Vigilance needed over MDR-TB in children, South African doctors warn

South African doctors have found a high prevalence of multi-drug resistant tuberculosis among children receiving care in a Johannesburg hospital, highlighting the need to screen for drug resistance and test for HIV in children with TB symptoms.

Among a cohort of children with culture-confirmed tuberculosis in Johannesburg, South Africa, a setting with high rates of HIV co-infection, 8.8% (95% CI: 4.8-14.6%) had multi-drug resistant tuberculosis (MDR-TB) and 14.2% (95% CI: 9.0-20.9%) were isoniazid resistant according to Lee Fairlie and colleagues in a retrospective cross-sectional study published in the January 26 edition of BMC Infectious Diseases.

Only four of the children with MDR-TB had contact with an adult with TB, none of whom had MDR-TB suggesting that MDR-TB occurred within the community as well as the household, the authors note. They add this probably indicates a significant number of undiagnosed drug-resistant TB among adult contacts of these children.

The authors add that in spite of high prevalence of HIV co-infection no association between HIV and MDR-TB was found.

Sub-Saharan Africa has a high burden of both HIV and TB. The majority of the estimated 2.5 million children with HIV live in sub-Saharan Africa, of which 10% are in South Africa. In 2007 of the 2.9 million new cases of TB in sub-Saharan Africa, 10% were in children; close to 40% of all cases were co-infected with HIV.

Global prevalence of MDR-TB in adults, meaning resistance to at least isoniazid and rifampicin was estimated to be 4.6% in 2006, the authors note. Data on the prevalence of drug resistance in children with TB in Africa are limited and estimated at 0.6-6.7%.

Children with TB, the authors note, usually represent cases of recently acquired infection so are an important measure of on-going adult disease and “the drug-susceptibility patterns and circulating strain diversity of TB in a community.”

The authors undertook an analysis of the prevalence of MDR-TB in children under 14 years of age with culture-confirmed TB diagnosed between January 1 and December 31, 2008 at Chris Hani-Baragwanath Hospital (CHBH) and Rahima Moosa Mother and Child Hospital (RMMCH) in Johannesburg, South Africa. They reviewed laboratory data including clinical and radiological outcomes at six and 12 months post-diagnosis.

1317 children were treated for TB in 2008 at CHBH and RMMCH. Of the 204 children with culture-confirmed TB, 148 (72.5%) underwent drug susceptibility testing (DST).

Among whom, the authors note, the high MDR-TB prevalence rate of 8.8% was higher than in previous African paediatric studies (2.3-6.7%). These results suggest a high prevalence of MDR-TB within this setting where contact tracing is poorly performed, they add.

16% had drug resistant TB. Of the 14% of children resistant to isoniazid 75% were HIV-infected. The high rates of isoniazid resistance could affect the recently recommended South African TB guidelines raising questions about the effectiveness of isoniazid preventive therapy (IPT) in this community, note the authors. It may partly explain why primary isoniazid prophylaxis is not effective in improving TB-free survival of HIV infected children in this setting, they add.

South African TB guidelines do not recommend routine DST in adults or children. However, the authors note that planning is in process to improve routine testing of all smear-positive adults and paediatric specimens using a line probe assay and so report drug sensitivity patterns more rapidly.

Of the 13 children with MDR-TB, ten (76.9%) were treated of which four (30.8%) died at a median of 2.8 months (0.1-4.0 months) after diagnosis. Three of the children were HIV-infected and on antiretrovirals and were started on MDR-TB treatment with no significant delay compared to those who lived. Previous reported paediatric studies in the Western Cape and in Lima, Peru had significantly lower mortality rates, 10% over a four year period with 50% HIV co-infected and 3.5%, respectively.

The authors note that all the children died in hospital highlighting the problems of infection control. In many general hospitals in South Africa appropriate isolation facilities are lacking. Even with efficient and swift diagnosis of MDR-TB there will be, nonetheless, an increased risk of transmission. The authors stress the urgency for infection control needs to be addressed in areas of increased MDR-TB rates and high HIV prevalence.

Limitations according to the authors include a possible overestimation of MDR-TB prevalence since those who attend referral hospitals may be at increased risk compared to those treated in primary care facilities. DST at the clinician’s discretion undertaken in 72.5% of children with culture-confirmed TB may have introduced bias.

The authors conclude the high prevalence of drug-resistant tuberculosis “likely represents a large burden of undiagnosed drug-resistant TB in household and community adult contacts of these children.” “Routine HIV [testing] and DST are warranted to optimise the management of childhood tuberculosis in settings such as ours.”