Ph2 h2a d dose ose sele selection b base ased on on the r he rela elationship b bet etween PK an and Tar arget Occup ccupancy cy CNS Session - EMA/EFPIA Workshop on the importance of Dose finding and dose selection London, December 5 2014 Mona Alameddine and Patricia Sanwald Ducray
Backgrou ground • Small molecule from the CNS portfolio • Phase 1 studies completed in healthy volunteers: – Single ascending dose – Multiple ascending dose – PET study • Uncertainty around the target occupancy (TO) needed at steady state for efficacy in the patient population • Which doses should be tested in Phase 2a to characterize the TO – Efficacy relationship? 2
Clin inica ical P l Pharmac macolo logy prop opos osals • To build a Population Pharmacokinetic model using Phase 1 data • To build a PKPD model for target occupancy using PET data • To simulate Target Occupancy time profiles at steady-state by dose • To compute the % of measurements where Target Occupancy was within given ranges over the dosing interval • To determine the optimal number of doses that should be tested in Phase 2a to characterize the TO – Efficacy relationship 3
PK PK and PK PKPD PD mod odels • A two-compartment Population PK model with first order absorption adequately described the PK profile Dos Dose Gut LAGTIME Ka Cen Central Per eripher heral Q/F /F V2/F /F V3/F /F CL/F CL • An Emax PKPD model adequately described the PK/Target Occupancy relationship 4
Si Simulated T Target Occ Occupancy ti time profiles a at Ste Steady State te Median Median 95% Prediction Interval 95% Prediction Interval Perc ercen entage of of hou ourl rly s simulated ed mea easure remen ents with thin TO ra ranges es TO TO Rang anges 10-30% 10 0% T TO 30-50% 30 0% T TO 50-80% 50 0% T TO Dose 1 1 52 1 0 5 10 48 42 Dose 2 2
Optimal c covera rage ge of Target O Occu cupan ancy cy achieved wi with 3 3 dose oses 6
Conclu clusio ions • Population PK and PKPD models were used to simulate Target Occupancy time profiles at steady state by dose • Doses that cover optimally the target occupancy ranges over the dosing interval were identified • Phase 1 data were leveraged using simple modeling methods that translated into powerful graphical representation to support dose selection • Team was open and valued the recommendation
8
Doing now what patients need next
Recommend
More recommend