Translational Pharmacokinetic-Pharmacodynamic Modeling and Simulation in the Development of Spectinamides, a Novel Class of Anti-Tuberculosis Agents
Date of Award
Doctor of Philosophy (PhD)
Bernd Meibohm, Ph.D.
Andrea N. Edginton, Ph.D. Richard E. Lee, Ph.D. P. David Rodgers, Pharm.D., Ph.D. Charles R. Yates, Pharm.D., Ph.D.
Dose-Exposure-Response, Intratracheal delivery, PBPK, PK/PD, Spectinamides, Tuberculosis
New chemotherapeutic agents are urgently needed to control the spread of multidrug-resistant (MDR) and extensively drug-resistant (XDR) forms of tuberculosis, which still remains an important public health challenge globally. Recently, spectinamides have emerged as a novel class of anti-tuberculosis agents that overcomethe native drug efflux. Spectinamides bind to the 30S bacterial ribosomal subunit which interferes with ribosomal translocation, and ultimately results in inhibition of protein synthesis. They have potent in vitro activity against drug resistant Mycobacterium tuberculosis (Mtb), and also demonstrated sustained efficacy in (Mtb)-infected mouse models. Pharmacokinetic (PK)/ pharmacodynamic (PD) analyses play a critical role in identifying the optimum dosing regimen for new treatments. In this dissertation, I hypothesized that the application of translational PK/PD modeling and simulation techniques would facilitate rational dosage regimen design for spectinamides.
To characterize the dose-exposure-response of Lee 1810, a dose-fractionation study was performed in BALB/c mice infected with a low dose aerosol of (Mtb). Dosing with different dosing regimens was continued for 4 weeks with two blood samples obtained from each mice in the last week, followed by a washout period after which the mice were sacrificed and the lungs removed for measurement of colony forming units (CFU). Drug concentrations in plasma were analyzed with a validated LC-MS/MS method followed by a population PK analysis which also included as anchor point the data of a PK study in healthy mice with intensive sampling. A model for natural bacterial growth in Mtb infection in untreated mice was built from data on the natural history of Mtb infection in mice obtained from previously performed studies and from the literature. Based on the individual post hoc estimates from the population PK modeling, a sequential PK/PD analysis was performed by linking the PK model with the bacterial growth model via an exposure-dependent bacterial kill function that included a sigmoid Emax model for describing the overall rate of change in lung CFU with different dosing regimens. A two-compartment model with first-order absorption was used to describe the pharmacokinetics of Lee 1810. The average absorption rate constant (Ka), clearance (CL/F), volume of the central compartment (Vc/F), intercompartmental clearance (Q/F), and volume of the peripheral compartment (Vp/F) was estimated to be 2.31 h-1,1.17 L/h/kg, 0.435 L/kg, 0.0191 L/h/kg, and 0.161 L/kg, respectively. The inter-individual variability in CL/F was estimated as 19.9 %. The pharmacokinetics of Lee 1810 was found to be different between healthy and infected mice with the later having 56.5% lower CL/F, 69% lower Vc/F and 69.6% lower Q/F. The two-subpopulation model could successfully describe the natural bacterial growth. The replication rate constant (Krep) of Mtb was calculated as 0.0327 h-1 which is consistent with values reported in the literature. The death rate constant induced by the immune system (Kir) was 0.00303 h-1, cell countof fast growing population at the initiation of the infection (N1,0) was 1.93 Log CFU and maximum number of bacteria (Nmax) was 6.44 Log CFU. The inter-individual variability in Krep and Nmax was estimated as 70.8 % and 54.7%, respectively. The bacterial kill induced by the drug was described using a sigmoid Emax model. The drug effect parameters (EC50), maximum kill rate (Emax) and Hill coefficient (y), were estimated as 239 μg/mL, 11.9 h-1 and 2.40 respectively. A Hill coefficient substantially greater than 1 is a typical characteristic of concentration-dependent killing. The concentration dependent killing characteristic of Lee 1810 supports its intermittent dosing.
Poor permeability of spectinamides across the gut limits its oral use. Additionally, since the lungs are the main site of infection in pulmonary TB, the efficacy of lead spectinamide Lee 1599 was evaluated after intratracheal (IT) administration in a mouse model of Mtb infection. A dose of 200 mg/kg TIW (3 days a week) for 28 days resulted in excellent efficacy with 2.2 Log CFU reduction in the lungs. Based on these observations, a comparative biodistribution study of Lee 1599 was performed after IT and SC administration in mice. Plasma and tissue samples were collected at pre-specified time points. The drug was extracted from plasma and homogenized tissues after protein precipitation and analyzed with an LC-MS/MS assay. The rate and extent of absorption was almost two times higher with IT as compared to SC administration. As expected, the highest exposure of Lee 1599 after IT administration was attained in the lungs, which was 2.5 times higher than in plasma. This is highly desirable as lungs are the main site of infection in pulmonary tuberculosis. Overall, this study supports the pulmonary route as a potential pathway for the treatment of tuberculosis with Lee 1599.
Physiologically-based pharmacokinetic (PBPK) modeling and simulation is a powerful methodology used in support of dose selection for first-in-human studies. The objective was to develop a PBPK model for describing pharmacokinetics of Lee 1599 in rats and mice, and to extrapolate this PK behavior to humans. 10 mg/kg of Lee 1599 was administered intravenously to rats and 200 mg/kg subcutaneously to mice. The PBPK model was developed based on the observed rat plasma concentrations, physicochemical properties of Lee 1599, and in vitro data from its metabolism, protein binding and permeability. The concentration-time profile of Lee 1599 in rats was well described by the optimized PBPK model. The model was prospectively qualified by PBPK scaling from rats to mice and comparing predicted murine concentration-time profiles to observed plasma concentrations. This model was also successful in predicting murine PK with observed PK parameters within two-folds of predicted values. The model predicted, weight normalized human clearance of 0.25 L/h/kg was as expected less than the values in rats (0.666 L/h/kg) and mice (1.25 L/h/kg). The PBPK model predicted, a dose of 7.5 mg/kg and 27.5 mg/kg administered once daily via intravenous administration will be required to attain similar exposure as observed in mice after subcutaneous administration of 50 mg/kg and 200 mg/kg respectively. This model suggests that an efficacious systemic exposure can be achieved with daily doses feasible in humans, and may be useful during drug development for understanding the dose requirements for future human studies.
In conclusion, translational PK/PD approaches have been successfully used for the further development and characterization of spectinamides leads Lee 1599 and Lee 1810. The results from the above studies will be helpful in identifying and optimizing the dosing regimens which can strike a balance between bacterial reduction, adverse effects, and emergence of resistance.
Rathi, Chetan (http://orcid.org/0000-0003-1015-5199), "Translational Pharmacokinetic-Pharmacodynamic Modeling and Simulation in the Development of Spectinamides, a Novel Class of Anti-Tuberculosis Agents" (2017). Theses and Dissertations (ETD). Paper 423. http://dx.doi.org/10.21007/etd.cghs.2017.0428.
Other Pharmacy and Pharmaceutical Sciences Commons, Pharmaceutics and Drug Design Commons
One year embargo expires February 2018.