Pietro Laddomada1,2, Salvatore D'Agate1,2, Umberto Villani1,2, Marie Sylvianne Rabodoarivelo3, Rebeca Bailo3, Jordana Galizia3, Santiago Ramón-GarcÃa3,4,5, Oscar Della Pasqua1,2
1Consiglio Nazionale delle Ricerche (CNR), 2Clinical Pharmacology & Therapeutics Group, University College London, 3Department of Microbiology, Pediatrics, Radiology and Public Health, University of Zaragoza, 4Spanish Network for Research on Respiratory Diseases (CIBERES), Carlos III Health Institute, 5Research and Development Agency of Aragón (ARAID) Foundation
Introduction The persistent occurrence of Mycobacterium tuberculosis (Mtb) strains showing low susceptibility or resistance to some or all available antibiotics creates the need for the development of novel combination regimens that exhibit efficacy against drug-resistant bacilli[1,2,3]. To support this process, the establishment of a paradigm that can inform the selection of candidate drugs most likely to be efficacious in the clinic is required. An important factor to consider when selecting combinations is the characterization of the susceptibility of the different bacteria subpopulations to each drug in the combination. Here, we propose a model-based approach that uses in vitro data integrating different experimental conditions with different bacterial markers. Objectives The aim of this work was to characterise the effect of antitubercular drugs in monotherapy and combinations in different experimental conditions considering multiple bacteria subpopulations. Methods A time-kill assay (TKA) was performed to evaluate the in vitro efficacy of bedaquiline (BDQ), pretomanid (PTD) and linezolid (LZD) against bacterial cultures in both exponential and stationary bacterial growth phases, and in standard and cholesterol media. A two-compartment model was developed to describe bacterial growth rate and drug effects. This model integrates colony forming unit (CFU) data (reflecting solid media) and the difference between most probable number (MPN) and CFU (approximating liquid media), allowing for the characterisation of potential differences in drug susceptibility across culture conditions, due to the influence of different experimental setups on the metabolic phenotypes of Mtb. Logistic and Gompertz functions were tested to describe bacterial growth dynamics. Bactericidal activity was characterised using a sigmoidal Emax model, including a delay parameter to account for observed onset dynamics. Data below the limit of quantification (BLQ) was handled using the M3 method. Combination drug effects were assessed by selecting a backbone drug and modelling the impact of companion drugs as a discrete covariate on the EC50 parameter of the backbone drug’s potency curve. In the current analysis, BDQ was chosen as the backbone drug[4]. The impact of adding PTD and LZD on BDQ potency was then estimated. Results A two-compartment model was able to describe the growth rate and then the drug effect on CFU and on the difference between MPN and CFU (MPN-CFU). LZD monotherapy demonstrated no effect within the tested concentration range. A logistic model provided the best fit for bacterial growth dynamics, showing different growth rates for multiple experimental conditions (for stationary phase, 0.206 1/h and 0.223 1/h for CFU and MPN-CFU, respectively). The concentration-effect relationship was characterized by a sigmoidal concentration-effect curve, defined by Emax and EC50, whereas a lag time was required to describe the delayed onset of bactericidal activity of BDQ. Briefly, EC50 and Emax for BDQ were 0.554 µg/mL and 0.861 1/h for CFU; 0.92 µg/mL and 1.41 1/h for MPN-CFU. Analogously, for PTD these estimates were 4.99 µg/mL and 1.87 1/h for CFU; 10.6 µg/mL and 2.48 1/h for MPN-CFU. Significant differences in the EC50 shift of BDQ were observed across different experimental conditions when PTD and LZD were added. The 2-way combination of BDQ and PTD resulted in a quantifiable EC50 shift, indicating increased antibacterial activity. For the 3-way combination (BDQ, PTD, and LZD), the observed increase in bactericidal effect was primarily driven by increasing concentrations of PTD, which led to a greater shift in BDQ potency. In contrast, while LZD altered the potency of BDQ in the 3-way combination, this shift was not concentration-dependent for the range tested. Conclusion This study demonstrates the strength of an integrated model-based approach for the evaluation of bacterial growth dynamics and antibacterial effect of antibiotics on replicating and non-replicating bacteria under different experimental conditions in an in vitro protocol. Estimation of the shift in drug potency provides a pragmatic way to select and rank promising candidate compounds. Future work will consider the expansion of this framework with the addition of novel bacterial markers (e.g., RS ratio and MBLA) and information about potential drug degradation, further enhancing our ability to predict and optimise combination regimens. This work has received support from the Innovative Medicines Initiatives 2 Joint Undertaking (grant No 853989).
[1] World Health Organisation. Global tuberculosis report 2024 (accessible at: https://www.who.int/teams/global-tuberculosis-programme/tb-reports/global-tuberculosis-report-2024). [2] Dartois VA, Rubin EJ. Anti-tuberculosis treatment strategies and drug development: challenges and priorities. Nat Rev Microbiol. 2022; 20(11):685-701. [3] Muliaditan M, Davies GR, Simonsson USH, Gillespie SH, Della Pasqua O. The implications of model-informed drug discovery and development for tuberculosis. Drug Discov Today. 2017;22(3):481-486. [4] Muliaditan M, Della Pasqua O. Evaluation of pharmacokinetic-pharmacodynamic relationships and selection of drug combinations for tuberculosis. Br J Clin Pharmacol. 202; 87(1):140-151.
Reference: PAGE 33 (2025) Abstr 11387 [www.page-meeting.org/?abstract=11387]
Poster: Drug/Disease Modelling - Infection