Transmission

Transmission of M. tuberculosis is generally from person to person and occurs via inhalation of mucous droplets that become airborne when an individual with pulmonary or laryngeal TB coughs, sneezes, speaks, laughs, or sings. After drying, the droplet nuclei can remain suspended in the air for hours. Only small droplets (<10 µm in diameter) can reach alveoli. A video from the Centers for Disease Control and Prevention illustrates the transmission process as well as the pathogenesis of the disease, and is available at https://youtu.be/9112brXCOVc . Droplet nuclei can also be produced by aerosol treatments, sputum induction, aerosolization during bronchoscopy, and through the manipulation of lesions or processing of tissue or secretions in the hospital or laboratory. Transmission occurs rarely by direct contact with infected body fluids or fomites.

Many factors are associated with the risk for acquiring M. tuberculosis infection, including the extent of contact with a contagious person, the burden of organisms in the person’s sputum and the frequency of a person’s cough. Adults with pulmonary TB and bacilli present on acid-fast staining of sputum are more likely to transmit infection. Recent evidence from adult patients suggests that the magnitude of the inhaled dose of M. tuberculosis as measured by cough aerosols can predict an individual’s risk of TB progression following exposure. In addition, the risk for transmission correlates directly with the closeness of contact and amount of time spent with a contagious case. Most transmission to children occurs in the home. Markers of close contact such as urban living and overcrowding correlate with acquisition of infection. An increased risk for infection has been demonstrated in several institutional settings, including nursing homes, schools, correctional institutions, and homeless shelters. A growing problem concerns TB transmission in refugee and orphanage settings. There is some evidence that the risk for acquiring infection increases with age from infancy to early adulthood, likely because of increasing likelihood of contact with infectious persons.

It has been noted for decades that children with TB rarely infect other children or adults. In the typical case of childhood pulmonary TB, tubercle bacilli in endobronchial secretions are sparse, and when young children with TB cough, they lack the tussive force of adults. Sputum production is rare in children, and collected specimens usually do not show acid-fast bacilli (AFB) upon staining, indicating a low concentration of organisms. However, specimens from young infants with extensive TB infiltrates, children and adolescents with cavitary lesions, or intubated children with TB are potentially infectious. The few documented cases of transmission from children have been in individuals with typical findings of adult-type TB, with lung cavities and sputum production, or infants with congenital TB who have a large burden of organisms in the lungs. When transmission of M. tuberculosis has been documented in schools, orphanages or children’s hospitals, it has almost invariably been from an adult or adolescent with undiagnosed pulmonary TB. In fact, when a child is suspected clinically of having TB disease, the adults who accompany the child should undergo urgent testing for TB (usually by chest radiograph or AFB) sputum smear to be sure they do not spread infection in the facility.

A few classic studies have investigated the factors that influence whether or not an infected person will develop TB disease. It is clear that the risk for disease is highest shortly after initial infection and declines thereafter. From infancy to age 10 years, age is inversely associated with the risk for developing disease. Children under the age of 2 are at high risk of progressing from infection to disease (25%–30%, and this risk is even higher for children under 1 year of age, 40%–50%). Furthermore, children less than 1 year of age are more likely to develop severe forms of disease, such as miliary or meningeal TB. For unknown reasons, there is a second peak in the risk for developing disease during late adolescence and early adult life. Most young children who develop TB disease do so within the first year after infection, with most cases occurring within 6 months of transmission of the organism. There is also increasing evidence that certain strains of M. tuberculosis create an increased risk of progression to disease and an increased risk of severe disease.

Incidence and Prevalence

The WHO estimates that 2 billion people are infected with M. tuberculosis. This reservoir of TB infection leads to approximately 10.4 million new TB cases and 1.8 million deaths annually. Estimates of childhood TB disease did not exist prior to 2012. Several reasons contributed to the lack of reporting: (1) children not coming to health services, (2) the difficulties in diagnosis of TB disease in resource-poor settings, (3) the lack of age-disaggregated data for many countries and (4) underreporting of clinically diagnosed childhood cases. Once the WHO first started reporting global childhood TB estimates in 2012, the efforts were plagued by methodologic limitations.

Mathematical modeling has since provided figures that are probably more accurate. Within the 22 countries that carry 80% of the world’s burden of TB disease, it is estimated that 7.6 million children acquire TB infection each year, while 53 million children harbor the infection at any given time; of these, at least 650,000 develop TB each year. Systematic review with modeling to adjust for underreporting has led to an estimate of 850,000, or 1 million children with incident TB disease annually. In 2015, with increasing numbers of countries reporting age-disaggregated data and using similar methodology to the aforementioned studies, the WHO estimated that 1 million children develop incident TB disease annually, and 210,000 children die from TB. Despite the improvement in methodology and reporting, the true number of cases remains unknown. The majority of childhood cases occur in resource-limited areas where diagnostic tests include only acid-fast staining of sputum, which is positive in up to 70% of adults but in less than 10% of children with pulmonary TB. As a result, when microbiologic confirmation is required, only one-third of the estimated number of childhood cases are ever reported to national TB programs. In regions where improved case finding has been initiated, children may represent up to 39% of all cases, with a skew toward more serious and complicated cases.

