2Whole genome sequencing
The 100,000 Genomes Project
6.The launch of the 100,000 Genomes Project was announced by then Prime Minister, David Cameron, in 2012, with the objective of sequencing 100,000 genomes from NHS patients by the end of 2017.10 A new body, Genomics England, was established in July 2013 by the then Department of Health as a wholly owned limited company to deliver the 100,000 Genomes Project in England. The initiative had five main aims:
- to benefit patients by providing clinical diagnosis and, in time, new or more effective treatments;
- to provide new scientific insights and discovery;
- to accelerate the uptake of genomic medicine in the NHS;
- to stimulate and enhance UK industry and investment; and
- to increase public knowledge and support for genomic medicine.11
7.As of February 2018, just over 50,000 whole genomes had been sequenced as part of the Project.12 Following advice from Genomics England, the Department of Health and NHS England, the Government agreed to extend the sequencing element of the programme to the end of 2018.13 Accounting for this delay in completing the 100,000 sequences by 2017, Genomics England explained that the project is “working at the edge of known science”.14 Giving evidence to our predecessor Committee in February 2017, Professor Mark Caulfield, of Genomics England, described challenges in preserving cancer samples and building streamlined systems for handling genomic and related data as the two main obstacles that had delayed the project.15 He indicated that these challenges had at that stage been largely overcome and that Genomics England “anticipate finishing the 100,000 Genomes Project by the end of 2018”.16
8.Despite the delays, the Life Sciences Industrial Strategy highlighted the 100,000 Genomes Project as one of two examples of large healthcare infrastructure projects that have put the UK in globally leading positions, highlighting that Genomics England “has already set the global standard for healthcare genomic data in rare disease and now, increasingly, in cancer”.17 Professor Sir John Bell, the author of the Life Sciences Industrial Strategy, told us that the UK is now “multiple years ahead of the rest of the world in handling whole genome data”.18 Industry groups reported similarly positive feedback,19 while other witnesses praised the ambition and progress of the programme, though also highlighting some of the remaining challenges. The Wellcome Sanger Institute told us that the project was a “ground-breaking, world-leading genomics initiative”, but added that “although an excellent programme, there are areas where recurring concerns have been raised”.20 University Hospitals of Leicester told our predecessor Committee that the project was a “bold and ambitious project that promises exciting benefits for the country, for communities and individuals”, but which also “gives rise to several significant educational, social, cultural and economic challenges that require careful research and analysis”.21 Dr Hilary Burton, of the PHG Foundation, told us:
The way all the genomic medicine centres have been set up, so that they all have the facility to recruit patients, get the necessary clinical information together and submit them for sequencing, with the processes for doing the sequencing and interpretation and feeding back diagnostic information, has been a fantastic leap forward and will set the way for how we eventually use it. Our point is that there is still a long way to go and we still have to put substantial resources into making it happen across the whole country for all the specialties where it is relevant.22
9.Professor Sir John Bell told us that setting up the 100,000 Genomes Project as a separate entity from the NHS was key to its development:
Had we done the 100,000 Genomes Project in the NHS from its pilot phase it would have failed. The only way this works is if you take it out and set it up as an independent company wholly owned by the Department of Health. It was free to operate, employ people and do stuff in a way that the NHS structure would not allow. Once you know it works, you drop it back into the healthcare system.23
University Hospitals of Leicester, on the other hand, told our predecessor Committee that the project had missed opportunities to collaborate with existing projects, in particular the Clinical Research Network. They believed that this had led to duplication of “staffing and IT infrastructures, education, advertising, study design, communication and managerial governance arrangements”,24 and to competition between the projects:
