Answer to Q1
We need further information on the source case. If there is known contact with an infectious adult with pulmonary TB, the child will most likely be infected with the same strain as the source case. This is not only true for children but also for adults. In studies of household contacts of multidrug-resistant (MDR) cases who developed TB, 75-88% had MDR-TB as well. (1-4)
In this case, the family with whom the child’s parents stayed was a known family with several extensively drug-resistant (XDR)-TB cases over a period of 2-3 years. The current source case was discharged from hospital after therapy for XDR-TB failed. She was given the necessary advice and counselling regarding the contagious risk to other people. Despite this she took in boarders, including the patient and his family. There was no other known source case, the parents did not have TB therefore it was presumed that the child was infected with XDR-TB. It is important to remember that young age is one of the most important risk factors to develop disease after infection. In figure 1 we illustrate the age related risk to develop disease after infection and the risk of developing disseminated forms of disease.(5)
Figure 2: Age related TB disease risk after M. tuberculosis infection in HIV-uninfected children ten years and younger (5)
We need to obtain specimens from the child to try to confirm XDR-TB.
Confirming the diagnosis of TB through culture or molecular techniques is more difficult in infants and young children than in adults. Culture confirmation is only possible in 10-20% of children investigated for TB in community settings and 30-40% of children in hospital settings, the latter because the TB disease is usually more advanced once referred to hospital. The reason for the low yield is the paucibacillary nature of primary TB and the difficulty of obtaining sputum specimens from children.
The majority of children with TB are diagnosed on a history of contact, suggestive symptoms, and a CXR. This is acceptable practice for drug susceptible TB contacts. However, where there is any concern of drug resistance all efforts should be made to obtain specimens for bacteriological confirmation and drug susceptibility testing (DST) from the site(s) of disease prior to starting therapy if there is no immediate risk to the child, such as in children with tuberculous meningitis or miliary TB. Respiratory specimens should be obtained by gastric aspirate, induced sputum or, if bronchoscopy is done, by bronchoalveolar lavage. Multiple specimens are preferable. Xpert MTB/RIF could be done, but the sensitivity of Xpert is only 60-70% in smear-negative culture-positive pulmonary TB cases in children (a new, more sensitive Xpert Ultra has been developed). It also only provides rifampicin (RIF) DST and further DST cannot be performed on such a specimen. (6-9) If, however the Xpert confirms Mycobacterium tuberculosis and RIF resistance in a child with a known history of contact with a drug-resistant source case, it rapidly brings us closer to the correct diagnosis. Culture and DST is mandatory in these cases. Genotypic DST, such as the line probe assay (e.g. GenoType MTBDRplus, Hain Lifescience, Nehren, Germany) will yield results for both isoniazid (INH) and RIF resistance, as well as inform us on the mutation conferring INH resistance. The latter is important because in case of an inhA mutation conferring INH resistance, cross resistance with ethionamide is evident, but there is only low-level INH resistance, while if it is a katG mutation, high-level INH resistance is likely but ethionamide may still be an active drug. (10) With a positive culture available, further phenotypic, and possibly genotypic DST, for second-line antituberculosis drugs, especially the fluoroquinolones and second-line injectable agents can be done to confirm XDR-TB.
This child presented with clinical airway obstruction which was confirmed on CXR. The extent of airway obstruction and the possible need for surgical intervention can only be truly assesed by bronchoscopy with or without computed tomography (CT) scan. In most cases, corticosteroids will first be given together with antituberculosis treatment, but with lobar or lung collapse or hyperinflation of a lung due to a ball-valve effect, surgical decompression of the nodes may be indicated.(11)
In our child the Xpert on respiratory specimens was negative, but we confirmed XDR-TB on culture from gastric aspirates, induced sputums as well as bronchoalveolar lavage. INH resistance was conferred by a katG gene mutation. Bronchoscopy confirmed airway compression, which responded to prednisone therefore surgical decompression was not indicated.
Question 2: How would you treat this child and is there any other information needed to decide on an effective treatment regimen?
