The Development of an Age-Appropriate Fixed Dose Combination for Tuberculosis Using Physiologically-Based Pharmacokinetic Modeling (PBBM) and Risk Assessment

Abstract

Background/Objectives: The combination of isoniazid (INH) and rifampicin (RIF) is indicated for the treatment maintenance phase of tuberculosis (TB) in adults and children. In Brazil, there is no current reference listed drug for this indication in children. Farmanguinhos has undertaken the development of an age-appropriate dispersible tablet to be taken with water for all age groups from birth to adolescence. The primary objective of this work was to develop and validate a physiologically-based biopharmaceutics model (PBBM) in GastroPlus^TM, to link the product's in vitro performance to the observed pharmacokinetic (PK) data in adults and children. Methods: The PBBM was developed based on measured or predicted physico-chemical and biopharmaceutical properties of INH and RIF. The metabolic clearance was specified mechanistically in the gut and liver for both parent drugs and acetyl-isoniazid. The model incorporated formulation related measurements such as dosage form disintegration and dissolution as inputs and was validated using extensive literature as well as in house clinical data. Results: The model was used to predict the exposure in children across the targeted dosing regimen for each age group using the new age-appropriate formulation. Probabilistic models of efficacy and safety versus exposure, combined with real world data on children, were utilized to assess drug efficacy and safety in the target populations. Conclusions: The model predictions (systemic exposure) along with clinical data from the literature linking systemic exposure to clinical outcomes confirmed that the proposed dispersible pediatric tablet and dosing regimen are anticipated to be as safe and as effective as adult formulations at similar doses.

Keywords: PBBM; efficacy; formulation; modeling; pediatrics; safety; simulation; tuberculosis.

Figure 1

Overview of the modeling strategy.

Figure 2

Structure of INH (left), Ac-INH (middle), and RIF (right).

Figure 3

Prediction of INH AUC (A), RIF AUC (B), INH C_max (C), and RIF C_max and plasma concentrations (D) across all the validation clinical datasets for the adult and pediatric studies.

Figure 4

C_max predicted for populations of pediatric subjects for INH (upper panel) and RIF (lower panel) by age group according to the dosing schedule of Table 1. The horizontal line shows the minimum threshold for efficacy according to Kiser et al. [57] for INH (upper panel) and Pasipanodya et al. [68] for RIF (lower panel).

Figure 5

AUC predictions for pediatric populations of INH SA (A), INH RA (B), INH IA (C), and RIF (D) according to the schedule of Table 1. The horizontal lines show the minimum AUC for efficacy and maximum adult AUC for INH DILI according to Zheng et al. [69]. The horizontal green line for panel (D) shows the threshold for efficacy according to [68].

Figure 6

PK profile prediction for 600 mg RIF administered in the fasted state (A), fed state (B), and following ARAs (C). The PK data are reported by Peloquin et al. [80]. Panel (D) shows the log degradation half-life for RIF in the fasted state, fed state, and following ARA administration.

Figure 7

Evolution of percent abnormal liver markers in children treated prophylactically or with active pulmonary TB as a function of INH dose from Donald [78], compared to predictions resulting from this work using the average Brazil genotype reported in [86] and risk exposure thresholds reported by Zheng et al. [69].