Thursday, November 11, 2010

Genetic Testing: How Much Do We Want—or Need—to Know?

Technology is rapidly expanding the amount of data that can be extracted from genetic testing. Harold Pollack provided a thought-provoking article on the subject in the ‘American Prospect.” He asks the question: “if a genetic test indicates a potentially negative life outcome, but it is an outcome over which you have little or no control, then what is the benefit of having taken the test?”

Pollack focuses on issues related to screening at birth. His questions are important because, as he points out, that is the only opportunity you have to do large-scale genetic screening of the population. The amount of screening performed on newborns is expanding, with even more being recommended.
“If you've recently had a child, you might be surprised by the number of rare conditions for which she was screened in the hospital nursery. All 50 states test newborns for sickle-cell disease and for cystic fibrosis. The emergence of a single technology, tandem mass spectrometry, now allows newborn-screening programs to simultaneously test for dozens of traits for roughly $10 per blood sample. This dramatically expanded the scope of newborn screening. In 1995, the average state mandated newborn screening for five conditions. By 2005, this number had increased to 24. The American College of Medical Genetics now recommends that babies be screened for more than 50 primary and secondary disorders.”

“Yet knowledge of genetic markers does not always bring clinical benefits. Many optimistic accounts of the future of genetics fail to consider the full costs and implications of widespread screening: How will these technologies be regulated? How should public-health authorities and health-care providers -- not to mention patients and their families -- respond when unsettling results are found? Will patients and their families even understand the complex options available to them? Research suggests that patients often make poor decisions based on genetic information and that doctors don't do much better. We face an embarrassing mismatch between our lofty aspirations of personalized genomic medicine and the everyday capacities of our medical-care system. As Hank Greely, director of Stanford's Center for Law and the Biosciences, told The Washington Post, ‘Information is powerful, but misunderstood information can be powerfully bad’."
Pollack uses the example of a genetic defect called fragile-X syndrome as an example where screening provides information that is difficult to interpret, and could cause considerable harm if misinterpreted. This defect has a broad range of outcomes, and is more likely to produce severe handicaps in males than in females. Only a small percentage of males, those with a completely defective gene, seem certain to have significantly diminished mental capacity. Many females in the same situation will demonstrate no symptoms. Blanket genetic screening for this defect will provide information that is worrisome to many people, but of what value is it to them? The only clear diagnosis for screened children applies to a very small number of males.

The author raises this as a question that needs to be considered and discussed. It is a good example of the types of dilemmas that can arise as biotechnology progresses. The author gives us this to ponder.
“Prospective parents can now be tested before pregnancy, and those found to be carriers for serious conditions have the option of in-vitro fertilization, whereby embryos can be pretested for genetic markers associated with Alzheimer's disease, hemophilia, muscular dystrophy, Tay-Sachs disease, and more. Many of these same markers can also be detected by do-it-yourself genetic-testing kits, which are beginning to appear on the Internet and on drugstore shelves. Walgreens may soon sell a cheap home test that covers 37 genetic conditions. (Sales are postponed pending approval by the Federal Drug Administration.) You will soon be able to buy a test kit online, mail a finger-stick blood sample or a saliva swab to a lab, and receive an e-mail containing your detailed DNA workup a few days later. In the not-too-distant future, researchers note, you will be able to purchase your complete genetic sequence encoded onto a chip for $1,000, or maybe even as little as $100.”
I thought a more compelling—and heartbreaking—example of the dilemma one faces when presented with genetic information arises for those carrying a certain mutation of the BRCA1 and BRCA2 genes. According to the National Cancer Institute:
“According to estimates of lifetime risk, about 12.0 percent of women (120 out of 1,000) in the general population will develop breast cancer sometime during their lives compared with about 60 percent of women (600 out of 1,000) who have inherited a harmful mutation in BRCA1 or BRCA2. In other words, a woman who has inherited a harmful mutation in BRCA1 or BRCA2 is about five times more likely to develop breast cancer than a woman who does not have such a mutation.”
Assuming these numbers are accurate, this presents a truly frightening prospect for the affected women to face. The probability of breast cancer and early death is high. Many women have, preemptively, had both breasts surgically removed in an attempt to improve their odds. How does one know whether or not this is the best course of action?

Consider this article posted by CNN.
“Women who have gene mutations that increase the risk of breast and ovarian cancer can substantially reduce their chances of developing -- and dying from -- those cancers if they have their breasts or ovaries removed preemptively, according to a new study.”

“The study, which appears in the Journal of the American Medical Association, confirms that preventive mastectomy and ovary removal can slash the risk of cancer in women carrying the BRCA-1 or BRCA-2 gene mutations, and it suggests that surgery is more effective than rigorous screening at preventing future cancer.”

“Researchers at 22 cancer centers in the U.S. and Europe followed nearly 2,500 women with BRCA gene mutations for about four years. None of the women who underwent preventive mastectomy developed breast cancer during the study, whereas 7 percent of the women who opted against the surgery did. (The women who did not have surgery were put on an intensive screening schedule.)”
How would an affected woman interpret these results? They imply that preventive mastectomy is a certain path towards avoiding breast cancer, while “intensive screening” does no good. The first claim may be true. The second claim is a non sequitor.

