Genotype discrimination: the complex case for some legislative protection.

AuthorGreely, Henry T.

INTRODUCTION

Substance and process, process and substance. They dance through law; they dance through life. The big news from the dueling publications of "the" human genome was that humans have only about one-third more genes than round worms; we share versions of about ninety-nine percent of our genes with mice. The substance of the genes is important to the creature they produce, but so is when those genes are turned on, how high, and for how long. In genomes, the expression is as important as what is expressed. The substance of legislation is important; so is the process--the public discussions, debates, and understandings--through which it is made and implemented. The federal government should prohibit some uses of genetic information in decisions by employers and health insurers, but how those prohibitions are argued and implemented may be more important than the fact of their adoption.

This piece begins, in Part I, by demonstrating that genetic discrimination is not likely to have serious effects on a substantial number of people. It then points out, in Part II, weaknesses in both many of the justifications for legislation banning genetic discrimination and many of the methods suggested for doing so. It ends, with Part III, by supporting carefully crafted, and narrowly argued, federal legislation limiting genotype discrimination. There is a good case for such laws, but there just is not a case that is both good and simple. And to argue the case without using the shades of gray, to debate point by (counter) point, can do more harm than good.

  1. THE LIMITED REALITY OF GENETIC DISCRIMINATION

    Genetic discrimination is a much greater threat in people's fears than it is in reality, today or in the foreseeable future, for both scientific and social reasons. Most of the discussion that follows concerns the science behind fears of human genetics; I have tried to make it easily understandable to legal audiences. An accurate understanding of the power, and the limits, of genetic prediction is crucial to a useful analysis of genetic discrimination.

    1. Genes and Disease

      Start with some definitions. People do not have "genes" for disease. As far as we know, all humans have the same set of genes, about 32,000 of them,(1) except for those few genes on the Y chromosome, found in men and not in women.(2) Those genes come in many variations; indeed, for the most part, the human genes are just human variations of genes found in other primates, mammals, animals, or broader sets of life forms. Any two humans, on average, will be identical in their DNA sequences 99.9% of the time and 99.99% of the time in the regions of genes that contain the genetic code for constructing proteins (the so-called exons). The few differences make up genetic variations or different "alleles" of the genes. Traits or diseases are considered "genetic" when a person with a particular allele has a much greater chance than average of having a particular trait or disease--more broadly, when the genotype is correlated with a particular "phenotype."

      Progress in human genetics has led to fears about genetic discrimination, particularly in insurance and employment, because of the perceived power of human genetics to make predictions about people's future lives--and hence future health insurance risks, time of death, and employment productivity. The popular vision of a genetic disease is something that is caused only by a particular allele, that cannot be avoided by a person with that allele, that cannot be treated, and that leads to an inevitable death. And, indeed, sometimes human genetics can lead to predictions about people that are powerful, both because they are highly likely to come true and because they would have substantial consequences on those persons' lives.

      Huntington disease,(3) neurodegenerative disorder, may be the paradigm for this kind of genetic condition. It is only the product of a genetic variation. As far as we know, the only way to get Huntington disease is to have an excessive number of repeats of three bases of DNA--cytosine, adenine, and guanine, or CAG--in a particular stretch of a gene found near the end of the short arm of Chromosome 4. Most people have 10 to 20 repeated CAGs; only people with 36 or more ever develop Huntington disease. No one with more than 38 repeats is known to have avoided the disease, except by dying first from something else. There is no significant treatment for Huntington disease; death follows inevitably after about fifteen years of progressive physical and mental disability.(4) Many childhood genetic diseases, like Tay-Sachs disease, are similarly implacable.

