Ethical guidelines urgently needed for collecting, processing, using and storing human genetic dataParis/Montreal - Human genetic data is everywhere. Its presence in biological samples (of blood, tissue, saliva and sperm etc.) provides answers to questions from police and courts of law about paternity or the identity of sex offenders.
Scientists are also using it to detect diseases we may be programmed to get, thus raising the hope for new cures.
Genetic data use has produced spectacular results. Montreal University's magazine Forum reported (January 29, 2001) that 76 people on death row in the United States had been cleared of their crimes since 1987 after looking at their genetic imprint. But many are worried about the risks involved in some of these promising uses. Jacques Testart, the "father" of the first French test-tube baby, said in his book Des hommes probables: (Of Probable Men) "It is unclear whether modern societies can avoid new abuses, such as genetic racism which would be science-based, replacing racism based on skin colour or ethnic origin."
In this fast-developing field, the need for ethical guidelines is obvious and UNESCO, which drew up the Universal Declaration on the Human Genome and Human Rights (adopted in 1997) has since 1993 encouraged debate on the issue through its International Bioethics Committee (IBC), which has produced two reports - Confidentiality and Genetic Data (June 2000) and Human Genetic Data: Preliminary Study by the IBC on its Collection, Treatment, Storage and Use (May 2002).
The IBC's Drafting Group has for the past few months been drawing up an international instrument on human genetic data. The entire Committee will discuss the outline of the text for the first time at its 9th session in Montreal (Canada) November 26 to 28. Further consultations will follow and a final version will be presented for adoption to UNESCO's General Conference next autumn.
We may not always realize it, but the use of human genetic data is becoming an ever larger part of our lives, and not just where medicine is concerned. Some countries - such as Iceland today, Estonia soon and Latvia and Tonga a bit later - have undertaken genetic data censuses of their populations. Genetic databanks are springing up all over the place, ever more numerous and bigger. Their number is unclear, since nowadays even very small hospitals have a stock of processed genetic data or at least a collection of DNA samples waiting to be processed. There are thought to be about 50 databanks with more than a million items each.
Why collect so much? So we can see the structure of genetic sequences in given populations and variations among individuals and different populations (genetic polymorphism), the scientists reply. Both basic and applied research require such data to help explain developments or open the way to new treatments. To understand diseases, such as diabetes for example, scientists need samples from families where they occur particularly often. They may also choose to look at samples from groups that are unaffected by it.
Doctors are prescribing more and more genetic tests nowadays, not just when taking decisions regarding reproduction. Tests on adults are increasing and the results are swelling the mass of data kelp by hospitals and research centres. These tests are mainly used to study genetic mutations that can trigger diseases. For example, genetic factors are involved in 5 to 10% of cases of breast cancer, which affects nearly one woman in 11, and the mutant BRCA1, 2 and 3 genes have been identified as the culprits. While their presence does not guarantee getting the disease, it represents a significantly increased. In the United States, thousands of women have asked their doctors for such screening.
Industry is naturally interested in human genetic data as well. The legal battle between several European institutions, including France's Institut Curie, and the US firm Myriad Genetics shows this. It concerns screening for breast cancer and ovarian cancers both of which are linked to the presence of the BRCA1 gene. The Europeans are challenging Myriad's patents that give it an unofficial monopoly. The Europeans also say that because the firm refuses to grant manufacturing licences, all DNA samples will have to be sent to the Myriad Genetics headquarters in Salt Lake City for processing, providing the company with a unique databank about people at high risk.
The interest of the pharmaceutical industry is going to grow, since some people react badly (or not at all) to some drugs (20% of high blood pressure patients, for example) and firms hope to discover all the genetic differences between people that explain this. Perhaps one day we will have genetic tests so we can receive optimal treatment. If customized medical treatment is to become possible, it will be thanks to the genetic databanks currently being built up.
Moreover, genetic studies of entire populations are underway. They involve the collection of vast quantities of data assembled to show up the incidence of certain genetic traits, and also for non-medical uses by police and the courts. Even governments are increasingly interested in such data and some have already started a genetic census of their population, from which they can gain vital information that can help them improve their public health policies. If such a census finds a strong presence of a genetic mutation causing cardio-vascular disease, for example, a prevention programme can be launched and priority given to treatment, notably by funding research.
The stock of human genetic data is sure to continue increasing. So we have to think about possible misuses. At all stages - whether collecting, processing, using or storing - ethical problems arise and some of the abuses that would threaten human rights and human dignity can easily be imagined.
At the collecting stage, there is the problem of consent, which is not new to the medical profession. "Free, informed and express" consent is not always self-evident. Suppose researchers in rich countries decide to obtain raw genetic data from people living in countries with less developed economies and legal protection systems, with no legislation about genetic data or even basic information about it, what kind of consent can they give?
There is also the issue of withdrawing consent and the right of a donor to be informed or not (dealt with in article 5c of the Universal Declaration on the Human Genome and Human Rights). Some people might not want to have their genes tested or know the results. Someone whose father has developed Huntingdon's chorea, a neurological disease that affects one person in every 10,000, has an even chance of developing it too. A test for it exists but those taking it might want to avoid at all costs the trauma of a bad result for them and their families. Might they be practically obliged to take such a test if they wanted to take out life insurance, for example? The insurance company might ask them either to produce negative test results or to pay a much higher premium.
At the processing stage, confidentiality is the major problem. Should the samples or data be tagged with the person's name? If so, how will confidentiality be maintained? Some data (such as police specimens but also genetic tests) needs to be identified or identifiable in this way, yet everything must be done to respect confidentiality. Quite apart from possible abuses, we have to recognize the dilemmas. Anybody taking a test must be able to rely on confidentiality but family members could invoke their right to access the test results because they might be affected by the data. This dilemma could be particularly hard to resolve if the first symptom of the disease in question were blindness, for example, and one of the relatives of the person tested were an air traffic controller responsible for other people's lives.
