The increasing use of clinical data is rapidly improving our understanding of many clinical conditions. By studying the natural history of disease progression and identifying objective disease biomarkers, we will move closer to identifying potential new treatments or develop a targeted strategy for the use of existing therapies in patient subgroups in order to obtain maximal benefit.
Our understanding of the value of “big data” in the development of new disease treatments has evolved significantly in the recent past with the development of combined data sets to ultimately improve patient outcomes in clinical practice and new drug development. Two years ago I, in collaboration with a number of colleagues, identified both the need and potential for such an approach to gain a greater understanding of Parkinson’s disease (PD). Other diseases have undergone efforts to standardise and integrate relevant data, which have advanced therapeutic trial designs and enabled model-based drug development and personalised medicine strategies.
A number of data sharing models had already been developed which could provide a model for the sharing of PD data
- The Alzheimer’s Disease Neuroinaging Initiative (ADNI)
- The UK Medical Research Council Dementia Platform (DPUK)
- The European Medical Information Framework (EMIF)
- Sage Bionetworks DREAM Data Challenge
A meeting was organised by Parkinson’s UK in collaboration with the Critical Path Institute to bring together all of the key players in the field . These included academic and clinical researchers, industry representatives, government agencies and regulatory authorities and resulted in a summary publication.
The first aim of the meeting was to identify the key gaps that existed in PD research. These included:
- The need for regulatory approved endpoints, trial designs, and modeling tools
- Identification of early diagnostic tools to maximize the impact of neuromodulatory therapies
- Development of reliable biomarkers to monitor disease progression, particularly to assess agents that may modify the course of the disease
- Understanding disease subtypes to enable the stratification of patients to allow for more efficient clinical trials and the development of a personalized medicine therapeutic strategy.
With these in mind, the meeting focused on the information that would be required to address these questions and in particular how the existing clinical data could be integrated into a combined platform to allow for a precompetitive data-sharing approach. This would allow the key questions to be addressed while reducing duplication and increasing the ultimate effectiveness of clinical research.
There are many different PD clinical datasets in a variety of formats and the challenge is to identify the key common data elements that can be combined in order to address the key questions that had been identified. The key questions to be addressed are:
- Data transferability
- Remote data accessibility
- Privacy and consent issues
- Data remapping to agreed standards
- Data integration
But there are also challenges to data sharing including:
- Different data formats
- The need for reliable longitudinal (rather than single point) data
- Access to datasets
- Data protection and “ownership”, with particular reference to patient approval
- Incentives and recognition for the researchers who have generated the data
- The development of integrated infrastructures that will allow for the ready access of data
- The cost of the maintenance and updating of a common data source
- An understanding of new sources of data such as the use of remote monitoring devices
While these are not insurmountable, all of the potential barriers need to be highlighted and an appropriate strategy put into place. This will ensure that any initiatives will allow for the maximum benefit of the data sharing and that obstacles will be identified, wherever possible, in advance.
But we need to remember that the key stakeholder group in such a project is the patients. Informal discussions that I had with people with PD suggested that the vast majority were happy for their clinical data to be made available under certain conditions – primarily that it would be in an anonymous format and that it would be available for research free of charge. Their view was that, although it may not be of direct benefit to them, it will help to develop new therapies that will impact on future generations of people with PD. In particular, we will get to the stage where the treatments address the condition rather than the symptoms, as at present. But in order to achieve this, we need a much better understanding of the condition and the identification of biomarkers to objectively monitor the progression of the condition. The public view on data accessibility is also highlighted in the AllTrials campaign which seeks openness in the availability of clinical trial data to allow for an objective understanding of data obtained from drug trials. People who have participated in clinical trials would expect no less.
A plan is now being developed to establish a global database for PD. Specifically, it will require:
- Identification of the key current databases
- Agreement on data standards and common data elements
- Establishment of guidelines for data sharing
- Engagement of all stakeholders
Finally, patient datasets and registries are also being considered in the context of the future clinical trial development to maximise the benefit of patient datasets. The European Medicines Agency has established a Cross-Committee working group on patient registries. This will evaluate the potential use of existing and planned patient registries in the design of clinical trials. This will be a key step forward and it will provide yet another benefit for patients from the data that has been collected.