Tuberculosis in childhood differs from the adult clinical form and even has been suggested that it is a different disease due to its differential signs. However, prevention, diagnostics, and therapeutic efforts have been biased toward adult clinical care. Sensibility and specificity of new diagnostic approaches as GeneXpert, electronic nose (E-nose), infrared spectroscopy, accelerated mycobacterial growth induced by magnetism, and flow lateral devices in children populations are needed. Adequate and timely assessment of tuberculosis infection in childhood could diminish epidemiological burden because underdiagnosed pediatric patients can evolve to an active state and have the potential to disseminate the etiological agent
Tuberculosis is the leading cause of death worldwide, with over 1.5 million deaths per year. This disease is caused by
The natural history of TB in children and pediatric patients follows a series of steps. Phase 1 occurred 3–8 weeks after primary infection. This is the end of incubation period and the initiation of well-defined signs: fever, erythema nodosum, a positive tuberculin skin test response, and formation of the primary complex visible on chest radiography. Phase 2 occurred 1–3 months after the phase 1. In this period, the bacillus can migrate to other parts of the body via the blood and represented the period of the highest risk for the development of tuberculous meningitis and miliary tuberculosis in young children. This is the phase where dissemination of the bacillus most frequently occurs. Phase 3 occurred 3–7 months after primary infection. This is the period of pleural effusions in >5 years old children and bronchial disease in <5 years old children. Phase 4 presents after 1–3 years of the phase 1. In this period, the osteoarticular tuberculosis in children with 5 years or less appears. Phase 5 occurs up to 3 years after phase 1 and it is presented after calcification was completed. Until this phase, manifestations of classical adult tuberculosis appear [
From a tuberculosis control point of view, programs place an almost exclusive emphasis on adults with sputum smear positive. Tuberculosis in children remains a neglected area of research despite considerable morbidity and mortality [
One interesting issue about the disease in children is that age determines the progress to disease; infants are at the highest risk than older children, so it is possible to categorize all children with 3 years or less in high risk [
The risk of infection depends on the duration of exposure, the closeness of contact, and the microbial load of the source case [
In developing countries, the risk for TB infection and disease is relatively uniform in the population; annual rates of infection often exceed 2% depending mainly upon exposition to infected persons [
The diagnosis of tuberculosis in children is complicated mainly because (i) TB can mimic many common childhood diseases, including pneumonia, generalized bacterial and viral infections, malnutrition, and HIV [
Clinical similarities and differences between adult and childhood TB with relevancy to successful diagnosis.
Feature | Adults | Children |
---|---|---|
Typical signs | Radiological features and a positive sputum smear | TB can mimic many common childhood diseases. The clinical symptoms in older children are cough, fever, wheezing, fatigue, and failure to gain weight, and in pediatric children are pulmonary parenchymal disease and intrathoracic adenopathy, lymphadenopathy, and central nervous system involvement |
X-rays findings | Classical cavitation in lungs | Enlargement of hilar, mediastinal, or subcarinal lymph nodes and lung parenchymal changes, hilar lymphadenopathy with or without a focal parenchymal lesion |
TST | Cross-reaction with BCG vaccination and exposition with other mycobacteria | |
Sampling | Easy sputum and blood sampling | Difficulty to expectorate, blood sampling usually painful in pediatric children |
Bacillary load | High bacillary load, easy to find the bacillus when technician is skillful | Lower bacillary load and is usually smear negative even with fluorescent dyes |
Bacillus growth in culture | High yields of 90–100% | Confirmation by culture rarely exceeds 30–40% |
Tropism of | Commonly localized infection in the lungs | Commonly extrapulmonary, disseminated |
Most children with TB are classified as smear-negative pulmonary TB (PTB) for the reasons mentioned above (difficulty to expectorate, lack of equipment to gastric lavage, etc.) which is an inappropriate term as a smear or culture has not usually been done. This leads to difficulties in determining the true extent of PTB in children in different areas and circumstances. Extrapulmonary TB (EPTB) accounts for up to 20–30% of the total caseload of TB in children, and the diagnosis is usually easier than PTB because of the characteristic clinical features like lymphadenopathy with or without scrofula, spinal deformity, disseminated disease, meningitis, effusions (pleural or pericardial), or painless ascite [
The value of the classic diagnostic is named: (1) exposure to an adult index case; (2) chronic respiratory symptoms that do not respond to broad-spectrum antibiotics; (3) documented weight loss or failure to thrive; (4) a positive tuberculin skin test (TST); (5) the presence of suggestive signs on the chest radiograph (CXR), which is greatly reduced in endemic areas where exposure to and/or infection with
Pros and cons of most common diagnostic tests for childhood TB.
