By utilizing accumulating data to modify the operating characteristics of an active trial in accordance with pre-specified rules, adaptive designs can make clinical trials more flexible, efficient, and informative than fixed-sample designs. Adaptive design approaches can be applied across all phases of clinical development, including early oncology studies. These designs introduce real-time flexibility while a trial is underway, including the capability to select biomarker subgroups that identify patients more likely to respond to treatment, allow dynamic adjustment of dose schedules, adjust the size of the trial, or even combine two separate trial phases into a single seamless trial.
Examples of adaptive designs
Early oncology studies that target tumor mutations rather than tumor types, for example, may be well-suited to an adaptive approach known as a basket study. This design involves studying treatment effects among a group of patients who share the same biomarker. Response to treatment can be evaluated using a two-stage adaptive approach:
- Stage 1: In this stage, a small number of patients are exposed to the drug with the goal of making a preliminary assessment of the viability of the treatment. Viability is defined as whether or not the treatment achieves some defined minimum threshold for tumor response. The trial only proceeds to the second stage if the treatment is deemed viable
- Stage 2: This stage involves all remaining patients and focuses on an objective assessment of the magnitude of the treatment effect
The primary benefits of this two-stage design are that it minimizes patient exposure while also enabling an early decision to terminate if the treatment is found to be futile.
Another useful adaptive design in early oncology trials is a response-adaptive design, where a range of biologically plausible doses below the maximum tolerated dose is evaluated in parallel rather than in sequence. Based on the estimated likelihood of response at each dose, adaptations can be made to drop or add a dose level or make other changes that have been pre-defined in the protocol.
Planning an adaptive design trial
The key to the success of an adaptive design is careful planning. Regulatory agencies require sponsors to specify all pre-planned changes, as well as the intended analysis strategy, in the study protocol. Pre-planned changes might include, but are certainly not limited to:1
- Refinements in sample size
- Dynamic adjustment of dose schedules, or even dropping of treatments or doses
- Changes in treatment arm allocations
- Narrowing down to those patients most likely to benefit from the treatment
- Early stoppage
Sponsors should also consider how to handle logistical challenges that may come with executing any changes in data monitoring and data management processes that might be required by proposed adaptations.
In November 2019, the U.S. Food and Drug Administration finalized its guidance on adaptive clinical trial designs for drugs and biologics, which provides important principles for designing, conducting, and reporting the results from such studies. This final guidance is largely the same as the draft version released in September 2018, though the FDA has clarified its recommendations for Bayesian adaptive designs and provided flexibility around its expectations for the extent of pre-specification required for proposed adaptations. With regard to pre-specification, the agency acknowledges the “monitoring committee recommendations might occasionally deviate from the anticipated algorithm based on the totality of the data.” If sponsors wish to retain the ability to adjust these deviations to their pre-specified plan, they should acknowledge the possibility of deviations, identify factors that might lead to such deviations, and propose analysis methods that are not contingent upon strict adherence to the anticipated algorithm.
The FDA suggests that sponsors also review the draft guidance on complex, innovative trial designs.
Despite the potential benefits that come with adaptive designs, adoption of these approaches is still relatively low. The FDA has indicated it expects to receive about 20 drug and biologic marketing applications that rely on data from adaptive design studies each year.
To learn more about how adaptive designs can be used in phase 1 and phase 2 oncology studies, download our white paper Adaptive Trial Designs in Early Oncology: Minimizing Risk & Accelerating Timelines.
 Pallmann P, et al. Adaptive designs in clinical trials: why use them, and how to run and report them. BMC Med. 2018;16(1):29.
 Chow SC, Chang M, Pong A. Statistical consideration of adaptive methods in clinical development. J Biopharm Stat. 2005;15:575–591.