“An adaptive clinical trial is a clinical trial evaluating a medical procedure or treatment by monitoring patient results (and possibly other steps, such as side effects) on a specified schedule and changing the trial protocol parameters in compliance with those findings.
The adaptation process, as specified in the trial protocol, usually continues during the trial. Changes can include dosage, sample size, trial drug, patient selection criteria, and a combination of “cocktails.”
What are Adaptive Design Clinical Trials?
An adaptive design is defined as a design that allows modifications to the trial and/or statistical procedures of the trial after its initiation without undermining its validity and integrity.
The purpose is to make clinical trials more flexible, efficient and fast. Due to the level of flexibility involved, these trial designs are also termed as “flexible designs.
Adaptive clinical trial designs have been widely used in medical device production, and the lessons gained are now being applied to drug development.
Adaptive designs can make clinical trials more flexible by the use of results that accumulate in the trial to adjust the direction of the trial according to pre-specified guidelines.
These trials are often more effective, comprehensive and ethical than traditional fixed design trials , as they also allow better use of resources such as time and money and may involve fewer people. Adaptive designs can be implemented in all stages of clinical study, from early dose initiation to confirmatory testing.
Adaptive design will reduce the number of patients in a trial, as well as the total number of trials. This could also yield the outcomes of more insightful trials.
The added flexibility provided by adaptive design could also improve acceptability of stakeholders.
Also, these trials are especially useful in clinical safety and efficacy studies which are closely monitored. Adaptive design can also be very useful in early-stage research and exploratory studies as well as in subsequent post-marketing tests.
Using adaptive design in an exploratory setting would allow assessment of a broad range of doses, regimens, populations, and so on, with the ability to discontinue evaluation of suboptimal choices. This approach has the ethical advantage of exposing smaller amount subjects to suboptimal procedures, in addition to increased flexibility.
What are the Types of Adaptive Design Trials?
Based on adaptations employed, commonly considered adaptive design methods in clinical trials include
- an adaptive randomization design,
- a group sequential design,
- a sample size re-estimation design,
- a drop-the-loser design,
- an adaptive dose finding design,
- a biomarker-adaptive design,
- an adaptive treatment-switching design,
- a hypothesis-adaptive design,
- an adaptive seamless phase II/III trial design and
- a multiple adaptive design.
Another way to classify adaptive design clinical trials is by categorizing them according to different rules, usually four by number.
The allocation rule defines how the participants in a trial will be assigned to different arms and involves sensitive randomization and covariate adaptive allocation.
Sampling rule defines how many subjects will be sampled at the next stage and consists of sample size re-estimation design (both blinded and unblinded) and drop-the-loser design.
Stop rule defines when to stop the trial and consists of group sequential design and adaptive treatment-switching design.
Decision rule includes changes not covered by the other three categories and consists of hypothesis-adaptive design and changes to the primary endpoint or statistical method or population design of patients.
Sometimes, a fifth rule is added consisting of multiple adaptations, also comprising adaptive seamless phase II/III trial designs.
Adaptive vs. Non-Adaptive Clinical Trial Design
What is adaptive design, and how can it be used effectively?
In November 2019, FDA released its final guidance on “Adaptive Designs for Drug and Biologic Clinical Trials” to help address this issue. This document provides guidance to sponsors and applicants submitting investigational new drug applications (INDs), new drug applications (NDAs), biologics licensing applications (BLAs), or supplemental applications on the appropriate use of adaptive designs for clinical trials to provide evidence of the effectiveness and safety of a drug or biologic.
The Guidance for 2019 describes an adaptive design as “a design for clinical trials that allows for prospectively designed changes to one or more elements of the design based on the accumulation of subject data in that trial. Non-adaptive trial designs do not include such opportunities for modification.
Adaptive and Non-Adaptive Clinical Trial Examples
In early stage dose escalation studies, a common illustration of the adaptive design is present. Such research often employ prospectively designed interim analyzes by a Pharmacokinetic and Safety Data evaluation committee, which then takes decisions on how to proceed. For these situations, the protocol should describe the committee membership simply and prospectively, and the conditions for stopping dosing, repeating the previous dose or continuing at a higher dose.
