5. Healing Of Phthisis. 1900 Longmans, Green, And Co, losis: how bed rest might have helped consumption. Am J London. Respir Crit Care Med 2003. 168: 1029-1033. 50. Im, J. G., H. Itoh, K. S. Lee, and M. C. Han. CT-pathology 33. Broch, B. L. Chapter 14, The role of the bronchial tree in the correlation of pulmonary tuberculosis. Crit Rev Diagn Imaging pathogenesis of pulmonary tuberculosis. In The Fundamentals 1995. 36: 227-285. of Pulmonary Tuberculosis and its Complications for Students, 51. Lee, J. Y., K. S. Lee, K. J. Jung, J. Han, O. J. Kwon, J. Kim, and Teachers and Practicing Physicians. E. Hayes, ed. Charles C T. S. Kim. Pulmonary tuberculosis: CT and pathologic correla- Thomas, Springfield. 1949. 245-253. tion.J Comput Assist Tomogr 2000. 24: 691-698. 34. Ehrhardt, J., R. Werner, A. Schmidt-Richberg, and H. 52. Kondo, S., T. Miyagawa, and M. Ito. [Reevaluation of patho- Handels. Statistical modeling of 4D respiratory lung motion genesis of epituberculosis in infants and children with tubercu- using diffeomorphic image registration. IEEE transactions on losis]. Kekkaku 2007. 82: 569-575. medical imaging 2011. 30: 251-265. 53. Fish, R., and W. Pagel. The morbid anatomy of epituberculosis. 35. Du, K., J. E. Bayouth, K. Cao, G. E. Christensen, K. Ding, and J Path Bact 1938. 47: 593-601. J. M. Reinhardt. Reproducibility of registration-based mea- 54. Amita, R., S. Sandhyamani, and M. Unnikrishnan. Extensive sures of lung tissue expansion. Medical physics 2012. 39: unilateral pulmonary tuberculosis with segmental atresia of 1595-1608. principal bronchus. Lung India : official organ of Indian Chest 36. Lam, C. W., J. T. James, R. McCluskey, and R. L. Hunter. Society 2014. 31: 94-95. Pulmonary toxicity of single-wall carbon nanotubes in mice 7 55. Ahonen, A., and K. Valavirta. Ultrastructure of cilia in pulmo- and 90 days after intratracheal instillation. Toxicol Sci 2004. nary tuberculosis. European journal of respiratory diseases. 77: 126-134. Supplement 1983. 128 (Pt 2): 460-463. 37. O’Garra, A., P. S. Redford, F. W. McNab, C. I. Bloom, R. J. 56. Reed, G. W., H. Choi, S. Y. Lee, M. Lee, Y. Kim, H. Park, J. Wilkinson, and M. P. Berry. The immune response in tubercu- Lee, X. Zhan, H. Kang, S. Hwang, M. Carroll, Y. Cai, S. N. losis. Annu Rev Immunol 2013. 31: 475-527. Cho, C. E. Barry, 3rd, L. E. Via, and H. Kornfeld. Impact of 38. Korf, J., A. Stoltz, J. Verschoor, P. De Baetselier, and J. Grooten. diabetes and smoking on mortality in tuberculosis. PloS one The Mycobacterium tuberculosis cell wall component mycolic 2013. 8: e58044. acid elicits pathogen-associated host innate immune responses. 57. Davies, P. D., W. W. Yew, D. Ganguly, A. L. Davidow, L. B. Eur J Immunol 2005. 35: 890-900. Reichman, K. Dheda, and G. A. Rook. Smoking and tubercu- 39. Peyron, P., J. Vaubourgeix, Y. Poquet, F. Levillain, C. Botanch, losis: the epidemiological association and immunopathogene- F. Bardou, M. Daffe, J. F. Emile, B. Marchou, P. J. Cardona, C. sis. Trans R Soc Trop Med Hyg 2006. 100: 291-298. de Chastellier, and F. Altare. Foamy macrophages from tuber- 58. Farver, C., A. E. Fraire, and J. F. Tomashefski. Dail and culous patients’ granulomas constitute a nutrient-rich reservoir Hammar’s Pulmonary Pathology, Volume 1: Nonneoplastic for M. tuberculosis persistence. PLoS Pathog 2008. 4: Lung Disease 1998 Springer Verlag New York e1000204. 