Following the launch of the Inquiry into this very important area and "call for evidence" we sought advice from our Fellows with the appropriate expertise and interest. A number of very important issues have been raised and we enclose copies of three responses from the following which are of particular relevance:

Professor Bernadette Modell FRCOG.
Professor Dian Donnai FRCOG.
Professor Patricia Jacobs FRCOG.

  1.  The haemoglobin genes are the best understood of all human genes, and detailed knowledge is available on the biological and clinical implications of variation in their sequence. Haemoglobin disorders (mainly thalassaemias and sickle cell disorders) are the commonest group of human single gene disorders. Screening for carriers of these disorders is standard practice in many countries. Antenatal and neonatal screening for haemoglobin disorders is included in the national plan for the health service.

  2.  The UK Register of Prenatal Diagnosis for Haemoglobin Disorders and the UK Thalassaemia Register are held for clinical and audit purposes. They are continually evolving, and the information they contain overlaps extensively. It is planned to merge them into a single diagnosis register in the future, if continued funding can be identified.

  3.  The objective of screening for carriers of haemoglobin disorders is to identify carrier couples (who may have a one in four risk of an affected child in every pregnancy), provide them with risk assessment and genetic counselling, and offer prenatal diagnosis, before they have had an affected child. Risk assessment and prenatal diagnosis depend on DNA analysis. Only three UK centres provide DNA diagnosis for haemoglobin disorders: University College London Hospitals (UCLH)[7], King's College Hospital (KCH)[8] and the Institute of Molecular Medicine at Oxford[9].

  4.  Excess DNA remaining following the diagnosis is routinely stored at these three centres for future reference. From time to time, samples from patients or groups of patients are examined for other DNA variants that may be relevant to the clinical care of patients, or counselling of at risk couples (eg polymorphisms adjacent to the globin genes, or insulin, collagen of haemochromatosis gene variants).

  5.  The UK Register of Prenatal Diagnosis for Haemoglobin Disorders is a collaboration between the three DNA diagnosis laboratories. The register includes clinical and DNA data on couples at risk, any affected children they already have, and all fetuses with a prenatal diagnosis. Thus the register is a guide to stored DNA samples from patients at all three UK diagnostic laboratories.

  6.  The UK Thalassaemia Register is a patient register held by Bernadette Modell, Maren Khan and Matthew Darlison at the Department of Primary Care and Population Sciences, Royal Free and University College Medical School. The register contains clinical data for almost all patients with a diagnosis of a major thalassaemia ever resident in the UK. It has become apparent that a detailed DNA diagnosis is highly relevant to the course and clinical management of the thalassaemias. Genetic data on haemoglobin gene variants in patients and their parents, derived from protein and DNA studies, is therefore collected from clinicians and the three UK diagnostic laboratories for inclusion in the register. It is planned to develop this aspect of the register in the near future into an ongoing nation-wide genotype/phenotype study—provided continued funding can be identified.

7.  The patient register is limited to thalassaemias because it is maintained with research funding: the aim is to demonstrate the potential of such registers for national service audit, research, and quality patient care. With appropriate funding, the register could be turned into a national register of patients with haemoglobin disorders, including the far larger (but still unknown) number of patients with sickle cell disorders.

  8.  Patient ethnicity is recorded in detail in both the above registers. Though patients with stored DNA have one or two thalassaemia or sickle cell mutations, they otherwise represent random samples of the different at risk populations in the UK. These DNA sample collections are therefore representative collections from UK population "not of Northern European origin". The registers provide a mechanism for tracking available samples.

  9.  Why are these genetic databases being assembled? The UK register of prenatal diagnosis for haemoglobin disorders aims to provide national service audit by following the number and origins of referrals, and to maintain the highest possible laboratory standards by fully sharing information.

  10.  The UK thalassaemia register aims to promote best possible care for a very unevenly distributed group of patients with a rare (in the UK) genetic disorder with complex management. The register is in contact with all clinicians with thalassaemic patients (including 76 with only one such patient). It both regularly requests information on patient status, and provides a service by circulating reports, research results, clinical guidelines and patient information materials directly to the clinicians involved.[10]

  11.  Taken together, the registers demonstrate a powerful mechanism for national audit of services for both treatment and prevention of a genetic disorder, as they include information on every conception affected by a major thalassaemia in the UK, whether the outcome is a live-birth or a termination of pregnancy.[11]

  12.  How are these activities funded? The UK register of prenatal diagnosis for haemoglobin disorders was set up with a short-term audit grant from DoH, but now has no special funding. All current work is funded by the Wellcome Trust as part of their support for BM as a Principal Research Fellow.

  13.  What practical considerations will constrain developments?

  14.  Funding. Wellcome research funding has now ended (a) because BM has reached retirement age, and (b) because the Trust considers that the registers have shown their value for service audit and R&D, and should be transferred to the NHS.

  15.  Lack of recognition within the NHS of the need for national audit mechanisms for services for genetic disorders, and so of dedicated long-term funding.

