Childhood intra-thoracic tuberculosis : addressing the diagnostic dilemma

Date
2006-04
Journal Title
Journal ISSN
Volume Title
Publisher
Stellenbosch : University of Stellenbosch
Abstract
ENGLISH ABSTRACT: Children contribute little to disease transmission and the maintenance of the tuberculosis epidemic, but they constitute a significant proportion of the total tuberculosis (TB) caseload and experience considerable morbidity and mortality in endemic areas, despite the availability of cheap and effective treatment. The difficulty of diagnosing childhood tuberculosis is one of the major obstacles that hinder the provision of antituberculosis treatment to children in endemic areas. The diagnosis of childhood tuberculosis is complicated by the lack of a practical gold standard, as bacteriologic specimens are difficult to collect and the yield is low. In non-endemic countries the diagnosis of childhood tuberculosis is based on the triad of: 1) exposure to an adult index case, 2) a positive tuberculin skin test, and 3) suggestive radiographic signs. However, the triad has limited value in endemic areas where exposure to and/or infection with Mycobacterium tuberculosis is common, and chest radiography is rarely available. The objective of this dissertation was to address the diagnostic dilemma faced by health professionals in endemic areas with limited resources, where children currently have poor access to chemoprophylaxis and antituberculosis treatment. We first clarified basic disease concepts, through a critical review of the pre-chemotherapy literature that documented the natural history of childhood tuberculosis. Three central concepts were identified; 1) the importance of accurate case definition, 2) the relevance of risk stratification, and 3) the diverse spectrum of disease, which necessitates accurate disease classification. The importance of accurate case definition is illustrated by the fact that isolated hilar adenopathy, considered the principal radiographic sign of primary tuberculosis, becomes transiently visible in the majority of children following recent primary infection. Our analysis of the natural history of childhood tuberculosis allowed accurate quantification of the risk to progress to disease following primary infection with M.tuberculosis. This demonstrated that the risk depends mainly on the age and/or immune-status of the child, the time since primary infection occurred and the presence or absence of symptoms. After analysing these historic studies, we proceeded to document the burden of childhood tuberculosis in an endemic area. We first conducted a retrospective study to describe current diagnostic practices and demonstrated almost exclusive reliance on chest radiography. We then calculated the burden of childhood tuberculosis in a prospective descriptive study. The corrected tuberculosis incidence rate in children was 407/100 000/year and children with severe forms of disease, such as disseminated (miliary) tuberculosis and/or tuberculous meningitis, were rarely recorded in the TB treatment register used for routine community-based surveillance. An additional obstacle to progress in the field of childhood tuberculosis has been the lack of standard descriptive terminology. Following a careful review of the literature, we proposed a radiological classification of childhood intra-thoracic tuberculosis and explored the different pathologic mechanisms that underlie these diverse disease manifestations. We then conducted a prospective descriptive study to document the disease spectrum in children treated for tuberculosis in an endemic area. The disease patterns observed were consistent with those described in the pre-chemotherapy literature. In addition, we demonstrated that bacteriologic confirmation may be achieved in the majority of children with intra-thoracic tuberculosis, in highly endemic settings. Finally we developed a novel symptom-based approach to diagnose pulmonary tuberculosis in children from endemic areas with limited resources. We followed a step-wise approach by first conducting a community-based survey to document the prevalence of symptoms traditionally associated with tuberculosis in a random selection of children from an endemic area. The survey demonstrated that poorly defined symptoms offer poor diagnostic value. The second step was to evaluate the diagnostic value of well-defined (persistent, non-remitting) symptoms in a small prospective study. Well-defined symptoms demonstrated good diagnostic value, but these promising results required further validation. As a final step, we validated the diagnostic value of a novel symptom-based approach in a large prospective, community-based study. In this study, a simple symptom-based approach diagnosed childhood pulmonary tuberculosis with a remarkable degree of accuracy, particularly in HIV-uninfected children older than 3 years of age. This novel diagnostic approach offers the exciting prospect of extending antituberculosis treatment to children in endemic areas with limited resources, where current treatment access is poor.
