Guest Column | May 26, 2015

Adapting to Adaptive Trials

adaptive clinical trials

By Dr. Sy Pretorius, Chief Scientific Officer at PAREXEL

The draft FDA guidance on adaptive designs (2010) defines an adaptive clinical trial as “a study that includes a prospectively planned opportunity for modification of one or more specified aspects of the study design and hypothesis based on analysis of data (usually interim data) from subjects in the study”.1 Comparing the traditional way we run clinical trials to the adaptive approach outlined above is somewhat analogous to comparing the flight of a cargo plane without windows to a passenger plane with windows. In the example of the cargo plane, we need to wait until the end of the study to lock the database before we review what happened during the study (i.e., we need to open the doors of the plane to learn what happened inside the plane during the flight).

An adaptive trial is more like a passenger plane with windows, where we have the opportunity during the course of a trial to look at and evaluate data, as well as make predefined changes or course adjustments. Examples of adaptive trial designs include: designs where we stop the trial or an arm because the treatment or dose is not working; sample size re-estimation where we modify incorrect assumptions on effect size and population variance; and seamless Phase 1/2 or Phase 2/3 trials where we combine studies in order to save time and cost. Adaptive clinical trials should not be confused with adaptive monitoring (i.e., risk-based monitoring) and/or adaptive licensing (i.e., the adaptive regulatory pathway/conditional approval approach proposed by the European Medicines Agency for diseases with a high unmet medical need). In this article, we will focus primarily on the execution of adaptive trials in Phase 2 and to a lesser extent, Phase 3 trials.

There is growing evidence of an increased use of adaptive design in clinical trials. A survey on adaptive trials conducted by Industry Standards Research in February 2015 found that more than 80 percent of the 98 respondents indicated that adaptive trials are gaining momentum in their organizations.2 Additionally, a recent publication by the Tufts Center for the Study of Drug Development (CSDD)3  indicates that roughly 20 percent of clinical trials today are employing adaptive design. Additional publications have noted that up to 35 percent of trials at least examine the potential of using adaptive designs.

These designs are increasing in popularity for several reasons. For one, a number of regulatory agencies – including the FDA in the U.S. and EMA in Europe – are emphasizing the use of adaptive trials in a positive manner. Both the FDA and EMA have vocalized their support in comments regarding the use of adaptive designs where they are appropriate. Secondly, as the pharmaceutical industry faces significant pipeline and financial challenges, it is searching for ways to improve clinical trial efficiency, such as reducing costs and shortening timelines. Adaptive trials provide an attractive and practical offering in this regard. Thirdly, there have been several recent advancements in the operational execution environment and processes to run these trials, including statistical and information technology advances that support these trials. Finally, there is growing interest in personalized medicines and master protocols – disease versus drug-driven study protocols – for which adaptive designs are very well aligned. It is noted with particular interest that promoting the use of adaptive designs in clinical trials has been included in a recent draft bill submitted to the U.S. House of Representatives – 21st Century Cures Act – on how to accelerate the discovery, development, and delivery of treatments.

The Benefits Of Adaptive Designs

Important benefits of adaptive clinical trials include: improving the chances of success in a clinical trial and increasing clinical trial efficiency. To improve probability of success, it is critical to ensure the trial has the right population(s), correct dose(s), and appropriate sample size to measure a drug’s efficacy. . The best drug in the world will fail if it is tested using incorrect criteria. If a compound is not working, adaptive methods enable the trial to end as soon as possible, saving precious resources and shielding patients from exposure to potentially harmful treatments. Secondly, adaptive methods reveal the same information about how a drug is performing using fewer subjects and taking less time. We refer to this as providing more information per dollar invested. In other words, they are not necessarily cheaper or faster than traditional trial designs; however, information gained per dollar invested is typically a lot richer and more valuable in these trials because you have the opportunity to adjust course and make corrections. Some of the largest benefits of adaptive designs accrue across the portfolio of trials for a compound, as opposed to in just one single trial.

There is also an important ethical benefit to adaptive trials. Under the traditional trial paradigm, patients would sometimes be enrolled in a treatment arm that is ineffective. Adaptive trials provide the flexibility to either stop the study or reassign patients to a more effective treatment arm. The biggest advantage offered by adaptive trials is the ability to address one of the top concerns associated with clinical trials today: the high failure rate of Phase 3 clinical trials. Approximately 50 percent of trials in Phase 3 today fail primarily because of incorrect dosage, which has tremendous economic and social implications. Several commonly used adaptive designs seek more detail on the dose-response relationship by using accumulating data to adjust randomization allocations to the “best” doses (i.e., those yielding better results). As a result, sponsors can more confidently select the right dose in Phase 3 and, subsequently, increase the probability of success in Phase 3 trials.

The recent ISR survey2 further reveals that respondents do not perceive adaptive trials to be faster or more cost efficient when compared to traditional trials. What’s more, the majority attributed a higher risk profile to adaptive trials than to traditional trials. So why consider these? The answer lies in understanding differences in the execution environment required for adaptive trials. My personal view is that, as the industry gains more experience in their execution, efficiencies inherent in running this type of trial will become clearer. .

Barriers To Adoption

For all the benefits of the adaptive trial design approach, there still remain barriers to widespread adoption. A roundtable discussion facilitated by the Tufts Center for the Study of Drug Development (CSDD) identified the following challenges3:

  • operational concerns: Examples include delays and disruptions in trial execution, patient participation, and distribution of clinical supplies
  • risk aversion to adopting sophisticated adaptive design approaches due to the belief that more clarity from regulatory agencies is needed
  • lack of adaptive trial design experience
  • questions and concerns about how to monitor data without introducing bias
  • gaps in infrastructure and technology to implement more sophisticated adaptive designs
  • limited capacity of independent data monitoring committees
  • organizational resistance to change

Before executing Phase 2/3 adaptive trials, it is important to recognize that change and resistance management tend to be the biggest hurdles encountered and to plan for and mitigate this proactively. In addition to the complex statistical nature of these trials, the amount of operational effort and technology enablers required to establish an environment where these trials can be executed successfully should not be underestimated. These barriers are imminently solvable in part through proper training of employees, developing and implementing adaptive-specific processes and by adopting integrated technology systems to minimize operational bias and uphold clinical trial integrity and validity.

The following approach and steps can help ensure an organization is adaptive-ready:

  1. Embrace a functional workstream approach to assess current state, and perform a gap analysis and action plans for closing these gaps.
  2. Establish action plans focused on a variety of areas, which include actions for addressing people gaps (i.e., missing skill, experience, training), process gaps, and technology gaps (build or buy).
  3. Manage internal and external stakeholder perceptions and communications.

While adaptive trial designs can have several benefits as highlighted previously, they are not necessarily faster or cheaper than traditional trials, and there can be many barriers to success. As such, it is paramount that pharmaceutical companies choose their clinical development partner carefully. The following checklist offers suggestions to increase chances of adaptive trial success:

  • Assess the partner’s experience and approach to adaptive trials.
  • Determine the level of investment in developing the statistical skills, operational processes, and enabling technologies required to run these trials successfully.
  • Evaluate the level of scalability and global reach that is often needed to run these trials in later phases.

References

1. FDA. Draft Guidance for Indsutry - Adaptive Design Clinical Trials for Drugs and Biologics. The United States Food and Drug Administration, Rockville Maryland; February 2010.

2. ISR. Adaptive Trials: Market Dynamics and Service Provider Benchmarking. www.ISRreports.com: Industry Standards Research (ISR), February 2015.

3. Development TCftSoD. The Adoption and Impact of Adaptive Trial Designs. Boston, MA, USA: TUFTS University, 2013.