Medidata Solutions Worldwide The Interplay Between Adaptive Clinical Trials and E-Clinical Architecture A Coin Toss From Fair to Biased Hugh Levaux VP, Product Strategy About half of all Phase III clinical trials fail to prove the key clinical hypothesis under evaluation; the odds of proving a hypothesis are no better than a fair coin toss. Despite major scientific advances in the development of new compounds, the process of evaluating these compounds is not only becoming cost prohibitive, but is also widely regarded as inefficient. With the cost of bringing a compound to market approaching the $1 billion mark, pharmaceutical, biotechnology, medical-device, and diagnostic companies are looking for multiple ways to improve their R&D throughput, leading up to and through marketing approval by the regulator. Adaptive Clinical Trial Protocols One of the most promising new avenues in achieving breakthrough efficiencies is to design adaptive clinical trial protocols. The briefest definition of an adaptive trial design is one whereby an essential design feature — arm, dosage, or inclusion/exclusion criteria — can be modified during the course of the trial based on the results of data collected and analyzed during the trial. In other words, critical decision making occurs while the trial is under way, rather than at its conclusion. Anecdotal observation suggests that an increasing number of pharmaceutical companies are seriously evaluating adaptive protocol designs. Seamless Integration In many respects, the science and the technology behind adaptive clinical trials are relatively well-known and utilized. The challenge lies more in implementation than in developing a new science. Fundamentally, adaptive trials require the seamless integration of data and operations at the analytical, operational, and transactional levels. Even more challenging, this integration needs to occur multiple times during the trial as the data are reviewed periodically by a third-party data monitoring committee. From an analytical standpoint, given the dynamic nature of adaptive trials, there needs to be a close interaction and a positive feedback loop between study design and simulation, data capture and management, and biostatistics and analysis. As the protocol is written, a large number of trial scenarios must be generated to select the most comprehensive design. To reduce cost and potential for errors, data collection forms and related validation checks should be as automated as possible. Finally, data captured and managed within the central clinical database need to be seamlessly analyzed during interim reviews with suggested modifications integrated back into the protocol and related data capture screens. Connecting Operations Operationally, adaptive trials connect typically disconnected areas of clinical operations. Clinicians, statisticians, data managers, clinical monitors, and logistics experts all need to work jointly on planning and executing the trial. Traditional trials allow the various groups to work in relative isolation, while adaptive trials require constant communication and adjusting of the entire clinical trial process flow. One change in the treatment allocation, for instance, can affect every aspect of the trial and may require midstudy adjustments. Finally, analytical and operational challenges can only be met with a strong central transactional infrastructure, such as the one provided by integrated electronic data capture (EDC) and clinical data management systems (CDMS). This e-clinical infrastructure needs to capture, manage, and memorialize every step of the trial as well as provide information in real time to all trial participants. Needless to say, various participants have different data and informational needs. The reports, outputs, and extracts of the data must be specific to the user requesting the information — blinded or not, subject or site specific, read-only, or action-oriented, etc. Only the connectivity offered by the Internet can offer the platform for such capture, management, and reporting of study data on a global basis, 24/7. From a semantic and interoperability standpoint, the various systems — analytical, operational, and transactional — offer a seamless platform to conduct adaptive trials only when they share and exchange the metadata that define the trial, the data to be collected, and the various statuses that help characterize the state of given data. Integrated metadata models can make the coin toss, at last, a biased one and reduce the proportion of failed trials. CLInical-trial solutions Critical decision making occurs while the trial is under way, rather than at its conclusion. Anecdotal observation suggests that an increasing number of pharmaceutical companies are seriously evaluating adaptive protocol designs. Medidata Solutions Worldwide, New York, helps the world’s leading pharmaceutical, biotechnology, medical-device, and research organizations maximize the value of their clinical research investments. For more information, visit mdsol.com. October 2007 VIEW on E-Solutions

Download PDF