The Long Road to the Cystic Fibrosis Gene
When I first became interested in human genetics in the early 1970s, most persons with cystic fibrosis (CF) died before their 20th birthday, many at a much younger age. For the more severely affected children, CF was a parent's worst nightmare. For them childhood unfolded in a world of chronic illness. As lung function deteriorated, each hospitalization became a battle to snatch a beloved child from the hands of death, and each victory was temporized by the certainty that another battle loomed ahead. The losses were horrible. One of my most painful memories of medical school occurred during my rotation in pediatrics. When his 12-year-old daughter with cystic fibrosis died of pneumonia in the intensive care unit, a father tried to commit suicide in the hospital.
As recently as 1974, a leading medical text described CF as "not only one of the most common disorders of childhood, but... also one of the most enigmatic." Even its name reflects the mystery. The term "cystic fibrosis" was coined more than 60 years ago to describe the scarring found in the pancreas at autopsy, yet the far more devastating effects are on the lungs, in which the ravages of recurrent, serious pneumonia are evident at a glance. Most strange is the protean nature of the disease. Some children are severely ill from birth, others become ill only gradually over four or five years, and a few are quite well until young adulthood. Most patients have both lung disease and severe gastrointestinal problems, but some are burdened only with comparatively mild lung disease.
Scientists demonstrated that CF was an autosomal recessive disorder (the result of being born with two defective copies of the same gene, which if present in only one copy does not cause illness) in 1949, but they remained ignorant of the cause for 20 years. In 1967 a researcher showed that the serum of persons with CF contains a chemical factor that inhibits the movement of the tiny cilia in rabbit trachea. This suggested that the ravages of CF might arise from a defect in how some factor flowed into or out of lung cells. By the mid-1970s, scientists had compiled enough evidence to be confident that the injuries caused by CF could all be explained by a defect in the transport of ions across cell membranes, but they did not know what caused this to occur.
To a geneticist, one of the strangest aspects of CF is that it is such a common disorder. About 1 in 2500 white children is born with CF (the disease is less common in other racial groups). That means that about 1 in every 25 whites carries a single copy of the CF-causing allele. The odds of 2 carriers marrying is 1/25 X 1/25 which is 1/625. The chance that an at-risk couple will bear an affected child is 1 in 4 with each pregnancy. 1/625 X 1/4 equals 1/2500. But how could such a "bad" allele become so common, especially since persons with CF virtually never have children (affected men are infertile and affected women are often not healthy enough to risk a pregnancy)?
One of the first appealing hypotheses was proposed 30 years ago. Dr. Alfred Knudsen, who now works at the Fox Chase Cancer Center in Philadelphia, calculated that the unusually high prevalence of the CF allele could be explained if healthy persons who carried one copy of it were on average just 2% more fertile than were those who did not. That would be enough to maintain the allele at high levels despite the deaths of affected children. What aspect of the CF allele might confer a reproductive advantage on carriers? There have been several creative guesses, such as that by a scientist who suggested that in women who are carriers the vaginal fluids might have a different viscosity that would facilitate the movement and survival of sperm, increasing the chances of becoming pregnant. Recently, studies in India have shown that CF carriers are more likely than are non-carriers to survive cholera. Since cholera has been sweeping through human populations for millennia (it devastated the United States in 1832, 1849, and 1866), this is a plausible (though still unproved) explanation for the high frequency of the CF allele.
When gene mapping tools became available in the early 1980s, many top research teams took up the challenge of finding the CF gene. They studied the families burdened with CF, seeking to identify which of their ever-growing number of DNA markers were present in children who had the disease and absent in the brothers and sisters who had not inherited the disease. The DNA markers that met that test would have to be near the hidden CF gene. In 1985 a group that included academic and commercially based research teams announced that they had used a linkage approach to localize the CF gene to the long arm of chromosome 7. The CF gene was the first to be mapped by what was for a time called "reverse genetics" (now called "positional cloning"). The term was meant to reflect the fact that such studies permitted researchers to find genes even when they did not have a clue as to their function. In less than a year, other groups had found markers on each side of the gene, narrowing the area on chromosome 7 where it had to lie. Even with this exciting advance, which led to the first prenatal test for cystic fibrosis, the road to elucidating the CF gene remained long and arduous. It took several major scientific groups nearly four years to hunt through the 1,490,000 base pairs of DNA in that region of chromosome 7 until one team won the race in the fall of 1989. One of the leaders of the team that cloned the CF gene was Francis Collins, then a young professor of human genetics at the University of Michigan. This discovery propelled him to his current position as Director of the National Human Genome Research Institute.
