Background. Infants and toddlers often present with disseminated and lymph node tuberculosis, in which Mycobacterium tuberculosis (Mtb) is predominantly intracellular. Linezolid, used to treat tuberculosis in adults, has not been formally studied in infants. Infants clear linezolid 5 times faster than adults and achieve lower 0- to 24-hour area under the concentration-time curves (AUC0–24).
Methods. To mimic intracellular disease, we infected human-derived THP-1 macrophages with Mtb and inoculated hollow fiber systems. We performed dose-effect and dose-scheduling studies in which we recapitulated the linezolid half-life of 3 hours encountered in infants. Repetitive sampling for linezolid pharmacokinetics, Mtb intracellular burden, viable monocyte count, and RNA sequencing reads were performed up to 28 days.
Results. The linezolid extracellular half-life was 2.64 ± 0.38 hours, whereas intracellular half-life was 8.93 ± 1.30 hours (r2 = 0.89). Linezolid efficacy was linked to the AUC0–24 to minimum inhibitory concentration (MIC) ratio (r2 = 0.98). The exposure associated with maximal Mtb kill was an AUC0–24/MIC of 23.37 ± 1.16. We identified a 414-gene transcript on exposure to toxic linezolid doses. The largest number of genes mapped to ribosomal proteins, a signature hitherto not associated with linezolid toxicity. The second-largest number of differentially expressed genes mapped to mitochondrial enzyme inhibition. Linezolid AUC0–24 best explained the mitochondrial gene inhibition, with 50% inhibition at 94 mg × hour/L (highest r2 = 0.98).
Conclusions. We identified the linezolid AUC0–24/MIC target for optimal efficacy against pediatric intracellular tuberculosis, and an AUC0–24 threshold associated with mitochondrial inhibition. These constitute a therapeutic window to be targeted for optimal linezolid doses in children with tuberculosis.