Further to your letter of 20 July, I have pleasure in enclosing SmithKline Beecham's (SB) response to the above enquiry.

  Healthcare systems across Europe are a substantial but underused research resource. They have much to offer in epidemiology, technology assessment, outcomes research and population genetics and there are many ways in which industry together with healthcare providers and academic groups can use healthcare information databases. As one of the world's leading research based healthcare companies, SB is committed to exploring these opportunities. We therefore welcome the opportunity of responding to this enquiry as it promises to broaden the debate around a vital element of healthcare databases, namely the potential offered by the collection of genetic sequence information.

  I hope that the enclosed submission outlining SB's approach to genetic databases in the UK represents a useful contribution to this debate. I would obviously be delighted to elaborate on any of the issues raised in it, if the sub-Committee felt that might be helpful.

Tadataka Yamada

Chairman, Research & Development

SmithKline Beecham Pharmaceuticals

EXECUTIVE SUMMARY

  There are many ways in which industry, together with healthcare providers and academic groups, can use information generated from genetic databases to meet the challenges of a new era of medicine and healthcare provision.

  SB has several collections of DNA samples in the UK for use in the search for new disease genes suitable as targets for drug development and to validate existing targets using genetic approaches.

  The main constraints on SB's work in this area are the time taken to find suitable clinical cohorts with a high quality of phenotypic data and ethnically matched controls that can be used in association studies. Finding academic centres willing and able to collaborate with industrial partners is also a constraint.

  SB acknowledges the requirements of the EC Data Privacy Directive in relation to the collection, storage and analysis of patient data.

  SB does not believe that the potential of bigger and more numerous genetic databases raises any new intellectual property issues for industry, academia or the NHS.

  SB treats all DNA samples held in UK databases in the strictest confidence and implements, where necessary, appropriate organisational and technical measures against unauthorised or unlawful use of such data.

  The UK should be aiming to establish a non-restrictive legislative environment which protects the individual while promoting the science.

  SB believes that the concerns expressed about industry involvement in genetic database initiatives can be assuaged by incorporating the best practice developed by companies such as SB.

  The basic infrastructure to realise the potential offered by large-scale genetic databases is already in place. What is now required is the political will to tackle the issue of public acceptability.

INTRODUCTION: THE VALUE OF GENETIC DATABASES

  1.  The science and technology of genetics is advancing at a remarkable rate. The completion of human genome mapping, the development of a high-density Single Nucleotide Polymorphisms (SNP) map and associated technologies over the next one-two years will further the identification of disease susceptibility genes for common diseases and the identification of genetic markers which can be used to predict an individual's response to a medicine (pharmacogenetics). In general, genetic databases represent an important new resource that will:

2.  Facilitate the identification of additional disease susceptibility genes: For common diseases such as cardiovascular disease, asthma, osteoarthritis, migraine and Alzheimer disease there are a number of variants of susceptibility genes which interact with environmental factors to cause disease. Identification of these genes involves the study of individuals with and without the disease and is recognised as an increasingly important aspect of research to find new medicines.

3.  Facilitate an understanding of susceptibility genes in disease: Genetic databases will enable us to study the impact of carrying certain variants of susceptibility genes on disease (eg impact on disease severity and age of onset) and can also record lifestyle variables which will greatly facilitate an understanding of the environmental factors which interact with susceptibility genes to cause disease.

4.  Identify genetic markers to predict responses to medicines: As genetic databases develop, they will include information about medicines prescribed and patients' responses. Comparing the DNA of patients who responded in a particular fashion (eg with an adverse event) with suitably matched controls who did not experience the adverse event, may enable the identification of DNA markers to predict that response.

5.  Provide a basis for healthcare providers and governments to estimate more precisely pharmacoeconomic consequences of healthcare and its management: By collecting information on the medical management of patients with disease over time and their outcomes, the value of particular medical interventions can be evaluated to provide important information to estimate the pharmacoeconomic consequences of healthcare management. Because of this tremendous potential in helping to deliver better healthcare, SB welcomes the proposal of the Wellcome Trust and the MRC to establish a new genetic database. We believe that it is highly appropriate for industry to be involved in this, and other, national initiatives, based upon our shared desire to deliver better healthcare and the expertise we have already developed in this area. Outlined below are SB's responses to those questions raised by the Sub-Committee based upon the principles and practices we adopt when collecting DNA samples in the UK.

What current projects involve collecting genetic information on people in the UK? What other projects are about to start? Are there collections of material (eg tissue samples) that could be used to generate databases of DNA profiles?

