EMA EFPIA workshop Break-out session no.3 Case Study Title: M&S support to the bridging of a drug with Ethnic PK-differences Disclaimer The view and opinions expressed in these slides are my own and do not necessarily represent the views of AstraZeneca
Background, Scope • First in class for treatment of neuropathic pain • How can a clinical development plan (CDP), aiming to bridge from Western to Japanese, be designed - Assumptions, sensitivity to assumptions - Results from Phase I modeling - Modeling to support program design • Considerations 2 Matts Kågedal 29-30 November 2011
Assumption Framework ADME Disease (Progression) / Safety ADME Disease (Progression) / Safety Bridge to Bridge to Bridge to Bridge to Bridge to Bridge to new new new new new new population population population population population population 3 Matts Kågedal 29-30 November 2011
Assumption Framework ADME ADME Sensitivity to ethnic factors: PK: Elimination via Metabolism Bridge to Bridge to (Enzyme uncertain). new new population population Do PK in Japan early 4 Matts Kågedal 29-30 November 2011
Assumption Framework Disease (Progression) / Safety Disease (Progression) / Safety Sensitivity to ethnic factors: (Intrinsic and extrinsic) Biomarkers: - No biomarkers available. Efficacy/safety: Bridge to Bridge to Bridge to Bridge to - First in class! new new new new population population - Therapeutic index likely to be population population narrow Dose response data for Efficacy and safety needed in Japan to bridge results from the West. 5 Matts Kågedal 29-30 November 2011
Japanese vs Caucasian AUC based on last dose data in MAD and J-MAD Observed and model predicted AUC/dose • Higher AUC in Japanese as for Japanese and Caucasian subjects. compared to Caucasian. • Not explained by body weight • No ethnic difference in protein binding. - Additional studies may be needed to understand mechanism behind PK-difference to improve Japanese model predictions. prediction Caucasian model prediction Matts Kågedal 29-30 November 2011
Cmax vs probability of AE Day 1 Day 6 Day 12 0.9 0.9 0.9 Proportion with AE 0.8 0.8 0.8 0.7 0.7 0.7 0.6 0.6 0.6 0.5 0.5 S imulated prop 0.5 S imulated prop Prob for typical indiv Prob for typical indiv 0.4 0.4 0.4 Obs Prop J ap Obs Prop J ap Obs Prop C au Obs Prop C au 0.3 0.3 0.3 0.2 0.2 0.2 0.1 0.1 0.1 0 0 0 0 500 1000 1500 2000 2500 3000 3500 0 500 1000 1500 2000 2500 3000 3500 0 500 1000 1500 2000 2500 3000 3500 Cmax Probability of typical Japanese West Model predicted individual proportion Model is based on logistic regression with tolerance development. Figure showing proportion of subjects with AE in each dose-grope versus the mean Cmax in that group and the model predicted proportion based on simulation. - Similar PK-safety relationship for West and Japan, but limited data (MAD in western and Japanese subjects) - Tolerance development to side effect indicated. 7 Matts Kågedal 29-30 November 2011
Phase II and III program options (How M&S is used to inform Drug Development) Development options for Japanese population Assumptions Assumptions 8 Matts Kågedal 29-30 November 2011
Empirical support for assumptions in transition between phases Program 2 -S mall strata of Japanese in global trials Japan + Japan+West Japan+West Japan + Regulatory West West Regulatory Phase III West West Phase III Phase IIa File Phase IIa File Phase IIb One dose Phase IIb One dose Data to support assumptions in the transition between phases Important IIa -> IIb IIb -> III III->File Assumptions CL JAP <CL W SAD, MAD SAD, MAD, IIb SAD, MAD, IIb, III Similar SAD+MAD SAD, MAD, IIb SAD, MAD, IIb, III PK-safety (limited data) Similar None IIb IIb, III PK-Efficacy 9 Matts Kågedal 29-30 November 2011
Model based assessment of program Evaluation conditions (what if): 1. Similar exposure response in Caucasian and Japanese 2. Higher potency in Japanese ( e.g. half exposure -> same effect) Model: 1. PK based on SAD + MAD 2. Placebo model built based on data from previous trials (in- house and literature). 3. Assumed exposure response model based on literature, preclinical data and target product profile Decision criteria for phase III: 1. Is there an ethnic difference in exposure response Criteria to judge adequacy of program: 1. Precision and bias in estimation of optimal dose for Japanese and Caucasians 2. Power to detect an ethnic difference in exposure response for Efficacy and Safety 10 Matts Kågedal 29-30 November 2011
Simulation of program 2 assuming higher potency in Japanese Example with 1000 simulated programs * Ethnic N=35 Ethnic diff? N=285 Same difference? Same No dose dose Simul Ph III Simul Ph III Same Estim. Same Estim. exposure (IIb+III) exposure (IIb+III) No N=250 J(20%)+W J(20%)+W Ethnic N=1000 Yes Different Different difference dose dose detected in Simul. Estim. Simul. Estim. 95% of Ph IIb Ph IIb Ph IIb Ph IIb simulated Ethnic N=15 N=715 programs diff? Same Same Simul Ph III Simul Ph III Yes dose No dose Larger Jap Larger Jap Estim. group Estim. group (IIb+III) Dose can (IIb+III) Dose can N=700 * Made up example for differ differ illustration. No simulations Different Different Yes J(40%)+W were actually performed. J(40%)+W dose dose 11 Matts Kågedal 29-30 November 2011
Precision of estimated dose - program 2 (example for illustration) Two fold difference in potency: No difference in potency: Result: Power=95% 5% of studies with detected difference True Japan Est West Density Density Optimal dose Optimal dose Repeat for program 1 and 3 and compare outcome, cost and time • Different endpoints can be evaluated. E.g. precision of minimum • effective dose. 12 Matts Kågedal 29-30 November 2011
Important considerations Important considerations - The required precision in the ethnic comparison needs consideration - All ethnic groups are important. - Modeling can allow estimation of the individual components of intrinsic/extrinsic factors, using the data more effectively - Not all groups will have empirical support. Models built on biological principles can improve predictions. (Eg data on Koreans in Korea and the west can support prediction of response in Japanese in the West) - Acceptance to base dose selection on target exposure is needed (Phase III dose may not have been studied in one or both populations) 13 Matts Kågedal 29-30 November 2011
Backups 14 Matts Kågedal 29-30 November 2011
Precision of estimated dose CDP 3 (example for illustration) Two fold difference in potency: No difference in potency: True Est Japan West Density Density Optimal dose Optimal dose Program is more costly and require more patients. Better precision if there is a 2-fold difference and worse if there is no difference. (Grey curves= precision based on program 2.) 15 Matts Kågedal 29-30 November 2011
Similarity West-Japan Evidence build PK, Efficac y, Safety Uncertainty Development goal mitigate Severity/importance of consequence 16 Matts Kågedal 29-30 November 2011
Recommend
More recommend