The cloning of the CF gene and the elucidation of the complete amino acid sequence of the protein for which it codes was a gigantic leap for molecular medicine. By studying the protein's structure and location, scientists were able to show definitively that its primary function was to transport chloride ions across cell membranes, thus confirming the long-held hypothesis. It was named the CFTR (cystic fibrosis transmembrane conductance regulator) gene. This discovery raised exciting new approaches for research. For example, by testing patients and studying the correlation between various mutations that occur at different sites on the gene with the severity of the disease, scientists began to understand how CF could express itself in such varied ways.
Among whites of northern European ancestry, about 70% of all CF al-leles are missing the same three base pairs of DNA, causing the protein to lack just 1 amino acid (a phenylalanine at position 508) out of more than 1000. Because this mutation is so common, it was hoped at first that just a few more mutations would account for the other 30%, which would make it easy to develop clinical tests to identify persons who carried the CF gene. But the CF gene is large; its coding region contains 24 different regions
(exons) and uses more than 6000 base pairs of DNA. We now know that there are hundreds of different possible mutations, some found only in a single family (nicknamed "private" mutations). This made the task of developing a clinical test to identify carriers more difficult, especially for people of southern European extraction, a group in which the 508 deletion accounts for only about 30% of all CF alleles. It also greatly complicated the job of correlating the severity of disease with the pair of mutations (since each parent could contribute any one of hundreds) with which a patient was born.
The chore of correlating genotype with phenotype, a task that must be repeated every time investigators find a gene that contributes to a disease, has already taught some important lessons. In delineating the CF gene, geneticists, who had been lumpers, became splitters. For example, they found some adults with uncommon mutations who had such mild symptoms of lung disease that no physician had ever guessed they had CF. Even more astounding was the discovery that among otherwise healthy infertile men there are a small number who have mutations in each of their CF genes. Undiagnosed cystic fibrosis may account for 1-2% of male infertility. Such discoveries greatly enlarged the clinical spectrum of cystic fibro-sis. We are not yet at the point where we can use prenatal diagnosis and DNA analysis to predict the severity of the disorder in a fetus, but a rough picture of the relationship between the presence of the common mutations and the expected severity of disease is slowly emerging.
The most exciting dividend of the cloning of the CF gene has been the tremendous impact it has had on driving creative approaches to new treatments. Thankfully, the health and life expectancy of patients with CF has, because of new antibiotics and better management of the risk for pneumonia, improved steadily for the last 30 years. Children born with CF today have an even chance to live into their late 30s, but advances have slowed, the disability can be severe, and the costs of care are high. For example, patients with end-stage cystic fibrosis are the people who most often undergo lung transplants or heart-lung transplants.
Because the disease is so common and because the life-shortening effects of CF arise as a consequence of recurrent damage to the lung, which is relatively accessible, CF is a focal point of gene therapy (see Chapter 20). Less than two years after the gene was cloned, scientists showed that they could use a virus to deliver an artificially constructed copy of the normal
CFTR gene into cultures of cells derived from a CF patient in which the gene was crippled and correct the defect in ion transport. Of course there is a huge difference between fixing cells in a test tube and treating humans, but hope is high. There are more than a dozen research projects under way in which the goal is to develop techniques to get normal CF genes into the lung tissue of affected children. Most focus on using relatively benign viruses that are known to penetrate human lung cells as "bio-missiles" to carry a payload of normal CFTR genes to the target organ. Some progress has been made. A decade from now there may well be an easy way to treat the disease, quite possibly with a "gene inhaler," a DNA spray that works much like the little pumps that persons with asthma use to deliver preset doses of medicine to their airways. But the discovery of the gene has also created sometimes vitriolic debate about genetic testing.
In the fall of 1989, only weeks after the CF gene had been cloned, leaders of the American Society of Human Genetics publicly urged that the knowledge should not be used to screen the general population to identify carriers who, if they married a fellow carrier, would have a 1 in 4 risk in each pregnancy for conceiving a child with the disease. After two days of at times rancorous debate, a group of experts convened by the NIH in the late spring of 1990 to conduct a more thorough review of the issues reached the same conclusion. The scientific rationale was that it would not be helpful to screen the general population with a test that could identify less than 90% of those who actually carried a CF mutation. This is because such a test creates situations in which one partner definitely learns that he or she carries the CF allele while the true status of the other (who has tested negatively) remains unknown. Such couples have about a 1 in 1000 chance of having a child with CF. Some experts argued that couples in this situation might feel substantial anxiety in pregnancy. The experts also feared that if the test came into wide use too early most physicians in general practice would be unable to offer proper counseling about the implications of the result. Many on the NIH panel were also worried that carriers might be unfairly denied access to health or life insurance at reasonable rates.
