Certara, Inc. · NASDAQ:CERT

Certara — Technical Presentation (September 2013)

37 slides from Certara's technical presentation, September 2013

  1. Slide 1: Powering Pediatrics Studies: A Seamless Approach for Meeting the FDA's Pharmacometrics Criteria

    Powering Pediatrics Studies: A Seamless Approach for Meeting the FDA's Pharmacometrics Criteria Using Phoenix NLME and Trial Simulator Samer Mouksassi, PharmD PhD FCP September 2013

  2. Slide 2: Pediatric Drug Development: Challenges and Opportunities

    Pediatric Drug Development: Challenges and Opportunities Clinical Need: Current Status: • Safe/Effective Dose • Empirical/Off-label • Easy to administer • Linear extrapolation • Easy to compute (mg/kg or mg/m2) Challenges • Data collection (sampling, PD evaluation, recruiting,…) • Ethical issues • U

  3. Slide 3: What leads to Pediatric Drug Development?

    What leads to Pediatric Drug Development ? EMA Pediatric FDA Investigation Plan Written requests (PIP) Justification and Rationale to run a Pediatric Protocol © Copyright 2012 Certara, L.P. All rights reserved. 3

  4. Slide 4: FDA Written Request Sample:

    FDA Written Request Sample: © Copyright 2012 Certara, L.P. All rights reserved. 4

  5. Slide 5: FDA Written Request Sample: Bridging Information

    FDA Written Request Sample: Bridging Information © Copyright 2012 Certara, L.P. All rights reserved. 5

  6. Slide 6: How Many Samples by Patient? Precision?

    How Many Samples by Patient ? Precision ? © Copyright 2012 Certara, L.P. All rights reserved. 6

  7. Slide 7: Pediatric Study Design Endpoint

    Pediatric Study Design Endpoint • Design that has ≥ 80 % Power to meet the precision criteria of having the 95 % CI of the geometric mean within 60-140% © Copyright 2012 Certara, L.P. All rights reserved. 7

  8. Slide 8: Pediatric Study Design Endpoint

    Pediatric Study Design Endpoint 𝑊𝑊 0.75 𝐴𝐴𝐴 𝐻𝐻𝐻𝐻 CL = θ × × 70 𝑇𝑇𝑇𝐻𝐻𝐻𝐻 + 𝐴𝐴𝐴𝐻𝐻𝐻𝐻 Typically we have an estimate and standard error for θ and for the Hill parameter How do we compute Clearance and its precision for 6-months old ? What body weight should be used ? © Copyright…

  9. Slide 9: Items to be Discussed with the Health Authority

    Items to be Discussed with the Health Authority • Doses – Fixed vs mg/kg ? – Weight or age cutoffs ? • Number of subjects The key methodology to answer all of these • Number of samples questions is clinical trial simulation/Refitting • Data Analysis Method – NCA vs. POP PK – Combine prior data with

  10. Slide 10: Framework to Optimize The Protocol Design

    Framework to Optimize The Protocol Design PK/or PK/PD Model Realistic Covariate Model 15 PRED 95% GIRLS PRED 50 PRED 5% + + + + + + + + Optimal Design + + + + + + + + + + + + + + + + 10 + + + + + + Weight (kg) + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + Doses, N of pa

  11. Slide 11: Practical Considerations and Software Applications: Simulating the protocol

    Practical Considerations and Software Applications • Simulating the protocol – User-friendly interface that helps in coding simple and complex trial designs • Missed samples • BLQ • Time deviations – Can read a file to use as input for covariate distributions – Can generate fitting ready datasets Li

  12. Slide 12: Practical Considerations and Software Applications: Fitting the Simulated Data

    Practical Considerations and Software Applications • Fitting the Simulated Data – User-friendly interface to fit NLME models by Replicate using the sorting functionality – Convenient secondary parameter functionality to compute the RSE % at specific weight and age stparm(Cl = tvCl * (WEIGHT/70)^dCld

  13. Slide 13: Practical Considerations and Software Applications

    Practical Considerations and Software Applications • Simulation/Refitting may require several iterations Design does not have enough power  Increase the number of samples  Other software we need to regenerate data template which is more time consuming Design not feasible in the clinic  Reduce the

  14. Slide 14: Framework to Optimize The Protocol Design

    Framework to Optimize The Protocol Design PK/or PK/PD Model Realistic Covariate Model 15 PRED 95% GIRLS PRED 50 PRED 5% + + + + + + + + Optimal Design + + + + + + + + + + + + + + + + 10 + + + + + + Weight (kg) + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + Doses, N of pa

  15. Slide 15: Methodology for Optimizing Pediatric Study Design

    Methodology for Optimizing Pediatric Study Design Adults data and model are available • Design Optimization Step (initial design) – Customize Adult Model to be used for Simulation • In-house data • Literature data • Adult Model allometrically scaled and potentially taking into account maturation…

  16. Slide 16: Practical Considerations and Software Applications

    Practical Considerations and Software Applications – User-friendly interface to code simple and complex models – Fast customization of the models to add allometric scaling and or maturational functions – Fast and efficient model comparisons – (Live Visual Demonstration) © Copyright 2012 Certara, L.P

  17. Slide 17: Methodology for Optimizing Pediatric Study Design

    Methodology for Optimizing Pediatric Study Design • Determine Dose that matches adult exposure (by age groups) Target Window Mouksassi MS, Marier JF, Cyran J, Vinks AA Clinical trial simulations in pediatric patients using realistic covariates: application to teduglutide, a glucagon-like peptide-2 a

  18. Slide 18: Methodology for Optimizing Pediatric Study Design

    Methodology for Optimizing Pediatric Study Design • Design Optimization Step (initial design) • Covariate distribution Model (PMA and Weight) The CDC charts are based on modeled distributions that can be used to simulate age-matched weight. LHS or GAMLSS models are typically used. The raw data is fr

  19. Slide 19: Methodology for Optimizing Pediatric Study Design

    Methodology for Optimizing Pediatric Study Design • Design Optimization Step (initial design) • Covariate distribution Model (PMA and Weight) 15 PRED 95% GIRLS PRED 50 PRED 5% + + + + + + + + + + + + + + + + + + + + + + + + 10 + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +

  20. Slide 20: Flexible Model to Simulate Weight in Various Populations

    Flexible Model to Simulate Weight in Various Populations 120 Normal Growth Population 100 80 Weight (kg) 60 40 20 0 5 10 15 Age (years) 20 © Copyright 2012 Certara, L.P. All rights reserved.