M. tuberculosis can develop resistance to therapeutic drugs if individuals are improperly treated, as discussed later in the chapter. Once an isolate has developed resistance to isoniazid (INH) and rifampin (RIF), it is considered a multidrug-resistant (MDR) organism, and when individuals develop disease from a MDR organism, it is referred to as MDR-TB. The WHO estimates that 3.3% of new TB cases in 2015 were MDR-TB cases. India, China and the Russian Federation have the majority of cases of MDR-TB, but the areas with the highest rates of MDR-TB cases are clustered in eastern Europe and Central Asia, where more than 18% of new TB cases were estimated to be MDR.

Children usually develop MDR-TB after inhaling already resistant organisms from an older individual with contagious TB who has been treated improperly (secondary resistance) instead of developing MDR-TB during his or her own treatment (primary resistance). Because it is difficult to isolate the organism, especially from children, on which to perform drug susceptibility testing (DST), it is difficult to estimate the number of children with MDR-TB. Two mathematical modeling studies have estimated that, globally, between 25,000 and 32,000 children develop MDR-TB annually.

In high-resource settings, such as the United States, Canada, Australia, the United Kingdom, and many western European countries, TB incidence declined upon the advent of effective therapy and prevention. In such settings, foreign-born individuals and racial and ethnic minorities bear a disproportionate amount of TB infection and disease. Screening of immigrants and high-risk populations with connection to appropriate evaluation and treatment are key to TB control in these settings.

The Epidemiology of Tuberculosis in Individuals Infected With Human Immunodeficiency Virus

The United Nations Program on Human Immunodeficiency Virus (HIV)/Acquired Immunodeficiency Syndrome (AIDS) estimates that there are 3.2 million children living with HIV. More than 91% of these children live in sub-Saharan Africa in high-burden TB settings. Despite dramatic gains in the fight against HIV owing to successful prevention of mother-to-child transmission and the robust rollout of antiretroviral therapy, TB is the leading cause of death in people living with HIV. In 2014, HIV-associated TB deaths among adults accounted for 25% of all TB deaths (among HIV-uninfected and HIV-infected people and one-third of the estimated 1.2 million deaths from HIV/AIDS). In high-burden settings, HIV-infected children are at increased risk of acquiring TB infection. TB is a major cause of mortality in HIV-infected children, who have a high risk of progression to TB disease and death following TB infection. HIV-infected infants and young children are at least 24 times more likely to develop culture-confirmed TB than are HIV-uninfected infants and young children. Other populations to consider are HIV-uninfected and HIV-exposed but uninfected children living in a household with HIV-infected individuals. As adult caregivers with HIV infection are at risk for TB disease, all children living in HIV-affected households are more likely to be exposed and to consequently develop TB disease. Maritz has shown that HIV-infected and HIV-exposed children in a high-burden setting have similar rates of TB exposure, with TB disease being found in 19% of the HIV-infected children and 8% of the HIV-exposed children. Although Madhi and colleagues have shown a high incidence of TB disease in HIV-exposed but uninfected children (41 cases per 1000 child-years), TB incidence was threefold greater in HIV-infected children (121 per 1000 child-years in the same high-burden setting).

Immune-compromised HIV-infected children are prone to developing disease manifestations indicative of poor organism containment, such as cavitation of the Ghon focus and disseminated (miliary) disease. This presentation is notably similar to that of very young non–HIV-infected children, suggesting that the immune system of both populations has difficulty controlling M. tuberculosis. HIV-infected individuals are also at risk for other HIV-related lung pathology, such as lymphocytic interstitial pneumonia (LIP), which can present with similar symptoms and radiographic findings (see Chapter 66 ). The nonspecific clinical and radiographic presentations of TB disease in HIV-infected children make this a challenging diagnosis.

Evolution of Clinical Disease in Children

A review of information available from the prechemotherapy era provides a rich understanding of the natural evolution of clinical disease in children. Following infection, all children progress through an asymptomatic incubation period generally lasting 3–8 weeks. The subsequent development of clinical disease is determined by the interaction of the host and the organism and is highly age-dependent ( Table 29.1 ).

The various manifestations of TB tend to occur according to a predictable timetable. Disseminated TB and TB meningitis tend to be early manifestations, often presenting 2–6 months after initial infection has occurred. The primary complex and its complications become apparent most often 3–6 months after infection. It is common for untreated primary complex TB to result in calcification of the lung parenchyma and/or regional lymph nodes, a process that occurs at least 6 months after infection ( Fig. 29.2 ). Although pleural and lymph node TB often develop within 3–9 months after infection, other extrapulmonary forms of TB, especially skeletal and renal disease, may not develop for several years.

A calcified parenchymal lesion and lymph node in a child with tuberculosis infection. These lesions contain a small number of viable Mycobacterium tuberculosis organisms.

Children younger than 2 years of age are at greatest risk for both the development of disease and severe manifestations of disease. Children between 2 and 5 years of age have the next highest risk. Children between 5 and 10 years of age are at the lowest risk of progressing from infection to disease; this time span is often called the “favored age.” The risk increases again in adolescence, when children are more likely to manifest adult-type disease, including cavitary disease.

Intrathoracic Tuberculosis

Following inhalation of M. tuberculosis and the formation of the primary complex (see Fig. 29.2 ), a child can develop intrathoracic TB, which is the most common manifestation of TB disease. Intrathoracic TB consists of a spectrum ranging from subclinical to severe manifestations, which are described in the following sections.