There were missed opportunities for joint working (with no cross-subsidisation of funding) when patients were eligible for Clinical Research Network and Genomics studies […] The demands on clinical genetics departments to lead and deliver the 100,000 Genomes Project, especially those resulting from complex and challenging inclusion criteria, has reduced time to concentrate on CRN study delivery and recruitment. This is compounded by genomics projects often competing for the same patient groups.25
Subsequently, the Life Sciences Industrial Strategy has recommended the creation of a Health Advanced Research Programme (HARP), and described Genomics England as one of two “excellent examples for future large-scale HARP projects”.26
10.Our predecessor Committee highlighted, as an issue for further inquiry, “whether patients are being invited into the 100,000 Genomes Project due to perceived long-term research and commercial benefits, at the expense of more immediate benefits to their health”.27 Genomics England explained that this was not the case.28 Professor William Newman, of the British Society for Genetic Medicine, assured us that:
In the 100,000 Genomes Project, one of the eligibility criteria is that patients should have already had their routine clinical care, routine assessment and tests, such that they are not being disadvantaged in any way. The application of whole genome sequencing should be something additional to try to find the answer for their condition.29
Professor Sue Hill, of NHS England, confirmed that “with cancer, currently any patient being entered into the 100,000 Genomes Project is having their standard diagnostic testing done, and that includes any genetic-based tests and other tests that help direct their therapies”.30
11.Concerns had also been raised in our predecessor Committee’s inquiry about a perceived lack of planned evaluation for the 100,000 Genomes Project.31 Dr Hilary Burton, of the PHG Foundation, had hoped that evaluation processes would have been built into the 100,000 Genomes Project to prospectively collect information on the overall results and clinical impact of genomic testing across the whole Project,32 but told us that her “understanding is that that has not happened”.33 The CMO rejected Dr Burton’s concerns, telling us that Dr Burton “is clearly not close enough to the project”.34 Nonetheless, concerns about a lack of evaluation of the Project persist. Dr Edward Blair, of the Oxford NHS Genomic Medicine Centre, told us that “there has been no significant review of outcomes since only small numbers of completely analysed and clinically validated genomes have been provided”.35 Dr Burton told us that “some of the more routine learning about how [whole genome sequencing] would be implemented in normal patient pathways has not been available, but we have to get on with it now”.36 The PHG Foundation emphasised that “to ensure that the legacy of the 100,000 Genomes Project is maximised, we strongly recommend a formal evaluation of its different work programmes from inception to 2017–18. Results will be critical for informing the design and implementation of future NHS healthcare services”.37 The Association for Clinical Genomic Science made a similar plea.38
12.The 100,000 Genomes Project is an ambitious project that has helped put the UK in a world-leading position on whole genome sequencing and genomic medicine. As the 100,000 Genomes Project approaches the completion of its sequencing target, the Government should formally evaluate it to inform the wider introduction of whole genome sequencing in the NHS (which we explore further below). The 100,000 Genomes Project could be a model for future ‘Health Advanced Research Programme’ projects, as suggested in the Life Sciences Industrial Strategy. If so, HARP projects should have processes and resources put in place from the start to allow their subsequent evaluation, and should explicitly take account of how existing NHS initiatives and resources will be complemented or absorbed.
Whole genome sequencing
13.The CMO’s report focused on whole genome sequencing, describing it as “the pinnacle of a pyramid of molecular diagnostics”.39 Other genetic diagnostics, including ‘single gene’ diagnostics and ‘panel’ and ‘exome’ sequencing,40 are already used by genetics laboratories. In 2016, NHS England stated that it intended to “commission whole genome sequencing and embed genomic medicine into routine care pathways where it is clinically and cost effective to do so”.41 We examine below those clinical-effectiveness and cost-effectiveness issues in turn.