59. Pagel W, and P. W. Zur Histochemie der Lungentuberkulose, 40. Korf, J. E., G. Pynaert, K. Tournoy, T. Boonefaes, A. Van mit besonderer Berucksichtung der Fettsubstanzen und Oosterhout, D. Ginneberge, A. Haegeman, J. A. Verschoor, P. Lipoide. (Fat and lipoid content to tuberculous tissue. De Baetselier, and J. Grooten. Macrophage reprogramming by Histochemical investigation.). . Virchows Arch Pathol Anat mycolic acid promotes a tolerogenic response in experimental 1925. 256: 629-640. asthma. Am J Respir Crit Care Med 2006. 174: 152-160. 60. Johnson, J. L., and J. J. Ellner. Chapter 57. Cavitary Pulmonary 41. Melo, R. C., and A. M. Dvorak. Lipid body-phagosome inter- Disease. In Clinical Infectious Diseases: A Practical Approach. action in macrophages during infectious diseases: host defense R. K. Root, ed. Oxford University Press, New Hork. 1999. or pathogen survival strategy? PLoS Pathog 2012. 8: e1002729. 539-549. 42. North, R. J., and Y. J. Jung. Immunity to tuberculosis. Annu 61. Akinosoglou, K. S., K. Karkoulias, and M. Marangos. Rev Immunol 2004. 22: 599-623. Infectious complications in patients with lung cancer. 43. Nuermberger, E. L., T. Yoshimatsu, S. Tyagi, W. R. Bishai, and European review for medical and pharmacological sciences J. H. Grosset. Paucibacillary tuberculosis in mice after prior 2013. 17: 8-18. aerosol immunization with Mycobacterium bovis BCG. Infect 62. Hunter, R. L. On the pathogenesis of post primary tuberculo- Immun 2004. 72: 1065-1071. sis: the role of bronchial obstruction in the pathogenesis of 44. Kondo, A., N. Oketani, M. Maruyama, Y. Taguchi, Y. cavities. Tuberculosis (Edinb) 2011. 91 Suppl 1: S6-10. Yamaguchi, H. Miyao, I. Mashima, M. Oono, K. Wada, T. 63. Pattnaik, M. K., P. C. Majhi, A. K. Nayak, and D. Senapati.A Tsuchiya, H. Takahashi, and S. Abe. [Significance of serum rare presentation of a huge mature mediastinal teratoma with surfactant protein-D (SP-D) level in patients with pulmonary right lung cavitation. BMJ case reports 2014. 2014. tuberculosis]. Kekkaku 1998. 73: 585-590. 64. Humphrey, D. M., and M. H. Weiner. Mycobacterial antigen 45. Chroneos, Z. C., K. Midde, Z. Sever-Chroneos, and C. detection by immunohistochemistry in pulmonary tuberculo- Jagannath. Pulmonary surfactant and tuberculosis. sis. Hum Pathol 1987. 18: 701-708. Tuberculosis (Edinb) 2009. 89 Suppl 1: S10-14. 65. Ridley, M. J., and D. S. Ridley. Histochemical demonstration 46. Wang, Z., U. Schwab, E. Rhoades, P. R. Chess, D. G. Russell, of mycobacterial antigen, specific antibody and complement in and R. H. Notter. Peripheral cell wall lipids of Mycobacterium the lesions of tuberculosis. The Histochemical journal 1986. tuberculosis are inhibitory to surfactant function. Tuberculosis 18: 551-556. (Edinb) 2008. 88: 178-186. 66. Pedersen, J. S., I. Clarke, and J. Mills. Improved detection of 47. Lovacheva, O. V., V. V. Erokhin, N. V. Chernichenko, G. V. mycobacteria species in formalin-fixed tissue sections. Evgushchenko, L. N. Lepekha, and O. A. Rozenberg. [Results Histopathology 2011. 59: 993-1005. of use of surfactant in complex therapy of patients with de- 67. Guerardel, Y., E. Maes, V. Briken, F. Chirat, Y. Leroy, C. structive pulmonary tuberculosis]. Probl Tuberk Bolezn Legk Locht, G. Strecker, and L. Kremer. Lipomannan and lipoarabi- 2006. 12-17. nomannan from a clinical isolate of Mycobacterium kansasii: 386 Annals of Clinical & Laboratory Science, vol. 44, no. 4, 2014