AFRIKAANSE OPSOMMING: Tuberkulose beheer programme plaas feitlik geen klem op die behandeling van kinders nie, omdat kindertuberkulose selde aansteeklik is en die persepsie bestaan dat kinders slegs in raar gevalle ernstig siek word. Tuberkulose lewer egter ‘n betekenisvole bydrae tot kindermorbiditeit en mortaliteit in endemiese areas, terwyl dit ‘n maklik behandelbare siekte is. Kindertuberkulose is moeilik om te diagnoseer en dit is ‘n belangrike faktor wat daartoe bydra dat kinders dikwels nie antituberkulose behandeling ontvang wanneer hulle dit benodig nie. Die diagnose van kindertuberkulose is moeilik, omdat die organisme selde aangetoon kan word. In nie endemiese areas word kindertuberkulose dikwels gediagnoseer na aanleiding van: 1) blootstelling aan ‘n volwasse indeks geval, 2) ‘n positiewe tuberkulien veltoets, en 3) die teenwoordigheid van radiologiese tekens suggestief van tuberkulose. Hierdie benadering het defnitiewe tekortkominge in endemiese areas, waar blootstelling aan en infeksie met Mycobacterium tuberculosis algemeen is. Gevolglik berus die diagnose van kindertuberkulose hoofsaaklik op die subjektiewe interpretasie van die borskasplaat, wat welbekende tekortkominge het en verder is radiologiese toetse dikwels nie beskikbaar in hierdie areas nie. Die doel van die navorsingsprojek was om die dilemma rondom die diagnose van kindertuberkulose in endemiese areas aan te spreek. Eerstens is basiese siektekonsepte uitsorteer deur ‘n kritiese oorsig van studies uit die pre-chemoterapie era. Hierdie kosbare studies het die natuurlike verloop van tuberkulose in kinders beskryf, nog voordat antituberkulose middels beskikbaar was. Drie sentrale konsepte is geidentifiseer; 1) die belang van akkurate siekte definisie, 2) die relevansie van risiko stratifikasie en 3) die diverse spektrum van patologie wat akkurate siekte klassifikasie noodsaak. Die belang van akkurate siekte definisie word geïllustreer deur die feit dat geïsoleerde hilêre adenopatie ‘n verbygaande verskynsel is in die meerderheid van kinders kort na primêre infeksie. Ons analise het daarop gefokus om die risiko om siekte te ontwikkel nadat primêre infeksie met M.tuberculosis plaasgevind het, te kwantifiseer. Die hoof risiko faktore was; 1) die ouderdom en/of immuunstatus van die kind, 2) die tydsverloop sedert infeksie, en 3) die teenwoordigheid van simptome al dan nie. Hierna het ons die siektelas wat tuberkulose vandag op kinders in endemiese areas plaas gedokumenteer. Ons het eers die huidige diagnostiese praktyke geëvaluaeer in ‘n retrospektiewe studie en toe ‘n prospektiewe beskrywende studie gedoen om die siektelas so akkuraat as moontlik te meet. Die insidensie van kindertuberkulose was hoog (>400/100 000/jaar), selfs na korreksie vir kinders wat ontoepaslik behandeling ontvang het. Verder is gevind dat die meerderheid van kinders met ernstige siekte toestande soos miliêre tuberkulose en/of meningitis, nie in roetine moniterings data reflekteer word nie. ‘n Bykomende struikelblok in kindertuberkulose is die gebrek aan standaard beskrywende terminologie. Om dit te bevorder ontwikkel ons ‘n nuwe radiologiese klassifikasie van intra-torakale kindertuberkulose en beskryf ons die verskillende patologiese meganismes onderliggend tot hierdie uiteenlopende siektebeelde. Daarna dokumenteer ons die volledige spektum van kindertuberkulose in ‘n endemiese area en demonstreer dat die siektepatrone wat ons vandag observeer soortgelyk is aan die wat in die pre-chemoterapie literatuur beskryf is. Ons toon ook dat bakteriologiese bevestiging moontlik blyk te wees in die meerderheid van kinders wat vir intra-torakale tuberkulose behandel word in endemiese areas. Nadat ons duidelikheid verkry het oor die basiese siektekonsepte, siekte klassifikasie en die siektelading in ons omgewing, kon ons op die ontwikkeling van ‘n simptoom-gebaseerde benadering tot die diagnose van kindertuberkulose fokus. Ons het ‘n stapsgewyse benadering gevolg. Die eerste stap was om die voorkoms van simptome wat gebruiklik met tuberkulose vereenselwig word te dokumenteer in ‘n ewekansige groep kinders. Die gemeenskapsopname het getoon dat swak gedefiniëerde simptome swak diagnostiese waarde bied. Die tweede stap was om vas te stel of verbeterde simptoom definisie die diagnostiese waarde kan verbeter. ‘n Klein prospektiewe studie het getoon dat goed gedefiniëerde simptome (persisterende simptome van onlangse aankoms) goeie diagnostiese waarde bied. Die finale stap was om hierdie belowende benadering formeel te toets in ‘n groot prospektiewe, gemeenskapsgebaseerde studie. Hierdie studie het getoon dat ‘n eenvoudige simptoom-gebaseerde benadering pulmonale tuberkulose met goeie akkuraatheid kan diagnoseer, veral in HIV-ongeïnfekteerde kinders wat ouer is as 3 jaar. Hierdie nuwe diagnostiese benadering bied die moontlikheid om antituberkulose behandeling te voorsien aan kinders in endemiese areas wat tans feitlik geen behandeling ontvang nie.
Description
Dissertation (PhD)--University of Stellenbosch, 2006.
Keywords
Tuberculosis in children, Tuberculosis in children -- Diagnosis, Mycobacterium tuberculosis, Theses -- Medicine, Dissertations -- Medicine, Theses -- Pediatrics, Dissertations -- Pediatrics
Citation