Methodology | Pros | Cons |
---|---|---|
Symptoms | No need for lab infrastructure, diagnostic value if appropriate risk stratification is applied | This criterion has been approved only in conjunction with the TST and suggestive chest radiography |
Traditional chest radiograph | The basic equipment is very common in hospitals and some research centers. | The images are not always clear and the lesions in children are often subjective |
Thorax CT scan | Enhanced visualization of small lesions not seen on chest radiograph. X-ray high-resolution computed tomography, it is the most sensitive tool currently available to detect hilar adenopathy and/or early cavitation can be used for follow-up | Costly; requires scanner which is not readily available in many settings |
Algorithms | They are very helpful and easy to use in countries with restricted technology | Is not commonly used due to lack of validation, it is based on responses of patients to which scores are given which are thought to be very subjective |
Gold standard for definitive diagnosis of adult TB | Culture usually takes weeks (or four days in accelerated culture), low sensitivity ( | |
Rapid | Very low sensitivity ( | |
Tuberculin skin test (TST) | Very common and cheap reagent, easy to use and to interpret the results | Inespecific, only indicates infection with a mycobacteria or prior BCG vaccination |
Polymerase Chain Reaction (PCR) | This is a rapid, sensitive, specific and affordable method | These tests are not performed correctly in all clinical laboratories. The cost involved, the need for thermocycler (or boiling pots at specific temperature), and scrupulous technique to avoid cross-contamination of specimens preclude the use of PCR techniques in many developing countries |
In-house nucleic acid amplification assays | Mean sensitivity of 60%, with a proper technique could be done efficiently | These assays are dependent of operator’s skill |
Adenosin deaminase | This method does not require sputum, only blood. Very high sensitivity and specificity | The report presents unclear case definition, exclusion of nontuberculous patients, and a relatively small TB patient population (20 with active TB) |
Serology and antigen detection | In this method, the sample is blood which is easier to obtain than sputum (in PTB). It is very rapid and does not require specimen from the site of disease | Sensitivity and specificity depend on the antigen used |
These methods can replace TST for detection of latent TB infection. Rapid test versions are inexpensive, and dozens of commercial kits are on the market; high specificity (98–100%) | The test may have impaired sensitivity for very young children, for whom it should not be used to exclude the presence of | |
GeneXpert MTB/RIF system | This requires minimal manipulation of sample and operator training. It utilizes real-time PCR technology to both diagnose TB and detect rifampicin resistance. Results in ~105 min. | Only one report in a children population from South Africa. There is a need to validate in other populations |
Gas sensor array electronic nose (E-Nose) | High specificity | Without data in children populations |
The difficulty to obtain samples for TB diagnosis in children has led researchers to create smart approaches as “
The use of well-defined symptoms improves diagnostic accuracy of PTB [
In general, there is a sense of skepticism regarding the potential diagnostic value of symptom-based approaches; however, the natural history of childhood tuberculosis demonstrates that symptoms may have diagnostic value if appropriate risk stratification is applied. Marais et al. in 2006, conducted a study to assess the ability to diagnose TB in HIV-negative children with symptoms and concluded that PTB can be diagnosed with a reasonable degree of accuracy in HIV-uninfected children (with a high degree of accuracy in the low-risk group), using a simple symptom-based approach. This offers the exciting prospect of improving access to antituberculosis treatment for children in resource-limited settings [
