By Seve Loudon
Since the first sequencing of the human genome, genomics has played an integral role in the development and implementation of novel medicines and treatment options for patients across the world. In doing so, genomics has demonstrated great potential to address deep-seated health inequalities through the bespoke targeting of individual needs. However, it also holds the potential to drastically exacerbate those same inequalities if access to genomics remains limited.
For these reasons, PPP brought together experts, political decision makers, diplomats, and clinical experts to explore the current global challenges faced in equitably utilising genomics. The conference was opened by the Rt. Hon. Stephen Dorrell, Executive Chairman of PPP, who highlighted how interlinked genomics is throughout the wider health and life sciences landscape.
Importance of international co-operation
In her keynote address, Shadow Minister for Business, Energy and Industrial Strategy, Chi Onwurah MP, spoke of the importance of life sciences in developing not just the health sector but the UK economy as a whole.
In line with this message, Onwurah stated that the UK cannot enact change alone and must act in concert with European and other international partners. PPP’s work on genomics has consistently stressed this point – that ensuring equitable access to genomics requires a concerted effort towards collaboration and co-operation on a global scale. Onwurah also stressed the importance of long-term strategies to harness the potential benefits of research, stating that, “a short-term approach to research projects is a barrier to an effective research base with diversification”.
- International collaboration is a necessity if the UK is to realise its potential as a life science superpower.
- Governments must provide an enabling environment for research to thrive through increased attention and funding into life science.
Integrating genomics into clinical settings
The first panel, chaired by Professor Sir Mark Caulfield, focused both on how clinicians are using genomics and on potential areas where improvement can be made for the integration of genomics into clinical settings. On a base level, the integration of genomics into clinical settings allows for more effective patient treatment options. In many instances, this can also allow for better diagnostics, leading to better outcomes for patients and clinicians alike. Panellists throughout the session detailed both the challenges to integration, as well as the clear benefits of doing so.
During the panel, Our Future Health Director of Clinical Feedback, Dr Cosima Gretton, highlighted the need to present data in ways that are simple and easy to understand. She also remarked that “there remain data blackholes. A patient in a particular hospital may be treated by clinicians who do not request genomic tests”. The lack of gathered genomic data from a patient may result in them missing out on potential clinical benefits.
One such benefit was highlighted by Professor Mina Ryten, Professor of Clinical Genetics at UCL Great Ormond Street Institute of Child Health, who stated that “if you can embed genetic testing in existing clinical pathways and structures, we will pick up on issues much earlier”.
Dr. Emma Stanton from Oxford Nanopore Technology (ONT) further reiterated this point by stating that “investing in the patient pathway to find the best point for genetic testing” is needed.
Dr. Angela George, Medical Director for the London Genomic Medicines Service highlighted that “vacancy rates for scientists is a huge problem. There are not enough senior scientists, or others across the board, to interpret the genetic tests”.
This panel highlighted the key touchpoints where genomics can truly make a difference in clinical settings. Further insights from the panel pointed to the importance of addressing staffing needs within clinical genetics labs. Additionally, the panel stated that there is a need to build a health economics case for the inclusion of genomics in clinical settings. Indeed, this is an issue that ONT – as Dr. Emma Stanton stated – has been working on together with Guy’s and St Thomas’ Hospital, as well as on a global level with conditions such as tuberculosis, in conjunction with the WHO.
- Genomic data must be presented to patients in a way that is clear to understand.
- There is a growing need to understand the patient pathway in ways that allow for genomics to be best utilised for patient care.
- Utilising genomics as a diagnostic tool is a significant opportunity for patients and clinicians alike in designing/delivering treatment options.
- Embedding technical expertise within hospitals is vital so that genomics can be properly and effectively utilised throughout care delivery.
Realising the potential of pharmacogenomics in clinical settings
Pharmacogenomics can often appear to be a complex issue. This panel sought to first clarify its role within health care delivery, articulating its benefits, before moving into the challenges and barriers to its increased usage in clinical settings. The benefit of pharmacogenomics speaks to the overarching ability to deliver effective treatment. This appears through the use and discovery of vaccines as well as in the ability to properly treat patients that have rare diseases. It also highlights how clinicians are able to treat diseases, in general, more effectively through a better understanding of patient reactions to specific drugs or modalities of treatment suited to a patient’s genetic makeup.