This article reeks of the kind of sloppy science that we learned about in Lies, Damned Lies, and Medical Science. The lowliest scientist or engineer couldn’t publish the viscosity of a household lubricant without quoting the measurement uncertainty. Yet, when it is a matter of life or death, all manner of assumptions, biases and statistical uncertainties can be ignored. What is the effect of only following people for only four years? We are dealing with small numbers here. The statistics of small numbers has misled many a researcher over the years. Might these results, with error bars applied, be consistent with a null result? The original article might address these issues, but for CNN that would complicate a good storyline.

Now consider this disclaimer provided by the National Cancer Institute:
“It is important to note, however, that most research related to BRCA1 and BRCA2 has been done on large families with many individuals affected by cancer. Estimates of breast and ovarian cancer risk associated with BRCA1 and BRCA2 mutations have been calculated from studies of these families. Because family members share a proportion of their genes and, often, their environment, it is possible that the large number of cancer cases seen in these families may be due in part to other genetic or environmental factors. Therefore, risk estimates that are based on families with many affected members may not accurately reflect the levels of risk for BRCA1 and BRCA2 mutation carriers in the general population. In addition, no data are available from long-term studies of the general population comparing cancer risk in women who have harmful BRCA1 or BRCA2 mutations with women who do not have such mutations. Therefore, the percentages given above are estimates that may change as more data become available.”
In other words, the quoted probabilities of getting breast cancer only apply if you are in a family with a history of breast cancer. That seems like rather critical information. Too bad CNN doesn’t have time to tell you that.

If a woman knows that she carries this mutation, she has three options: she can try to ignore it, she can carefully monitor her condition hoping to defeat any cancer that occurs, or she can have her breasts surgically removed. She should realize that the latter path is not a certain cure. She should also realize that having major surgery in a hospital is a pretty risky procedure. My (male) inclination would be to follow the close monitoring regimen. There is always the possibility that the “science” is wrong; current treatments might suffice if she should contract cancer; a treatment could be developed to negate the effect of the mutated gene; there is a significant probability that she will not contract breast cancer anyway.

Once the issue arises, does the woman really have the option of trying to ignore the possibility of breast cancer? For those who would not consider preemptive mastectomy, has the knowledge been of any benefit to them? That comes back to the question Pollack raised in his article. If the current interpretation of data holds for the general population, and a screening test was economically viable, we could screen all infants for this gene mutation. But should we? Is it really helpful to know that sometime in our life there is a 60% chance that we will contract cancer (rather than a 12% chance)?

People who worry about the economics of medical care often talk in terms of “quality-adjusted life years” (QALY) in assessing the cost effectiveness of a medical procedure. This manner of reckoning assumes that a medicine or a process that extends life but destroys the quality of life in the process should get a low ranking in terms of effectiveness. One could apply a similar criterion to genetic testing. Knowledge that one has a higher probability of contracting cancer is not something easily ignored. It will probably taint the remainder of one’s life, diminishing the “quality.”

Genetic testing must be done carefully and thoughtfully. At the moment, neither patients nor doctors are sufficiently informed on the issues. That has to change.

Steven Shapin provides a revealing description in a ‘New Yorker” article of how this worry about probabilities and outcomes can take control of a person’s life.
“’s also populated by legions of the screened and the tested, who become more and more aware of the dangers battering away at their cells from the external environment and lurking inside, encoded in their genes. This is the world of the cancer ‘risk factor’: of the Pap smear; the annual mammogram; the prostate-specific antigen test; the colonoscopy; the wait for the results of biopsies of polyps removed in the colonoscopy; the daily dose of Prilosec taken because frequent heartburn is thought to be a risk factor for esophageal cancer; even the world of knowing one’s personal genome and the world of the prophylactic mastectomy.”

“The risk-factor world holds out hope for avoiding cancer while recruiting masses of us into the anxious state of the ‘precancerous.’ The physician and historian Robert Aronowitz offers an acute illustration of the problem: a fifty-eight-year-old woman diagnosed with breast cancer has a lumpectomy, followed by local radiation and months of chemo. After that, she is put on the anti-estrogen Tamoxifen for five years. As she finishes that course of treatment, she weighs the decision whether to go on a different type of hormonal therapy and what type and frequency of M.R.I. or mammogram to get. She is an active member of a breast-cancer-survivor group, and she closely monitors the latest developments on the Web. Meanwhile, another woman, the same age, has not received a breast-cancer diagnosis. She has, however, taken supplementary estrogen pills for several years in connection with menopause, and her doctor now tells her to stop, because estrogen may constitute a risk factor for breast cancer. She has been getting annual mammograms since she was forty, and four years ago an abnormal mammogram was the occasion for an aspiration biopsy. This proved negative, but her anxiety increases. She surfs the Web for information about risk factors, and she is struck by direct advertisements for Tamoxifen to prevent the development of breast cancer, for which she now believes she is at serious risk. The first woman had cancer; the second woman does not have cancer. But their experiences eerily resemble each other.”

No comments:

Post a Comment

Lets Talk Books And Politics - Blogged