Clinical effectiveness
14.The optimism in the medical community at the potential for whole genome sequencing is clear. The Wellcome Sanger Institute believed that “genomics has the potential to dramatically improve patient care by improving specificity of diagnosis and helping stratify management and treatment”.42 The CMO told us that over the course of the 100,000 Genomes Project:
It has become clear that whole genomes are extraordinarily important […] we now know that, even if you want the exome, you are much better getting it from a whole genome, because it picks up inversions and quite complicated things and gives you a better-quality exome.43
15.Professor Sian Ellard, of the South West NHS Genomic Medicine Centre, told us how whole genome sequencing can benefit the diagnosis of rare diseases:
In the past we could provide testing only for patients who had the most common rare diseases, because you had to set up a test for each specific condition and it was only feasible to do those tests where there was sufficient volume. This technology means that potentially we can diagnose any rare disease for which the genetic basis is known. That is really exciting. We have seen a huge increase in the number of patients for whom we can provide a diagnosis.44
She added that whole genome data can be revisited as research discovers new genetic causes of very rare disorders.45 Lord O’Shaughnessy, Parliamentary Under-Secretary of State for Health, told us that it was such versatility, and the opportunity to replace a range of tests with whole genome sequencing, that had been “convincing to Government”.46
16.The optimal balance between whole genome sequencing and alternative genomic tests was not, however, always clear. The Scottish Genomes Partnership told us that, in comparison to targeted sequencing panels, whole genome sequencing: takes longer to return results to the patient; generates more data, increasing storage costs; has stricter demands on patient samples; and provides reduced sensitivity to specific genetic targets.47 In regards to rare diseases, they thought that choices between whole genome sequencing and alternative diagnostic tests were “complex”, but anticipated “the eventual mainstream delivery of whole genome testing within NHS Scotland”.48 In contrast, for cancer, they told us that “increasing numbers of scientists and oncologists […] are reaching the conclusion that targeted sequencing panels are a better choice [than whole genome sequencing] for the foreseeable future, for both the patient and the NHS”.49 Several other witnesses expressed a similar point of view.50
17.Patients with rare diseases may have spent years seeking a diagnosis. For cancer patients, however, Professor Sir Mike Stratton of the Wellcome Sanger Institute highlighted that whole genome data analysis is needed “within a couple of weeks in order to make the appropriate decisions with respect to choice of therapies”.51 Genomics England told us that “at this early stage of genomic medicine, it may take many months for results to come back”, although “in future, this is likely to get quicker”.52 Genomics England are currently running a fast track cancer analysis pilot aiming to return cancer reports within four weeks;53 Professor Sue Hill told us that “in those small numbers of samples, that is being done in around 20 days”, and that “in some instances that is better than standard care at the moment”.54
Cost effectiveness
18.Genomics England told us that sequencing costs around £600 per genome, which is “affordable for a healthcare test”.55 The CMO’s report noted that there are also costs associated with bioinformatics analysis, clinical interpretation and reporting, and that these costs were not falling as quickly as sequencing costs.56 In return for these costs, Professor Sue Hill, of NHS England, believed that whole genome sequencing could shorten the “diagnostic odyssey” that people with rare and inherited diseases can go through, which can last “12 years or longer”, and during which “some will have been in and out of the healthcare system, utilising healthcare resources”.57 She calculated that “we spend, often, more [on a series of tests] than doing whole genome sequencing”, providing the potential for it to save costs compared with current diagnostics.58
19.Professor Hill emphasised that “pharmacogenomic59 profiling associated with [whole genome sequencing] could drive the use of appropriate medicines, rather than a one-size-fits-all medicine approach”,60 which could save costs and reduce the risk of adverse reactions to treatment. A 2016 NHS England report on personalised medicine concluded that “key pharmaceutical interventions are effective in only 30–60% of patients due to differences in the way an individual responds to and metabolises medicines”, and that “1 in 15 hospital admissions in the UK are linked to adverse drug reactions”.61 Professor Sir John Bell highlighted how medicines developed for increasingly targeted (and hence smaller) patient populations reflect a more general trend away from ‘blockbuster’ medicines, with uncertain consequences for affordability.62
20.Whether there are net savings or costs from whole genome sequencing is, however, uncertain. Dr Magdalini Papadaki, of the Association of the British Pharmaceutical Industry, explained that:
All these transformative therapies, whether they are curative or it is early prevention so a disease or cancer does not manifest itself, have long-term healthcare burdens or uncertainties, because you now have groups of people who survive and then can get diseases later on, or in ageing. All those cost offsets from previous deaths, as tragic or controversial as it might sound, will now not exist. We do not know exactly how this cost-benefit balance will play out and what it will mean in the future.63
Dr Edward Blair, of the Oxford NHS Genomic Medicine Centre, also described the unknown economic impact of ‘additional findings’.64 An additional, or secondary, finding is one that is unrelated to the condition that led to the whole genome sequencing being conducted. He said that such findings could lead to additional healthcare activity such as clinical testing of patients and their relatives or prophylactic treatment of at-risk individuals, and saw the return of additional findings as a form of targeted ‘genetic screening’, which he believed should be formally assessed by the UK National Screening Committee before introduction.65