In this regard, the panel illustrated that there’s a general need to move to proactive pharmacogenomics. Meaning that, as Professor Matt Brown, Chief Scientific Officer at Genomics England put it, “we should be insisting that all clinical trials should include a genetic component”; thus, enabling a culture of prevention rather than retroactive treatment.
Ujwala Sangam, a pharmacist specialising in pharmacogenomics, pointed to the role that pharmacogenomics has in enabling effective patient treatment in a way that ensures that patients are receiving the right medicine that is, often, more effective for them.
While the potential is promising, implementing pharmacogenomics into clinical pathways will require novel approaches, particularly around the use of data. Dr John McDermott, NIHR Doctoral Research Fellow at the University of Manchester asserted that “the biggest challenge is how we get the data to clinicians in a clinically relevant format and in a clinically relevant timeframe, and that is context specific”.
Another key barrier was identified as clinical and public buy-in. New technologies will not be applied if they are not appropriately understood and it was expressed that instilling a greater understanding of pharmacogenomics’ benefits among clinicians could help to increase buy-in, both among clinicians and patients. Tonya McSherry from Oxford Nanopore highlighted that “we need to hear more of the story around the benefits of pharmacogenomics” when questioned from the audience as to why so little understanding of the field is held by the public, and even some clinicians.
In many instances, pharmacogenomics has great potential to impact patient care and overarching effective medical treatment through medicine. Pharmacogenomics, as a preventative and diagnostic tool, can be incredibly impactful. In terms of health economics, this also means that costs have the potential to decline through increased preventative treatment.
- Pharmacogenomics should be a standard part of clinical practice.
- Simpler terminology is a necessity around pharmacogenomics.
- More work is needed to demonstrate the benefits of pharmacogenomics in clinical settings.
AI and genomics: a new frontier?
Artificial intelligence (AI) is becoming increasingly topical as a major cross-cutting issue, nowhere more so than within the field of genomics. Though while genomics certainly requires the increased use of AI, it differs from most fields in the sense that it relies on the ability of AI to manage complex, large-scale datasets. In genomics, AI has an improved capability to predict disease-linked genetic variants located in the non-coding regions of the genome. However, AI models in genomics are trained on datasets that lack diversity and therefore have biases ingrained.
This panel discussed how to strike the balance between the promises and perils of AI, to harness its potential in genomics and prevent it from exacerbating health inequalities. From this standpoint, the conversation throughout the panel largely focused on how AI could best be utilised to enable more equitable practices with regard to genomic data. Contributions centred around the need to acknowledge that the use of AI does not exist in a vacuum and has real world impacts when it comes to research outcomes and health inequalities.
Speaking on a need for diversity in genomic data sets, Dr. Mike Vella, Director of Machine Learning Operations at Oxford Nanopore stated that “increased diversity [in datasets] improves analytical capacity across the board”. His point highlighted the fact that diversity in data needs to be included at every level. Meaning that even within an already “diverse” group, steps should be taken to ensure that intersectional characteristics within the group are also represented.
Dr. Matt Howard, International Healthcare Strategy Lead for AI in Healthcare at Amazon Web Services, spoke to a need to incorporate points around diversity from the outset of AI design and storage. He highlighted that these issues should be looked at proactively rather than attempting to address diversity of data issue retroactively; as by this point internal biases around data could already become ingrained within the data usage.
To achieve this, Dr. Neil Ebenezer, Medtech and Genomics Specialist Consultant, pointed to other international examples around AI training sets and regulations that could potentially be adapted for different contexts. This further highlighted the need for international co-operation around genomics in particular.
Also aligning behind the point that diversity needs to be linked to the outset of model design, Sasha Henriques, a genetic counsellor highlighted that although “the conversation around AI is incredibly mature, we must look at who is in those conversations and how those conversations can become more diverse”.
Professor Chris Yau, from the Big Data Institute at the University of Oxford further stated that “we forget that even though there’s lots of data, we forget where it comes from and who that data impacts”.
It is clear that AI is inextricably linked with genomics. To address areas of concern around diversity of data and a need to address health inequalities rather than exacerbating them, data must be as diverse as possible, and address the intersectionality within communities. This panel also showcased that AI is first and foremost a tool. The use of AI, and how it addresses key issues of inequality, are down to the human decisions of researchers and analysts in how they design and utilise AI models.
- There’s a need to understand where the data is coming from and to ensure that such genomic data is being collected in an equitable and ethical manner.
- Understanding and addressing biases in AI and AI systems is crucial in the further development of integration with the field of genomics.