For a non-adaptive clinical design trial, critical research parameters are formulated using the assumptions and the best estimates. This includes issues such as population mean and incidence rate, variability, size and location of the dose-response effect, and rates of discontinuation. This will work well if estimates and assumptions are accurate, but if they are not, problems will arise.
Non-adaptive designs for the analysis also contain elements to reduce the risks associated with uncertainty.
For example, if the analysis is intended to determine the dose-response, multiple randomized fixed-size groups might be included in the procedure to ensure a appropriate dose is taken. Such design decisions are based on the basis that suboptimal doses are likely to be used for various types of treatment.
But what if you didn’t have to compromise effectiveness in order to make sure the optimal dose is taken? Unlike conventional study designs, an adaptive design may select Phase 2 doses using a model-based approach.
The model could then be modified when intermediate data became available and used to make improvements to the design of the study to minimize the number of participants receiving suboptimal doses and to concentrate on doses with the potential for efficacy.
Adaptive design can also provide futility requirements that are useful in situations where the drug may not be effective. Insead sof waiting for the study to complete, an adaptive design could use interim data to determine whether further subject enrollment will result in a failed study. Another example of where adaptive design could be helpful is the selection of an acceptable sample size for research.
In a conventional trial, there is a threat of underpowering a study if estimates of variance and treatment effect are excessively enthusiastic; similarly, if these parameter estimates are too conventional, then you can end up with an excessively large study population.
When a study is underpowered then the evidence for the analysis would not be adequately accurate from a statistical point of view to draw valid conclusions. If the study population is too big then substantial amounts of time and money would be needed. The sample size can be adjusted using an adaptive approach based on the accumulation of data from the study in such a way as to avoid such unexpected effects.
In short, the core benefit of adaptive design is the ability to include prospectively developed opportunities for modifying research design elements and hypotheses based on interim data analyses. These improvements in the protocol need to be prepared prospectively and any intermediate analyzes for statistical bias need to be monitored.
Adaptive Design versus Conventional Trials
In conventional or classical clinical trials, one proceeds to accept/reject the null hypothesis in a well-defined population.
No modification in the design of trials or statistical procedures or population of patients is permitted after a trial has started without recording protocol changes and the approval of the Committee on Institutional / Independent Ethics, although pre-specified modifications are permitted in adaptive designs based on the interim review.
In classical trials, not more than two or three study arms are undertaken at a given time, while in adaptive design clinical trials, several treatment arms can be tested at the same time [Table].
|Features||Conventional trial||Adaptive design|
|Treatment arms||Maximum two or three||Many simultaneously|
|Hypothesis||Test the hypothesis under consideration||Fit data into a hypothesis|
|Modifications||Not allowed without protocol amendments||Pre-specified allowed|
|Phases||Phases I–II are well defined||Can be seamless phase 2/3 design|
|Statistical analysis||Use routine frequentists methods||Use complicated Bayesian approach|
|Organization||Much simple||Complicated, requiring simulations|
|Interim analysis||Not a routine||Done routinely and frequently|
|Role of IDMC||More once trial/phase is over||Proactive role throughout the trial|
|Regulatory view||Well endorsed||Still speculative|
Comparison between conventional trial and adaptive design trial
Adaptive Design Clinical Trials for Drugs and Biologics Guidance for Industry
This document offers guidance to sponsors and candidates submitting study new product applications (INDs), new drug applications (NDAs), biologics licensing applications (BLAs), or additional applications on the correct use of adaptive designs for clinical trials to provide proof of drug or biological effectiveness and protection.
Such guidance describe the principles required for the design, conduct and reporting of the outcomes of an appropriate clinical trial. The guideline also advises sponsors on the types of details to send to allow FDA review of adaptive design clinical trials, including adaptive Bayesian and complex trials that rely on computer simulations for their design.”