PI-092 - THE SUPER APPROACH AS AN ALTERNATIVE FOR ESTIMATING THE PKPD PARAMETERS OF DRUGS IN COMBINATION THERAPIES

N. de Castro Suarez¹, J. Ernest², E. Nuermberger³, R. Savic^4,5; ¹UCSF, UCSF - San Francisco, CA, USA, ²UCSF, San Francisco, California, USA, ³Johns Hopkins University, Johns Hopkins University, USA, ⁴University of California, San Francisco, San Francisco, CA, United States, ⁵UCSF Center for Tuberculosis, University of California, San Francisco, San Francisco, CA, United States.

Background: Phase IIa pulmonary tuberculosis (TB) trials typically assess early bactericidal activity (EBA) of monotherapy and dose-response within the first 14 days; however, few studies evaluate drug combinations, as optimal doses for monotherapy may not apply to combinations (1). Our lab has developed a pharmacokinetic-pharmacodynamic (PK-PD) translational platform with bacterial dynamics that predicts Phase IIa outcomes from preclinical mouse monotherapy data (2) and is now being extended to predict the EBA of TB drug combinations. The study aimed to validate the translational platform for predicting Phase IIa outcomes of various two-drug TB combinations.

Methods: PK data from mice treated with bedaquiline (B), pretomanid (Pa), pyrazinamide (Z), or linezolid (L) as monotherapy and sputum CFU counts for BPa, BZ, and PaZ in Phase IIa trial were collected from literature (2,3). BALB/c mouse lung colony-forming unit (CFU) counts after treatment were obtained for both monotherapy and five two-drug combinations (BL, PaL, BPa, PaZ, BZ) as efficacy data. To assess PD drug interactions, the empirical SUPER approach (5) evaluated overall efficacy by focusing on the shifted exposure-response of the SUPER drug in combination with the companion drug.The exposure-response parameters of each SUPER drug in a combination regimen were modeled using an Emax model. Translational prediction of clinical EBA was performed using mouse PK-PD relationships, informed by clinical PK models and species-specific protein binding.

Results: The model that best fit the data for B, Pa, L, and Z as the SUPER drug in two-drug regimens fixed EC50 while estimating Emax. Emax increased for L, Pa, and Z in their combinations, while the efficacy of B decreased when combined with L in BL and with Pa in BPa. The dose of the companion drug influenced the PK-PD parameters of the SUPER drug, with higher doses associated with increased Emax. The model reasonably predicted the clinical EBA for the BPa and PaZ regimens based on sputum CFU counts.

Conclusion: Our translational platform was expanded to include drug combinations, enabling the prediction of clinical EBA outcomes in Phase IIa trials. This empirical methodology demonstrates that EBA in TB patients can be predicted from preclinical experiments and offers a framework for evaluating the impact of companion drugs in combination therapies."