Houwert et al. in 1998, conducted a prospective evaluation of the WHO criteria to see if it was specific and predicted the situation of TB in children and they conclude that the diagnosis of tuberculosis must be more seriously considered when a child presents the triad of the above mentioned criteria [
The radiography became available after the First World War, and since that time, PTB detection became easier [
There are point-scoring systems to make a diagnostics classifications. Diagnostic algorithms were developed to deal with these diagnostic difficulties and provide the health care worker with a rational, stepwise tool to identify children in need of TB treatment. They are very helpful and very easy to use in countries with restricted technology, but a few of them are used now [
Microbiological confirmation of TB in young children is not routinely attempted in many high burden settings due to the difficulty in obtaining samples and the poor performance of smear microscopy [
All of these alternative ways of sampling have been made to increase yield because a positive culture is regarded as the “gold standard test” to establish a definitive diagnosis of TB in a symptomatic child [
Advances have been done in the performance of smear microscopy for the rapid detection of MTB, for example, the concentration of specimens by centrifugation or the change of the staining of carbol fuchsin (Ziehl-Neelsen or Kinyoun) for a fluorescent dyes (auramine-rhodamine), which both increases sensitivity and reduces the time for screening [
The Tuberculin skin test, or Mantoux TST, is based on the detection of a cutaneous delayed-type hypersensitivity response to purified protein derivative, a poorly defined mixture of antigens present in
With a TST, it is not possible to assert or deny the presence of TB, but it only indicates infection with a mycobacterium
Diagnostic PCR is a technique of
Sensitivity and specificity of commercial and in-house methods for TB diagnostics.
Methodology | Sensitivity | Specificity |
---|---|---|
Commercial tests | ||
AMTD Standard | 0.79 | 0.91 |
Smear positive | 0.98 | 0.55 |
Smear negative | 0.75 | 0.90 |
Gastric aspirate only | 0.73 | 1.00 |
Cut-off: 71,000 | 0.83 | 0.91 |
Cut-off: 7,300,000 | 0.90 | 0.85 |
Cut-off: 30,000 | 0.93 | 0.66 |
AMTD Enhanced | 0.89 | 0.98 |
Smear positive | 1.00 | 0.90 |
Smear negative | 0.83 | 0.98 |
Low suspicion of TB | 0.83 | 0.97 |
Intermediate suspicion of TB | 0.75 | 1.00 |
High suspicion of TB | 0.88 | 1.00 |
Amplicor COBAS | 0.72 | 0.99 |
Low pretest probability | 0.33 | 0.99 |
Intermediate pretest probability | 0.33 | 0.98 |
High pretest probability | 0.47 | 1.00 |
Smear positive | 0.91 | 0.50 |
Smear negative | 0.75 | 0.99 |
Amplicor manual | 0.68 | 0.94 |
Smear positive | 0.91 | 0.74 |
Smear negative | 0.57 | 0.90 |
LCx assay | 0.90 | 0.96 |
Smear positive | 0.98 | 0.10 |
Smear negative | 0.90 | 0.96 |
Amplicis Myco B | 0.92 | 0.85 |
Sputum only | 0.91 | 0.90 |
GeneXpert adult population | 0.95 | 1.00 |
Tanzanian adult population (sputum and smear positive, | 0.88 | 0.99 |
Children population | ||
Two induced sputum samples | 0.76 | 0.99 |
Smear positive | 1.00 | 0.99 |
Smear negative | 0.61 | 0.99 |
In house tests | ||
IS 986 | 0.90 | 0.95 |
Smear positive | 0.97 | 0.83 |
Smear negative | 0.75 | 0.47 |
Sputum only | 1.00 | 1.00 |
IS 6110 | 0.79 | 0.84 |
Chemical DNA extraction | 0.60 | 0.92 |
Simple boiling | 0.85 | 0.98 |
Smear negative | 0.90 | 0.92 |
Smear positive | 0.92 | 0.42 |
Bronchiestasis only | NA | 0.86 |
Upper lobe infiltrates | 0.67 | 1.00 |
Agarose gel electrophoresis | 0.90 | 1.00 |
Dot blot hybridisation | 0.92 | 0.98 |
ELISA | 0.90 | 1.00 |
MTP40 | 0.97 | 0.86 |
MTP40 and | 0.74 | 1.00 |
MPB70 | 0.98 | 0.50 |
Smear positive | 0.99 | 0.13 |
Smear negative | 0.96 | 0.53 |
2.4 kb DNA | 0.55 | 0.94 |
65 kDa | 0.84 | 0.85 |
Gastric aspirate only | 1.00 | 0.80 |
Sputum only | 1.00 | 0.84 |
MPB64 | 0.56 | 0.84 |
2.4 kb DNA and MBP64 | 0.98 | 0.70 |
2.4 kb DNA and MBP64 and 65 kDa | 0.98 | 0.70 |
