Plasma concentrations of rifampicin may be too low in children using currently recommended doses

The currently recommended dosage of rifampicin for children may be too low according to a fairly large pharmacokinetic study presented at the 39th Union World Conference on Lung Health held in Paris.

“We feel that children are getting much lower serum concentrations of rifampicin than adults,” said Professor Peter Donald of Stellenbosch University who presented the study, “and we think that the World Health Organisation recommendations should be increased.”

Background

Rifampicin is an essential drug in the management of tuberculosis, but there is very little data regarding the pharmacokinetics of the drug in children, and no data whatsoever on its pharmacokinetics in children with HIV.

So from 2004 to 2006, his team enrolled children into a pharmacokinetic study at Brooklyn Hospital for Chest Diseases in Cape Town, South Africa. The children, both HIV-positive and negative, were aged between 3 months to 13 years of age, and had severe forms of tuberculosis. A qualified dietician assessed their nutritional status (anthropometry) using standard methodology.

Pharmacokinetic evaluations occurred on two occasions, approximately one and four months after starting treatment. After dosing, sequential blood samples were taken at one, one and a half, three, four, and six hours to measure plasma concentrations, with children fasting from midnight onwards. 2 hour concentrations were calculated.

The children were given a fixed dose formulation of the standard TB regimen (made by Sandoz), which supplied a mean dose of 9.61-3 mg per kilogram bodyweight.

Participant characteristics

The study enrolled 60 children, 26 of whom were HIV-positive (with a mean age of 3.73 years) and 34 who were uninfected with a mean age of 4.05 years. 6 subjects dropped out.

Baseline diagnostic, clinical and radiological characteristics were similar between the HIV-positive and negative, with the exception of the TB skin test (Mantoux) which was significantly more likely to be positive in the HIV-uninfected. The HIV-infected children had a worse nutritional status however, with more children with marasmus (38% vs 9%, p<0.01) or marasmic kwashiorkor (19% vs 9%, p<0.04).

Anthropometry and serum CDR (mg/l) one month and four months after treatment

Results

Rifampicin serum concentrations were lower than anticipated, and tended to be slightly lower in HIV-infected children, though the confidence intervals overlapped.

One month after starting treatment, the mean rifampicin concentration in the six hours after dosing (AUC_0-6) was 14.9 and 18.1 µg h/ml in the HIV infected and uninfected children respectively. Four months after starting treatment, concentrations were higher at 16.52 and 17.94 µg h/ml respectively (p=0.59).

The mean two-hour concentration was 3.9 and 4.8 µg/ml respectively (p=0.20) at one month after starting treatment and 4.0 and 4.6 µg/ml (p=0.33) at four months after starting treatment.

These values are “very low” according to Professor Donald, as the scientific literature recommends aiming at concentrations closer to 8-20 µg/ml at two hours post dosing. And in order to have a bactericidal effect on mycobacteria, a concentration of around 10-15 mg/L is needed, which is generally accomplished by a dose of 10 mg/kg in adults.

Thus current WHO guidelines recommend that daily and intermittent dosing for adults and children should be around 10 mg/kg (range 8-12).

However, this study suggests that those dose recommendations may not be enough for children. “Close to half of the children in our study had calculated 2-hour concentrations of less than 4 µg/ml,” said Professor Donald, “so adjustment of dosages of rifampicin is necessary, particularly in young children.”

He noted that in other fields, body surface area is used to calculate the appropriate doses for children.

Another paper several years ago applied these calculations to rifampicin

The dose of rifampicin calculated by body surface area (BSA) (Ansel & Stoklosa, 2001).

Another doctor in the audience asked if all this was theoretical, since the currently recommended dose seems to work fine in most children.

Professor Donald replied that “many children with TB have really very minimal disease. The currently recommended doses may work fine for those children, but for more severe forms of childhood TB I believe we need a similar concentration as for adults. We tend to use 10-20 mg/kg in children by manipulating the fixed dose formulation.”

Professor Donald also suggested that the dose of isoniazid may be too low in children with severe forms of TB (this will be covered in more detail in an upcoming issue of HATIP).