Effectiveness and Pharmacokinetic Exposures of First-Line Drugs Used to Treat Drug-Susceptible Tuberculosis in Children: A Systematic Review and Meta-Analysis

Abstract

Abstract Background Optimal doses of first-line drugs for treatment of drug-susceptible tuberculosis in children and young adolescents remain uncertain. We aimed to determine whether children treated using World Health Organization–recommended or higher doses of first-line drugs achieve successful outcomes and sufficient pharmacokinetic (PK) exposures. Methods Titles, abstracts, and full-text articles were screened. We searched PubMed, EMBASE, CENTRAL, and trial registries from 2010 to 2021. We included studies in children aged <18 years being treated for drug-susceptible tuberculosis with rifampicin (RIF), pyrazinamide, isoniazid, and ethambutol. Outcomes were treatment success rates and drug exposures. The protocol for the systematic review was preregistered in PROSPERO (no. CRD42021274222). Results Of 304 studies identified, 46 were eligible for full-text review, and 12 and 18 articles were included for the efficacy and PK analyses, respectively. Of 1830 children included in the efficacy analysis, 82% had favorable outcomes (range, 25%–95%). At World Health Organization–recommended doses, exposures to RIF, pyrazinamide, and ethambutol were lower in children than in adults. Children ≤6 years old have 35% lower areas under the concentration-time curve (AUCs) than older children (mean of 14.4 [95% CI 9.9–18.8] vs 22.0 [13.8–30.1] μg·h/mL) and children with human immunodeficiency virus (HIV) had 35% lower RIF AUCs than HIV-negative children (17.3 [11.4–23.2] vs 26.5 [21.3–31.7] μg·h/mL). Heterogeneity and small sample sizes were major limitations. Conclusions There is large variability in outcomes, with an average of 82% favorable outcomes. Drug exposures are lower in children than in adults. Younger children and/or those with HIV are underexposed to RIF. Standardization of PK pediatric studies and individual patient data analysis with safety assessment are needed to inform optimal dosing.