  6.  SmithKline Beecham (SB) has several collections of DNA samples for use in population association studies in the UK. These include nearly 3,000 samples from a cohort of women in the age range of 45-55 from the Aberdeen region who have had two bone mineral density scans at five year intervals. In addition, we have over 5,000 anonymised control samples from the same region. We have a population collection of individuals with schizophrenia from Scotland and London, and a collection of controls and patients with psoriasis from Stoke and Bristol.

  7.  At present all future plans for further sample collection are on hold until after the merger to form GlaxoSmithKline, as we need to compare our portfolios of samples before further studies are initiated. The DNA samples could feasibly be used to generate additional genetic databases.

Why are these genetic databases being assembled? How are these activities funded? What practical considerations will constrain developments? Are there alternative ways of fulfilling the objectives?

  8.  These collections are being used in the search for new disease genes suitable as targets for drug development and to validate existing targets using genetic approaches. The work is funded by SB as part of its Discovery Research programme.

  9.  The work is constrained by the time taken to find suitable clinical cohorts with a high quality of phenotypic data and ethnically matched controls that can be used in association studies. Finding academic centres willing and able to collaborate with industrial partners is also a constraint.

  10.  Our strategy requires sampling from the human population to search for genes relevant to human disease processes especially those related to an ageing population, for example type 2 diabetes and many late-onset neurological disorders. There are no direct alternatives to studying the diseases in human populations.

What is the genetic information that is being collected? How is it being stored and protected?

  11.  Depending on the project, we collect basic phenotypic information on individuals including age, gender, racial origin plus medical information necessary for the disease under study. This might include weight, height and blood pressure. This phenotype information is stored in a separate database from any genotyping data generated using the DNA from patients. All DNA samples are bar-coded and no patient information is available within the laboratory environment. No names or other personal identifier information are ever stored in SB. All information is stored on separate servers and can only be accessed through a secure password system. There are only a limited number of named individuals who are able to access the databases with phenotype information and the genotyping data and this only occurs at the analysis stage. SB acknowledges the requirements of the EC Data Privacy Directive in relation to the collection, storage and analysis of patient data.

How do the organisations involved see their responsibilities regarding privacy; consent; future use; public accountability; and intellectual property rights?

  12.  SB has considerable experience in performing clinical trials under very closely regulated conditions which are designed to protect patients from misuse of their personally identified clinical data. The process of prior consent and practices to ensure patient privacy and confidentiality are central to patient protection. SB is therefore very conscious of our responsibilities for any genetic data that we hold, hence the development of security systems, like the one described above, which make it extremely difficult to identify anybody individually. We treat all genetic data in the strictest confidence and implement, where necessary, appropriate organisational and technical measures against unauthorised or unlawful use of such data. Individuals participating in studies give their prior consent in writing and our procedures meet with all the current guidelines in the Helsinki and UNESCO declarations plus the appropriate European laws relating to scientific research involving humans.

  13.  All individuals have the right to withdraw from studies and are now also, following implementation of the EC Data Privacy Directive, able to decide if their samples can only be used for specific studies or for a secondary purpose, for example in a broader analysis of the disease process.

  14.  The mere identification of the sequence of nucleotide bases in segments of DNA contained in databases does not in itself represent a patentable invention. However patents on DNA (genes) may nevertheless be obtained in appropriate circumstances. This is clear from the extensively debated European Biotechnology Directive (98/44/EC), Articles 1 to 11 of which were implemented nationally in the UK on 28 July 2000. While controversial, patenting genes is entirely appropriate since the rationale for the patent system is to stimulate R&D investment. Developing treatments for presently incurable diseases using the promising but hugely costly genomics-based approach is only possible within a climate of strong IP protection.

  15.  To be patentable an invention must be novel, non-obvious and have utility. Finding, isolating and purifying gene segments associated with disease will in most cases represent inventive activity, deserving of patent protection.

  16.  It is sometimes argued that patenting of genes and other research tools stifles research: if a company obtains a patent on an important gene, no-one else can do research on the gene and important medical advances may be delayed. This is a misconception. UK patent law permits non-commercial research on patented subject matter, so pure research by academic institutions in the UK is not affected by the existence of patents. Even in relation to commercial research it does not necessarily follow that others are irrevocably blocked as it will often be possible to negotiate a licence under the patent—or challenge its validity. Finally, it is always open to third parties to obtain "dependent patents"—that is to say patent a new use for an already patented gene. The original finder of the gene could not then commercialise the new use without a licence under the dependent patent. This situation, which tends to stimulate cross-licensing, is inherent in the patent system. It is frequently encountered with pharmaceuticals and is in principle no different in the genomics field.

