28 January 2019
Comments made by the UK Health Secretary Matt Hancock over plans for the NHS to offer paid-for genomic sequencing to healthy people who will agree to share the data produced have been widely reported this week.
The proposal to recruit paying ‘genomic volunteers’ was part of the Life Sciences Sector Deal 2 announced in late 2018. Following on from the 100,000 Genomes Project and launch of the new NHS Genomic Medicine Service, the ambitions for the UK to become the ‘home of genomic healthcare’ were boosted by a commitment to sequence one million whole genomes within five years (half ‘through the NHS’, half as part of the UK Biobank project) and a stated aspiration to reach five million whole genome sequences in the same period.
The Life Sciences Sector Deal went on to state that Genomics England – the Department of Health owned company set up to deliver the 100,000 Genomes Project and also charged with building a genomics research base and driving development of the UK genomics industry – would undertake development work with NHS England and public and patient groups on ‘a new service to enable genomic volunteers to pay for a personalised report on their unique genetic makeup’, adding that ‘with permission, the genetic data will be made available to researchers and scientists’. Presumably such volunteers could help build whole genome numbers towards the aspirational five million figure.
There has been little time for development of the concept since it was originally announced, especially since the NHS Genomic Medicine Service is only just getting going and Genomics England is in a state of apparent change, with the announcement on Friday that both the Chairman Sir John Chisholm and the Chief Executive Officer Prof John Mattick were stepping down. Sir John is succeeded as Chair by a fellow Board Member Jonathan Symonds, with chief scientist Prof Mark Caulfield stepping up as interim chief executive.
News sources such as the BBC report that the proposed new service for genome sequencing will provide purchasers with a health report that can ‘predict the risk of developing conditions like cancer or Alzheimers’. Building a larger genomic database does, of course, aid further biomedical research; Matt Hancock said that: “While healthy people should not have this service free on the NHS, there are huge benefits to sequencing as many genomes as we can…Every genome sequenced moves us a step closer to unlocking life-saving treatments".
Unsurprisingly given the lack of information about the service, many questions and concerns have been raised. Dr Anneke Lucassen, Chair of the British Society for Genetic Medicine, the professional society for NHS clinical geneticists, scientists and related experts, praised the ambitions to collect and collate data to create a resource for research but voiced concerns over the predictive ability of genetic testing in healthy individuals – which remains limited – and also over the potential to create a two-tier health system, with a service offering potential health benefits offered to paying customers (or paying genomic volunteers) only.
Much will depend on the detail about how such a service would work. In theory, measures to boost genome sequencing and the sharing of the subsequent data and analysis with robust linked clinical information (suitably anonymised) for research purposes is a huge plus; a rapidly increasing range of direct-to-consumer (DTC) genetic testing companies offer such services, with widely ranging quality and clinical utility – many promise health benefits such as the ability to predict disease and inform healthier lifestyles (such as tailoring diet or fitness plans), though few are clinically robust. Genomics England is, after all, a commercial company, albeit one owned by the Department of Health, and charged with generating revenue as well as global eminence in genomics. However, there is typically a very clear distinction between NHS and private medical services, so it is natural that there are concerns about new proposals.
Perhaps the most burning question is just what information might be returned to genomic volunteers. Would it be only clinically significant and actionable findings, such as identification of a serious inherited disease or risk thereof? Would it include less serious but potentially useful findings such as pharmacogenetics variants that affect how they may metabolise certain medicines? In these cases, would the information be available to their own GPs and other professionals to act upon – and what would the impact on NHS services be? If only important clinical findings are returned, then it would be appropriate for the NHS to act on them – but some would feel that this could indeed be supporting a two-tier system of access to services.
For less serious and robust results, the picture rapidly becomes more indistinct. Firstly, the implications of many genomic variants when identified in even an ill individual are of uncertain clinical significance. Even where a rare variant is detected that scientifically and medically could be involved with an existing disease, it is often very difficult to establish that it is definitely involved in the disease. This situation can and does change over time as more evidence linking a variant with a condition develops, but at a point in time many variants are classified as of ‘uncertain significance’ or as being probably, but not certainly, linked to disease or not. It is unlikely that information about the presence of such variants would be returned to genomic volunteers, but even so questions about the capacity or desire to re-contact these individuals if important new evidence emerged that suggested clinical action was needed.
Even considering robust, reliable links between genetic variants and disease risks, predictive genomics outside relatively rare conditions is a relatively young field and one where hype frequently exceeds reality. It is undoubtedly true that genomic analysis can improve disease risk prediction, and the development of polygenic risk scores offers promise even for common, complex diseases, but utility is presently limited. For example, even possession of two copies of the APOE4 allele, which is reliably associated with a 3-5 increased risk of Alzheimer’s disease and about as predictive as you are likely to get for most individuals, leaves plenty of uncertainty to grapple with.
Other questions of note include:
Ultimately, it is too soon to make any sensible assessment of this proposal, since there is no substance to them. However, Genomics England now has the opportunity to consult extensively with stakeholders to develop the plan for further consideration.