      But Huntington disease and other certain and inevitable links between genetic variations and disease are proving to be the exceptions, not the rule. In some genetic diseases, the physiological defect may follow inevitably from the genetic variation, but disease and death may not. For example, phenylketonuria ("PKU"), a genetic disorder affecting about one birth in 10,000, is caused by the inability to metabolize the essential amino acid phenylalanine. This leads to the build up of excessive amounts of phenylalanine, which then causes mental retardation. If a child is known from an early age to have PKU, however, an arduous diet, low in phenylalanine, prevents retardation so effectively that PKU screening at birth is now required in all U.S. states.(5) Even some of the more common genetic diseases--cystic fibrosis, sickle-cell anemia, and beta thalassemia--although not curable, are now susceptible to treatments that have expanded the quantity and improved the quality of the lives of those born with defective versions of the responsible genes.

      Even more importantly, many alleles associated with nonrare diseases, particularly common diseases, increase the bearer's risk--but not all the way. Having one copy of disease-related alleles of either of the genes BRCA 1 and BRCA 2 increase a woman's chances of being diagnosed with breast or ovarian cancer. Her lifetime risk of breast cancer increases from roughly 10% to somewhere between 50% and 85%. Her risk of ovarian cancer rises from about 1% to about 30%. (These ratios, the percentage of those with a given genotype who develop a particular phenotype, are known as the genotype's "penetrance.") No one knows yet what determines which women with these mutated alleles get the disease--variations in other genes, environmental influences, or just bad luck. It is clear, though, that not all women with these alleles get either disease; most women with the alleles do not get ovarian cancer.

      There clearly are genetic variations, still largely unknown, that are associated with higher (and lower) risk of asthma, diabetes, coronary artery disease, stroke, schizophrenia, and a host of other common diseases. But, for the most part, the change in risk associated with any given allele seems likely to be quite small. This should not be surprising. The strong associations between genetic variations and disease--the alleles with a high penetrance--are the easy ones to find. A gene for asthma with the near-perfect penetrance of Huntington disease, or even the penetrance of BRCA 1, would stand out in family studies and would be easy to find. More genetic links to common diseases are probably proving hard to find because they are not very powerful.

      For common diseases, the genetic story is likely to be extremely complicated. Alzheimer disease provides a useful example. An American has about a 10-15% chance of being diagnosed with Alzheimer disease, usually after the age of sixty-five.(6) Three genes--presenilin 1, presenilin 2, and the amyloid precursor protein gene--have now been found where unusual alleles lead almost certainly to the disease, and usually several decades earlier than usual. Mutations in presenilin 1 may be found in as many as 1 person in 1000; the other two alleles are vanishingly rare. Altogether, they may account for about 1% of the people who get Alzheimer disease.

      Another gene, APOE, is also associated with Alzheimer disease. Everyone has two copies of the APOE gene, one inherited from each parent. This gene has three common alleles, called APOE 2, APOE 3, and APOE 4.(7) The APOE 3 allele is the most common, making up about 70% of all the versions found. People who inherit two copies of the APOE 4 allele--about 2% of the population--have a very high risk of getting Alzheimer disease, although at the usual age. People with one copy of the APOE 4 allele and one copy of either APOE 3 or APOE 2 make up about 15% of the population. Their risk of Alzheimer disease is somewhere between 30% and 50%, 2 to 5 times higher than the general population risk. But someone with one APOE 4 allele is still less likely to be diagnosed with Alzheimer disease than to be diagnosed with it. On the other hand, a person with two APOE 2 alleles, about 1% of the population, has a very low chance of being diagnosed with Alzheimer disease.

      Thus, the same disease is strongly genetic for about 2% of the population, weakly genetic for about 15% of the population, and, genetically, nearly ruled out for about 1% of the population. The other 7/8 of the population, which will account for most people with Alzheimer disease, have, as far as we now know, risks that are neither increased nor decreased significantly by their genes. While other alleles of other genes are under investigation for association with Alzheimer disease, the logic of the discovery process means that the strong associations are likely to be rare; the less-rare associations are likely to be weak.

      The relationship between genetic variations and disease is thus complicated. For some few, unlucky people, possession of a particular allele is a very strong predictor of disease, which may or may not be treatable. Most people, though, are likely to have much less powerful genetically predicted disease risks. For example, my particular genetic variations might make...

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT