Winding Your Way Through DNA Symposium
San Francisco, California
Saturday Afternoon, September 26, 1992
Neil Holtzman, MD, MPH
Neil Holtzman, MD, MPH, Professor of Pediatrics at Johns Hopkins
Thank you, Rick, I suppose.... You've given the beginning of my talk,
so the audience will bear with me if there's some repetition.
Last night you heard that having a detailed map of his route across the
United States enabled David Botstein to find the wallet he had
inadvertently dropped out of an airplane east of Cheyenne. The scenario
may be improbable but the point David made was not. A detailed map of
the human genome will enable scientists to find all of the genes, all of
our genes. This is the goal of the Human Genome Project as
you've heard from Dr. Watson. It will increase our understanding of how
the body works and what goes wrong in genetic diseases.
At the moment the path for applying that information to improve the
outlook for those born with genetic disorders is more like a maze than a
map. A maze, like the one used in animal experiments, in which hazards
shock and are more plentiful than the reward. This afternoon I will
spend the first part of my talk considering what we can expect and not
expect from the Genome Project in improving the outlook for human
diseases. Then I will turn to the hazards of genetic testing to
individuals and society.
Let me begin, and here's where the repetition comes in, by describing
the two categories of disease in which genes in the germ line, the eggs
and sperm, play a role. First, diseases that play a role when a single
gene goes awry, as explained by Dr. Singer this morning, we call single
gene or Mendelian diseases, for the monk, Gregor Mendel, who first
discovered discrete units of inheritance. Second, diseases whose
occurrence depends on the presence of more than one gene, called
multi-factorial disorders. (Dr. Singer's ras story deals with somatic
cells in genetic diseases, you might consider cancer in that way, not
germ line disease and I will not discuss the somatic diseases further.)
As you've heard from Rick Meyers, there are over 5,000 human traits that
can be attributed to mutations in a single human genes, but some of
these traits simply represent normal variation.
This slide shows some single gene diseases for which the genes have
already been identified. I want to make three points about the diseases
on this slide.
First, they include the most frequently occurring single gene
diseases. Cystic fibrosis affects about one in 2,500 white children,
and sickle cell anemia affects about one in 500 black children.
Second, the genes involved in these diseases were either
discovered or were well on the way to discovery before the
Human Genome Project officially began.
Third, each of these 12 diseases can be diagnosed by laboratory tests
before symptoms appear, even in fetuses. But in only three
of them, growth hormone deficiency, hemophilia, and phenylketonuria,
can treatment correct or compensate for the genetic defect.
This pattern of being able to diagnose disease or predict its future
occurrence before being able to treat its symptoms effectively will be
repeated again and again as additional disease-causing genes are
identified. The reason is simple: once a gene is identified it can be
inserted into plasmids by recombinant DNA techniques and cloned, and
you've heard all that. The large amounts of DNA that are manufactured
in this way can then be used as a test probe to determine whether
a person has a disease-causing mutation in the gene. Finding a safe and
effective treatment, on the other hand, is much more difficult. I will
return to the ethical consequences of this temporal gap between
diagnosis or prediction on the one hand and treatment on the other in
the second part of the talk.
Now if the genes responsible for the most frequently occurring single
gene disorders have already been discovered, what is left for the Genome
Project? Most of the single gene traits that do cause diseases,
albeit rare ones, are waiting to be identified. Collectively they
affect more people than the few frequently occurring diseases like
cystic fibrosis and sickle cell anemia. Mapping the human genome will
accelerate the discovery of genes for all of the remaining rare single
Mapping the genome will also lead to the identification of genes that
predispose people to multi-factorial disorders. Many of these
disorders occur much more frequently than any of the single gene
disorders. We can view the discovery of these genetic predispositions
akin to the discovery of risk factors for disease such as high
cholesterol and cigarette smoking. If you have the risk factor your
chance of getting the disease is higher than if you don't but you can
have the risk factor and never get the disease or you can get the
disease even if you don't have the risk factor. In contrast to high
cholesterol and smoking, you can't get rid of your genetic
predisposition. You can lower your cholesterol by diet or drugs and
you can stop smoking, but you can't exchange bad predisposing genes for
good ones, either now or in the foreseeable future. Telling people of
their genetic predispositions may prompt them to get more frequent exams
so that the disease can be caught early, or knowing they are at
increased risk, people may be more willing to take steps to reduce their
chance of getting the disease. But we still don't know many steps that
will work or how frequently people will accept them.
The ability to map predisposing genes to specific locations on the human
genome is proving far more difficult than mapping single gene disorders.
One problem is that the multi-factorial diseases are not
discrete entities as are many single gene diseases. The chance
for a mistake in diagnosis is higher.
Another problem is that the same gene may not play a role in all
persons afflicted with a particular disorder. As a result of such
difficulties, reports of mapping genes for manic depressive or bipolar
disorder and schizophrenia have been retracted and an association of a
specific gene with alcoholism has been disputed.
