By Melanie Buitendyk, Director, Biostatistical Consulting
A Sponsor has a promising therapy and is in the process of planning their clinical development strategy. Key components of this planning include regulatory and corporate objectives, go-no-go decisions, and on a study level, includes study design, study objectives, and endpoints. Well-planned, well-designed, and well-written protocols are essential in accelerating the time it takes to get potential new therapies to patients.
When regulators provide comments and/or review issues on submitted protocols, the clinical and statistical reviewers typically comment on the choice of primary and secondary endpoint, robustness of the statistical analysis (e.g., alpha control, intercurrent events, and missing data), potential sources of bias, and choice of the control. Involving a statistician at the time of planning, designing, and writing of the protocol will mitigate these review issues, reduce the need for major protocol revisions and ultimately reduce the time to the first patient screened.
Statisticians and the Study Design Team
At the time of planning, it is typical to include medical, regulatory, biostatistics, clinical pharmacology, and clinical operations as the primary functions for discussion. It takes a team to achieve success and is best approached when multiple functions with varied perspectives are involved. A statistician is an integral part of the core team that is needed to efficiently design a protocol that meets corporate and regulatory objectives and is successfully moved to execution. A statistician will provide invaluable insight into the following important components required for success.
- Support defining study objectives to align with Sponsor priorities and goals
- Determine and define the appropriate endpoints to address study objectives
- Help determine whether the study will assess superiority, non-inferiority or equivalence
- Provide guidance towards estimand framework including consideration of potential intercurrent events and strategies to handle them
- Identify the primary statistical methods to manage missing data
- Identify sensitivity analyses to explore the impact of missing data on the robustness of the study conclusions
- Define data to be collected for each endpoint and frequency of assessments
- Identify circumstances where interim analyses and study adaptations are appropriate and how to align them with clinical development needs
- Reduce bias and variability in collection of endpoints
- Randomization strategies including stratification
- Consideration of potential data to be collected in current study to plan for future trials
In addition, the statistician is essential in ensuring the sample size properly aligns with the study phase and study objectives. Final decisions about several principal study design features depend on the sample size calculations and need to be carefully considered by the full study team. These include alpha, power, primary and key secondary endpoint selection and definitions, number and type of interim analyses, and study adaptations. The statistician will consider this multitude of factors when assessing sample sizes for various scenarios.
Key Methodology Considerations
When it comes to statistical methodology and choosing the appropriate analysis method for each endpoint, the statistician’s guidance is crucial. Robust analysis plans, appropriate for each study phase, need to be included in the study protocol. Methodology, assessment of statistical assumptions, baseline factors and covariates, and testing of each hypothesis is another way the statistician will contribute to protocol development.
Furthermore, to demonstrate to the agency that specific data issues have been well thought through and correctly mitigated, the protocol is expected to address the following:
- Identifying intercurrent events and the strategies for handling them
- Planning for missing data and sensitivity analyses to test the robustness of results
The statistician will discuss these key items with the core development team, come to an agreement, and provide detailed descriptions of the analyses/strategies in the protocol.
An Example: Superiority or Non-inferiority, Choosing the Primary Endpoint in a Phase III Study
Sponsor X is designing a Phase III study within a rare disease indication where a current therapy has been approved. One might immediately consider a non-inferiority approach. However, the size of the study required may be too large for the population. The consulting statistician, who has been a long-term member of the clinical development team, met with the clinical program lead to discuss options.
Multiple scenarios of study endpoints, including combination endpoints, were discussed. Non-inferiority and superiority were explored, and sample sizes for each scenario were estimated. Two potential paths were identified as plausible, one as a superiority study using a new endpoint in which ‘New Drug’ is hypothesized to be superior to current therapy and one as a non-inferiority study using the endpoint in which the current therapy was approved, hypothesized that ‘New Drug’ would provide similar efficacy as current therapy.
The statistician acted as a key member in supporting Sponsor X with endpoint consideration, sample size estimation, and study design, ensuring that the outcome of the study was well thought through, robust, and plausible for the rare disease indication.
Engaging the Right Team and the Right Time
The goal of all Sponsors is to develop a protocol expeditiously that gains approval by regulatory authorities so therapies can be provided to patients faster. To aid in this process, having a statistician as part of the core team during all stages of clinical development is of the essence. Our team of expert statistical consultants here at Allucent has long experience and in-depth expertise in a wide array of indications and study designs. We stand ready to support you.
Contact us today to learn more about our expert statisticians.
Share This
