1. What current projects involve collecting genetic information on people in the UK?

  There are a vast number of projects ongoing in academic and commercial laboratories which involve the collation of genetic data on people in the UK. Most of these studies are focused on identifying genes involved in particular diseases and involve the collection of both patient populations and families where an individual in the family suffers from a particular disease. Individually these are relatively small and focused in their structure, although some of them involve studying patient populations in excess of 10,000 individuals. In all cases the DNA being tested has come from people who have provided informed consent for the project. Obviously these collections of DNA could provide the material for broader DNA databases encompassing information about genetic variation at a large number of loci around the genome. This is unlikely to occur as these collections are in the hands of hundreds, if not thousands, of different scientists, the quality and quantity of DNA is variable and on the whole patients have not been consented for such a broad approach to genetic characterisation.

  In the future it is likely that large patient cohorts will have DNA collected and be consented for the general characterisation of DNA variants and their correlation with a whole range of potential phenotypes. This will occur both in large clinical trial populations and within prospective cohorts. I believe that most of the major pharmaceutical companies are now collecting DNA from patients undergoing large clinical trials. Academic laboratories undertaking clinical trials also are more routinely collecting DNA samples and there are a number of large projects that involve the characterisation of hundreds of thousands of individuals in the UK whose DNA could potentially be characterised and correlated with diseases that they develop. The purpose of these studies is to understand more accurately the relationship between genes and environment but also to understand individual variations to drug response. Funding for these activities comes from all sectors, both the commercial sector and the public sector including the Research Councils, the Wellcome Trust, the Cancer Charities, the British Heart Foundation, the British Diabetes Association and the Arthritis and Rheumatism Council.

  2.  The major practical considerations are the efficient collection at a very large scale of hundreds of thousands of samples and their characterisation by genotyping in a systematic and efficient way. These are logistic issues seldom confronted in the academic laboratory setting and require careful and thorough consideration. If we are to understand the genetic basis of common disease and in particular understand the relationship between genes and the environment in producing these disorders we will have to undertake such studies.

3.  What is the genetic information that is being collected?

  The genetic information is on the whole DNA polymorphism either in genes or in flanking regions. The variation that exists is not necessarily predictive of disease onset for several reasons. Firstly, the contribution of a single DNA variant to disease susceptibility may be modest (relative risk of two to three). In addition the DNA variant being studied may simply be in linkage disequilibrium with the disease variant further reducing the predictive value of such a test. On the whole the genetic variation being characterised will contribute only to a better understanding of risk of common disease rather than identifying causative genes which are highly predictive.

  The genetic material is normally stored with coded patient information. Generally, however, patients can be identified from these databases if necessary. The coding of patients in most databases would make it difficult, however, for an individual outside the study to readily correlate a genotype with a patient name.

  4.  Most of the organisations involved have been careful about obtaining full consent from patients for the use of their DNA. On the whole patients are anxious to know the cause of their disease even if it is only to help future generations who might suffer and hence are willing to support such research activities. Any single individual's DNA contributes virtually nothing to an intellectual property base. Firm conclusions about disease genes are the result of the very large numbers of samples being analysed and the characterisation of populations with and without disease. It would be difficult if not impossible to unpick the contribution of any single individual to such a complex activity as disease gene identification; as a result this issue has been seldom raised.

5.  The future

  The ability to characterise large amounts of genetic information efficiently and cheaply is likely to increase dramatically over the next five to 10 years. Although the amount of data being generated is likely to increase dramatically in the end the utility of that data is totally dependent on the analysis of the available information. One major threat is the premature and statistically naive approach to identifying disease genes with relatively small contributions to risk based on multiple testing protocols that are not clearly replicated in several populations. This has already substantially confused the genetic literature with numerous false positives and there are serious risks that this will endanger the reputation of the field over time.

  6.  The other major genetic database initiatives which are identifying and following large cohorts of patients over time include the Iceland Cohort and the Estonian Cohort. The Iceland Cohort has problems in that it is severely under powered from the nature of chronic diseases of adult life. The Estonian population is larger but it is not clear to me how useful the medical follow up data in the Estonian Health Service is likely to be for the follow up of these patient cohorts. The Icelandic model has been an entirely commercially driven activity and this has created anxieties in the population and internationally about the ownership of individuals' genetic information. Care must be taken not to pursue this approach in the UK. The critical issue for UK prospective databases should be the requirement to separate genotype phenotype analysis from the names of the patients and clearer patient identification. By having a firewall between these two activities most patient anxiety should be allayed. Consent also is critical and in order for these to be studied public confidence that these studies are productive and informative needs to be sustained. The benefits of genetics need to be delivered over the next five years to ensure that such studies can be continued with public confidence into the future.

John Bell MADM FRCP
Nuffield Professor of Clinical Medicine
University of Oxford

7 August 2000