A quantitative systems pharmacology model describing the cellular kinetic-pharmacodynamic relationship for a live biotherapeutic product to support microbiome drug development


This poster features a collaboration with Takeda.


  • Live biotherapeutic products (LBPs) are human microbiome therapies showing promise in the clinic for a range of diseases and conditions.
  • Describing the kinetics and behavior of LBPs poses a unique modeling challenge because, unlike traditional therapies, LBPs can expand, contract, and colonize the host digestive tract.
  • We have built a predictive model of the CK/PD relationship for microbiome therapeutics based on published data. The model incorporates published CK/PD data and extrinsic and intrinsic variables that impact the CK/PD relationship.
  • Aim: To establish a quantitative relationship between the cellular kinetics (CKs) and PDs of microbiome therapies to guide trial design and identify dosing regimens that provide the maximal therapeutic benefit.


Model Structure:

  • The model consists of nine species: LBP in the lumen and on the epithelial surface, host bacteria that do not compete with the LBP on the lumen and epithelial surface, host bacteria that compete with the LBP on the lumen and epithelial surface, vancomycin, and butyrate in the lumen and gut.
  • Model reactions include logistic growth of bacteria, killing of bacteria by vancomycin, binding and unbinding of LBP to the epithelial surface, production of butyrate by bacteria, clearance of vancomycin from the lumen, diffusion of butyrate into the gut, and clearance of butyrate.

Model parameterization and analysis:

  • Host bacteria carrying capacities, bacterial growth rates, butyrate production, diffusion, and clearance, and vancomycin clearance and compartment volumes were estimated from literature.
  • Remaining model parameters were calibrated and validated using published data from a phase Ia/b study of VE303 in healthy volunteers (Dsouza et al., 2022).
  • Global sensitivity analysis was performed to identify model parameters that have the greatest influence on bacterial engraftment. Partial rank correlation coefficients and linear regression were used as sensitivity measures.
  • To construct a virtual population, sensitive parameters were varied in a log-normal distribution centered around the nominal value. Standard deviations were estimated to match observed variability in the VE303 relative abundance data for 14 days of dosing with 8e9 CFU/day.


  • Simulations demonstrate a dose-dependent increase in butyrate with increasing doses and dose frequency. Loading doses of LBPs can accelerate engraftment of bacterial species.
  • The modeling and clinical data suggest that LBPs benefit from antibiotic preconditioning. However, antibiotic preconditioning leads to lower initial levels of microbial products like butyrate due to depletion of the host microbiome. A loading dose can accelerate butyrate concentrations returning to and exceeding baseline levels.
  • Early bacterial dynamics can have long-term effects on LBP abundance and butyrate levels, but prolonged or chronic dosing treatment may still be necessary for long-term efficacy


Download the Poster