Priority hierarchy
As pharma companies continued their extremely challenging search for new drugs to treat Alzheimer’s disease (AD), it became increasingly clear within the industry that there was little consensus on how to evaluate the benefit of potential clinical interventions. In other words, there was a need to know what value patients, public health authorities, regulators and bodies responsible for reimbursement (HTAs) would place on drugs that contribute to alleviating different negative outcomes of the disease. In order to get answers, industry partners sought a neutral, non-competitive platform where all parties with a stake in the outcome could collaborate.
Clinical trials are often designed using the clinical definition of the disease as endpoints. However, a clinically significant endpoint does not always translate through to an important impact on patients. ROADMAP wanted to establish which outcomes matter the most to patients, many of whom care about the practical and personal considerations that interfere with their ability to perform daily activities. Beginning the project with no preconceptions about what these priorities might be, the project partners studied the existing literature and carried out surveys, group consultations and interviews with patients, their carers and doctors. They defined a hierarchy of priority outcomes for each group, demonstrating that patients’ priorities are often centred around their interactions in their personal and family and community life, their level of access to information and services, and things like costs and loss of income due to missed work.
RWE gathering and challenges
In parallel, ROADMAP sought out data about AD from around Europe relevant for researchers working on developing disease-modifying AD treatments. The project partners used a number of technical solutions (including the IMI-EMIF data catalogue) to find and integrate data sources; including Alzheimer’s-related RWE derived from patients’ electronic health record data, clinical trial data, and data from AD cohorts. They mapped the critical ethical, legal and social issues that arise from creating a real-world evidence platform that re-uses existing health data pooled from different sources.
They found that the practicalities of using real-world evidence in research is challenging, and produced some valuable lessons for others who attempt such a task in the future. They found that despite the wealth of real-world evidence out there, it is not always in matching formats, nor does it contain all the necessary information. Challenges related to disease modelling include the fact that every dataset is designed for a particular purpose, and they all have the ‘legacy’ of the original purpose tied to it. While cohort datasets are designed to indicate aetiology, electronic health records were generated to make decisions about treatment. When these datasets are brought together, it becomes obvious that asking the data questions is very difficult.
An example of a challenge relating to real-world data is the potential mismatch between data concerning a patient’s own reported quality of health and their carer’s estimate of their quality of health (the carer might think the patient is fine, contrary to the patient's own claims). A project conclusion was that there is a pressing need for new data to generate fit-for-purpose real-world evidence, a consensus on study design and a consensus on standardisation of medical records.
They also concluded that it will become increasingly important to figure out ways to harmonise the collection and collation of data, as well as the use of standardised methods and analytical tools, as new interventions become available. The area in which much of the work in AD is done is at the later stage when the cognitive rate of change is faster. In order to make interventions at the early, preclinical stage when the rate of change is much slower, the solution will be to carry out studies that are either much longer or done on a much bigger scale, neither of which options are currently feasible. A conclusion from the project is that we need more collaboration across Europe in order to achieve the necessary scale.
Disease modelling and simulation
Despite these challenges, the consortium sought out and studied over 40 existing diseases models for mild cognitive impairment and dementia in order to come up with a new core disease model, tested against the existing datasets. The disease model, which describes the time course of disease status and tracks disease severity over time, allowed them to analyse disease trajectories and the effects of different interventions.
They produced a sophisticated and interactive data visualisation tool called the Data Cube that synthesises the data sources and different outcomes from around Europe, offering a landscape view. Users can switch between the perspectives of people with dementia, carers and health professionals, and the response is immediate. By helping researchers access more relevant information, the scientific community will be better equipped to identify and develop drugs that might be able to modify outcomes in ways that matter to the right people. It will help other stakeholders make decisions about reimbursement, the use of new technologies, as well as answer questions about better data capture.
Health economic impact models and regulatory and HTA engagement
ROADMAP carried out a review of the literature on health-related quality of life data of people with pre-dementia or dementia, and the resources used and costs incurred. With input from representatives of HTA bodies and external experts, they were able to define the specifications for an economic model for evaluating future interventions for AD.
An expert group made up of people with regulatory and HTA backgrounds provided non-binding recommendations and other guidance on the principles of the use of RWE in AD in the regulatory and HTA context. In particular, they emphasised the need for defined, validated and widely-accepted outcomes that capture the early stages of AD, and tools with which to measure these outcomes. They also insisted on the importance of capturing the outcomes for carers.
Going forward, the project concluded that seeing as therapies that target the earlier phases of the disease will leave many questions unanswered about the long-term impacts on patients, caregivers and healthcare systems, there is a clear need to establish an international consensus on the disease outcomes that will help regulatory and HTA decision makers. The challenge will be to ensure that data sources across Europe are able to generate the evidence needed to support this. Filling certain gaps in the ROADMAP Data Cube would provide a step change in preparing Europe’s healthcare systems for the future disease-modifying drugs that are so urgently needed.
The project is part of IMI’s Big Data for Better Outcomes programme, which aims to facilitate the use of diverse data sources to deliver results that reflect health outcomes of treatments that are meaningful for patients, clinicians, regulators, researchers, healthcare decision-makers, and others.