novel structural features and apoptosis-inducing properties.J 85. Birnbaum, M. Prof. Koch’s Method to Cure Tuberculosis Biol Chem 2003. 278: 36637-36651. Popularly Treated. In PROJECT GUTENBERG EBOOK 68. Chan, J., X. D. Fan, S. W. Hunter, P. J. Brennan, and B. R. METHOD TO CURE TUBERCULOSIS. http://www. Bloom. Lipoarabinomannan, a possible virulence factor in- gutenberg.net EBook #27181. 2008. volved in persistence of Mycobacterium tuberculosis within 86. Moreira, A. L., L. Tsenova, M. H. Aman, L. G. Bekker, S. macrophages. Infect Immun 1991. 59: 1755-1761. Freeman, B. Mangaliso, U. Schroder, J. Jagirdar, W. N. Rom, 69. Talaat, A. M., S. K. Ward, C. W. Wu, E. Rondon, C. Tavano, M. G. Tovey, V. H. Freedman, and G. Kaplan. Mycobacterial J. P. Bannantine, R. Lyons, and S. A. Johnston. Mycobacterial antigens exacerbate disease manifestations in Mycobacterium bacilli are metabolically active during chronic tuberculosis in tuberculosis-infected mice. Infect Immun 2002. 70: murine lungs: insights from genome-wide transcriptional pro- 2100-2107. filing. J Bacteriol 2007. 189: 4265-4274. 87. Medlar, E. M. The pathogenesis of minimal pulmonary tuber- 70. Rogerson, B. J., Y. J. Jung, R. LaCourse, L. Ryan, N. Enright, culosis; a study of 1,225 necropsies in cases of sudden and un- and R. J. North. Expression levels of Mycobacterium tubercu- expected death. Am Rev Tuberc 1948. 58: 583-611. losis antigen-encoding genes versus production levels of anti- 88. Wayne, L. G., and C. D. Sohaskey. Nonreplicating persistence gen-specific T cells during stationary level lung infection in of mycobacterium tuberculosis. Annu Rev Microbiol 2001. 55: mice. Immunology 2006. 118: 195-201. 139-163. 71. Bhalla, A., M. Mahapatra, R. Singh, and S. D’Cruz. Acute 89. Hurford, J. V., and W. H. Valentine. Viable tubercle bacilli in lung injury in miliary tuberculosis. Indian Journal of closed lesions. Tubercle 1957. 38: 194-198. Tuberculosis 2002. 49: 125-128. 90. Hunter, R. L., M. R. Olsen, C. Jagannath, and J. K. Actor. 72. Lincoln, E. M., and W. Grethmann. The potetial dangers of Multiple roles of cord factor in the pathogenesis of primary, tuberculin tests. J Pediatrics 1939. 15: 682-696. secondary, and cavitary tuberculosis, including a revised de- 73. Massaro, D., and S. Katz. Rapid Clearing in Hematogenous scription of the pathology of secondary disease. Ann Clin Lab Pulmonary Tuberculosis. Archives of internal medicine 1964. Sci 2006. 36: 371-386. 113: 573-577. 91. Rook, G. A., K. Dheda, and A. Zumla. Immune responses to 74. Trapnell, B. C., B. C. Carey, K. Uchida, and T. Suzuki. tuberculosis in developing countries: implications for new vac- Pulmonary alveolar proteinosis, a primary immunodeficiency cines. Nat Rev Immunol 2005. 5: 661-667. of impaired GM-CSF stimulation of macrophages. Current 92. Dheda, K., H. Booth, J. F. Huggett, M. A. Johnson, A. Zumla, opinion in immunology 2009. 21: 514-521. and G. A. Rook. Lung remodeling in pulmonary tuberculosis. 75. Carey, B., and B. C. Trapnell. The molecular basis of pulmo- J Infect Dis 2005. 192: 1201-1209. nary alveolar proteinosis. Clin Immunol 2010. 135: 223-235. 93. Chapel, H. M., and P. J. August. Report of nine cases of acci- 76. Ramirez, J. Pulmonary alveolar proteinosis. Treatment in a dental injury to man with Freund’s complete adjuvant. Clin case complicated by tuberculosis. Am Rev Respir Dis 1967. 95: Exp Immunol 1976. 24: 538-541. 491-495. 94. Freedman, D. H. Why Scientific Studies Are So Often Wrong: 77. Huang, X. Y., C. Yu, X. M. Xu, W. T. Yang, X. Zhang, P. P. The Streetlight Effect. Discover 2010. July-August. Liu, and L. X. Wang. Pulmonary alveolar proteinosis associat- 95. Hatipoglu, O. N., E. Osma, M. Manisali, E. S. Ucan, P. Balci, ed with tuberculosis and aspergilloma formation. Chinese A. Akkoclu, O. Akpinar, C. Karlikaya, and C. Yuksel. High medical journal 2012. 