Isolated Lymphadenopathy

The hallmark of childhood pulmonary TB is enlargement of the regional hilar, mediastinal, or subcarinal lymph nodes on chest x-ray ( Fig. 29.3 ). When thoracic adenopathy is isolated without other findings, it usually causes few or no clinical signs or symptoms. This manifestation is most common soon after exposure, usually as part of active case finding.

A young child with pulmonary tuberculosis who has developed cavitation of the lung parenchyma (asterisk). The solid arrow shows narrowing of left mainstem bronchus, while dotted arrows show splaying of the main bronchi.

Progressive Primary Infection

A primary focus, without or occasionally with cavitation, may be seen in symptomatic children who have weight loss, fatigue, fever, and chronic cough. If the host is unable to contain the tubercle bacilli, progressive caseation occurs in the lung parenchyma surrounding the primary focus. The area of caseation may discharge into a bronchus, resulting in the formation of a primary cavity with possible endobronchial spread ( Fig. 29.4 ; see also Fig. 29.3 ). The tubercle bacilli disseminate further to other parts of the lobe and can involve an entire lung. On rare occasions an enlarging primary focus ruptures into the pleural cavity, creating a pneumothorax, bronchopleural fistula, or caseous pyopneumothorax. Profound fever, cough, and weight loss accompany a severe progressive lesion. Before the advent of chemotherapy, 25%–65% of children with progressive primary disease died. However, with appropriate therapy, the prognosis is excellent.

Hilar adenopathy in a child with early pulmonary tuberculosis. Most children with this radiographic appearance have few or no symptoms.

Bronchial Disease

In some children, particularly infants who have smaller-caliber airways, infected lymph nodes continue to enlarge, resulting in lymphobronchial involvement, where the affected bronchus may become partially or totally obstructed because of nodal compression, inflammatory edema, polyps, granulomatous tissue, or caseous material extruded from ulcerated lymph nodes ( Fig. 29.5 ). Although chest radiography rarely delineates the specific pathologic process causing the abnormalities, it is often apparent if bronchoscopy or a computed tomography (CT) scan of the chest is performed. The most frequently affected lobes are the right upper, right middle, and left upper lobe. Symptoms vary according to the degree of airway irritation and obstruction but frequently include localized wheezing or persistent coughing, which may mimic pertussis. A common radiographic sequence is adenopathy followed by localized hyperinflation and then atelectasis of contiguous parenchyma, referred to as collapse-consolidation or segmental lesions ( Fig. 29.6A, B ). However, collapse-consolidation lesions rarely produce symptoms unless the area affected is quite substantial. The radiographic and clinical picture mimics foreign-body inhalation and other obstructive disorders.

A classic collapse-consolidation lesion with hilar adenopathy in a child with pulmonary tuberculosis.

(A) Hyperinflation of the right lower lobe in an infant with pulmonary tuberculosis and airway obstruction. At this point the child had respiratory distress. (B) A chest radiograph from the same child several hours later shows complete collapse of the right lower lobe. At this point the respiratory distress had ceased.

Additional pathology that may accompany bronchial disease includes airway allergic consolidation, caseating consolidation, and bronchopneumonia. Children with allergic consolidation typically experience high fevers, acute respiratory symptoms, and signs of consolidation on chest radiography. Children with caseating consolidation are very ill, with high undulating fevers, chronic cough, and occasional hemoptysis ( Fig. 29.7 ). These children are likely to have positive cultures. The affected airway is completely obstructed, and surgical intervention may be necessary to provide symptomatic relief. In some children, an acute secondary bacterial infection that occurs distal to the obstructed bronchus plays a role in the clinical picture. Children with secondary bacterial pneumonia often present with high fever, cough, and crackles. The clinical signs and symptoms often respond to antibiotics, but the chest radiographic findings do not clear because of the underlying TB.

Pronounced caseating consolidation in a young child with pulmonary tuberculosis.

Complicated Lymph Node Disease

Although segmental lesions and hyperaeration are the most common findings produced by enlarging thoracic lymph nodes, other problems may occur. Enlarged paratracheal nodes may cause stridor and respiratory distress. Subcarinal nodes may impinge on the esophagus and cause difficulty in swallowing, followed occasionally by the formation of an esophageal diverticulum, or the nodes may rupture directly into the esophagus and produce a bronchoesophageal fistula. Enlarged lymph nodes may compress the subclavian vein and produce edema of the hand and arm, or they may erode major blood vessels, including the aorta. They also may rupture into the mediastinum and point in the left, or more often the right, supraclavicular fossa. Compression of the left recurrent laryngeal nerve has been reported. Rupture into the pericardial sac causes tuberculous pericarditis. The late results of bronchial obstruction include the following possibilities: (1) complete reexpansion of the lung and resolution of the radiographic findings; (2) disappearance of the segmental lesion, with residual calcification of the primary focus or the regional lymph nodes; or (3) scarring and progressive contraction of the lobe or segment usually associated with bronchiectasis ( Fig. 29.8 ). Permanent anatomic sequelae result from segmental lesions in approximately 60% of all cases, even though the abnormality is usually not apparent on plain radiographs. Cylindric (rarely saccular) bronchiectasis, stenoses, and elongation or shortening can be demonstrated on bronchography. Fortunately most of these abnormalities are asymptomatic in the upper lobes. However, secondary infection may occur in the middle and lower lobes and cause the middle-lobe syndrome.

