Prof. H. Simon Schaaf, a clinical researcher from the Department of Paediatrics and Child Health, Faculty of Health Sciences, Desmond Tutu TB Centre, Stellenbosch University, South Africa has worked extensively on childhood TB and multi-drug resistant TB (MDR-TB) in children. Last year he was awarded an A-rating from the National Research Foundation, South Africa. A-rating is the highest award given to scientists by the Foundation. This award was in recognition of his excellence in research in paediatric TB and MDR-TB in children.
Prof. Schaaf explained to me by email on several issues connected with MDR-TB, including why drug-resistant TB should be considered and screened for when children respond poorly to drug-sensitive TB treatment.
Considering that children aged under five years have very few bacilli in the sputum, how is a diagnosis made for MDR-TB even when the index case in the same household has been diagnosed with MDR-TB?
Children in contact with a household index case with MDR-TB should be managed as if they are infected (or have disease) by the same organism (TB bacillus/TB bug) as the index case. Infection means they have the TB bug in their bodies, but do not yet have TB disease (no symptoms, clinically well, normal chest X-ray) while TB (TB disease) means the children have symptoms and/or signs of TB disease and/or an abnormal chest X-ray showing that the infection has developed into disease.
In case of infection after exposure to an MDR-TB case, the organism will usually not be found in the sputum of the child, but in case of disease, health-care workers can get samples of sputum (or gastric aspirates) to try and confirm that the child has the same TB organism as the adult index case.
Our experience in South Africa, a high TB incidence country, is that when we do find the TB organism in the child it is the same as in the identified MDR-TB case in 80 to 90 per cent of cases. Unfortunately, because of the fewer number of organisms in children, even if they do have TB disease, it is only possible to get positive cultures in about 40 per cent of cases. Therefore, if a child develops TB with a known MDR-TB index case, the child should empirically be managed as having MDR-TB — the same as in the adult case — until proven otherwise.
How common are facilities to collect sputum samples using gastric and bronchoalveolar lavages in South Africa?
The main specimen types collected from children with lung TB are gastric aspirates or gastric washings, induced sputum or expectorated sputum (the latter if they can cough and spit out sputum effectively — usually by those >6 years of age).
Bronchoalveolar lavage is only done in cases of complicated lung TB (as this means putting children under general anaesthesia). Collection of gastric aspirates and sputum (induced or expectorated) samples are available at all hospitals, but not usually at primary-care clinic level in South Africa, where many of the children are diagnosed with TB. In cases of suspected MDR-TB, children should therefore be referred to a hospital for specimen collection (especially those <5 years of age).
How many children aged under five years have MDR-TB in South Africa? What percentage of them have been contacts of an index case with MDR-TB in the same household? How many MDR-TB cases are found in children <15 years?
This data [<5 years] is not known. In a recent prospective study we identified 149 children [<15 years] at the two hospitals where I work that we started on MDR-TB treatment. Of these, 75 per cent had a known MDR-TB source case, a few had other source cases and the rest had no known source case.
In South Africa, what percentage of children aged under five years suffer from extra-pulmonary and pulmonary MDR-TB?
This is a difficult question to answer. In surveys done at Tygerberg Children’s Hospital in the Western Cape Province, 8.9 per cent of children diagnosed with culture-confirmed TB had MDR-TB in the period 2009-2011. The majority of children with TB, including MDR-TB, has pulmonary TB, but extrapulmonary TB may occur in approximately 25 per cent of these cases, with or without a pulmonary TB component.
Compared with drug-sensitive TB, do young children have the same chances of getting infected with MDR-TB when the index case in the same household has MDR-TB?
Yes, the infectiousness and ability to cause disease is not less in drug-resistant organisms compared with drug-susceptible organisms.
In South Africa, if the index case has been diagnosed with MDR-TB disease, will children <5 year from the same household with MDR-TB disease be started on treatment right away even before you get a bacteriologically confirmed diagnosis?
If they have a known MDR-TB contact and they are diagnosed with TB disease, we will take specimens for culture and drug susceptibility testing and then start them on MDR-TB treatment according to the index case’s isolate drug susceptibility test result (if that is the only known and recent index case). In more than 80 per cent of cases we are correct with this approach.
How quickly do children progress from an infected state to a diseased state in the case of MDR-TB?
This depends on many factors — young children and children who are immunosuppressed can develop disease after infection very rapidly — within a few weeks. However it can take much longer — the majority (90 to 95 per cent) of children who develop disease after infection will do so within 12 months after infection. There is, however, a lifelong small risk of progression to disease if infected as a child, especially if the immune status of the person deteriorates (e.g. with HIV or old age).
As in the case of drug-sensitive TB, are all asymptomatic children started on prophylactic MDR-TB treatment?
If policy guidelines were implemented, all children <5 years of age in contact with infectious drug-susceptible TB cases [from the same household] should receive isoniazid preventive therapy, but we know that this is poorly implemented, especially in high-burden countries.
With MDR-TB contacts, the guidelines vary widely. It is not yet confirmed that preventive therapy effectively protects high-risk contacts of MDR-TB cases against development of MDR-TB. However, more recent publications and data presented at conferences do indicate that the correct preventive therapy can prevent MDR-TB in these contacts.