125: 3191-3192. resolution computed tomographic findings in pulmonary tu- 78. Betancourt, S. L., S. Martinez-Jimenez, S. E. Rossi, M. T. berculosis. Thorax 1996. 51: 397-402. Truong, J. Carrillo, and J. J. Erasmus. Lipoid pneumonia: spec- 96. Collins, J., D. Blankenbaker, and E. J. Stern. CT patterns of trum of clinical and radiologic manifestations. AJR Am J bronchiolar disease: what is “tree-in-bud”? AJR Am J Roentgenol 2010. 194: 103-109. Roentgenol 1998. 171: 365-370. 79. Sulkowska, M., and S. Sulkowski. Endogenous lipid pneumo- 97. Karimi, S., M. Shamaei, M. Pourabdollah, M. Sadr, M. nia- and alveolar proteinosis-type changes in the vicinity of Karbasi, A. Kiani, and M. Bahadori. Histopathological find- non-small cell lung cancer: histiopathologic, immunohisto- ings in immunohistological staining of the granulomatous tis- chemical, and ultrastructural evaluation. Ultrastructural pa- sue reaction associated with tuberculosis. Tuberculosis research thology 1998. 22: 109-119. and treatment 2014. 2014: 858396. 80. Mustafa, T., S. Phyu, R. Nilsen, R. Jonsson, and G. Bjune. A 98. Corbett, E. L., A. Zezai, Y. B. Cheung, T. Bandason, E. Dauya, mouse model for slowly progressive primary tuberculosis. S. S. Munyati, A. E. Butterworth, S. Rusikaniko, G. J. Scand J Immunol 1999. 50: 127-136. Churchyard, S. Mungofa, R. J. Hayes, and P. R. Mason. 81. Chapman, R. C., and S. M. Claydon. Mycobacterium tubercu- Provider-initiated symptom screening for tuberculosis in losis: a continuing cause of sudden and unexpected death in Zimbabwe: diagnostic value and the effect of HIV status. Bull west London. Journal of clinical pathology 1992. 45: World Health Organ 2010. 88: 13-21. 713-715. 99. van Rie, A., R. Warren, M. Richardson, T. C. Victor, R. P. Gie, 82. Alzeer, A., A. Mashlah, N. Fakim, N. Al-Sugair, M. Al- D. A. Enarson, N. Beyers, and P. D. van Helden. Exogenous Hedaithy, S. Al-Majed, and G. Jamjoom. Tuberculosis is the reinfection as a cause of recurrent tuberculosis after curative commonest cause of pneumonia requiring hospitalization dur- treatment. N Engl J Med 1999. 341: 1174-1179. ing Hajj (pilgrimage to Makkah). J Infect 1998. 36: 303-306. 100. Huaman, M. A., C. T. Fiske, T. F. Jones, J. Warkentin, B. E. 83. Gonzalez Saldana, N., M. Macias Parra, M. Hernandez Porras, Shepherd, L. A. Ingram, F. Maruri, and T. R. Sterling. P. Gutierrez Castrellon, V. Gomez Toscano, and H. Juarez Tuberculosis and the risk of infection with other intracellular Olguin. Pulmonary Tuberculous: Symptoms, diagnosis and bacteria: a population-based study. Epidemiology and infec- treatment. 19-year experience in a third level pediatric hospital. tion 2014. 1-9. BMC infectious diseases 2014. 14: 401. 101. Ledru, E., S. Ledru, and A. Zoubga. Granuloma formation and 84. Eldour, A. A. A., M. Elfatih, R. A. A. Salih, and H. G. Ahmed. tuberculosis transmission in HIV-infected patients. Immunol Blood Coagulation Changes among Sudanese Patients with Today 1999. 20: 336-337. Pulmonary Tuberculosis. International Journal of Science and 102. Kenyon, T. A., T. Creek, K. Laserson, M. Makhoa, N. Research 2014. 3: 176-719. Chimidza, M. Mwasekaga, J. Tappero, S. Lockman, T. Moeti, Pathogenesis of Post Primary Tuberculosis 387