Bronchiectasis of the left lower lobe in a child with tuberculosis who was nonadherent to appropriate treatment.

In adults, symptomatic endobronchial TB has been treated with various interventional bronchoscopic techniques including endobronchial stenting, balloon dilation, and argon plasma coagulation. There are no data supporting these interventions in children. When children are symptomatic from endobronchial disease, corticosteroids are a useful adjunct to antitubercular therapy in order to decrease the amount of airway inflammation.

Adult-Type Disease

Chronic pulmonary TB, often called the adult type, or reactivation TB, occurs when pulmonary tissue has been previously sensitized to M. tuberculosis at primary infection and an isolated focus of organisms begins proliferating in that tissue. Chronic pulmonary TB rarely develops in children who acquired infection with M. tuberculosis before 10 years of age and is especially rare among children less than 2 years of age. It occurs more common when TB infection is acquired close to the onset of puberty. The preponderance of evidence supports the concept that most cases of reactivation TB result from endogenous reinfection with the dormant bacilli. However, reinfection with a different strain of M. tuberculosis leading to typical adult-type TB has been documented.

Reactivation pulmonary TB arises from the small round foci of organisms in the apices of the lungs (often called Assmann or Simon foci) resulting from the lymphohematogenous spread at the time of the initial infection. Fibronodular infiltrates in one or both upper lobe apices are most common, but more extensive pulmonary involvement leads to diffuse consolidation or cavitation ( Fig. 29.9 ). Involvement of thoracic lymph nodes is usually absent. Cough, remitting fevers, night sweats, chest pain, sputum production, and hemoptysis are the most common clinical manifestations.

A chest radiograph of an adolescent girl with chronic pulmonary tuberculosis.

Pleural Disease

Pleural TB results from the direct spread of caseous material from a subpleural parenchymal or lymph node focus or from hematogenous spread. Pleural TB is uncommon in children younger than 6 years of age and rare in those younger than 2 years of age. The presence of caseous material in the pleural space may trigger a hypersensitivity reaction, with the accumulation of serous straw-colored fluid containing few tubercle bacilli. This exudate has a high protein concentration and lymphocyte predominance; the number of polymorphonuclear cells depends on the acuteness of onset. Although direct microscopy is usually negative, culture yields may be as high as 40%–70%. However, molecular tests such as GeneXpert are not recommended for pleural fluid owing to their low sensitivity. Pleural biopsy often demonstrates caseating granulomas and increases the culture yield if ample tissue is collected. The clinical course associated with pleural involvement characteristically begins with acute chest pain, accompanied by high fever in the absence of acute illness, an ill-defined loss of vigor, and a dry cough. Active caseation in the pleural space may cause thick loculated pus containing many tubercle bacilli ( Fig. 29.10 ). The prognosis for children with tuberculous pleural effusion has always been good compared with other overt forms of TB, even before chemotherapy was available.

A large tuberculous pleural effusion in an 11-year-old girl.

Extrapulmonary Disease

Tubercle bacilli from the lymphadenitis of the primary complex are disseminated during the incubation period in all cases of TB infection. The clinical picture produced by lymphohematogenous spread is probably determined by host susceptibility at the time of spread and by the quantity of tubercle bacilli released. Three clinical forms of dissemination are recognized:

• 1.
  

  The lymphohematogenous spread may be occult, in which case it usually remains so, or it may be occult initially with metastatic extrapulmonary lesions appearing months or years later (e.g., renal TB).

  
  
• 2.
  

  So-called protracted hematogenous TB, rarely seen today, is characterized by high spiking fever, marked leukocytosis, hepatosplenomegaly, and general glandular enlargement, sometimes with repeated evidence of metastatic seeding in the choroid plexus, kidney, and skin. Calcifications may appear subsequently, often in large numbers, in the pulmonary apices (Simon foci) and in the spleen, thus attesting to the earlier dissemination of tubercle bacilli via blood. The TST is usually strongly positive. Bone marrow biopsy may confirm the clinical impression, but treatment must often be started on a presumptive basis. Although this type of TB often ended tragically in TB meningitis in past years, today it is completely treatable if diagnosed in time.

  
  
• 3.
  

  The third form of lymphohematogenous spread, analogous to sepsis with pyogenic bacteria, is miliary TB. It usually arises from discharge of a caseous focus, often a lymph node, into a blood vessel such as a pulmonary vein. It may be self-propagating, with repeated discharge arising at various sites. Most common during the first 2–6 months after infection in infancy, it can arise even in adults who have apparently well-healed calcified lesions.

The clinical picture of miliary TB varies greatly, probably depending on the number of bacilli in the bloodstream. Sometimes the patient is afebrile and appears to be well, and the condition is diagnosed by chance during contact investigation of another individual with infectious TB. The onset can be insidious, often occurring after the patient has had another precipitating infection. In rare cases, the onset is abrupt. Drowsiness, loss of weight and appetite, persistent fever, weakness, rapid breathing with a rustling sound on auscultation of the lungs, occasionally cyanosis and almost always a palpable spleen are the clinical manifestations that lead the clinician to obtain a chest radiograph.