Despite these problems, extravagant claims have been made by reputable
scientists for the importance of the Human Genome Project. Daniel
Koshland, a leading biochemist and the editor of Science, had the
following to say at a genome conference in 1989. Koshland argued that
no group would benefit more from the project than the homeless, since
many of them suffer from disorders that would eventually be prevented or
treated thanks to the Genome Project. This suggests that, contrary to
fact, that the vast majority of people who find themselves out on the
street have diseases like bipolar affective disorder and schizophrenia.
Despite the hype, the fact remains that by the time the Genome Project
is finished, all 100,000 human genes will be located on specific
chromosomes in sequence. This has a nice, egalitarian ring to it.
Regardless of which deleterious genes any of us carries, they'll all be
identified. But we have no assurance that people at risk for any
genetic or gene-influenced disease are equally likely to derive benefit
from the discoveries or that the benefits will exceed costs and risks.
I will consider three problems: applied research and
development, pricing, and access.
Funds earmarked for the Human Genome Project will not be used to
elucidate the function of genes as they are identified.
Consequently, the potential benefit to people at risk for some diseases
may be delayed or never realized. I doubt we'll see the same rush for
capitalizing on the discovery of genes for rare diseases as we have
for cystic fibrosis. It did not take long after the CF gene was
found, about August 1989, before the rush was on to develop a test that
could detect the one in 25 whites in the general population who are at
risk for having children who are at risk for having CF. This is from
February, six months later, 1990. Companies and a few medical centers
are already selling such tests although they are not capable of
detecting all CF carriers. Seldom will knowledge of the gene for a rare
disease prove lucrative enough to stimulate commercial interest. One
exception is the gene for growth hormone deficiency. Growth hormone is
used...to treat children with a rare deficiency but it could also be
used in healthy kids who are not tall enough to play basketball,
professional basketball in particular. This use of genetic research to
enhance normal function is one ethical problem that I will not comment
on further, but you should think about it.
The next problem is pricing. With patent protection,
manufacturers can fix prices artificially high on genetic tests and
therapies as has happened with growth hormone. Originally intended to
provide a reward for inventors, patents may not serve the public
well in the genetic area. The discovery of the genes for many diseases
which is an essential step in the development of genetic tests or
effective treatments, is being accomplished by the investment of public
funds in the Genome Project. The public does not hold the patents on
these genes and seldom will in existing patent law. In many instances,
Universities hold the patents as the recipients of public grants under
the Genome Project. Due to changes in the patent law in 1980, a
university can grant an exclusive license to a company to develop
a product based on the patent held by the university. As a result of
such a monopoly, the public, who through tax revenues paid for
the project in the first place, could pay more than the costs justify
when the product comes on the market. Unless health insurers agree to
reimburse for these high prices, only the affluent will be able to
afford the results. Any reduction in price that is possible, that a
manufacturer--a monopolistic manufacturer makes--will be to maximize
profit and not to make the products available to everyone.
This brings me to the final one of these three problems: access.
It, too, is tied up with reimbursement for health care in the United
States. Let us return to Professor Koshland's contention about the
benefit of the Genome Project to the homeless. Even if he is correct,
how the benefit would get to the homeless under the current
reimbursement scheme is puzzling at best. As I will consider shortly,
rather than finding the fruits of the Genome Project used for their
benefit, those alleged to have genetic predispositions may find that
they are used against them.
Given these problems, we have to ask whether three billion dollars
planned for the 15 year Genome Project, assuming no cost or time
overruns, is a reasonable allocation of scarce resources. Last night,
Dr. Watson said we need more money for this enterprise, not less. If we
can harness these discoveries to serve mankind without hazard and if
there were no competing needs to be satisfied, I could not disagree. I
also want to make clear that I'm referring only to the Human Genome
Project and not questioning the development of recombinant DNA
technologies which, as we've heard this morning, have had far-reaching
and very important applications well beyond genetic diseases.
If it were true that non-genetic diseases were no longer significant
problems we could argue with greater conviction that the investment in
expanding our knowledge of the human genome was justifiable. In recent
years, however, the greatest increase in age-adjusted deaths among
Americans has been AIDS, now the 11th leading cause, and infections in
the blood stream, now the 14th most frequent cause. Deaths from
homicide and legal intervention have also increased and are now the 10th
leading cause of death. The recrudescence of tuberculosis, the
emergence of other infectious diseases, public health problems such as
Lyme Disease and Toxic Shock Syndrome and the pernicious problem of
antibiotic resistance as has been described by Dr. Cohen and Dr. Bloom
indicate that we have not licked infectious diseases. As we heard from
Dr. Bloom, infectious disease looms large in the Third World but when
fewer than 10% of children in Houston have been immunized, as he told
us, it is still a problem here.
We come finally to the potential misuses of genetic tests. Toward the
beginning of the talk I indicated that tests to diagnose diseases or
predict their future occurrence will be available long before effective
treatments. What happens in this gap between detection and treatment?
Let us answer this first for tests that are intended to predict future
disease or an increased risk for disease in the person being tested.