21.The CMO’s report acknowledged that personalising medicines and looking for additional findings pose “economic risks”, but argued that:
These are issues for the health system to tackle eventually under any scenario, and the progressive awareness of patients of their own risks with support from appropriate health professionals could be key to managing these issues.66
Discussing the cost-effectiveness of whole genome sequencing compared to current genetic diagnostics, her report stated:
It is possible that because whole genome sequencing can be industrialised as a single common process it will become more cost-effective to draw panels or exomes from a whole genome sequence factory than invest in many bespoke diagnostics for particular conditions.67
The Association for Clinical Genomic Science, on the other hand, told us that the affordability of whole genome sequencing could be undermined if competition were weakened:
There is concern in the community that a monopoly may be created by the transitioning of a substantial proportion of testing to centralised whole genome sequencing that will not encourage competition between commercial providers in the genomic industry and that costs will remain too high to create greater access within the NHS.68
The CMO’s report also flags the potential for developments in genomic medicine to benefit the economy.69 We explore the opportunity for NHS England to capture the value of its genomic data in Chapter 4.
Evaluation
22.Professor Sue Hill told us that NHS England would base its assessment of the clinical- and cost-effectiveness of whole genome sequencing on systematic and ongoing reviews of the available evidence.70 Professor Patrick Chinnery, of the University of Cambridge, believed that there was already strong evidence to support the replacement of some existing diagnostics with whole genome sequencing; that “the diagnostic yield from whole genome sequencing is superior to anything else that preceded it”.71 While he thought whole genome sequencing will have “a more immediate impact” for rare cancers, he was less certain for common cancers, and did not expect whole genome sequencing to become routine NHS care in these cases for five to ten years.72
23.The CMO’s report cautioned that:
For whole genome sequencing to become part of the regular commissioned service it will have to demonstrate superior efficacy (and efficiency) to alternative sequencing regimes. While there are strong indications that all these conditions will be met as the technology develops, more progress will be required in the 100,000 Genomes Project to help provide the evidence.73
Most existing studies, the report noted, “tended to be of fairly low quality”, or used “very small patient samples, which is in total contrast to the data which the 100,000 Genomes Project will provide”.74 The CMO acknowledged that evaluation of the cost-effectiveness of whole genome sequencing “is in its infancy”.75
24.At this stage, there has been no overarching evaluation of the 100,000 Genomes Project. Genomics England has publicised the details of cases in which individuals have benefited from whole genome sequencing and their involvement in the 100,000 Genomes Project.76 Additionally, Professor Mark Caulfield, of Genomics England, told us that “in rare diseases, 20% to 25% of the participants in the pilot are now receiving potentially actionable diagnoses in the health system today”.77 Professor Sue Hill told us that in cancer, “we are seeing a greater number of potentially actionable changes—up to the level of 65% in some patients—emerging from whole genome sequencing”.78
25.The PHG Foundation, however, was sceptical about the evidence to date on the impact of whole genome sequencing on clinical outcomes. They believed that Genomics England’s evaluations do “not tell us what effect the [whole genome sequencing] test result had on clinical decision making and clinical outcomes for groups of patients with particular clinical presentations arising through routine practice”.79 In addition to recommending evaluation of groups of patients, rather than individuals, they urged that an evaluation span the entire ‘clinical pathway’, “from assessment, referral, acceptance, consenting, testing, receiving a result (diagnostic outcome) and ultimately to the patient level clinical consequences”.80
26.Despite the apparent importance of the 100,000 Genomes Project as a source of evidence on the clinical and cost-effectiveness of whole genome sequencing, Professor Sue Hill could not say whether there were plans for a formal evaluation of the 100,000 Genomes Project.81 She instead outlined a variety of other sources of evidence that NHS England has used to assess the results of whole genome sequencing.82 In the meantime, the charity Genetic Alliance UK was concerned at the lack of publicly available information regarding the introduction of the Genomic Medicine Service.83
27.There is great potential for whole genome sequencing to improve patient care, particularly for diagnosing rare diseases and for more personalised targeting of medicines and treatments. However, there is not yet sufficient unambiguous evidence gathered to demonstrate its benefit for routine care, in particular for common cancers. As the first large-scale whole genome sequencing exercise in the world, the 100,000 Genomes Project must be an important source of evidence to determine the technology’s clinical efficacy and cost-effectiveness across the whole ‘clinical pathway’, and at the level of patient populations, rather than individual patients. Such evaluation does not appear to have been conducted, or at least has not been made public. In advance of the launch of the Genomic Medicine Service, NHS England should undertake and publish a detailed evaluation of the 100,000 Genomes Project, to inform an assessment of the anticipated clinical- and cost-effectiveness of routine whole genome sequencing in the NHS.
28.Research and evidence-gathering needs to be a continuing process. The CMO’s report recommended that NHS England “embeds implementation research (including cost effectiveness) at all stages of service redevelopment and laboratory reconfiguration”.84 The PHG Foundation concurred, highlighting that “implementation research is a vital element to ensure that resources are used effectively, and especially to underpin the substitution of new technologies for existing redundant technologies”.85 Professor Sue Hill told us:
There will be an evaluation function within NHS England that will be based and built upon [the UK Genetic Testing Network]. It will include the expert standing committee recommended by the CMO in her report, and this will review evidence on an ongoing basis across all the tests that will be introduced into the NHS for both rare disease and cancer […] We will be working really closely together with the National Institute for Health and Care Excellence (NICE) to ensure that we have a seamless process that helps direct the commissioning system on the basis of the finance that is going to be available and the affordability. We will look at it through the lens of the five year forward view, and on quality, improving access, and then its affordability.86
29.NICE has already included genetic diagnostics in its clinical guidelines and in its Technology Appraisals and Highly Specialised Technologies programmes,87 although these have not yet covered whole genome sequencing.88 The 2016 independent Accelerated Access Review noted that NICE’s Technology Appraisals mostly focus on pharmaceuticals rather than diagnostics, and that positive guidance outside of the Technology Appraisal programme does not carry with it a funding requirement.89 It recommended that NICE “rebalance its work towards products which, accompanied by appropriate changes in clinical pathways, can improve system efficiency whilst delivering equivalent or better patient outcomes”.
30.Professor Sir John Bell told us that genomic sequencing is difficult for NICE to evaluate:
This is hard for NICE because it is not a domain that it really understands and knows about, but it has been quite responsive in helping to think through how to get the value proposition for genomics to work.90
Dr Mark Kroese of NICE told us, however, that the CMO’s ambition to mainstream genomics in the NHS “fits really well with the NICE work programmes”.91 He reported that NICE had not had experience of the whole genome sequencing yet, but he did not see “any challenges to our abilities to evaluate genomic tests in the diagnostic assessment programme”.92
31.As whole genome sequencing becomes approved for certain conditions, there will be financial pressure to remove alternative diagnostic tests from the ‘directory’. As Professor Bell put it:
To adopt innovation successfully you have to do two things. You have to invest in it to get it in the system, and then you have to work hard to extract the things you do not need to do any more to get the savings from the innovation ultimately to produce more efficient systems.93
Professor Lyn Chitty, of the North Thames NHS Genomic Medicine Centre, commented elsewhere, however, that existing diagnostic tests will need to continue alongside whole genome sequencing in the short term, to allow for comparison and validation of whole genome sequencing.94
32.The 100,000 Genomes Project will not be able to provide all of the evidence required to assess the effectiveness of whole genome sequencing for all conditions. Research and evidence-gathering will need to be continuing processes. We endorse the CMO’s recommendation for NHS England to embed implementation research at all stages of redevelopment and laboratory reconfiguration for the Genomics Medicine Service. Where more evidence is needed to approve whole genome sequencing for particular conditions, current diagnostics should be maintained alongside whole genome sequencing, as was done in the 100,000 Genomes Project, unless the genomic diagnostic has proved more accurate for that condition.
10 ‘Genomics England and the 100,000 Genomes Project’, Genomics England
11 ‘The 100,000 Genomes Project Protocol’, Genomics England (2017)
12 Genomics England, ‘The 100,000 Genomes Project by numbers’, accessed 1 March 2018
15 Oral evidence taken on 8 February 2017, HC (2016–17) 854, Qq16–20
16 Oral evidence taken on 8 February 2017, HC (2016–17) 854, Q16
17 ‘Life Sciences Industrial Strategy—A report to the Government from the life sciences sector’ (2017)
18 Q17
19 Association for the British Pharmaceutical Industry (GEN0040) and the BioIndustry Association (GNH0017)
22 Q101
23 Q15
26 ‘Life Sciences Industrial Strategy – A report to the Government from the life sciences sector’ (2017)
27 Science and Technology Committee, Sixteenth Report of Session 2016–17, ‘Genomics and genome-editing: future lines of inquiry’, HC 854
29 Q58
30 Q60
32 Q111
33 Q113
34 Q156
36 Q116
39 ‘Generation Genome’, Annual Report of the Chief Medical Officer 2016 (2017)
40 Different genetic diagnostics differ in which parts of the genome are sequenced. Single gene tests look for mutations in specific genes known to relate to the disease a patient is suspected of having; gene panels sequence multiple genes simultaneously, varying from a few genes known to relate to a particular condition through to all known genes with disease-related function; exome sequencing reads the entire exome—the 1% of DNA that provides instructions to the body on what proteins to produce.
41 NHS England Board Paper, 30th March 2017
43 Q155
44 Q101
45 Q110
46 Q163
50 For example, the Wellcome Trust, the Association of Medical Research Charities and Cancer Research UK (GEN0038), the Association for Clinical Genomic Science (GNH0036) and Roche Products Limited (GNH0038)
51 Oral evidence taken on 8 March 2017, HC (2016–17) 854, Q181
54 Q60
56 ‘Generation Genome’, Annual Report of the Chief Medical Officer 2016 (2017)
57 Oral evidence taken on 8 February 2017, HC (2016–17) 854, Q5
58 Q53
59 Pharmacogenomics entails using a patient’s specific genetic mutations to inform the choice of drug treatment, identifying which treatments will be most effective and which might provoke adverse reactions.
60 Oral evidence taken on 8 February 2017, HC (2016–17) 854, Q5
61 ‘Improving Outcomes through Personalised Medicine’, NHS England (2016)
62 Q40
63 Q40
66 ‘Generation Genome’, Annual Report of the Chief Medical Officer 2016 (2017)
67 ‘Generation Genome’, Annual Report of the Chief Medical Officer 2016 (2017)
69 ‘Generation Genome’, Annual Report of the Chief Medical Officer 2016 (2017)
70 Q65
71 Q158
72 Q160
73 ‘Generation Genome’, Annual Report of the Chief Medical Officer 2016 (2017)
74 ‘Generation Genome’, Annual Report of the Chief Medical Officer 2016 (2017)
75 Q156
76 For example, see Genomics England, ‘Participants from NHS Genomic Medicine Centres have shared their stories with us’, accessed 2 February 2018
77 Oral evidence taken on 8 February 2017, HC (2016–17) 854, Q57
78 Q60
81 Q55
82 Q55
84 ‘Generation Genome’, Annual Report of the Chief Medical Officer 2016 (2017)
86 Q81
88 Q78
90 Q34
91 Q78
92 Q80
93 Q30
94 ‘Implementing a National Genomic Medicine Service for the NHS: building on the legacy of the 100,000 Genomes Project’, joint event by the All-Party Parliamentary Health Group and the All-Party Parliamentary Group for Personalised Medicine, 7 November 2017
Published: 20 April 2018