Usually, within no more than 3 weeks after the onset of symptoms, tubercles, sometimes tiny and at times large, can be seen evenly distributed throughout both lung fields ( Fig. 29.11 ). In the initial stages, they are often detected best on a lateral view of the retrocardiac space. Recurrent pneumothorax, subcutaneous emphysema, pneumomediastinum, and pleural effusion are less serious but well-recognized complications of miliary TB. Cutaneous lesions—including painful nodules, papulonecrotic tuberculids and purpuric lesions—may appear in crops. In addition, choroid tubercules can be noted on ophthalmologic exam and are specific for the diagnosis of miliary TB. The diagnosis of disseminated TB is usually established by means of the clinical picture and a chest radiograph; sometimes by a liver or skin biopsy; by culturing M. tuberculosis from the gastric aspirate, sputum, urine, or bone marrow; or by fiberoptic bronchoscopy. Treatment is usually very successful.

The appearance of miliary tuberculosis in the chest radiograph of an infant.

Extrathoracic Tuberculosis

A complete description of extrathoracic TB is beyond the scope of this chapter. However, pulmonologists may encounter this possibility in evaluating children with pulmonary involvement and other manifestations. The most common forms of extrathoracic disease in children include TB of the superficial lymph nodes (scrofula) and the central nervous system. Other rare forms of extrathoracic disease in children are osteoarticular, abdominal, gastrointestinal, genitourinary, cutaneous, and congenital disease.

TB of the superficial lymph nodes (scrofula) is the most common form of extrathoracic disease. Although this disease is also associated with M. bovis acquired by drinking unpasteurized cow’s milk or cheese, most current cases occur after primary pulmonary infection with M. tuberculosis within 6–9 months of initial infection. Children between the ages of 5 and 10 years, followed by children over the age of 10 years, are most likely to have this presentation, but scrofula can occur in any age group. Scrofula occurs when primary lesions of the upper lung fields or abdomen extend to involve the supraclavicular, anterior cervical, tonsillar and submandibular nodes. In children, TB of the skin or skeletal system rarely leads to involvement of the inguinal, epitrochlear, or axillary lymph nodes. Early during infection, lymph nodes are discrete, firm, and nontender. The lymph nodes become fixed to surrounding tissues and feel matted as the infection progresses. Low-grade fever may be the only systemic symptom. Although usually present, a primary pulmonary complex is visible radiologically only 30%–70% of the time. TST or IGRA results are usually reactive. Although spontaneous resolution is possible, untreated lymphadenitis frequently progresses to caseating necrosis, capsular rupture and spread to adjacent nodes and overlying skin. Rupture through the skin results in a draining sinus tract that may require surgical removal. Excisional biopsy and culture of lymph nodes are often required to differentiate between TB and NTM disease.

Central nervous system disease is the most serious complication of TB in children and complicates approximately 0.5% of untreated primary infections. It is most common in children 6 months to 4 years of age and generally occurs within 2–6 months of primary infection. Central nervous system disease arises from the formation of a caseous lesion in the cerebral cortex or meninges that results from early occult lymphohematogenous spread. Lesions enlarge and discharge bacilli into the subarachnoid space, leading to an exudate that infiltrates the cortical and meningeal blood vessels. This results in inflammation, obstruction, and subsequent infarction of the cerebral cortex. The clinical onset of TB meningitis can be rapid or gradual. Infants and children are more likely to experience a rapid progression to hydrocephalus, seizure, and cerebral edema over several days ( Fig. 29.12 ). In most children, signs and symptoms progress over several weeks, beginning nonspecifically with fever, headache, irritability, and drowsiness. Disease often advances abruptly with symptoms of lethargy, vomiting, nuchal rigidity, seizures, hypertonia, and focal neurologic signs. The final stage of disease is marked by coma, hypertension, decerebrate and decorticate posturing, and eventual death. Rapid confirmation of TB meningitis can be extremely difficult, with wide variability in cerebrospinal fluid characteristics and nonreactive TSTs in 40% of cases. Although some older studies reported normal chest radiographs in 50% of cases, more recent reports indicate that 80% of young children with TB meningitis have significant abnormalities on chest radiography. Since improved outcomes are associated with prompt treatment, empiric antituberculosis therapy should be instituted for any child with basilar meningitis and hydrocephalus or cranial nerve involvement that has no other apparent cause.

A computed tomography scan of the head demonstrating the communicating hydrocephalus that is typical of tuberculosis meningitis.

Tuberculosis and Human Immunodeficiency Virus

Immune-compromised HIV-infected children have a higher risk of developing extrapulmonary disease manifestations indicative of poor organism containment, such as tuberculomas and disseminated (miliary disease). Their presentations are notably similar to those of very young non–HIV-infected children. Lack of an age-related difference in disease presentation suggests that immune maturation is less relevant in HIV-infected children. Because of the common presence of other HIV-related lung pathology on chest x-ray, such as LIP, it is often challenging to diagnose TB accurately, particularly disseminated (miliary) disease. These tremendous diagnostic difficulties result in a tendency to overdiagnose TB in this vulnerable group of children. When pulmonary TB develops in an HIV-infected child, the response to standard short-course TB therapy has lower cure rates and higher mortality, but HIV-infected children can still recover with appropriate therapy. However, these children can develop sequelae such as bronchiectasis if TB therapy and antiretroviral therapy (ART) are delayed. TB may further hasten the progression of HIV disease by increasing viral replication and depleting CD4 ⁺ T lymphocytes. The confirmation of TB disease in HIV-infected children is complicated by the low yield of culture and chronic HIV-related comorbidities.

