REVIEW URRENT C OPINION
Managing multidrug-resistant tuberculosis in children: review of recent developments H. Simon Schaaf a, Anthony J. Garcia-Prats a, Anneke C. Hesseling a, and James A. Seddon a,b
Purpose of review Childhood multidrug-resistant (MDR) tuberculosis is an emerging disease with increasing numbers being recognized. This review presents recent developments in childhood MDR tuberculosis. Recent findings New molecular-based diagnostic tests, although not optimal, have reduced the difficulty in confirming the diagnosis of MDR tuberculosis in children. However, the importance of making a diagnosis of probable MDR tuberculosis has been reaffirmed by contact tracing studies showing 80–90% of child contacts of MDR tuberculosis cases who develop disease have MDR tuberculosis themselves. Prevention of MDR tuberculosis in child contacts with appropriate preventive treatment regimens is supported by new observational data and deserves further study. When diagnosed and treated appropriately, outcomes for MDR tuberculosis and even extensively drug-resistant tuberculosis in children are good, despite limited pharmacokinetic data on second-line drugs. Novel anti-tuberculosis drugs and regimens are becoming available and should be studied in children for dose-finding and safety. Recording and reporting of MDR tuberculosis in children are frequently poor, leading to inaccurate estimates of disease burden and suboptimal resource planning. Summary Rapid diagnosis and appropriate treatment results in good outcomes in the majority of children with MDR tuberculosis. Additional research on optimal diagnosis, prevention and treatment of MDR tuberculosis in children remains a high priority. Keywords children, diagnosis, multidrug-resistant tuberculosis, prevention, treatment
INTRODUCTION Drug-resistant tuberculosis is an emerging problem but is under-recognized in children. This is mainly because of difficulties in confirming the diagnosis by culture and drug-susceptibility testing (DST) and the perception that children are not at risk for drugresistant tuberculosis. The number of all cases of multidrug-resistant tuberculosis (MDR tuberculosis; i.e. resistance to at least isoniazid and rifampicin) and extensively drug-resistant tuberculosis (XDR tuberculosis; i.e. MDR tuberculosis plus resistance to a fluoroquinolone and at least one of the secondline injectable drugs) are increasing worldwide; World Health Organization (WHO) estimates of new MDR tuberculosis cases increased from 250 000 in 2009 to 450 000 in 2012 [1–2]. Progress towards achieving WHO targets for the diagnosis and treatment of MDR tuberculosis is slow; the majority of high-burden tuberculosis countries
detected less than 25% of the patients estimated to have MDR tuberculosis in 2012 [2]. MDR tuberculosis surveillance data in children is poor. A recent evaluation of data from WHO found that only 35 out of 85 countries reporting MDR tuberculosis cases included children ( twice Xpert cost
More than Xpert
Used as comparisona
Time to DST result
2–6 weeks (depending on organism load)
Directly on specimen within hours; with cultured isolates dependent on culture time
Within 2 h
DST results
First and second-line DST on isolates
Provides both INH and RIF DST results as well as mutation conferring INH resistanceb If mycobacteria isolated, further DST can be done. LPA for second-line DST available but not as reliable as INH/RIF DST
Only provides DST result for RIF. No further DST possible on same specimen
Sensitivity [11]
Gold standard for culture
100% in smear-positive specimens; 58–60% in smear-negative, culture-positive specimens
100% in smear-positive specimens; 58–60% in smear-negative, culturepositive specimens
Type of specimen
Any specimen obtained
Any specimen obtained
Mainly sputum; recently used with several other specimens, e.g. gastric aspirates, fine needle aspiration, other body fluids and even stool (with special preparation) [12–15]
DST, drug susceptibility testing; INH, isoniazid; LPA, line probe assay; MGIT, mycobacterial growth indicator tubes; RIF, rifampicin. a If taken into account that Xpert MTB/RIF usually replaces smear microscopy and culture is still done, especially in children with mostly smear-negative tuberculosis disease, Xpert cost may be additional cost to culture and DST/LPA. b Provides information on level of INH resistance and possible co-resistance with ethionamide.
