The Clinical Trial and Imaging Subgroup of the European Medicines Agency (EMA) has released a report on the issues associated with the combination of data from different clinical trials into large datasets.
Data from clinical trials are the key influencers of most regulatory decisions, especially for the evaluation of marketing authorisations and variations. This report illustrates different data sharing and standardisation activities, and discusses the opportunities and challenges of combining data from several clinical trials to inform regulatory decisions.
The status of the implementation of big data related to clinical trials is summarised below:
- Clinical trial data sharing activities are currently relatively mature and they are already providing access to many thousands of clinical trials. However, it is not fully clear how these activities could be applied in regulatory or scientific procedures.
- Data standardisation activities are critical to ensure usability and applicability of data.
- Anonymisation in data sharing activities is a difficult balance between data utility and data privacy.
- Imaging has promising potential in a big data context, but specific expertise in this area is lacking within the national competent authorities (NCAs), and the use of these data in the regulatory context is rare.
- Data from single clinical trials including temporal high-frequency or high-dimensional data are not currently used efficiently for regulatory decision-making. Currently there are no guidelines or clear principles, for example, defining new type of outcomes based on the output from the devices (or imaging) generating high-frequency data.
However, there are many steps required for harnessing the potential of big data from clinical trials. Firstly, data from existing clinical studies need to be characterised to understand how the data can be combined with data from other studies. Secondly, data from all the different clinical studies need to be combined in a database. Finally, regulatory applications have to be explored.
The report recommends the following actions which could improve the points mentioned above:
- Combining clinical trial data offers unexplored and useful opportunities for improving decision making throughout the product life cycle, from drug discovery to safety surveillance. Regulators should encourage using approaches where individual patient data from multiple trials are used.
- Individual patient data (IPD) from clinical trials should be requested as part of the submission and assessed as part of the review of the marketing authorisation application. This would require establishing the legal basis for requesting IPD and agreeing on data format for submissions. Furthermore, NCAs may need to increase statistical capacity and skills for analysis of IPD.
- Data harmonisation, both in terms of format and variable definitions, is a key requirement to facilitate the combination of clinical trial data. Furthermore, in the era of multiregional clinical trials standards should be as globally aligned as possible. In order to harmonise terminology, the use of international standards such as the International Classification of Diseases (ICD) and the Medical Dictionary for Regulatory Activities (MedDRA) should be encouraged.
- Actions to create open data platforms for proactive sharing of clinical data should be supported. In particular, efforts targeting diseases with scarce data, such as “ultra-rare diseases” or high-risk populations should be promoted.
- Encouraging more extensive or exhaustive use of images, genomic data, data from wearables collected during a single clinical trial combined with big data analytics could benefit drug discovery and may further inform regulatory decisions. Guidance should be developed to efficiently use these data in the regulatory processes.
“In order to promote incentives to use big data from clinical trials, it is essential that support is given at European level to encourage applications of data sharing in the regulatory setting.”
To view the full report, click here.