At that NIH meeting I was among a minority that saw the matter differently. I argued that individuals have a right to ask questions about themselves, and that in an activity as crucial as childbearing there would be some couples who would be interested in learning about carrier testing for cystic fibrosis and who might well seek to be tested. I thought that the stance adopted by the majority was paternalistic, and that the patient's right to know should be inviolate. In addition, it seemed to me that the average family practitioner could readily be trained to provide the necessary information and support to patients who were thinking about whether or not to take the test.
During the 1990s there were many advances in the technical quality of the DNA-based test for cystic fibrosis, and it is available at reasonable price from many commercially and academically based labs. Currently, the most comprehensive test is offered by Genzyme Genetics, Inc., which in 1997 launched a test that could identify any 1 of 72 different mutations. Yet, as of 1999, CF testing is still not widely used. There were only about 75,000 carrier tests done that year in the United States (on a per capita basis, usage was somewhat higher in northern Europe). In general, most people who seek testing do so because they have a relative with the disorder or have symptoms that could be compatible with atypical cystic fibrosis, but the pattern of usage may soon broaden.
In April, 1997, another group of experts met at NIH to review progress in CF testing since the 1990 meeting and to reassess the place of carrier screening in medicine. I anticipated little change in their position, so I was quite surprised with their findings. In a cautiously worded statement, they supported offering the CF test to persons regardless of whether or not they had any family history of the disorder. This is a reversal of the position taken in 1990 and opens the doors to mass population screening. The 1997 position adopts an approach already in use in some prominent genetics clinics, such as that led by Dr. Arthur Beaudet at Baylor Medical Center in Houston, who advocates informing all patients, regardless of their family history, that a test is available to identify CF carriers.
One can imagine a number of different paths to develop routine population-based screening to identify CF carriers. Initially, the most efficient would probably be to offer the test with appropriate education and counseling to all young women who were planning soon to have families. Given the rapid shift to managed care, it would not be particularly difficult to incorporate a CF screening program into routine visits. If one assumes that 1 in 25 white women is a carrier, one can reasonably anticipate that about 1 in 28 white women would test positive for a mutation (the number is different because the test would not find about 10% of carriers). The next phase of testing would be to offer the CF test to the partners of the women who test positive. Again, about 1 in 28 of these men will be found to be a CF carrier.
Let us assume that a large managed care group adopted a CF carrier screening program and that 10,000 women in their 20s or 30s took the test in a year. Among them the test would find about 360 carriers. Among their 360 husbands, testing would identify about 13 carriers. Thus, by conducting 10,360 tests, the managed care group would identify 13 couples who are at 1 in 4 risk of having a child with CF. Without the program it is likely that among them, these 13 couples would have 3 or 4 first-born children with CF.
How does one decide if population-based CF screening is worthwhile? Assuming that it costs about $100 to test, educate, and counsel each person who decided to check her or his CF carrier status and to follow up with at-risk couples, the tab to identify four affected fetuses is about $1,000,000 or about $250,000 per case detected. The cost of finding a fetus with CF is considerably lower if one factors in savings from avoiding the birth of second affected children in some of the families, which often happens because of the delay in diagnosis that frequently occurs. Then the case finding cost might be closer to $100,000. This is still very expensive. Furthermore, either because of their religious beliefs or because they are impressed with the tremendous progress in caring for people with CF, about one-third of couples will not avoid at-risk pregnancies or abort an affected fetus.
One fact emerges from this back-of-the-envelope analysis. Until the unit cost of DNA testing drops by a factor of 2-5, it is unlikely that we will see a significant effort to screen millions of young Americans to identify CF carriers. The trump card that could refute this prediction is malpractice litigation. If even a single couple who gave birth to a child with CF successfully sued their physician for failing to alert them to the availability of the carrier test, CF testing would rapidly become widespread. During the late 1970s and early 1980s, lawsuits or fear of them accelerated the pace of prenatal screening for Down syndrome and for spina bifida (see Chapter 8).