Consider Huntington's disease, a single gene neuropsychiatric
disorder that begins insidiously at about 40 years of age and follows a
progressively cruel course to death in ten years or so. There is no
treatment but there is the predictive test. Would a person at
risk, that is, someone who has a mother or father with the disease, want
to know that he or she will get Huntington's? A positive test could be
like a death sentence for some while for others it could relieve the
anxiety generated by uncertainty. A positive result also means that
each of the person's children has a 50% chance of having the disease.
Because the disease is untreatable, people with Huntington's disease who
have a positive test result for it cannot obtain health insurance in the
United States. Employers who know a person is at risk are unlikely to
employ that person, partly because care will raise the cost of the
employer's health benefits. A person with a negative result, on the
other hand, may not be able to get health insurance whereas he or she
was uninsurable when they knew the risk was 50%. Insurance underwriters
call this "fair" discrimination. A person who has a greater chance of
getting sick or dying soon, they say, has to pay more for protection.
There may be some merit to this for life insurance but if we are all
entitled to health care it is hardly fair for medical insurance. For
Huntington's disease and many other conditions for which tests will be
available, the person with a positive result will be denied insurance or
have excluded from coverage the disease for which he or she will incur
the greatest cost. This problem is not unique to genetics. Health
insurers for many years refused to insure people because the occupation
in which they worked increased the risk of disease.
Getting group insurance is progressively harder as this story talks
about, for those who are or have been sick. Now, 60 years ago when Blue
Cross started, private insurance provided broad coverage to many people
because their individual risks could not be predicted. Almost everyone
could be charged the same amount for health insurance. As it became
increasingly possible to distinguish risks, some by predictive tests,
others by epidemiological information on hazards of different
industries, private insurance became affordable for fewer and fewer
people. That is what we see today and we need something better.
I will now turn to the gap between detection and treatment for
tests that are used to predict untreatable disease, not in the person
being tested but in unborn children by prenatal diagnosis.
Although couples may elect to continue the pregnancy when the fetus is
found to be affected, I think it's fair to say that the technologies for
prenatal diagnosis would not have been and continue to be developed if
the alternative option of abortion was illegal. Under the Supreme
Court's decision in Roe v. Wade, which still stands, the mother has the
right to decide whether she wants to terminate the pregnancy for any
condition prenatally diagnosed by about the 24th week. One concern is
that couples will use prenatal diagnosis and choose abortion for
disorders that do not appear until late in life, are not very severe, or
for which the prenatal diagnosis cannot be made with certainty. As a
result of the proliferation of tests as the genome is mapped, there will
be more and more disorders in those categories.
In our culture, women's acceptance of prenatal diagnosis and abortion,
from some of our studies and the work of others, correlates with the
severity of disease and the certainty of prenatal diagnosis. Prenatal
diagnosis, for instance, to select the sex of one's fetus and not in
situations dealing with X-linked disease is seldom requested.
Some scientists and others believe that women should not have this right
to bring children with disabilities into the world, if that's what they
choose to do. And the reasons they give are either because their life
will be too miserable to be worth living, a contention that is disputed
by many people with disabilities, or because the support and care for
such people is too costly, although it seems unlikely that abortion for
severe genetic defects could be mandated in today's climate. A woman's
right to choose freely can be eroded if insurance companies refuse to
pay for the care of infants with diseases that could have been detected
prenatally, or if public programs to assist programs in caring for
children with disabilities are dismantled. In addition, if
research-defined treatments for such disorders are not supported,
prenatal diagnosis and abortion as an interim solution until treatments
are found could become the final solution.
In conclusion, the Genome Project has been accompanied by a resurgence
of interest in genetics. This is evidenced by the headlines that
illustrate this talk. No doubt, the Project will greatly improve our
understanding of how the human organism functions but the Project and
the hype surrounding it may lead to exaggerated expectations that could
undermine our social values. I've already mentioned the portrayal of
homelessness and alcoholism as genetic problems and I'm now going to
talk about something that apparently a number of people in the audience
had raised this morning.
Here is a front page New York Times report of Department of Justice data
claiming a family link to criminality. Although the article pointed out
that the link need not be genetic, according to a more recent story in
the Times, scientists working in the government are developing and I
quote, "a violence initiative, which among other goals would seek to
find biological correlates of violence in males, detecting them early in
life and seeking to prevent their expression." Do we really believe
that if there were genetic predispositions to homelessness, alcoholism
or criminal behavior and companies marketed tests to detect them, that
we could eradicate the problems? By looking for genetic solutions we
are coming dangerously close to the eugenics of the early 20th
century in this country and elsewhere. The emphasis on genetic causes
is victim-blaming of the worst sort. It absolves the rest of us of
The Human Genome Project will provide a map for progress and also
introduce a maze of problems. How the Project is conducted, who will
profit from it, what might be foregone as a result of the public
investment, and how individuals in society could suffer, are some of the
questions I have raised. We are fortunate to have the opportunity to
consider these questions at a relatively early stage in the quest for
the complete genome. We should not take this past lightly.