Correlates of risk in children exposed to multidrug-resistant tuberculosis

Lead Research Organisation: Imperial College London


About ten million people become ill with TB each year and many of these people live in a house with children. If a child breathes in TB bacteria there is a chance that they will develop TB infection. This means that the TB bacteria are living in their body but not causing any problems. Roughly sixty million children globally have TB infection. Each year, in about a million of these children, the TB bacteria overcome the immune system and cause the child to become sick. This is then called TB disease. Of the children who get TB disease, a quarter die. This makes TB one of the greatest causes of death in children worldwide. If you give a medication to children who have TB infection it reduces their chance of developing TB disease. However, you would have to treat a large number of children to prevent each case of TB disease. In most places in the world, even though it is the policy to give this preventive medicine, it is not given. If it were possible to identify which children were at the highest risk of developing TB disease after being exposed to TB in their house, then you could be more selective and give the medications just to them. It would also be useful to find out what happened to the immune system of children when they are given TB infection treatment, as this might help both to make decisions about clinical care as well as for studies of new treatments. Finally, we know that a virus called cytomegalovirus (or CMV), switches on certain parts of the immune system in children. These parts have been associated with the development of TB disease. CMV is a very common virus and it will be important to find out whether having CMV increases the risk of developing TB infection and disease.

1. Is it possible to predict, from a blood sample taken from a child soon after they have been exposed to someone with TB, whether they are at high risk of developing TB disease themselves?
2. Is it possible to identify, from a blood test, a change in the immune system in children with TB infection when they are treated?
3. Does CMV infection switch on parts of the immune system that are associated with the development of TB disease?
4. Does CMV infection increase the risk of getting TB infection and disease after you have been exposed to someone with TB disease?

Although we know how to treat children with TB infection if they have been exposed to normal TB, we do not know what to do if they have been exposed to drug-resistant TB. A clinical trial has already been funded (called TB-CHAMP) that plans to answer this. TB-CHAMP will recruit 1500 children, from four sites in South Africa, who have been exposed to drug-resistant TB and randomise them to receive either an antibiotic called levofloxacin each day for six months or a placebo (inactive, dummy drug). This trial provides an amazing and unique opportunity to study what happens to children following exposure, who are not given treatment. As we know how to treat children exposed to normal TB, it would not be appropriate to monitor a group of them carefully without treatment. As part of the trial all children will have regular blood tests taken. In this Fellowship, I plan to use these blood samples to answer the questions above by doing laboratory and data analysis.

The majority of the work will be done by myself in South Africa with regular support and collaboration from South African experts at Stellenbosch University. However, I will also benefit from collaborating with clinical and basic scientists in London who will provide advice, input and training.

If it were possible from a blood test to identify which children were at the greatest risk of getting sick, after they have been exposed to someone with TB, this could dramatically change how we manage children who have been exposed to TB and might reduce the number who become unwell with TB disease.

Technical Summary

Of the estimated 60 million children with TB infection, 1 million each year develop TB disease. Of these one quarter die. If it were possible to identify, for example by using a blood transcriptomic signature, which children with TB infection were at highest risk of TB disease progression, treatment could be targeted. In addition if it were possible to identify a change in transcriptomic signature with drug therapy, this may have implications for patient care and for drug development. Finally, the highly immunogenic virus, CMV, generates a type I interferon response, a pattern of immune response also seen in patients with TB disease. Exploring the role of CMV in the risk and pathogenesis of childhood TB would provide insight into the immunology of dual infection but also potentially contribute to TB vaccine design and alternative interventional strategies.

1. To identify a blood RNA transcriptomic signature that can identify which MDR-TB-exposed children are at increased risk of TB disease progression
2. To identify a blood RNA transcriptomic signature in children with TB infection that changes with antimicrobial therapy
3. To evaluate the impact of CMV viraemia on the RNA transcriptomic signature of children with TB infection
4. To evaluate the impact of CMV infection on the risk of developing TB infection and disease in young children following exposure to TB

The TB-CHAMP trial has already been funded (MRC/WT/DfID) to investigate the management of child MDR-TB contacts. The trial will recruit 1500 children and follow them up for 2 years. Children will be randomised to daily levofloxacin or placebo for six months. Regular blood samples will be taken. For this Fellowship samples will be analysed for their RNA expression profile and CMV status. Bioinformatic analysis will be used to identify baseline signatures that are associated with disease progression and changes on treatment. The impact of CMV will be explored.

Planned Impact


1. Children exposed to TB
2. The families of children exposed to TB
3. Communities with high TB burdens
4. Healthcare workers looking after children exposed to TB
5. Clinical and basic science investigators at Imperial and Stellenbosch
6. The wider academic community
7. Industry involved in diagnostics and vaccine development
8. Policy-makers
9. Ministries of Health and National TB Programmes
10. Advocates for Childhood TB


1. This research has the potential to make a dramatic difference to the lives of children exposed to TB. If we could identify those at high risk of disease progression, infection treatment could be targeted, with improved implementation and adherence, and reducing toxicity and pill burden. The number of children developing TB disease would fall, along with a fall in TB mortality and morbidity. Identifying a signature that changes with treatment may allow for the development of more effective TB infection treatments and reduce the time required to evaluate new regimens. A better understanding of the role of CMV in TB pathogenesis in children could lead to interventions to reduce the impact of CMV, or give further impetus for vaccine development.
2. Currently many children are treated for TB infection to prevent each TB disease case. These treatments are challenging to provide for families and are associated with high pill burden and toxicity. Children require regular follow up, necessitating parents taking time off work. By providing treatment only to the children who need it would reduce unnecessary treatment.
3. Managing children exposed to TB can put a significant strain on communities with high TB burden. The community impact of children becoming unwell or dying of TB is hugely damaging. If research findings suggest it is possible to predict which children are at the highest risk, treatment could be targeted with benefit to the community.
4. Currently it is challenging for healthcare workers to manage a child who has recently been exposed to TB. The current tests of TB infection are imperfect and are not good at predicting who will progress to TB disease. Long courses of treatment are often prescribed, without knowing whether it is beneficial for that individual child. By identifying a marker of increased risk, treatment can be targeted to those most in need, freeing up time for other priorities.
5. In the conduct of this research, the academic and technical staff will learn new skills, develop new collaborations and experience how research is conducted in a large, cutting-edge environment. There will be significant capacity strengthening in laboratory and bioinformatic staff in Cape Town.
6. The findings of this research will have an impact on the academic staff carrying involved with research into immunology, bioinformatics, clinical trials, virology, epidemiology and others as described in the academic beneficiary section.
7. Organisations and commercial companies that develop diagnostics and vaccines for TB will be interested in the results of the research as the immunological insights could provide future directions for product development.
8. For childhood TB policy-makers, this research could impact on recommendations for treatment of TB infection and how to manage CMV infections. Although positive findings would need to be translated to point-of-care tests, the potential to use a predictive test would be operationally attractive and wide implementation would be anticipated.
9. Findings may reduce the burden of childhood TB, reducing costs to National TB Programmes, costs which be redirected to other priorities.
10. This research could provide a powerful tool for childhood TB advocates to communicate the message that although children are vulnerable to TB disease progression following infection, those at the highest risk can be identified and treatment can be targeted.


10 25 50