Diagnostic Imaging

Various imaging modalities can be used to evaluate for intrathoracic TB. Ultrasonography is increasingly used as a point-of-care test in many settings to detect mediastinal lymphadenopathy and pleural fluid because of it has the advantages of lack of radiation, ease of interpretation, potential for portability and rapid results. However, this requires significant performer training. CT is another option that is available in some settings. It provides high definition of the lung parenchyma and mediastinal structures, which is useful if the chest x-ray findings require further delineation, such as determining if a density is due to prominent vasculature or lymphadenopathy. However, CT uses a significant amount of radiation, requires intravenous contrast and is costly. These issues limit CT use in low-resource settings.

Chest Radiography

Chest radiography was the first diagnostic test outside of stain and culture that assisted in the diagnosis of TB. Radiography services are available in high-burden low-resource settings but are more likely to be located at referral centers than at decentralized providers. Many of the pathognomonic findings of TB can be seen on plain chest radiographs. Posteroanterior and lateral images have the most utility. Parenchymal disease, such as that found early in progressive primary infection or in acute pneumonia, appears to be identical to other airspace diseases as a nonspecific infiltrate. These areas can calcify and subsequently appear as hyperlucencies on imaging. If parenchymal disease is untreated, the area can develop a cavity, which manifests itself as a hyperlucent ring devoid of airspace markings on the interior (see Fig. 29.3 ).

The hallmark of pediatric TB is intrathoracic lymphadenopathy, which can manifest in a multitude of ways on radiography. Lymph nodes appear as a soft tissue density on radiography. Because these nodes commonly occur in the hilum and mediastinum, it is often difficult to differentiate these densities from the other soft tissues, such as the thymus and major blood vessels. It is often easier to see the consequence of lymphadenopathy rather than the lymph node itself. Once the lymph nodes become enlarged, they can affect nearby structures in several ways. Large lymph nodes can exert mass effect on the airways. For example, enlarged subcarinal nodes can push on the mainstem bronchi and cause the bronchi to appear splayed (see Fig. 29.3 ). Lymph nodes can also impinge on the airway, causing a narrowing of the airway that is notable on imaging ( Fig. 29.13 ), with distal hyperinflation from a ball-valve effect (see Fig. 29.6A ), or they may cause secondary collapse of the distal segment, manifesting as a fan-shaped consolidation (see Fig. 29.6B ). When not exerting a mass effect, hilar lymph nodes can also be noted on posteroanterior images as enlargement of the mediastinal silhouette ( Fig. 29.14A ). On lateral images, lymphadenopathy densities can produce what is known as the “donut sign” in the posterior mediastinum, comprising the great vessels of the heart superiorly and the hilar lymphadenopathy inferiorly ( Fig. 29.14B ).

Posteroanterior view of a child with right-lower-lobe infiltrate and narrowing of the right mainstem bronchus (solid arrow).

(A) Posteroanterior view of a child with right hilar lymphadenopathy (solid black arrow). (B) Lateral view of same child with a “donut sign” within the dotted circle comprising the great vessels superiorly and subcarinal lymph nodes inferiorly.

Interpretations of chest radiography findings can vary between clinicians. Du Toit compared three experienced pediatric radiologists’ interobserver and intraobserver interpretations of chest radiographs and found that the radiologists agreed on the presence of lymphadenopathy with a kappa of 0.40, which indicates only moderate agreement. Agreement between a radiologist’s first read of an image and a second read after weeks of delay was higher but still offered only moderate agreement (a kappa of 0.55). This illustrates the difficulties in assessing intrathoracic lymphadenopathy, and consequently TB, in pediatric films. Several standardized approaches have been recommended for radiologic interpretation of childhood TB, but these are not always utilized.

Tuberculin Skin Test

The TST remains the most widely employed test for the diagnosis of TB disease and infection in children. The sensitivity and specificity of the TST are significantly affected by several factors; these are reviewed in this section.

Infection with M. tuberculosis produces a delayed-type hypersensitivity reaction to specific antigenic components of the bacilli that are contained in extracts of culture filtrates called tuberculins. A batch of purified protein derivative (PPD), called PPD-S, produced by Siebert and Glenn in 1939, serves as the standard reference material worldwide. In response to the antigen, previously sensitized T cells release lymphokines that induce local vasodilation, edema, fibrin deposition and recruitment of other inflammatory cells. The reaction begins 5–6 hours after injection and reaches maximal induration at 48–72 hours, the time when the test should be interpreted. Variability of the results of the TST may be reduced by careful attention to details of administration and reading, which should be done by trained professionals. The diameter of induration should be measured transversely to the long axis of the forearm and recorded in millimeters. Use of the ballpoint pen method developed by Sokal minimizes interobserver variability. In some individuals, the reaction may peak after 72 hours, and the largest reaction size at any time after 48 hours is considered the result. Vesiculation and necrosis occur rarely. In these cases, the result is always considered positive and repeat tuberculin testing should be avoided. A video from the Grey Bruce Health Unit in Ontario demonstrates TST administration and is available at https://www.youtube.com/watch?v=bR86G-itrTQ .