- While the maturity of the conversations around AI has improved drastically, there remains a need to look at who’s in those conversations and address any potential inequalities from the outset.
Strengthening genomics co-operation between the UK and Europe
Throughout the Global Genomics programme, international collaboration has remained a vital cornerstone of discussion. This panel sought to contribute to that discussion by focusing on ongoing and potential future co-operation between the UK and Europe. Within the past decade, co-operation on many fronts between the UK and Europe has encountered barriers – not least of all through the UK withdrawal from the EU. However, strides have been made to enhance collaboration; the UK has an agreement in principle to rejoin the Horizon research programme, is continuing to participate in the European Health Data Space and has signed bilateral agreements with European nations – chief among them the Swedish-UK Memorandum of Understanding (MoU) on Life Sciences.
Marika Amartey, Science and Innovation Counsellor at the Embassy of Sweden to the UK, spoke directly to this collaboration. She highlighted the Swedish-UK MoU as an example for how co-operation between the UK and European nations and, wider industry can look in a post-Brexit world. She stated that the MoU is “not really about the governments, it’s about the community that we’re doing this for”.
Speaking towards other forms of collaboration around data and the necessity for international co-operation in this area, Dr. Thomas Keane, Team Leader for the European Genome Phenome Archive at EMBL-EBI highlighted that “we can all generate data, but actually making it discoverable, interoperable, being able to access it, it is just so important. And the project is fundamentally about interoperability between the different countries.”
However, as Dr. Mark Bale, Independent Genomics Consultant pointed out, collaboration cannot exist solely for collaboration’s sake. For collaboration and co-operation to work “it has to be a genuine partnership with genuine benefits”.
Dr. Julia Wilson, Associate Director at the Wellcome Sanger Institute pointed to some of the biggest challenges to international co-operation around genomics. She stated that “the biggest barriers to cooperation currently are that we need a frictionless movement of people and scientists to collaborate with each other. We also need the safe, responsible, and frictionless sharing of genomic data in order that we all benefit from genomics research.”
In her role working with the European Health Data Space – a source for ongoing co-operation between Europe and the UK – Dr. Sarion Bowers, Head of Policy at the Wellcome Sanger Institute stated that a significant challenge is in the ability to align legislative actions and priorities. She highlighted the growing list of international legislation in this field and stated that having the UK present in those conversations will ultimately “allow [the UK] to influence our own legislation which will then help us in our own activities and engagement with partners”.
Throughout the discussion, members of the panel highlighted the need to continue and reinforce collaborative efforts around genomics and life sciences.
- Work on genomics, and progress towards that end, is only truly possible through international collaboration and co-operation. It cannot happen within solely one country or even only within academia or industry.
- Collaboration must be a two-way approach that elaborates on the benefits for each party.
- Genomics requires international input. Even while the UK is a significant actor in this regard, in order to create lasting impact, more data is necessary which in turn can only come through co-operation.
Driving innovation through collaborative partnerships
This session acted as a closing keynote that narrowed in on the role that industry plays – and can play – within public-private partnerships. Led by Professor Dame Sue Hill, Chief Scientific Officer at NHS England, this panel touched why industry partnerships with national health systems and other public sector institutions are critical to driving innovation. Importantly, this session highlighted a need for the UK to invest in its own genomics ecosystem to drive collaboration and innovation within the sector. The panel specifically focused on the UK genomics industry and the critical role that the NHS plays within the sector. The overarching theme within this session highlighted a critical statement: collaboration is essential. It is through collaboration that more investment can be obtained and put to use to drive forward progress on genomics, within the UK in particular, but also globally.
- Investment is key; coming in many forms such as human or financial investment, more needs to be done to drive progress.
- Collaboration and co-operation between public and private sector institutions are necessary for real progress to be made.
The Global Genomics Conference consistently highlighted how collaboration between a range of stakeholders is necessary within the field of genomics. This co-operation was discussed in various terms, whether on an international level or between public and private institutions. It demonstrated that tools, such as AI, are inextricably linked to genomics. Not only this, but researchers must take care to incorporate intersectionality and diversity within the models rather than rely on AI to fill the gaps to avoid exacerbating health inequalities. Through such tools, clinicians and researchers alike have the capability to treat and prevent a wide range of disease more effectively. Genomics has a critical role to play within the life science sector and throughout global health systems. It remains up to those within the sector to ensure that equality and diversity are at the forefront of their practices and policies.