Pharmocokinetics of azithromycin in severe malaria bacterial co-infection in African children

Severe malaria killed an estimated 475,000 African children in 2013. Fast-acting effective antimalarial drugs are now used in most hospitals, but a large number of children still die (~1 in every 10). To reduce this number, we need to find better ways to manage these sick children. Some children with severe malaria infection also have a higher chance of also having infections caused by bacteria at the same time. These bacterial infections increase the risk of children with severe malaria dying in hospital even more (to ~1 in 4 chance). Around one-third of all severe malaria deaths in African children are thought to be due to these bacterial infections. The problem is that most African hospitals are not able to grow the bacteria from blood to work out which children really have these bacterial infections. So there are two options: no one gets antibiotics, or everyone gets antibiotics. The problem with giving all these children antibiotics is that most of them don't need them, and using antibiotics for all children can increase the risk of resistance in the community (meaning antibiotics stop working for people who really need them). There is no agreement on which antibiotics, at what dose or for how long, they should be used in children with severe malaria. The main bacteria responsible for these infections come from the gut, because the gut becomes 'leaky' in severe malaria so these bugs can cross over into the blood. These bacteria are frequently resistant to, or are not treated by, currently recommended and commonly available antimicrobials. What is needed now is to examine one of the antibiotics that can be given by mouth which has the potential to treat most common causes of infections and to find out what is the correct dose to give (to treat infections) in order to progress to the next step which will be a larger trial comparing different types of antibiotics to improve both short term and longer term outcomes.

We plan to examine 3 doses (10, 15 and 20mg/kg) of an oral antibiotic called azithromycin given for 5 days to find the optimal dose for curing infection in 105 Uganda children hospitalized with severe malaria that have the greatest risk of bacterial co-infection. We have chosen azithromycin because it is not used commonly to treat other infections, so using it in many children with severe malaria should not stop it working for these other conditions. Previous studies have also suggested that azithromycin could help the body fight infections and maybe helpful itself against malaria. To find out which is the right dose we will measure the levels of azithromycin in samples sent to a specialist laboratory in Nijmegen and pharmacological data will be compared to the clinical and infection (microbiological) outcomes of the children in the study in order for us to select the optimal and safest dose for future clinical trials.

We also want to find out whether we can identify children with severe malaria who are at risk or bacterial infection and those who are not so antibiotics could be targeted in future. We will also recruit a small number of children (n=50) hospitalised with non-severe malaria to examine whether special blood tests that could be done at the patient's bedside in combination with clinical signs could predict which children with severe malaria really have bacterial infections as well as malaria and so need antibiotics. We want to do this so that we can target antibiotics better in future to children who really need them, and reduce the total amount of antibiotics we use. This is to stop the spread of antibiotic resistance. Using cheap tests like this would help resource-limited hospitals across many parts of Africa where microbiological services to grow bacteria are poorly developed or non-existent.

African children with severe malaria are susceptible to gram-negative bacterial co-infection resulting in a substantially higher rates of in-hospital and post-discharge mortality. Current evidence for treating co-infection is lacking, and there is no consensus on the dosage or length of treatment required.

To provide supporting data to inform a future trial we propose to (i) establish the appropriate dose of oral dispersible azithromycin as an antimicrobial treatment for children with severe malaria and (ii) investigate whether antibiotics can be targeted to those at greatest risk of bacterial co-infection using clinical criteria alone or in combination with a rapid diagnostic biomarker tests. We will conduct a Phase I/II trial in 105 Uganda children with severe malaria at highest risk of bacterial co-infection comparing three doses of azithromycin (10, 15 and 20 mg/kg prescribed for feasibility by weight-bands) given over 5 days in order to generate relevant pharmacokinetic (PK) data by sparse sampling during dosing intervals. The goal is to determine, via population PK modelling, the optimal azithromycin dose in severe malaria, and investigate associations between azithromycin exposure and potential mechanisms (PK-pharmacodynamics, PKPD) using change in C-reactive protein, a putative marker of sepsis at 72 hours (continuous) and microbiological cure (7-day) (binary), alone and as a composite with 7-day survival, while providing preliminary data on longer-term survival (to day-90). We also aim to establish whether a mixture of clinical and rapid diagnostic tests can be used to predict which children with severe malaria are at greatest risk of having bacterial co-infections so that the results may be generalizable in future where microbiological services are poorly developed or non-existent both in children enrolled to the trial and in a comparator cohort (n=50) of children hospitalised with severe malaria at low risk of bacterial co-infection.

In Africa, clinical syndromic management of childhood illnesses means that without evaluation of current treatment recommendations in clinical trials, polypharmacy may become the norm in the management of severe malaria. The direct beneficiaries therefore will be African children hospitalised with severe malaria. In the broader context, research conducted to address this key public health concern and progress in this area will have wider downstream benefits, given the high incidence of severe malaria in many tropical communities, since the indiscriminate use of antibiotics in all children presenting with severe malaria would be costly both financially and with regard to potential development of antibiotic resistance.

At present there are no ongoing trials addressing this major research gap and data to estimate plausible effect sizes are completely lacking. One major barrier to progress in this field is that most African hospitals lack the microbiological facilities so are therefore not able to inform or tailor antimicrobial treatment. Nevertheless, despite the fact that current evidence for treating co-infection in children in these settings is lacking, WHO treatment guidelines indicate 'all children with suspected severe malaria should receive broad spectrum antibiotics in areas of moderate and high transmission until a bacterial infection is excluded'. The definition for severe malaria are very broad, incorporating hyperparasitaemia as a single criterion, and thus potentially applicable to a large proportion of paediatric admissions in such regions. Moreover, there is no consensus on the type, dosage or duration of antibiotic treatment required. There are three potential options either recommend no-one gets antibiotics, or everyone with suspected severe malaria receive antibiotics or to conduct trials to inform treatment policy for hospitals with and without microbiological service.

The multi-facteted approach we propose to adopt will be the first step to provide data on the optimal azithromycin dose in severe malaria to help inform the rational design of future trials. Our outcome measures for this study include microbiological cure (7-day) alone and as a composite with 7-day survival and longer-term survival (to day-90) that will also help to inform the sample size for future trials. In addition, the sub-study investigating clinical and point of care diagnostic biomarkers will also enable us to identify whether azithromycin could be further targeted to subgroups that would benefit most, which is essential given the global challenge of antimicrobial resistance. We envisage that the design of future Phase III trial could have two strata: the first stratum comparing a pharmacologically-informed dosage of oral azithromycin to standard-of-care (largely 3rd generation cephalosporins) in children at highest risk of bacterial co-infection and second stratum comparing standard of care versus no antibiotics in children with severe malaria but minimal risk of bacterial co-infection to establish whether a policy for targeted antibiotic therapy could substantially reduce malaria-associated mortality while minimising the risks of excess antibiotic prescribing.