The Programme

Registration

Opening remarks from the Chair

Keynote introduction

Regulatory considerations for adaptive clinical trials

• Key points to note in the CHMP reflection paper on methodological issues in confirmatory clinical trials planned with an adaptive design (CHMP/EWP/2459/02)
• Experience of adaptive designs in regulatory submissions and proposed development programmes
• What does the future hold for adaptive designs in regulatory submissions?

Setting the scene with key industry case studies

A continually adapting design for acute migraine

• Determining the study design
• Confronting the logistical challenges
• Engaging with patients

Seamless trial designs using a short-term survival endpoint in the learning phase

• Using a short-term survival endpoint in the learning phase for a long term survival endpoint in the confirmatory phase
• Building on regulatory endpoints required for prostate cancer drugs
• Discussing designs with senior management and potentially with the authorities
• Our constant search to develop more efficiently: design and assessment pros and cons compared to more classical approaches

Morning refreshments

Implementing a complex Bayesian adaptive design in drug development: a case study

• The importance of understanding the dose–response for successful drug development
• Efficient characterisation of dose-response
• Planning the study: including regulatory interactions
• Study results and learnings

Real case studies exploring the critical features of data collection strategy: adaptive trial design…as simple as “collect, think and act”?

• Data acquisition options: paper or eDC?
• Data management options: data cleaning in real-time
• Data review: concurrent data visibility, statistical analysis and re-modelling
• Decision-making: collect, think and act
• Trials adaptation: amending your data acquisition solution in real-time
• Realising the potentials of adaptive trial design: achieving your data reporting goals on time and within budget

Planning for an effective adaptive trial

• Selecting the right technology solutions
• Considerations for real-time implementation of design adaptations
• Options for overcoming increased trial supply overages
• Pre- / mid-study estimation of supply requirements

Planning for an adaptive design in an acute setting

• Is an adaptive design appropriate to test an acute treatment?
• Case study: planning an adaptive trial in acute ischaemic stroke

Cautionary thinking in adaptive trials

• Is there too much emphasis on statistical significance relative to clinical importance?
• The limits of adaptive trials: why they are not always appropriate
• Revisiting the benefits or adaptive trials: expense, ethics, enrolment, safety and potential faster regulatory approval

Statistical methodology in multiarmed adaptive trials

• Sources for alpha inflation
• Procedures proposed in the literature
• Designing issues: which parameters influence the statistical performance?
• Software for multi-armed designs

Estimation following treatment selection in confirmatory adaptive designs

• Estimation of treatment effects at the end of a flexible clinical trial where treatments are selected and sample sizes are reassessed at an adaptive interim analysis
• Flexible design where single-step adjusted p-values are used for the multiplicity adjustment and the construction of simultaneous confidence intervals is possible
• What price do we have to pay for the construction of simultaneous confidence intervals?
• Discussing and quantifying statistical properties of the resulting confidence intervals
• Comparing single-step flexible designs to usual step-down flexible designs in a simulation study

Opening remarks from the Chair

Adaptive trial design: true innovation or just the buzzword of today?

• Adaptive trials: from upper management's viewpoint
• Adaptive trials: from the statistician's viewpoint
• Two examples of semi-adaptive trials including some words on the use of biomarkers
• The classical question: how do we perform a long term trial in a hurry?

Scaling up: moving adaptive clinical trials in to the mainstream

• How do we get the company to take adaptive trials seriously?
• What does the company need to do to take adaptive trials seriously?
• Running an adaptive trial was fun but a lot of work. How can we cope with 10+ a year?
• Scaling up the use of adaptive trials across trial selection, design and execution

A complete EDC database as a pre-requisite for adaptive decisions in an ongoing trial

• Which data need to be available in the EDC database for an interims analysis before an adaptive trial decision?
• How to encourage site staff to fill in e-CRFs in a timely manner
• Supervising the completeness of the e-CRF during the active trial

Analyse this! Practical issues arising from a large adaptive group sequential design in oncology

• Adaptive versus traditional group sequential designs
• Case study: an adaptive group sequential design as the pivotal trial in an oncology NDA
• Regulatory concerns with this design and how they were addressed
• Statistical issues with inference after this design and how they were addressed

A critical appraisal

• Using simulation to assess overall performance
• Application 1: sample size modification
• Application 2: switching to a patient subpopulation
• Application 3: treatment selection and testing in a combined phase 2/3 design
• Conclusions and recommendations

Rescuing the classic Bauer-Kohne adaptive design

• Classical analysis is not sound: proving a positive effect despite a negative effect on average
• Are these designs convincing in any situation?
• New alternative method that puts a safeguard in the analysis
• How to perform sample size re-estimations that are as powerful as with Bauer-Kohne method

Estimating the probability of success in phase 3

• What defines a successful phase 3 trial?
• The statistician's role in estimating whether a trial will be successful
• Using Bayesian methods to combine historical data with phase 2 trial data to estimate the probability that a phase 3 trial will be successful
• A case study: methods and outcomes of this approach

Are we really that blind?

• Blinded (namal) data can be used continuously to assess within group variance and absolute treatment effects. How will this affect us?
• Do you need to be unblinded to do sample-size re-estimation?
• Cons of randomly permuted blocks as opposed to other randomisation techniques

Panel discussion: how far have we come?

• Which adaptations are safest to attempt?
• The latest responses and initiatives from the industry and authorities
• Comparing and contrasting the relative positions of the EU and US
• Realistic expectations for 2015