  21. Slide 21: Flexible Model to Simulate Weight in Various Populations

    Flexible Model to Simulate Weight in Various Populations 120 e.g., malnutrition, short-bowel syndrome… 100 80 Weight (kg) 60 40 20 0 5 10 15 Age (years) 21 © Copyright 2012 Certara, L.P. All rights reserved.

  22. Slide 22: Flexible Model to Simulate Weight in Various Populations

    Flexible Model to Simulate Weight in Various Populations 120 e.g., obese pediatrics 100 80 Weight (kg) 60 40 20 0 5 10 15 Age (years) © Copyright 2012 Certara, L.P. All rights reserved. 22

  23. Slide 23: Practical Considerations and Software Applications

    Practical Considerations and Software Applications • Realistic Covariate Distributions – In-house data – Literature data – Customize normal population model to disease specific population – Data available from CDC – Phoenix Connect conveniently allow you to run R/ SAS/ S-PLUS / NONMEM from within th

  24. Slide 24: Methodology for Optimizing Pediatric Study Design

    Methodology for Optimizing Pediatric Study Design • Determine the initial optimal sampling points (D-optimal design) – Assumptions • Perfect trial execution • Design optimized and evaluated at specific values of age and or weight • Theoretical relative standard errors percent • Revaluate the design

  25. Slide 25: Practical Considerations and Software Applications

    Practical Considerations and Software Applications • Initial Protocol Design – Feasible number of subjects by age cohort – Feasible number of samples by subject – Determine the dose that matches the adult exposure • Expert opinion • Simulation PFIM/ WinPOPT/ POPED BSVCL BSVVc – User-friendly interfa

  26. Slide 26: Practical Considerations: Providing Report Outputs

    Practical Considerations: Providing Report Outputs • Summarizing the Outcome Model 2 Model 1 Statistics on the Precision (Theoretical Allometric (Empirical Model) of PK Parameters Scaling Model) CL V CL V Number of runs with 169/200 169/200 177/200 177/200 reported standard errors Min 4.6 6.5 4.6 6.

  27. Slide 27: Practical Considerations and Software Applications

    Practical Considerations and Software Applications • Simulating an actual trial outcome (Future version of TS2 will be under the Phoenix platform for better integration and refreshability) © Copyright 2012 Certara, L.P. All rights reserved. 27

  28. Slide 28: Why Sufficient Precision?

    Why Sufficient Precision ? • Once PK parameters are characterized with sufficient precision within each age group, simulations can be performed to find the best dosing schedule by age group • We do not need to re-run a trial with the alternative dosing regimen that may meet adult exposure (efficacy/

  29. Slide 29: Simulation Based Approved Dose!

    Simulation Based Approved Dose ! Clin Pharmacokinet. 2011 Oct;50(10):627-35. © Copyright 2012 Certara, L.P. All rights reserved. 29

  30. Slide 30: Simulation Based Approved Dose!

    Simulation Based Approved Dose ! Clin Pharmacokinet. 2011 Oct;50(10):627-35. © Copyright 2012 Certara, L.P. All rights reserved. 30

  31. Slide 31: Analyzing Actual Protocol Data

    Analyzing Actual Protocol Data PK Model • Fit all collected data simultaneously • Provide estimate for secondary parameters and their uncertainty (secondary, bootstrap) • Provide estimates and descriptive stats for PK parameters (Cmax, AUC, etc.) Vc • Sometimes we only have data from older age group

  32. Slide 32: Analyzing Actual Protocol Data: Secondary Parameters

    Analyzing Actual Protocol Data: Secondary Parameters © Copyright 2012 Certara, L.P. All rights reserved. 32

  33. Slide 33: Analyzing Actual Protocol Data: Bootstrap

    Analyzing Actual Protocol Data: Bootstrap © Copyright 2012 Certara, L.P. All rights reserved. 33

  34. Slide 34: Bootstrap 95% CI Met the Criteria

    Bootstrap 95 % CI Met the Criteria      © Copyright 2012 Certara, L.P. All rights reserved. 34

  35. Slide 35: Actual FDA response and impact on the Label!!!

    Actual FDA response and impact on the Label !!! © Copyright 2012 Certara, L.P. All rights reserved. 35

  36. Slide 36: The methodology is continuously evolving

    The methodology is continuously evolving • Goals – Faster trials – Reduced sample size – Best dosing regimen in the label • Modeling and Simulation tools are now pivotal in bridging the information between prior knowledge and newly generate data • Pediatric data is hard to obtain and a good planning

  37. Slide 37: Acknowledgments

    Acknowledgments • Clients/Partners • Health Authorities • PCS colleagues and co-authors: • JF Marier & Alexander Vinks • Nastya Kassir, Nathalie Gosselin, Leon Pheng and Thomas Peyret Posters (Page and Acop): • N Kassir, MS Mouksassi, NH Gosselin, JF Marier. A Modeling and Simulations Framework to O

Certara — Technical Presentation (September 2013) on 6ix