Tutorial- Tuesday at 1:00 PM EDT
Title: Sample Size Re-estimation in the Context of Dual Endpoints Using a Promising Zone Approach – Illustrated with Two Pharmaceutical Case Studies
Presenters: Parvin Fardipour, Cytel

Abstract:
"Responding to the number of underpowered studies conducted across the industry, the Promising Zone approach (Mehta & Pocock, 2011) was introduced to the statistical audiences. The new design made it simpler to optimize study samples while de-risking a clinical trial. This method also opened up new avenues for clinical trial financing, with study sponsors able to offer investors returns based on an asset’s risk profile, measured during the course of a clinical study. The Promising Zone Design became the most popular sample size re-estimation technique in the pharmaceutical industry.
In this tutorial, I will present key concepts associated with the Promising Zone Design and illustrate their application in two case studies/examples drawn from the pharmaceutical industry.

Case Study 1: The sponsor had previously conducted a randomized, double-blind, placebo-controlled Phase 2 study. To move forward into a confirmatory Phase 3 clinical trial, the Sponsor needed to design a trial that efficiently made use of the prior information obtained from the Phase 2 study.

• Given the scant data available, how can we identify the best sample size for the Phase 3 trial to appropriately detect the effect size?
Learning:
• The attendees will learn how to combine Bayesian Decision-Making with Frequentist Final Analysis in a Phase 3 Oncology Trial where primary endpoint is the overall survival (OS).
• The attendees will learn how to use a surrogate biomarker endpoint (early readout, progression free survival (PFS)) to make a go/no-go decision at interim analysis
• The attendees will learn how the trial is designed to stop for efficacy or futility
• The attendees will learn that even when the trial is not stopped for efficacy, how the sample size adaptation can be considered based on a promising zone design

Key concepts: Conditional Power (CP), Bayesian Posterior Probability, Sample Size Re-estimation, Promising zone, correlation between the biomarker and the final endpoints

Case Study 1 Extended: The client just received good news from the NMPA. The regulatory authorities in China gave them a green light to submit their trial with the secondary endpoint results and if it is accepted, then the study can be approved in China. Now, the sponsor wants to change the design and consider success on both PFS and OS endpoints.

How could the sponsor design a trial to address their new research questions?
• How do we design a trial that claims success of both PFS and OS endpoints?
Learning:
• The attendees will learn how to set up the CP for both for primary and key secondary endpoints either through a Fixed sequence procedure or a Fallback procedure
• The attendees will learn how to use a “weighted combinations Test to combine all stages of trial design at the final analysis (stage1 is before the interim analysis, stage2 is after the interim analysis)

Key concepts: dual endpoints approach, CP for dual endpoints, CHW Combination Test
Case Study 1 Extended to address Trial Efficiencies: Some practical approaches to enhancing the trial efficiencies with be addressed through performing simulations, timing of interim analysis, and logistic requirements for trial monitoring"

Bio:
Parvin Fardipour " has over 30 years of experience in Pharmaceutical and Biopharmaceutical industry and has been working the past 16 years
in the adaptive design space to bring innovative approaches to clinical development and facilitate better and earlier decision making. Designing, implementing, and executing adaptive designs to enable real-time learning, Parvin has extensive experience applying innovative methods across different therapeutic areas in both drugs and devices development, including utilization of surrogate endpoints, population enrichment and indication finder with potential benefits across indications. Parvin also supports clients in regulatory meetings across the clinical development lifecycle.

Parvin started her career in the UK at SmithKline, Sandoz and then Wyeth where she managed the team of statisticians and programmers with focus on European submissions.

In 2001, Parvin was offered a position in the US to lead the inflammation and internal medicine statistics group. In 2006, Parvin joined the Quantitative Science team at Wyeth, focused on innovative design development, and worked on the development of response adaptive designs for several compounds. Parvin joined ICON plc Innovation Center team in 2012 and then Cytel in 2020 to continue bringing innovative approaches to the industry.

During her journey in drug development, Parvin has successfully completed several New Drug Applications (NDAs) in many therapeutic areas including Neuroscience, Pulmonary and Respiratory, Woman’s Health, Cardiovascular, Musculoskeletal, Transplantation, Infectious Disease, Endocrinology, Oncology, Gastroenterology, and Vaccine. Parvin has written several publications and white papers on adaptive clinical trials that are available upon request.

She has a PhD (System’s science) – Thesis (Modelling and Simulations of Human Cardiovascular Systems) - University College London and B.Sc. (Mathematics and Statistics) - 1st class honors, King’s college, London"