and N. Binkin. Risk factors for transmission of Mycobacterium during mycobacterial infection of murine macrophages. tuberculosis from HIV-infected tuberculosis patients, Microbiology 2003. 149: 2049-2059. Botswana. Int J Tuberc Lung Dis 2002. 6: 843-850. 121. Kan-Sutton, C., C. Jagannath, and R. L. Hunter, Jr. Trehalose 103. Wang, C. S., H. C. Chen, C. J. Yang, W. Y. Wang, I. W. 6,6’-dimycolate on the surface of Mycobacterium tuberculosis Chong, J. J. Hwang, and M. S. Huang. The impact of age on modulates surface marker expression for antigen presentation the demographic, clinical, radiographic characteristics and and costimulation in murine macrophages. Microbes Infect treatment outcomes of pulmonary tuberculosis patients in 2009. 11: 40-48. Taiwan. Infection 2008. 36: 335-340. 122. Behling, C. A., B. Bennett, K. Takayama, and R. L. Hunter. 104. Dannenberg, A. M., Jr., and F. M. Collins. Progressive pulmo- Development of a trehalose 6,6’-dimycolate model which ex- nary tuberculosis is not due to increasing numbers of viable plains cord formation by Mycobacterium tuberculosis. Infect bacilli in rabbits, mice and guinea pigs, but is due to a continu- Immun 1993. 61: 2296-2303. ous host response to mycobacterial products. Tuberculosis 123. Retzinger, G. S., S. C. Meredith, R. L. Hunter, K. Takayama, (Edinb) 2001. 81: 229-242. and F. J. Kezdy. Identification of the physiologically active state 105. McGrath, J. W. Social networks of disease spread in the lower of the mycobacterial glycolipid trehalose 6,6’-dimycolate and Illinois valley: a simulation approach. Am J Phys Anthropol the role of fibrinogen in the biologic activities of trehalose 1988. 77: 483-496. 6,6’-dimycolate monolayers. J Immunol 1982. 129: 735-744. 106. Musser, J. M., A. Amin, and S. Ramaswamy. Negligible genetic 124. Retzinger, G. S., S. C. Meredith, K. Takayama, R. L. Hunter, diversity of mycobacterium tuberculosis host immune system and F. J. Kezdy. The role of surface in the biological activities of protein targets: evidence of limited selective pressure. Genetics trehalose 6,6’- dimycolate. Surface properties and development 2000. 155: 7-16. of a model system.J Biol Chem 1981. 256: 8208-8216. 107. Rook, G. A., and R. Hernandez-Pando. The pathogenesis of 125. Schabbing,R.W.,A.Garcia,andR.L.Hunter.Characterization tuberculosis. Annu Rev Microbiol 1996. 50: 259-284. of the trehalose 6,6’-dimycolate surface monolayer by scanning 108. Fujita, Y., Y. Okamoto, Y. Uenishi, M. Sunagawa, T. Uchiyama, tunneling microscopy. Infect Immun 1994. 62: 754-756. and I. Yano. Molecular and supra-molecular structure related 126. Bekierkunst, A. Acute granulomatous response produced in differences in toxicity and granulomatogenic activity of myco- mice by trehalose-6,6-dimycolate.J Bacteriol 1968. 96: bacterial cord factor in mice. Microb Pathog 2007. 43: 10-21. 958-961. 109. McMullen, A. M., S. A. Hwang, K. O’Shea, M. L. Aliru, and 127. Bekierkunst, A., and E. Yarkoni. Granulomatous hypersensi- J. K. Actor. Evidence for a unique species-specific hypersensitivity to trehalose 6,6’-dimycolate (cord factor) in mice infect- tive epitope in Mycobacterium tuberculosis derived cord factor. ed with BCG. Infect Immun 1973. 7: 631-638. Tuberculosis (Edinb) 2013. 93 Suppl: S88-93. 128. Geisel, R. E., K. Sakamoto, D. G. Russell, and E. R. Rhoades. 110. Okamoto, Y., Y. Fujita, T. Naka, M. Hirai, I. Tomiyasu, and I. In Vivo Activity of Released Cell Wall Lipids of Mycobacterium Yano. Mycobacterial sulfolipid shows a virulence by inhibiting bovis Bacillus Calmette-Guerin Is Due Principally to Trehalose cord factor induced granuloma formation and TNF-alpha re- Mycolates. J Immunol 2005. 