We should be able to take for granted the reliability and quality of processed data in view of the professional standards practitioners, but there is a special risk when it comes to research concerning behaviour genetics. Just as some scientists once believed they could identify criminals by the shape of their skulls, some today might be tempted to track down genes for such things as crime, alcoholism or violence. The media are usually keener on this kind of shortcut than scientists, but some studies have already proved controversial. In fact the problem is less with the data itself than with its interpretation, but the danger of looking for "high-risk" individuals or hunting for "bad genes" is too great to be overlooked.
At the using stage, the first question is what for? The IBC Drafting Group's outline only spells out the purposes of medical and scientific research, health care, forensic medicine and the needs of the judiciary in civil or criminal proceedings. The use (even just the collection) of genetic data to discriminate against a section of the population is not acceptable, any more than is use or collection for eugenic purposes.
But "racial" discrimination is not the only kind of discrimination and genetic data can be used to discriminate in other ways. For example, a bias against employing certain people might be presented, at least ostensibly, as based on very sound motives. An oil company might ask its workers to take a test to identify those running a particular risk of falling ill because of contact with solvants. Such screening fits nicely into preventive medicine in the interest of the employee. But the test - suggested or imposed, the line is a thin one - can put some candidates for a job at a disadvantage. Employers more interested in productivity than health or justice might exclude employees sensitive to certain products from services where these substances are used in order to reduce financial losses due to sick leave.
Some people suspect the members of the insurance industry of wanting to introduce such discrimination. Genetic tests are a godsend for insurers because they can enable them to weed out "bad customers". The great temptation is to ask life insurance applicants to be screened for certain illnesses and penalize (through higher premiums) those who refuse or who turn out to be at risk for certain illnesses. This means that some will be lucky in the genetic lottery - with little likelihood of being ill and thus benefiting from low premiums - while others will lose out. As well as being likely to develop illness, insurers will treat them as pariahs.
This may not yet be the case, mainly because tests are only available for a few illnesses and those that are available are not always reliable and produce only strongly indicative results (except in the case of Huntingdon's chorea) with no proof that a person will develop the disease. However, except in a few countries where restrictions have been imposed (Austria, Belgium, Norway, the Netherlands and parts of the United States), the danger does exist. In some countries there is no legislation about such practices and in some cases insurers have decided themselves not to use genetic testing for the time being. Though their moratorium is, in general, very short term, usually five years only. They may use genetic tests in the future, especially once tests become more accurate.
When speculating about possible scenarios, we can imagine what might happen in education. At first, it would obviously be in children's interest that school doctors took genetic samples. The next step would be to try to use the data to steer children through their schooling. After the "crime gene", would there be a hunt for the one for mathematics? Schools eager to score high could even devise an admissions policy that favoured "geniuses," who would be defined as those able to prove they had the "best genes."
There is also the problem of the time element in collecting human samples as collections might come to be used for new ends. If genetic research was needed on lung samples from victims of a century-old flu epidemic, who would give permission for the samples to be used in a way not originally intended and obviously not authorized at the time? This possibility is considered in the outline the Drafting Group will present in Montreal. The text proposes to allow such use without consent on condition that the samples are of undeniable interest for medical and scientific research or for public health and that the anonymity of the data providers is made irreversible.
The Drafting Group thinks that as a rule one purpose should not exclude another. Human genetic data collected for scientific and medical research, health care, forensic medicine and judicial purposes in civil or criminal proceedings "can be used for another purpose", the Group's version says, "as long as prior, free, informed and express consent of the person concerned is obtained again."
Use of genetic data implies expected beneficial results. At the collection stage, the problem of "shopping" for genetic data in poor countries by researchers and firms in rich countries has been raised. This issue of fairness reappears at the results stage. If genetic data gathered from a group of people is used to the advantage of the acquirers in the form of tests and new drugs, should the "donor" population not also benefit from this?
There are also difficult issues at the storing stage. What kind of human genetic data management can ensure adequate protection of human rights and basic freedoms? Which samples should be kept and which discarded, especially in the case of those collected in a criminal investigation or lawsuit?
Cross-linking data presents the storage aspect of the purpose problem already mentioned. Since one purpose should not entail another, cross-linking, in some case must not be allowed. The Drafting Group's version says that "human genetic data collected for scientific and medical research and health care shall not be cross-linked with data collected for judicial purposes in civil or criminal proceedings." There can be no question of using data gathered for medical purposes, or even in a national genetic census, to track criminals.
The very technical issues raised by genetic data that the IBC will consider in Montreal are very complex, largely because such data groups both medical and personal information. On the one hand, people's genetic traits can determine or predispose them to many illnesses, on the other, as each person's genetic imprint is unique, forensic scientists and the law can use it for identification purposes. All this data is sensitive yet long-lasting (remaining relevant throughout a person's lifetime). It also provides information about siblings, descendants and more broadly about the population group the person belongs to.
In view of this complexity, the IBC Drafting Group chose at its first three meetings to take a practical approach. It opted for a non-binding instrument - a declaration rather than a convention - the better to adapt to a constantly changing environment and to reach consensus more easily. The discussions in Montreal should allow for further progress.
Contact: Pierre Gaillard
Bureau of Public Information, Editorial Section
Tel.: +33 (0)1 4568-1740
During the IBC meeting: +33 615 695 372
For photographs: Ariane Bailey
Tel: +33 (0)1 45 68 16 86