The size of the area of induration should be interpreted within the view of the child’s risk of having acquired TB infection and the risk of progressing to TB disease. TST interpretation is dependent on the clinical setting. In the United States and Canada, 5 mm is used as the smallest area of induration that can be positive for the highest-risk groups, such as children with HIV infection, signs of TB disease, or recent exposure to a known TB case. An area of induration 15 mm or greater is positive even for an individual with no risk factors for disease. Guidelines for TST interpretation in the United States are available in Box 29.1 . The United Kingdom now uses a universal size of 5 mm as positive. The WHO recommends use of a cutoff of 10 mm for most children and a cutoff of 5 mm in immunosuppressed children. WHO guidelines are followed in most countries. A nonreactive TST does not exclude TB infection or disease. Several factors can diminish tuberculin reactivity resulting in a false-negative reaction ( Box 29.2 ). The administration of live-attenuated vaccines results in immune system suppression that appears more than 48 hours after vaccination. Tuberculin skin testing may be performed on either the same day as vaccination with live virus or 4–6 weeks later. Studies have demonstrated that up to 10% of immunocompetent children with reactive anergy tests and culture-confirmed pulmonary TB have false-negative reactions to tuberculin testing.

In many of these children, TST conversion occurs after several months of treatment, suggesting that the infection was recently acquired or resulted in suppression of the immune response. Skin testing results were available in nearly all (95.4%) children reported to have newly diagnosed TB in the United States between 1993 and 2001. Of these children, 11% had negative skin test results as defined by CDC guidelines. Of note, children who were diagnosed with disseminated or meningeal TB were less likely to have a positive TST result (57.6% and 54.6%, respectively) than those with pulmonary TB (90.6%).

A number of factors have been associated with false-positive tuberculin reactions and decreased tuberculin test specificity. Because some antigens in PPD are shared with other mycobacteria, false-positive reactions can occur in children who have been infected with other mycobacteria or who have been vaccinated with BCG. Exposure to NTM varies geographically and generally results in smaller, transient indurations than those caused by M. tuberculosis. The degree of BCG cross-reactivity is dependent on a number of factors, including strain of BCG employed, repeated BCG vaccination, age and nutritional status at vaccination, frequency of skin testing, and years since vaccination. In most studies of children who received a BCG vaccine during the newborn period, only 50% reacted to tuberculin testing at 12 months, and 80%–90% lose reactivity within 2–3 years. Although BCG vaccination of older children or adults results in greater initial and more persistent cross-reactivity, most of these individuals lose cross-reactivity within 10 years of vaccination. Most guidelines currently suggest that the TST should be interpreted in the same way for patients who have and have not received a BCG vaccination; however, this will lead to some children with false-positive TST results being treated. A growing body of evidence now demonstrates that the impact of infant BCG vaccination on TST responses in children exposed to TB wanes with age, but in young BCG-vaccinated children, a TST cutoff of 5 mm is associated with poor specificity. Nevertheless, as young children have the greatest risk of progression from infection to disease and are also the most susceptible to severe disseminated forms of TB disease, test sensitivity remains more important than specificity in evaluating children in this age group.

Interferon-Gamma Release Assays

The identification of genes in the M. tuberculosis genome that are absent from M. bovis BCG and most NTM has supported the development of more specific tests for the detection of M. tuberculosis. M. bovis BCG has 16-gene deletions, including the region of difference 1 (RD-1) that encodes for early secretory antigen target-6 (ESAT-6) and culture filtrate protein 10 (CFP-10). ESAT-6 and CFP-10 are strong targets of the cellular immune response in patients with M. tuberculosis infection and disease. In people with TB infection or disease, sensitized memory/effector T cells produce IFN-γ in response to M. tuberculosis antigens, forming the biologic basis for both the TST and IGRAs. Research over the past two decades has resulted in the development of two commercial brands of IGRAs that are approved for use in Europe and the United States. In brief, the Quantiferon assays (QFTs; Cellestis Limited, Australia) are enzyme-linked immunosorbent assay (ELISA)–based whole blood assays measuring the amount of IFN-γ produced in response to M. tuberculosis –specific antigens. In contrast, the enzyme-linked immunospot (ELISPOT)–based test (T.SPOT.TB; T.Spot; Oxford Immunotec, Abingdon, Oxfordshire, United Kingdom) uses peripheral mononuclear cells to detect the number of INF-γ–producing T cells. Of note, both commercial IGRAs also measure a response to a negative control and a nonspecific positive control stimulant, such as phytohemagglutinin (PHA). Results are then interpreted on the basis of a comparison of the magnitude of the response to the M. tuberculosis –specific stimulus compared with the negative controls. A lack of response to the positive control is interpreted as an indeterminate (Quantiferon) or invalid (T-SPOT.TB) result. Similarly, a robust response to the negative stimuli is considered an invalid or indeterminate result.