MTB 10 and MTB 11 | 0.94 | 0.94 |
Ag 85 | 0.90 | 0.94 |
groEL | 0.82 | 0.81 |
Meta-analysis for tuberculous pleuritis | 0.92 | 0.90 |
ADA | 0.80 | 0.84 |
Pab | ||
E-nose | ||
Culture positive | 0.89 | 0.91 |
EN Rob | 0.68 | 0.75 |
EN Walter | 0.75 | 0.67 |
Values were obtained from the average of different studies [
This is a rapid, sensitive, specific, and reasonable-cost [
Studies in children have obtained better sensitivity by PCR than by culture. In 2001, Gomez-Pastrana et al. [
Montenegro et al. in 2003, reported a heminested PCR assay which specificity was 67%. This was significantly higher than Löwenstein-Jensen culture (54%) or AFB stain (42%) for children with highly probable tuberculosis. PCR detection rates for culture-positive specimens were 100% for smear-positive samples and 76.7% for smear-negative samples. The specificity of PCR was 100% in control children. Compared with culture, PCR showed a sensitivity of 90.4%, a positive predictive value of 89%, a specificity of 94%, and a negative predictive value of 95% [
These assays are highly dependent of operator’s skills. Performance is also influenced by the choice of target sequence and DNA extraction method. Interpretation of the performance of these assays in pediatric TB suspects is confounded by the lack of a sensitive and specific reference standard. When compared with culture, the sensitivity of NAA for the diagnosis of childhood TB is typically low (40–83%). However, it appears, at least from some reports, that NAA identify a group of children who are clinically diagnosed with TB but in whom mycobacterial culture is negative. This means that with a proper technique it could be done efficiently [
Considering the low yield of smear and culture in PPTB, nonmicrobiological methods may provide new tools for diagnosis. Adult studies have shown increased levels of adenosine deaminase (ADA) in pleural TB and TB-caused meningitis, both paucibacillary forms of TB, and have advocated its use in diagnosis. Due to this evidence, a serum ADA has already been evaluated in a childhood population with a very high sensitivity (100%) and specificity (90.7%) for pulmonary TB. This study demonstrated the great potential of this technique because it has significant difference in serum ADA levels between children with disease and infection. However, there were several weaknesses in the study design, including unclear case definition, exclusion of nontuberculous patients, and a relatively small TB patient population (20 with active disease) [
In the case of extrapulmonary TB, ADA measurement can be helpful, but its sensitivity and specificity varies widely and has been lower than multiplex PCR using primers for IS6110, dnaJ, and hsp65 [
In absence of good diagnostic method for tuberculosis, the interest in serodiagnosis has been increased [
Hussey et al. in 1991 [
Serology has found little place in the routine diagnosis of tuberculosis in children, even though it is rapid and does not require specimen from the site of disease. Sensitivity and specificity depend on the antigen used, gold standard for the diagnosis of tuberculosis, and the type of tubercular infection. Though most of these tests have high specificity, their sensitivity is poor because several factors can alter the results such as age, exposure to other mycobacteria, and BCG vaccination [
A rapid and accurate tool for diagnosing childhood TB would be highly beneficial. Much attention has been focused on immune-based assays that do not rely on sputum but can be done with blood [
Dogra et al. in 2007, conducted a study in India in which he compared the results between the TB-gold Quantiferon and TST which obtained comparable results even in malnourished children [
Bakir et al. in 2008, reported the prognostic value of IFN-
GeneXpert includes the development of integrated DNA extraction and amplification systems. This requires minimal manipulation of sample and operator training. It utilizes real-time PCR (rt-PCR) technology to both diagnose TB and detect rifampicin resistance. The test amplifies a region of the