  16.  Lack of NHS clinicians trained in the public health aspects of medical genetics.

  17.  Lack of training and support for NHS clinicians to manage electronic databases. This type of work requires some senior physicians trained in public health aspects of genetics, with formal national responsibility, and supported by cutting-edge informatics—that is, recognition of the discipline of community genetics.

  18.  Rapid turnover and low pay for NHS clerical staff makes it very difficult to obtain and retain staff capable of consistently entering data into databases—even when this is held as a patient record.

  19.  Are there alternative ways of fulfilling objectives?

  20.  There is no alternative approach for collecting national data on phenotype/genotype relations or conducting national service audit. We see automated data aggregation within electronic records as a more cost-effective approach for tracking availability of genetic samples and genetic information.

  21.  The genetic information being collected is restricted at present to variants in the regions of DNA where globin genes are located, and in some cases on other genes eg haemochromatosis, osteoporosis, HLA type and diabetes, where variation may be relevant to the clinical management of the patients.

  22.  The data is stored (a) in the patient's notes, (b) as paper records in locked filing cabinets and (c) in a password-protected electronic database. Data is accessible only to register staff. There are formal conventions on who the data can be released to, and in what format.

  23.  The organisations involved: how they see their responsibilities regarding privacy; consent; future use; public accountability; and intellectual property rights? The organisations are University College London Medical School, University College London Hospitals, King's College Hospital and the Institute of Molecular Medicine at Oxford (NHS-supported unit within an academic unit).

  24.  Privacy is considered the same as for medical records.

  25.  Consent procedures vary at the three institutions. The prenatal diagnosis consent form used at UCLH and KCH contains a clause giving broad consent for samples to be used for research related to treatment and prevention of haemoglobin disorders. Consent at the Oxford centre is obtained by the clinician sending the sample for analysis, and is unlikely to include broad consent for research. Consent for data to be held on the UK Thalassaemia Register is via the clinician caring for the patients.

  26.  Future use—we see existing permission as giving consent for use of patient records and DNA samples for the benefit of the family or for patients in general. For permission for other research uses of the samples or data, we visualise that we should go back to the patient or the family. Families in the UK prenatal diagnosis register could be contacted directly by the UCLH or KCH centres, which have direct patient contact, or through their doctor by the Oxford laboratory. Families in the UK Thalassaemia Register could be contacted through their doctor.

  27.  Public accountability. The existence, methods and objectives of the registers have been widely publicised, and are approved by the UK Thalassaemia Society (the patient support association) and the UK Forum on Haemoglobin Disorders. We share information arising from the registers with doctors and patients, and have an ongoing consultation with stake-holders.

  28.  Intellectual property rights. We visualise that these are owned by the curators of the data, UCL, UCLH and the Wellcome Trust (though there is some vagueness here).

  29.  How will our activities in the area of genetic databases develop in the future? What advances in sequencing, screening and database technology do we anticipate?

  30.  In the future, genetic testing and screening will become far more widespread, and significant aspects of health care are likely to be based on it. Our experience shows that the genetic information should be mutation-specific to be fully useful at the clinical level.

  31.  We therefore anticipate widespread use of an electronic health care record that will be able to handle genetic and family information properly—and will have the potential to link directly with genetic databases, for audit purposes.

  32.  Information is the therapeutic intervention in predictive genetics. It will therefore become necessary to provide everyone with a genetic diagnosis, with mutation-specific information. We therefore visualise large-scale linkage between databases of mutation-specific information including appropriate patient information materials, and electronic health care records, to underpin appropriate service delivery.

  33.  Registers of genetic diagnoses will also be needed to audit delivery of, and patient responses to, genetic information.

  34.  In conclusion, haemoglobin provides the only presently available model of the clinical application of precise genetic information on a population scale. The two registers described here demonstrate the integral importance of precise genetic databases for patient services.

  I can give evidence from two perspectives, namely as director of a Regional Genetics Laboratory that provides both cytogenetic and molecular diagnostic services for a population of approximately 2.5 million and as an established researcher involved with a number of human genetic research projects many of which are population based.

  1.  Diagnostic laboratories are sent specimens from patients most of whom are suspected of having a genetic abnormality detectable by cytogenetic or DNA analysis. Our laboratory handles about 8,500 diagnostic specimens a year and of these about 5,000 have DNA extracted and stored for an indefinite period. In most of the remaining specimens, fixed cells are stored for a number of years from which DNA could be extracted, albeit with difficulty. Similar tissues must be stored in pathology laboratories from which DNA could be extracted. For a substantial minority of the diagnostic specimens we have detailed pedigree information.

  In the course of our research we have obtained collections of DNA and/or information from patients with a variety of conditions including mental retardation, behavioural abnormalities, diabetes, reproductive difficulties, cytogenetic abnormalities, specific malignancies etc. We also utilise DNA provided by a large regional cohort, namely the Avon Longitudinal Study of Parents and Children (ALSPAC).

  I presume our experiences as a diagnostic and research laboratory are similar to many others in this country and together they present a large (and often overlooked) collection of DNA and genetic information on the British population.

  2.  These genetic databases are assembled for diagnostic purposes, funded by the NHS, or for research purposes, funded by research councils and charitable organisations. I cannot see any reasons why the development of these types of genetic databases should be constrained, nor do I see any alternative ways of fulfilling the diagnostic or research objectives that underlie their development.

