Assessing End of Phase 2 Decision Criteria

Abstract

The objectives of the Phase 2 stage in a drug development program are often to evaluate the safety and tolerability of different doses, select a promising dose range, and look for early signs of activity (i.e., to establish proof of concept). At the end of Phase 2 (EOP2), a decision to initiate Phase 3 studies is made that involves the commitment of considerable resources. One of the key factors in this decision is the expected efficacy and the associated predicted probability of success (PoS) in Phase 3. Making a decision based upon the PoS requires decision makers to select a benchmark (or PoS criterion) for the PoS, which if achieved would enable a “go-to-Phase 3” decision. However, appropriately choosing the criterion requires knowledge of the operating characteristics associated with the decision criteria. A key operating characteristic that a funder/sponsor requires to make such a choice is to understand the predicted conditional probability of making a go decision and subsequently failing in Phase 3. In this article, we show how such risks can be informed through the use of clinical trial simulation. We also highlight through a worked example in pancreatic cancer how this simulation exercise can help to decide between different development strategies for a specific indication. Supplementary materials for this article are available online.

Key Words:

• Clinical trial simulation
• Decision criteria
• End of Phase 2
• Pancreatic cancer
• Probability of success.

Acknowledgment

The authors thank Amgen for the opportunity to spend time researching this methodology.

Additional information

Notes on contributors

Tony Sabin

Tony Sabin is Senior Biometrics Team Leader, AstraZeneca Ltd., Melbourn Science Park, Royston, UK; MRC Clinical Trials Unit at University College London, London, UK (E-mail: [email protected]). James Matcham is Head of Early Clinical Development Biometrics, AstraZeneca Ltd., Melbourn Science Park, Royston, UK (E-mail: [email protected]). Andrew Copas is Reader in Statistics (E-mail: [email protected]), and Mahesh K. B. Parmar is Professor (E-mail: [email protected]), MRC Clinical Trials Unit at University College London, London, UK.

James Matcham

Tony Sabin is Senior Biometrics Team Leader, AstraZeneca Ltd., Melbourn Science Park, Royston, UK; MRC Clinical Trials Unit at University College London, London, UK (E-mail: [email protected]). James Matcham is Head of Early Clinical Development Biometrics, AstraZeneca Ltd., Melbourn Science Park, Royston, UK (E-mail: [email protected]). Andrew Copas is Reader in Statistics (E-mail: [email protected]), and Mahesh K. B. Parmar is Professor (E-mail: [email protected]), MRC Clinical Trials Unit at University College London, London, UK.

Andrew Copas

Tony Sabin is Senior Biometrics Team Leader, AstraZeneca Ltd., Melbourn Science Park, Royston, UK; MRC Clinical Trials Unit at University College London, London, UK (E-mail: [email protected]). James Matcham is Head of Early Clinical Development Biometrics, AstraZeneca Ltd., Melbourn Science Park, Royston, UK (E-mail: [email protected]). Andrew Copas is Reader in Statistics (E-mail: [email protected]), and Mahesh K. B. Parmar is Professor (E-mail: [email protected]), MRC Clinical Trials Unit at University College London, London, UK.

Mahesh K. B. Parmar

Tony Sabin is Senior Biometrics Team Leader, AstraZeneca Ltd., Melbourn Science Park, Royston, UK; MRC Clinical Trials Unit at University College London, London, UK (E-mail: [email protected]). James Matcham is Head of Early Clinical Development Biometrics, AstraZeneca Ltd., Melbourn Science Park, Royston, UK (E-mail: [email protected]). Andrew Copas is Reader in Statistics (E-mail: [email protected]), and Mahesh K. B. Parmar is Professor (E-mail: [email protected]), MRC Clinical Trials Unit at University College London, London, UK.