tuberculosis source cases diagnosed as rifampicinsusceptible using Xpert. Drug-resistant tuberculosis is essentially a microbiological diagnosis, but bacteriological confirmation is typically possible in less than half of children. Traditionally, clinicians and tuberculosis programmes were reluctant to treat children for drug-resistant tuberculosis without bacteriological confirmation. Recent studies have demonstrated the importance of empirical MDR tuberculosis treatment, mainly in the context of clinical disease combined with exposure to an MDR tuberculosis source case. Of child MDR tuberculosis contacts who develop tuberculosis, 80–90% have MDR tuberculosis [22–25], although complete concordance of DST patterns is sometimes relatively low (
Managing multidrug-resistant tuberculosis in children: review of recent developments H. Simon Schaaf a, Anthony J. Garcia-Prats a, Anneke C. Hesseling a, and James A. Seddon a,b
Purpose of review Childhood multidrug-resistant (MDR) tuberculosis is an emerging disease with increasing numbers being recognized. This review presents recent developments in childhood MDR tuberculosis. Recent findings New molecular-based diagnostic tests, although not optimal, have reduced the difficulty in confirming the diagnosis of MDR tuberculosis in children. However, the importance of making a diagnosis of probable MDR tuberculosis has been reaffirmed by contact tracing studies showing 80–90% of child contacts of MDR tuberculosis cases who develop disease have MDR tuberculosis themselves. Prevention of MDR tuberculosis in child contacts with appropriate preventive treatment regimens is supported by new observational data and deserves further study. When diagnosed and treated appropriately, outcomes for MDR tuberculosis and even extensively drug-resistant tuberculosis in children are good, despite limited pharmacokinetic data on second-line drugs. Novel anti-tuberculosis drugs and regimens are becoming available and should be studied in children for dose-finding and safety. Recording and reporting of MDR tuberculosis in children are frequently poor, leading to inaccurate estimates of disease burden and suboptimal resource planning. Summary Rapid diagnosis and appropriate treatment results in good outcomes in the majority of children with MDR tuberculosis. Additional research on optimal diagnosis, prevention and treatment of MDR tuberculosis in children remains a high priority. Keywords children, diagnosis, multidrug-resistant tuberculosis, prevention, treatment
INTRODUCTION Drug-resistant tuberculosis is an emerging problem but is under-recognized in children. This is mainly because of difficulties in confirming the diagnosis by culture and drug-susceptibility testing (DST) and the perception that children are not at risk for drugresistant tuberculosis. The number of all cases of multidrug-resistant tuberculosis (MDR tuberculosis; i.e. resistance to at least isoniazid and rifampicin) and extensively drug-resistant tuberculosis (XDR tuberculosis; i.e. MDR tuberculosis plus resistance to a fluoroquinolone and at least one of the secondline injectable drugs) are increasing worldwide; World Health Organization (WHO) estimates of new MDR tuberculosis cases increased from 250 000 in 2009 to 450 000 in 2012 [1–2]. Progress towards achieving WHO targets for the diagnosis and treatment of MDR tuberculosis is slow; the majority of high-burden tuberculosis countries
detected less than 25% of the patients estimated to have MDR tuberculosis in 2012 [2]. MDR tuberculosis surveillance data in children is poor. A recent evaluation of data from WHO found that only 35 out of 85 countries reporting MDR tuberculosis cases included children ( twice Xpert cost
More than Xpert
Used as comparisona
Time to DST result
2–6 weeks (depending on organism load)
Directly on specimen within hours; with cultured isolates dependent on culture time
Within 2 h
DST results
First and second-line DST on isolates
Provides both INH and RIF DST results as well as mutation conferring INH resistanceb If mycobacteria isolated, further DST can be done. LPA for second-line DST available but not as reliable as INH/RIF DST
Only provides DST result for RIF. No further DST possible on same specimen
Sensitivity [11]
Gold standard for culture
100% in smear-positive specimens; 58–60% in smear-negative, culture-positive specimens
100% in smear-positive specimens; 58–60% in smear-negative, culturepositive specimens
Type of specimen
Any specimen obtained
Any specimen obtained
Mainly sputum; recently used with several other specimens, e.g. gastric aspirates, fine needle aspiration, other body fluids and even stool (with special preparation) [12–15]
DST, drug susceptibility testing; INH, isoniazid; LPA, line probe assay; MGIT, mycobacterial growth indicator tubes; RIF, rifampicin. a If taken into account that Xpert MTB/RIF usually replaces smear microscopy and culture is still done, especially in children with mostly smear-negative tuberculosis disease, Xpert cost may be additional cost to culture and DST/LPA. b Provides information on level of INH resistance and possible co-resistance with ethionamide.
tuberculosis source cases diagnosed as rifampicinsusceptible using Xpert. Drug-resistant tuberculosis is essentially a microbiological diagnosis, but bacteriological confirmation is typically possible in less than half of children. Traditionally, clinicians and tuberculosis programmes were reluctant to treat children for drug-resistant tuberculosis without bacteriological confirmation. Recent studies have demonstrated the importance of empirical MDR tuberculosis treatment, mainly in the context of clinical disease combined with exposure to an MDR tuberculosis source case. Of child MDR tuberculosis contacts who develop tuberculosis, 80–90% have MDR tuberculosis [22–25], although complete concordance of DST patterns is sometimes relatively low (
