Traditional trials apply a frequentist strategy, which rely on the accuracy of pre-defined design assumptions (or inputs) to construct an effective design that yields robust final trial results. From first-patient-in to last- patient-out, trial execution proceeds without change, following a black box approach. While this strategy is common in clinical research, it is not well suited to rare oncology studies, which typically contend with very small patient populations that are often difficult to access and to treat.
In contrast, adaptive trial designs offer greater flexibility with preliminary design assumptions and allow for ongoing assessment of accumulating data from the trial during its execution — an inside-the-box paradigm. This can lead to lower trial failure rates and perhaps better outcomes for study participants.
What Is Adaptive Design?
The FDA defines an adaptive design as “a study that includes a prospectively planned opportunity for modification of one or more specified aspects of the study design and hypotheses based on analysis of data (usually interim data) from subjects in the study.” This includes the potential for trial modification after a study is underway but before unmasking/unblinding. This strategy allows for greater flexibility, letting study designers “hone-in” on approaches that may not be working and adjust them.
Typically, interim data comes from an aggregate analysis that maintains the blinding of treatment groups. An adaptive design does not involve retrospective, ad hoc changes after outcomes are already known and it is not an attempt to salvage a failed trial. With adaptive studies, the trial design plan must be outlined in detail prospectively in the protocol or statistical analysis plan and submitted to regulatory agencies for review and approval.
Adaptive designs in rare cancer research can be seen as an effort to make each and every participant count, maximizing the amount of useful data that can be collected.
Adaptations in rare oncology studies may include:
- Alterations to eligibility criteria
- Sample size adjustments
- Dropping, adding or changing treatment arms for maximum efficiency
- Changes to statistical design parameters
- Changes to study duration
It is important to note that to ensure scientific integrity in blinded studies, the study sponsor must remain blinded. However, an independent unblinded group, such as a separate statistical analysis team or an independent data monitoring committee, may be designated to conduct/review the interim analysis.
Some specific advantages of adaptive designs for rare oncology research include:
- Improved study power
- Possible reduction of required sample size and overall costs
- Greater patient protection, with fewer patients likely to receive ineffective treatments
- Improved ability to identify treatment effects in patient subgroups based on biomarker profiles
- Potential for shorter time to market for new treatments
Although a great tool for rare oncology research, adaptive designs are not without drawbacks. The greater amount of complexity that comes with some adaptive design types may exacerbate logistical difficulties. Adaptive design studies also have an unknown duration, with some trials possibly going longer than expected.
The question as to whether or not an adaptive design approach is best suited for a particular study, is a complex and multifaceted decision not always easy to make. Sponsors are advised to involve a biostatistician experienced with adaptive designs. For a more insights, as well as other strategies for expediting rare cancer research, check out our recent webinar.