Phase-1 Interim Analysis

About

In this example, we continue to work with the model and analysis from Immuno-modulation in Chronic Inflammation: Clinical Candidate Selection. We are targeting a membrane bound protein, which is often found to be involved in the pathogenesis of Immune thrombocytopenic purpura and Rheumatoid Arthritis, using a monoclonal antibody. The competitor molecule is still ahead in the clinic, dose administration and frequency are still the major Go/No-go criteria, and thus far we have observed high non-linear PK (IV and SC dosing) and high PK variability with SC dosing.

In this case study, we look at whether the non-linear PK and SC PK variability can be explained. Will it be possible to achieve >90% target inhibition in patients and should this program be discontinued?

Results

  1. The model was updated with clinical data.  The model predicted non-linear PK and target occupancy as well as variability in SC dosing.

     

  2. The model results were used to amend Ph1 protocol, to prepare Medicine and Marketing for counterintuitive Ph1 results and to obtain regulatory approval to change Ph2 trial design.

     

  3. Customer’s best-in-class molecule is now positioned to be first-in-class as well (competitor postponed clinical trials).

 

The Model

The systems pharmacology model was based on first principles as a system of elementary mass-action, mechanistic PK/PD, ordinary differential equations.  The model parameters and reactions include compartment volumes, ligand concentration and turnover rates, cell numbers and turnover rates, drug administration, target-mediated drug disposition on two cell types, and endogenous drug elimination.

Model diagram

Phase-1 interim analysis

Conclusion

  1. The preclinical model was updated with cynomolgus monkey data and parameters were varied to match physiological ranged for healthy volunteers.   The model accurately predicted the non-linear PK and target occupancy.

  2. Model parameters were varied to match patient ranges for two indications. The model predicted that the highest SC dose may not provide the required >90% target inhibition for indication 1, but likely will for all patents in indication 2. A higher dose is required to test proof of clinical concept in both indications.

  3. Ph1 study was amended with specified higher dose in support of Ph2.

 

Download the Full Case Study