  3.  As well as DNA, tissues and cell suspensions, we collect information on the immediate or extended family, relevant clinical information and the results of the various tests being undertaken. The data are stored in computers and are protected by restricted access, safeguarded by appropriate codes, and the professionalism of our staff.

  4.  The privacy of our laboratory databases is sacrosanct and information identifiable with a patient is given only to other diagnostic laboratories or clinicians on an individual basis where the information is necessary (i) to make a genetic diagnosis (eg part of a family being investigated by two different laboratories) (ii) to undertake an audit or (iii) to conduct ethically approved research. Furthermore such information is only provided with the written permission of the referring doctor. Written consent is obtained from all research subjects and all research protocols are passed by the relevant MREC or LREC.

  We use diagnostic specimens in two ways: (i) to diagnose conditions for which the specimen was submitted, (ii) anonymously as control material for other investigations. We use research specimens only (i) to undertake research on the condition for which the specimens were obtained or (ii) anonymously as control material for other investigations. We plan to continue to use the stored specimens ensuring privacy protection for the individuals from whom they have been obtained. Neither public accountability nor intellectual property rights have been issues with these types of databases but could be addressed if necessary.

  5.  I foresee the continuation of these types of diagnostic and research databases as inevitable given the type of work that we undertake. I do not anticipate that advances in sequencing, screening or database technology will substantially alter the nature of our databases, save to make our work more streamlined and perhaps easier.

  6.  I am unaware of genetic database information in other countries that will be helpful to us, in maintaining or expanding our own databases.

Patricia A Jacobs DSc FRS

Director

  The inquiry is of some concern to those of us involved in providing clinical services to individuals and families with, or at high risk of single gene disorders. There is a risk that DNA samples collected and stored from and for these families might be caught up in regulations developed because of the public concern about large scale collections of DNA where the aim is to look for genetic polymorphisms which might predispose to common diseases or to responsiveness to particular medications.

  To amplify the above I would make the following comments:

1.  Collections that are part of the infrastructure of clinical services

  Many genetic service laboratories have large DNA banks linkable to clinical information from patients and families referred to genetic services. Samples may be from:

(a)  Individuals known to be affected with a genetic condition where a molecular diagnostic test has been undertaken and an aliquot for the sample stored.

(b)  Samples from individuals with a genetic condition where scientific knowledge or NHS service development does not allow molecular diagnosis currently. When the relevant gene for their condition is identified other family members may want testing to see if they have inherited the family mutation; without material from an affected person it is impossible to interpret negative mutation screening results.

(c)  Samples from individuals with undiagnosed abnormalities (particularly babies with congenital abnormalities who may not live) stored in the hope that future research may allow a definitive diagnosis and hence accurate counselling for the family.

  All these collections are usually funded as part of the NHS service and linkable to NHS notes on the individual with the usual NHS arrangement for access. Formation of such collections is not subject to research ethics approval. In most centres consent forms for DNA analysis and storage are not completed since they are regarded as any other blood sample taken for clinical investigation and care. However this may change, particularly with regard to storage.

2.  Research collections of material from people with specific genetic conditions

  Very many individual research projects depend on collecting modest numbers of samples from people with a particular condition and often from unaffected relatives. These projects will normally require approval from a research ethics committee which will cover questions of use, access, confidentiality etc and will deal with the issue as to whether subjects or their GP will be informed of their test results. Contributions to such research will come from service departments who often have samples from individuals and families with rare genetic conditions (see above) or the researcher will seek out suitable subjects and ask them to take part eg through patient support groups.

3.  Large scale collections of samples used for prospective studies

  In general these studies will be population based and will be looking for genetic polymorphisms which may predispose or be protective with regard to the development of later-onset common diseases such as diabetes, hypertension and coronary heart disease. Similar studies may be done to identify factors which may affect drug response. Such studies would be without value if clinical outcomes were not linked to samples. Studies such as ALSPAC (the Avon longitudinal study) have identified a cohort of children with pregnancy, birth and later follow-up data. These sorts of studies are usually not undertaken by clinical genetic centres who, at the present time, focus more on those disorders which are due to single genes of high penetrance.

Professor Dian Donnai MB FRCP FRCOG FRCPCH
Consultant Clinical Geneticist

25 September 2000

8   Former head, Dr Mark Layton, Dept of Haematological Medicine, King's College Hospital. Back

9   Head of prenatal diagnosis laboratory, John Old, Institute of Molecular Medicine, John Radcliffe Hospital. Back

10   For example, survival data collected at the end of 1998 showed that patient survival has improved far less on a national level than had been anticipated from results obtained at specialist centres. This "early warning" information, together with advice on patient referral, was circulated to all doctors on the register at the end of January 1999. It was not published (in the Lancet) until June 2000. Fewer deaths have been reported for 1999 and 2000 to date. Back

11   For example, a collaboration of the registers with the National Confidential Enquiry into Genetic Counselling showed wide regional variations in the quality of national audit of quality of carrier screening and counselling, with a selective failure of service delivery to British Asians. Back