Population Pharmacokinetics of Isoniazid in Children with Pulmonary Tuberculosis
Simbarashe P. Zvada , Ulrika S.H. Simonsson , Paolo Denti , Peter R Donald , H Simon Schaaf , Pete J. Smith , Helen M .McIlleron 
 Division of Clinical Pharmacology, Department of Medicine, University of Cape Town, South Africa,  Department of Pharmaceutical Biosciences, Uppsala University, Sweden,  Department of Paediatrics and Child Health, Faculty of Health Sciences, University of Stellenbosch, and Tygerberg Children’s Hospital, South Africa
Background and Objectives: There is limited isoniazid pharmacokinetic information in children, and daily doses (4-6 mg/kg) derived from adult doses have been reported to be inadequate . The World Health Organisation’s revised guidelines (2010) recommend a dose of 10 (10-15) mg/kg/day in children across weights . The aim of our study was to develop a model that describes population pharmacokinetics of isoniazid in children with tuberculosis.
Methods: Previously published isoniazid concentration-time (530 observations) data from 56 children hospitalized for treatment of tuberculosis were used to describe the population pharmacokinetics of isoniazid . The median age for the children was 3.2 years (interquartile range: 1.6-5.4 years) and the median isoniazid dosage was 5 mg/kg/day (range: 2.9-15.6 mg/kg/day). Twenty-two patients were HIV infected and 18 had kwashiorkor (13 being females). The children were treated with daily doses of rifampicin, isoniazid and pyrazinamide for 2 months before pyrazinamide was stopped. “Pediatric” dispersible fixed dose combination tablets were used. Isoniazid concentrations were measured in plasma samples at 1 month and 4 months after starting treatment and analysed using NONMEM 7. Various models including single and multi compartment models as well as enzyme maturation, linear or nonlinear elimination of isoniazid were tested. Body weight was included through allometric scaling on all clearance and volume parameters. Median body weight of 12.5 kg was used as reference. Acetylator genotype information, known to influence isoniazid PK, was available and included in the model.
Results: The pharmacokinetics of isoniazid was best described by a two-compartment model with first-order elimination. Interindividual variability was supported on oral clearance and volume of central compartment, while interoccasional variability was significant on relative oral bioavailability and rate of absorption. Oral clearances in slow, intermediate and fast acetylators were estimated to 5.7, 9.92 and 15.0 L/h, respectively, in a 12.5 kg child. Using these values of CL, if the children were given 10 mg/kg doses in accordance with the new WHO guidelines, we estimate a median AUC0~∞ of 21.9, 12.6 and 8.3 mg•h/L in slow, intermediate and fast acetylators, respectively. These AUCs are low when compared to the value of 32.5 mg•h/L (interquartile range: 22.5-42.4 mg•h/L) obtained in an ethnically similar but adult population .
Conclusion: Our model adequately described the pharmacokinetics of isoniazid in children taking doses ranging from 2.9 to 15.6 mg/kg/day, and suggests that, even with the new target dose (WHO guidelines - 2010); children, especially the fast acetylators, may be under-dosed compared to adults.
 McIlleron et al., CID 2009; 48:1547-1553
 Report of the meeting on TB medicines for children. World Health Organization, Geneva 2008 (http://www.who.int/selection_medicines/committees/subcommittee/2/en/index.html).
 Schaaf HS, et al. BMC Infect Dis 2007;7:140  McIlleron et al., AAC 2006; 50(4):1170-1177