174: 5007-5015. lease. Microb Pathog 2006. 40: 245-253. 129. Yarkoni, E., and H. J. Rapp. Toxicity of emulsified trehalose 111. Middlebrook, G., R. G. Dubos, and C. Pierce. Virulence and 6,6’ dimycolate (cord factor) in mice depends on size distribu- morphological characteristics of mammalian tubercle bacilli. J. tion of mineral oil droplets. Infect. Immun. 1978. 20: Exp. Med. 1947. 86: 175-184. 856-860. 112. Bloch, H. Studies on the virulence of tubercle bacilli: Isolation 130. Syed, S. S., and R. L. Hunter, Jr. Studies on the toxic effects of and biological properties of a constituent of virulent organisms. quartz and a mycobacterial glycolipid, trehalose 6,6’-dimyco- J. Exp. Med. 1950. 91: 197-219. late. Ann Clin Lab Sci 1997. 27: 375-383. 113. Hunter, R. L., N. Venkataprasad, and M. R. Olsen. The role of 131. Ashenafi, S., G. Aderaye, A. Bekele, M. Zewdie, G. Aseffa, A. trehalose dimycolate (cord factor) on morphology of virulent T. Hoang, B. Carow, M. Habtamu, M. Wijkander, M. M. tuberculosis in vitro. Tuberculosis (Edinb) 2006. 86: Rottenberg, A. Aseffa, J. Andersson, M. Svensson, and S. 349-356. Brighenti. Progression of clinical tuberculosis is associated with 114. Goren, M., O. Brokl, B. Das, and E. Lederer. Sulfolipid I of a Th2 immune response signature in combination with elevat- Mycobacterium tuberculosis, strain H37Rv. Nature of the acyl ed levels of SOCS3. Clin Immunol 2014. 151: 84-99. substituents. Biochemistry 1971. 10: 72-81. 132. Young, D. Animal models of tuberculosis. Eur J Immunol 115. Andreu, N., I. Gibert, M. Luquin, P. E. Kolattukudy, and T. 2009. 39: 2011-2014. Sirakova. Neutral red staining of cells of a sulfolipid-deficient 133. Kubler, A., B. Luna, C. Larsson, N. C. Ammerman, B. B. Mycobacterium tuberculosis pks2 mutant proves that sulfolip- Andrade, M. Orandle, K. W. Bock, Z. Xu, U. Bagci, D. J. ids are not responsible for this cytochemical reaction.J Clin Molura, J. Marshall, J. Burns, K. Winglee, B. A. Ahidjo, L. S. Microbiol 2004. 42: 1379-1380. Cheung, M. Klunk, S. K. Jain, N. P. Kumar, S. Babu, A. Sher, 116. Goren, M. Cord Factor Revisited: A tribute to the late Hubert J. S. Friedland, P. T. Elkington, and W. R. Bishai. Bloch. Tubercle 1975. 56: 65-71. Mycobacterium tuberculosis dysregulates MMP/TIMP bal- 117. Goren, M. B., and P. J. Brennan. Chapter 4. Mycobacterial ance to drive rapid cavitation and unrestrained bacterial prolif- lipids: chemistry and biologic activities. In Tuberculosis. G. P. eration. J Pathol 2014 Sept epub ahead of print. Youmans, ed. W.B. Saunders, Philadelphia, PA. 1979. 63-193. 134. Frick, M. The TB Vaccines Pipeline. In HIV.HCV.TB Pipeline 118. Indrigo, J., R. L. Hunter, Jr., and J. K. Actor. Influence of tre- Report. Treatment Action Group http://www.pipelinereport. halose 6,6’-dimycolate (TDM) during mycobacterial infection org, New York. 2013. 263-283. of bone marrow macrophages. Microbiology 2002. 148: 135. Kondratieva, T., T. Azhikina, B. Nikonenko, A. Kaprelyants, 1991-1998. and A. Apt. Latent tuberculosis infection: What we know 119. Spitznagel, J. K., and R. J. Dubos. A fraction of tubercle bacilli about its genetic control? Tuberculosis (Edinb) 2014. possessing primary toxicity.J Exp Med 1955. 101: 291-311. 120. Indrigo, J., R. L. Hunter, Jr., and J. K. Actor. Cord factor trehalose 6,6’-dimycolate (TDM) mediates trafficking events