Correlations Between the Hollow Fiber Model of Tuberculosis and Therapeutic Events in Tuberculosis Patients: Learn and Confirm

BACKGROUND: The hollow fiber system model of tuberculosis (HFS-TB) is designed to perform pharmacokinetics/pharmacodynamics (PK/PD) experiments, and hence the design of optimal doses and dose schedules for the treatment of tuberculosis. To determine if this model is useful for deriving PK/PD data relevant to clinical outcomes, we compared its quantitative output to that from clinical trials. METHODS: We performed a PubMed search to identify clinical studies performed with antituberculosis therapy in which PK/PD data and/or parameters were documented or a dose-scheduling study design was employed. The search period was from January 1943 to December 2012. All clinical studies were published prior to HFS-TB experiments. Bias minimization was done according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses. Clinical publications were scored for quality of evidence, with 1 as the highest score (randomized controlled trials or meta-analyses of such studies), and 4 as the lowest score. RESULTS: We identified 17 studies that examined the same parameters as in 8 HFS-TB studies. Fifteen of 17 studies had a quality-of-evidence score of 1. The sterilizing and bactericidal effect rates for isoniazid, rifampin, pyrazinamide, and ethambutol were the same in the HFS-TB as in patients. Time to emergence of resistance for monotherapy was the same as in patients. The PK/PD indices associated with efficacy were the same in HFS-TB as in patients for all drugs examined. CONCLUSIONS: The HFS-TB model is highly accurate at identifying optimal drug exposures, doses, and dosing schedules for use in the clinic.