Current policy in most countries recommend clinical follow-up of such contacts for two years, and starting them on MDR-TB treatment if they develop TB disease. In some countries a combination of drugs to which the source case is susceptible will be given to the children <5 years (or HIV-infected) for 6-12 months to prevent development of MDR-TB. This is what we do in the Western Cape province of South Africa.
Since preventive chemotherapy for MDR-TB is not a global policy as enough scientific evidence is not available, how are you doing it?
There is no uniform global recommendation, but following the European CDC expert meeting on this issue, it seems as if there is a swing of the pendulum towards favouring preventive therapy for MDR-TB contacts. U.S. already has such a policy, ECDC gives a choice (follow-up with or without preventive therapy) and others say only follow-up (WHO current policy). The evidence is however increasing that preventive therapy is effective in MDR-TB contacts.
Can the absence of a global policy be reason enough for India or any other MDR-TB high-burden country not to take any preventive action?
I think that countries must decide what they think is correct for themselves. Follow-up and correct treatment if contacts develop disease are most important, but as I said above, the evidence is increasing that preventive therapy may be effective in preventing MDR-TB in high-risk individuals including young children <5 years of age.
Which anti-MDR-TB drugs do you give as a prophylactic to children, including those aged under five years? What is the dosage?
In the Western Cape province of South Africa we use a combination of high-dose isoniazid (INH; 15-20 mg/kg daily), ethambutol (20-25mg/kg) and a fluoroquinolone (now levofloxacin 15-20mg/kg/day) for 6 months. We recently published an article where we showed this to be quite safe and most likely effective (Seddon et al, CID 2013). We are thinking of doing a trial leaving out the ethambutol and giving only the other two drugs. In cases of resistance to the fluoroquinolones we give only high-dose INH, as in most of our local cases we have inhA promoter region mutation causing INH resistance, which confers low-level resistance to INH.
What is the duration of prophylactic treatment in the case of children <5 years when the index case has been diagnosed with MDR-TB?
As said, this varies — we use six months, but in other studies, they used 12 months.
What percentage of asymptomatic/infected children are put on preventive therapy? What is the default rate?
The usual adherence to isoniazid preventive therapy [IPT] in contacts of drug-susceptible index cases is poor (our experience 20 to 30 per cent of the low number of children started on IPT will complete the six months). The low number refers to the fact that very few children are screened as contacts of drug-susceptible TB cases, therefore a few are even considered for IPT. However, in MDR-TB contacts in the study we did, [James A. Seddon et al., Clinical Infectious Diseases, October 23, 2013], adherence to the six-month treatment was 75 per cent, but this is under study circumstances.
What are the commonly seen adverse effects in this age group of children when put on MDR-TB preventive therapy? Are they different from those seen in adults?
The adverse effects are the same in type, but are usually less common in children compared to adults. Again I refer you to the above paper — our experience with MDR-TB child contacts on our regimen for MDR-TB prevention is that serious adverse effects are very low.
For how long is the index case infectious before and after starting treatment for MDR-TB?
Before starting treatment — this depends on how long it takes from developing TB until time of starting treatment — [index case could be infectious for] months. After starting treatment — it is somewhat uncertain.
But if the index case is started on the correct MDR-TB treatment and the treatment is taken, the period of infectiousness should not be more than 1-2 months — and some experts believe immediately non-infectious, but I find that hard to believe. For safety’s sake, [the index case] is considered infectious until the individual is at least smear-negative for acid-fast TB bacilli twice over a period of >1 month. If these individuals interrupt treatment, they may become infectious again!
Which age groups are more prone to getting infected with MDR-TB even within the under-five age bracket? Is it like the one seen in the case of drug-sensitive TB?
Yes, [it’s the] same — younger and immunocompromised [children have] higher risk of disease. Extent of contact with index case (infectiousness — how many organisms coughed up, closeness and duration of contact) will dictate risk for infection.
Have there been instances when young children in South Africa are infected/diseased with MDR-TB even when the index case in the household has drug-sensitive TB?
This question can have a long explanation, but in essence: the child will develop disease from the organism the child has been infected with. In the same household (or community) there may be more than one source case spreading TB and some may not yet have been diagnosed. So, yes, the known source case may have drug-sensitive TB while the child has MDR-TB because the real source case has not yet been identified. In older children it is possible to develop drug-resistant TB if they take their treatment poorly or if they are prescribed incorrect treatment.
Are there instances when doctors wait for clinical outcomes after starting on first-line anti-TB drugs before testing or starting treatment for MDR-TB in children?
Yes. Sometimes the source case is not known — and not all children are started on TB treatment with cultures having been done (majority of TB is diagnosed without TB cultures in children). If children respond poorly, drug-resistant TB should be considered and screened for.
Are young children closely monitored during the entire course of MDR-TB preventive treatment? How important is monitoring and what are they monitored for?
Monitoring occurs, but differs widely from situation to situation due to resources, knowledge, number of TB cases etc. We do follow our known MDR-TB contacts closely for a minimum of 12 months to see whether they develop the disease or not.
(The Correspondent is a recipient of the 2013 REACH Lilly MDR-TB Partnership National Media Fellowship for Reporting on TB.)