The cost of DNA analysis may well drop by a factor of 5-10, probably in less than 5 years, and almost certainly in less than a decade, and it will continue to drop. To that virtual certainty add the fact that DNA testing will develop into an analytical system which is capable of asking thousands of questions about each blood sample regardless of what prompted the in dividual to seek information. Indeed, it might turn out (due to automation) to be easier and cheaper to ask thousands of questions about a sample than to ask only a few.
Assuming that the cost of DNA-based population screening for CF mutations falls to a few dollars per sample or less, the cost benefit analysis of finding the at-risk couple becomes compelling. It would then make good economic sense to identify those at risk, and to counsel them about the nature of the disease and their family planning options. This is the scenario that disturbs many individuals in the bioethics and disabilities communities. They fear that when costs of technology no longer constitute a buffer to widespread use, socioeconomic forces will drive consumers to use DNA testing and, for those who are positive, to avoid at-risk pregnancies or to electively terminate affected fetuses. Some see this as a sort of "neo-eugenics" (see Chapter 23) and warn of the return of a social policy in which the mantle of the state is taken up by the men who manage the flow of health care dollars.
In bioethics circles there is an apocryphal tale that officials at one HMO told the parents of a child with CF that should they decide to have another child, the HMO would only provide coverage of that child if the mother had prenatal testing for cystic fibrosis. In the early 1990s I was consulted on behalf of the couple that had reported this threat by their HMO. I was told that a HMO physician on one occasion did strongly suggest to a woman with a CF child that she should seek prenatal testing for future pregnancies, and that she perceived the suggestion to be a threat about loss of coverage. A single phone call to the HMO from the consulting geneticist caused the HMO to quickly distance itself from even a hint of coercive behavior.
Population-based screening to identify CF carriers raises thorny ethical questions about selective abortion. Is it right to provide DNA-based prenatal diagnosis to couples at risk so that they may have the option of pregnancy termination if the fetus is affected? Even though women have (since Roe v. Wade in 1973) a right to privacy that guarantees this option, a large fraction of the American population is opposed to it. In a 1997 poll of 819 adults conducted for USA Today, 49% said that they felt it was morally wrong to use prenatal testing to find and abort fetuses with cystic fibrosis. In the minds of many people the decision to terminate a pregnancy because one does not want to have a child is less offensive than is the decision to end a pregnancy because one is not satisfied with its anticipated condition.
Would some couples who learn that they are carrying a fetus destined to develop CF terminate the pregnancy? Yes. One survey of young adults who had no personal experience with CF indicated that about two-thirds would terminate. On the other hand, a survey of couples who had a child with CF found that less than one-quarter would electively abort an affected fetus. As the prospects for treating CF improve, interest in terminating such pregnancies will diminish.
Would young couples with no family history of CF seek carrier testing and, if they turn out to be carriers, avoid having a pregnancy that carries a 1 in 4 risk? They could circumvent the risk in a number of ways, including remaining childless, adopting, using a donor egg or donor semen, or using preimplantation genetic diagnosis (PGD). In PGD, CF testing is performed on a single cell taken from an 8-cell embryo conceived in a test tube with the couple's egg and sperm. This is possible because the laboratory can quickly make millions of copies of that cell's CFTR gene and scan it for mutations. If the embryo does not have CF (i.e., did not inherit two CF mutations), it is transferred to the woman by the same techniques used in standard treatments for infertility. The prospects for preimplantation diagnosis look excellent, but the cost is high, and at this stage the accuracy of the test results is sufficiently uncertain that physicians will only perform it if the pregnant woman agrees to undergo amniocentesis in order to obtain fetal cells to confirm that he or she does not have two disease alleles. Of course, this is a promise that the doctors cannot enforce.
Would mass population screening significantly decrease the number of infants with cystic fibrosis? A growing body of research suggests that once they learn about CF, the genetics of its inheritance, and the existence of a high-quality screening test, a sizable fraction of women want to be tested. However, we do not have a clue as to what impact mass population screening would have on the number of children born with the disorder. A few of the larger managed care organizations have started to offer CF screening routinely to young women, but it is uncertain how far or how fast this practice will spread.
Cystic fibrosis carrier testing may be among the first DNA-based tests for which population-based screening on a massive scale creates an early warning system for couples. However, the fact that genetic testing is so closely connected to selective abortion will continue to deeply trouble many people and cast a long shadow over molecular medicine.
A pedigree of a family burdened with hereditary breast and ovarian cancer.
Was this article helpful?