In the absence of a gold standard for infection, studies have assessed the performance of IGRAs in children with bacteriologically confirmed and clinically diagnosed TB disease. A systematic review of these pediatric studies estimated the specificity of commercially approved IGRAs for detecting M. tuberculosis infection at 91% for the ELISA-based tests and 94% for the ELISPOT-based test, compared with 88% for the TST (positive defined as 10 mm). Estimates of test sensitivity were similar for the three tests: 83% (QFT assays), 84% (T.SPOT) and 84% (TST), respectively. Similarly, a second pediatric systematic review found the performance of the TST and QFT assays to be no different but noted that a qualitative review of four pediatric studies suggested that the QFT assays were more specific for the detection of TB infection in children compared with the TST. Like the TST, IGRAs cannot differentiate between TB infection and TB disease.

The lack of a gold standard for the detection of TB infection complicates studies of diagnostic accuracy, many of which have employed surrogate measures of infection to serve as the reference standard for TB infection. The association between IGRA positivity and the presence of TB exposure is well described in low-burden countries. Results from high-burden countries examining IGRA positivity and the presence of TB exposure are conflicting and demonstrate both advantages of the T.SPOT and no difference in T.SPOT performance compared with TST and QFT. A few key studies have illustrated that exposure may be quantified to support direct comparison of tests of TB infection. The largest of these studies, including over 1300 children, demonstrated that IGRAs were more strongly associated with quantified TB exposure than the TST. Pooled analysis of studies employing a measure of TB exposure also demonstrates that positive IGRA results are more strongly correlated with quantified TB exposure than the TST results.

There is limited and conflicting evidence regarding IGRA use in young or impoverished children. Studies have demonstrated variable association between indeterminate IGRA results and young age, with some studies showing an association and others showing no association. A seminal study of European immigrant children found that indeterminate IGRA results were more frequent among young children but occurred at clinically insignificant rates of 1.8% and 1.6 % for the QFT and T.SPOT, respectively. Most studies have reported indeterminate rates less than 10%. Emerging evidence further demonstrates that a number of factors related to poverty (malnourishment, micronutrient deficiency and helminth infection) may lead to diminished IGRA sensitivity.

Limited data are available regarding IGRA performance in immune-compromised HIV-infected children, in whom the performance of the TST is impaired. Two studies suggest that test failures described as indeterminate or invalid results are common among pediatric oncology patients. Two studies have shown a noncommercial IFN-γ ELISPOT assay to have higher sensitivity for detecting TB infection compared with the TST in HIV-infected children. A comparison of the QFT–gold assay and the TST for the detection of TB infection in 36 young HIV-infected children with culture-confirmed TB disease found comparable sensitivity in children with CD4+ cell counts below 200/mL; unfortunately indeterminate QFT-gold results were reported in 25% of children tested. In a head-to-head comparison among 130 HIV-infected and 120 HIV-uninfected children, the TST and IGRAs performed similarly for the detection of TB infection in well-nourished HIV-uninfected children, but test performance was differentially affected by chronic malnutrition, HIV infection and age. In a study of over 1300 children, indeterminate QFT results were more frequent in HIV-infected (4.7%) than HIV-uninfected children (1.9%), while T.SPOT invalid results were rare (0.2%) and were not affected by HIV infection. Conversion, reversion and operational measures were not associated with HIV status. Among HIV-infected children, test sensitivities declined as malnutrition worsened. As a conclusion from the available evidence, clinicians should take age, nutritional status, and HIV status into consideration in interpreting IGRAs.

Both the TST and IGRAs have limitations, but use of the tests in targeted populations with increased risk of TB infection or TB disease maximizes the impact of testing. IGRAs are commonly used in children living in upper-income, low-burden countries and have influenced clinical decision making. National guidelines continue to change and vary dramatically among countries in light of rapidly emerging evidence. Pragmatic approaches to the use of IGRAs and TST have emerged through clinical practice and are emerging in formal guidelines.

The purpose of the TST and IGRAs is to determine whether the person is infected with M. tuberculosis, which can serve as a useful tool in the diagnosis of TB, and guide subsequent care for infection or disease. Attributes of the child and the purpose of the testing should inform the choice of which diagnostic test to use ( Fig. 29.15 ). Per the American Academy of Pediatrics (AAP), “Only children who have risk of TB exposure, underlying health conditions that require immune suppression, or suspected TB disease should be tested for TB infection in TB low-burden settings.” In low-risk, BCG-vaccinated or NTM-exposed children in whom the TST will result in a considerable number of false-positive results, high specificity is desired, making the IGRAs preferable. In young or immune-compromised children with significant risk of TB progression following TB infection, testing strategies should optimize sensitivity. Although it may compromise specificity of the testing strategy, both an IGRA and a TST should be performed in a child with a high suspicion of TB disease or a high risk of progression to TB disease. In these instances, a positive result with either the TST or IGRA should be considered evidence of TB infection. Although this approach may result in overtreatment, the benefits of avoiding disease progression generally outweigh the risk of overtreatment, particularly since children tolerate treatment of TB infection with few side effects.

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

Share this:

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

Related

Related posts:

1. Drug Administration by Inhalation in Children
2. Pneumonia in Children
3. Lung Injury Caused by Pharmacologic Agents
4. Pediatric Lung Transplantation

Stay updated, free articles. Join our Telegram channel

Join