  17.  In this way, SB does not believe that the potential of bigger and more numerous genetic databases raises any new intellectual property issues for industry, academia or the NHS.

How do they see their activities in the area of genetic databases developing in the future? What advances in sequencing, screening and database technology are they anticipating?

  18.  It is clear that databases will continue to develop in the next few years as more work focuses on efforts to understand the mechanisms involved in the susceptibility to and progression of complex disease in an increasingly ageing population. There will be growing emphasis on obtaining specific quantitative phenotypic measures, for example biochemical markers, as surrogates of the disease process. This will increase the demands on individuals participating and lead to increased costs and calls for improved vigilance in the security of information obtained. The completion of the human genome sequence and the current efforts to identify single nucleotide polymorphisms will lead to increased efforts to find genes involved in the more complex disorders, for example degenerative diseases. We would expect to see DNA sequencing technologies advancing in the next 10 years, to a level that enables sequencing a substantial proportion of an individual's genome in a short time-frame. This will lead to the ability to diagnose diseases or the potential risks to health on an individual basis based on the variants occurring in a particular genome. Thus the goal of personalised medicine will start to be possible leading to increased demands on the health service for both screening and preventative therapies.

What lessons should be learnt from genetic database initiatives in other countries?

  19.  Countries such as Denmark have a long history of developing population-based databases for genetic epidemiology. They have one of the longest experiences in Europe of the issues around such databases of privacy and access to data. These started with the twin registries and have developed into more population-based systems of data and blood collection. The Danes have handled such issues with sympathy maintaining a balance between personal privacy as well as enabling science to continue. The UK should be aiming to mirror this approach with a non restrictive legislative environment which protects the individual while promoting the science.

  20.  Important lessons can also be learned from an example of successful industry involvement in a public-private approach to a genetic project: the SNP Consortium. This international research initiative, by all measures a successful collaboration, between industry (initially 10 pharmaceutical companies including SB) and the Wellcome Trust, contracts with academic institutions to perform the work that is placed in the public domain. It is successful because it represents the generation of fundamental pre-competitive information that also can be used by all parties for developing standards in order to create the research framework for the provision of better healthcare.

  21.  Closer to home, the British NHS represents a singular but under-utilised resource for population genetics, and healthcare informatics more generally. It has the potential to offer unparalleled access to areas of sample acquisition, such as across primary care, that is not possible in more fragmented health systems or in the smaller cohorts studies hitherto. A national structure could provide homogeneity of data acquisition that is essential for large-scale studies.

  22.  The likely benefits accruing to the NHS (and the UK as a whole) include:

—  Progress in understanding disease at the public health level.

—  Provision of new resources to support NHS R&D.

—  Stimulating production of novel therapeutics, diagnostics and the better targeting of treatment.

—  Attracting inward investment by companies.

  23.  Concerns have been expressed by some on industry involvement as a partner with public bodies in genetic database initiatives. There is often an implicit assumption that genetic research conducted by industry is, by virtue of its commercial nature, ethically questionable. In fact, clinical research studies performed in academic institutions by academic researchers using public funds can often discover data of commercial value. Consequently there is a significant ethical problem in that the original informed consents for such possible commercialisation may not have disclosed the possible financial consequence: informed consent for commercial uses should not be assumed to be implicit for academic investigators and institutions and explicit only for industry-supported research.

  24.  We believe that the other concerns expressed about industry involvement in genetic database initiatives can be assuaged by incorporating the best practice developed by companies such as SB:

(i) Protection of patient privacy and confidentiality and using the "opt-in" approach to participation based on informed consent.

(ii) Correcting the widespread misunderstanding that raw gene sequence information can be patented and that the patent holder in some way owns that sequence as it exists in individuals.

CONCLUSION

  25.  Success at a national level within the UK will undoubtedly require a radical strategy which seeks to identify and mobilise the appropriate scientific and clinical skills, to build large-scale computational infrastructure and to address public concern over many of the ethical issues touched upon in this paper in the context of medical privacy, use of anonymous data and consent issues. This will not be easy. It will involve significant funding and an unprecedented working relationship between the public and private sector. However, these are not insurmountable barriers. The basic infrastructure is already in place, as witnessed by the collaborations already underway, on a smaller scale, within companies like SB. What is now required is the political will to tackle the issue of public acceptability. To date the UK population has been willing to participate in clinical research. It is important that this willingness to participate is not lost by confusing issues of genetic modification of foods with advances in medical diagnostics and therapy.

4 October 2000