What are Adaptive Clinical Trials?
The concept of adaptive designs in clinical trials has been around for over 50 years now. However, the Critical Path Initiative released by the United States Food and Drug Administration (FDA) in 2004 began the gradual push for the incorporation of this approach by pharmaceutical and biotechnology sponsors. Adopting an adaptive design allows researchers to plan for potential modifications to any number of variables within a study based on an analysis of data collected while a trial is still ongoing. These modifications can be made to variables such as treatment dosages, sample size, eligibility criteria, or the comparative treatment arms, as well as many more.
The objective of pursuing an adaptive clinical trial is to identify the safest and most effective doses or treatment arms more quickly. Compared with traditional trial designs, the added flexibility of adaptive clinical trials can enable sponsors to make critical decisions with greater efficiency and speed. Although this type of clinical trial is not yet widely embraced by ClinOps professionals, the FDA maintains the value of adaptive trials for carefully weighing options to optimize clinical outcomes.
Predicting Success with Biostatistics
Whether a clinical trial is designed to be traditional or adaptive, one challenge both have in common is utilizing a method for reasonably assessing the probability of success before sponsors invest significant resources in start-up. Recently, the expected Net Present Value (eNPV) method has become more favored for comparing cost, duration, and likelihood of success across multiple different indications or trials. eNPV calculations draw on values inputted from four central variables, then provide a risk-adjusted numerical output for data-driven clinical decision-making.
eNPV gathers data from activity-based benchmarked timelines, cost data from activity-based benchmarked costs, commercial data from revenue and commercial costs, and risk data using the Probability of Technical Success method. eNPV can quantify how these data points change and affect each other when even the most basic adaptive features are introduced into a trial design. This enables project planners to weigh different scenarios and use a data-based method for predicting the feasibility and success of their study.
The CHAMPION PHOENIX Trial: A Case Study of Data-Driven Decision-Making
In the Cangrelor versus Standard Therapy to Achieve Optimal Management of Platelet Inhibition (CHAMPION) PHOENIX controlled clinical trial, patients undergoing percutaneous coronary intervention (PCI) for coronary insufficiency were either randomized to receive the antiplatelet agent cangrelor intravenously or clopidogrel orally. To mitigate the risk of reducing the study’s statistical power, an adaptive approach was incorporated by planning for a sample size re-estimation at the interim analysis when 70% of the expected 10,900 patients had been enrolled.
If the percentage of decrease in relative risk fell in the “unfavorable zone” (risk lowered by <13.6%) or “favorable zone” (>21.2%), the sample size would not be changed; in the former, the difference would be too low to make re-estimation worthwhile, whereas, in the latter, the difference would already be high enough that no change is needed. However, if this percentage fell in the “promising zone” (≥13.6% to ≤21.2%), the sample size would be increased in a predetermined method because there could be a substantial improvement in risk by re-estimating. In this case, the adaptive method is advantageous because it enabled the researchers to make a decision only after their data was reviewed and shown to be promising. Because the CHAMPION PHOENIX trial’s interim results fell in the favorable zone, the sample size was not increased. The final analysis showed significant differences with cangrelor, which led to its approval in the US and European Union for use in PCI.
Looking to the Future: Will Adaptive Trials Be More Prevalent?
To achieve a future where adaptive clinical trials are widely accepted, their design implementation will need to be disseminated in more depth and researchers must embrace the greater complexity of its statistical approaches. However, data-driven methodologies like eNPV and the analysis used in the CHAMPION PHOENIX trial ultimately rely on the availability of quality data. Advances in data capturing and visualization technology will need to be leveraged alongside developments in adaptive designs to maximize the benefit of data-driven decision-making in the drug discovery process.
Vial CRO’s mission as a next-generation, tech-enabled partner for sponsors is to deliver more affordable and efficient trials. Visit https://vial.com/ to learn how our ClinOps leaders can help innovate and reimagine your clinical trial. Connect with a team member today!