Recently, Nicol et al. in 2011, reported the application of this method in 452 hospitalized children from South Africa, with or without HIV, with a median age of 19.4 months, and suspected of having TB. Two Xpert tests doubled the case detection rate compared with smear microscopy (76% versus 38%), identifying all smear-positive and 61% of smear-negative cases, the specificity was 98.8%. The sensitivities for smear-negative TB were 33.3% and 61.1% when testing one or two samples, respectively. The samplings were induced sputum and they detected three quarters of culture-confirmed tuberculosis with very high specificity; the yield of this method was twice that of smear microscopy. This could suggest the possibility of replacing the microscopy for this type of methodology which has greater sensitivity especially with a second sample [
The potential to detect different
Recently, it has been produced llama (
Tuberculosis is the leading cause of death worldwide in both adults and children, and the incidence and prevalence in the latter are underestimated due to the difficulty to collect samples and the lack of efficient methods to detect reaction specificity with no cross-infection with other mycobacteria. There are several methods currently being used for TB diagnosis in children, some classical approaches comprise mycobacterial culture, microscopy, TST, IGRAs, and more recently high-tech diagnostics approaches as real-time PCR (GeneXpert), E-nose, and infrared spectroscopy.
After the above mentioned, we can propose the identification of mycobacteria in children by two different methods. The first would be for use in places where access to infrastructure and technology is very limited, in this type of scenario, the best option would be a rapid test based on lateral flow device to be used in serodiagnosis, that could be implemented on a test strip to confirm or rule out the presence of bacilli in the patient. In case of having the necessary infrastructure, the GeneXpert would be the best option because it has shown no cross-reactions and also in studies with children have shown a higher sensitivity and specificity than culture methods, but it is clear that a gold standard for childhood tuberculosis is still needed, because to date is not possible to be sure about the presence or absence of the bacillus with a single test in this widely neglected but relevant epidemiological population.
Regarding the sample type, to date, in pulmonary TB, the sputum induction would be optimal since it is an approach that does not affect the results (as seen in the capsule as stomach acids can inhibit the growth of mycobacteria) and is painless and minimally invasive (unlike gastrointestinal and nasopharyngeal washings). Specifically, when sampling for extrapulmonary TB, blood would be the best option, however, in patients with systemic TB, mycobacteria have been found also in the urine, and this kind of sample, the least invasive to date (not including breath’s patient yet), has been poorly explored in adult TB, and even less has been done in children TB research.
Recent developments for TB diagnostics as infrared spectroscopy (serum) and gas sensor technology (breath or sputum) seem promissory fields due to they are fast and minimally invasive, but their drawbacks are that they must be validated in diverse populations and improved according to the patient’s needs. With the above-mentioned ideas, it is clear that we are starting a 1000-mile walk in childhood tuberculosis, but surely the next generation of TB diagnostics tools holds great promises.
This review is dedicated to TB researchers working in developing countries. This work was partially supported by the Fondos Sectoriales de Ciencia Básica SEP-CONACYT (CB-2008-01-105813, Mexico). G. G. Lόpez Ávalos is a recipient of a Ph.D. scholarship from CONACYT, and E. Prado Montes de Oca is a recipient of a research stimulus from the Sistema Nacional de Investigadores (SNI 41290, CONACYT). The authors declare no conflicts of interest.