Meeting on Personalised Medicine

I was very happy to participate in the REA Cluster Event on Personalised Medicine at the Instituto de Salud Carlos III in Madrid on the 5th-6th November, on behalf of the AffecTech project. The meeting was organised by Laurence Marrama-Rakatoarivony. As well as hearing about the variety of innovative work being carried out across a number of Marie Sk?odowska-Curie Actions (MSCA) Innovative Training Networks (ITN’s), a series of roundtable discussions between policy makers and researchers explored key topics relating to the future of personalised medicine.

As rapporteur for the discussion on Innovation and Regulatory structures, I provide here a short overview of a very lively and interesting discussion. As my own research is focussed on digital health interventions, certain aspects of the discussion may have passed me by – I record here what I took from it!

Influencing policy

The first point made was to recognise the importance of researchers and innovators engaging with policy makers. While the importance of influencing Science and Research policy is now more widely recognised, the importance of considering regulatory structures governing both research and implementation is perhaps something that fewer researchers have considered, and the avenues for doing so may not be obvious. The point was also made that it is vital for younger researchers to learn how policy is made, and how they can contribute to the process. A related point, and a theme which came up in several ways, was the importance and value of communications and public engagement in supporting attempts to influence policy.

Patient data, consent and access

Patient consent for use of their data is the foundation of Personalised Medicine, and so the community of patients, healthcare professionals, researchers and policy makers need to shape practical, implementable solutions. While anticipating future use in consent can help avoid repetition of many of the problems currently being experienced, the question of how people can be encouraged to share for a public good needs to be addressed. It was observed that it is easier to get consent from people with acute need, but predictive approaches for prevention and early intervention require much broader participation. It was noted that some people are more eager to share, but the current situation is one in which many EU citizens don?t have any practical access to their own Electronic Health Records. Why should someone share data they cannot see themselves, and how can they make informed decisions about the sharing of such data?

Obtaining access to existing data

All efforts and successes in Personalised Medicine are contingent on access to high quality data. Researchers getting to grips with this emerging field will often wish to access data that has already been gathered. To do so, they need to learn how to identify and access high quality data sources, to understand the roles and gatekeepers involved (e.g. in health care organisations), and learn how to overcome barriers.

An uncertain landscape

At this point in time, many researchers and innovators are very unsure of what they can do within current regulatory structures. When people encounter barriers, it is often unclear how much is due to the rules (e.g. GDPR regulators), and how much is due to organisations themselves, or over-conservative interpretation of rules. While regulations are inspired by important and real privacy concerns, examples were given of consented historical samples for rare diseases being unused due to legal concerns, and opportunities to help patients with these rare conditions being lost as a result.

Legal interpretations

For some research directions, such as linking to health service delivery data, these concerns are heightened, limiting research that could have direct real-world benefit. It was also noted that the laws themselves are a source of uncertainty until extensively tested, and that uneven implementation across Europe has the unintended consequence of providing advantage to researchers in some countries in terms of innovation.

Organisational capacity

For many participants, the root of their own problems was that the Institutions and data holders themselves don?t know what is allowed. In particular, it was felt that staff making decisions on access to data often don?t want to take responsibility due to the potentially severe legal and reputational consequences of making an error. While unnecessarily refusing access does harm to patient populations over the long term, it is the safer option for the individual concerned. This grey area was felt to be a greater problem for individual P.I.?s and SME?s in particular.

Data holders/Data hoarders

It was noted that at the current time, there is no practical incentive for researchers or healthcare organisations to share data. Often there is a strong sense of ownership of data by organisations, rather than patients. Important data which could provide benefit to patients can be effectively lost when projects end or staff retire. To combat these issues, avenues for providing recognition and funding for producing sharable data should be explored.

Trust

The future of personalised medicine requires careful nurturing and protection of trust from the public, by researchers and organisations deserving of this trust. It was noted that larger scale questions regarding the operation of large companies based outside the EU with fewer controls had the potential to undermine this trust, and particular examples were given of genetic testing companies who ?own? the data provided by patients, and sell on access to this data.

Implementation barriers

Many practical issues important for bringing Personalised Medicine products to market were discussed, such as the need to create technical documentation for regulatory approval throughout the process, and anticipate other regulatory requirements on equipment and testing. Uncertainty in some emerging regulatory areas (such as Software as a Medical Device), makes this more difficult. All innovation within healthcare is dependent on an appropriate environment for deployment, and researchers need to understand health service structures such as referral pathways for products to have a chance to succeed. Researchers also discussed many practical implementation barriers such as procurement rules and reimbursement codes. It was also noted that deployment will depend on clinician skills for personalised medicine, and that this topic is under-investigated.

New regulatory pathways

It was also noted that the appropriate regulatory paths for personalised medicine and individual clinical trials have not yet been developed in the EU, and that there is again a danger of competitive disadvantage when compared to other regulatory regimes which have been adapted to account for new developments.

Recommendations

A number of recommendations emerged from the discussion. All efforts are dependent on the research and patient community working together to explore and push solutions, and differences between countries need to be accommodated.

  • Clarity from authorities on what comprises an acceptable interpretation of the regulations.
  • Promoting public awareness of the benefits of research on personalised medicine, and how anonymised data is used.
  • Develop appropriate and practical consent procedures for patients that allow for supervised reuse. Biobanks were seen as potential exemplars for consent and data sharing.
  • Study best practice (recent efforts in Finland linking the main databases, supported and specified in legislation harmonised with GDPR and local laws, were seen as an example of this).
  • Training and materials for data holders and researchers, as well as clinicians for individual personalised medicine.
  • Establish a path to patient from the outset in order for projects to meet regulatory requirements, and support production of the required technical documentation. Due to the timescales involved in bringing innovations to market, this may also require longer projects or chaining multiple projects together
  • Innovation in Personalised Medicine may benefit from additional support structures to provide expertise – no researcher can be an expert in everything. Example actions might be to provide example documentation and expert guidance to projects which intend to produce innovations which may be subject to regulation, or direct support for submitting applications to regulatory agencies.
  • There may be a need to support best practice and experience sharing for funding agencies, at national and sectoral level.
  • There was a very strong feeling regarding the need to influence roadmap for regulatory agencies.

Successes

The starting point for discussion was the many innovative projects presented from the ITN?s. It was also noted that a variety of EU initiatives aim to address some of the issues raised. These include the Innovation Radar, which intends to identify, support, and promote high potential innovations from research projects. The European Open Science Cloud and EOSC Life initiatives are focussed on cloud research infrastructure, and are working towards issuing guidelines and recommendations. Initiatives such as PERMIT are also addressing research methodologies for personalised medicine, with the aim of reaching consensus and publishing guidelines, which can inform efforts to develop appropriate regulatory pathways.