Access to paid content on this site is currently suspended due to excessive activity being detected from your IP address Please contact the publisher to request reinstatement.
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
January 3, 2001

Genetic Disease, Genetic Testing, and the Clinician

Author Affiliations

Not Available


Copyright 2001 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2001American Medical AssociationThis is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Not Available

JAMA. 2001;285(1):91. doi:10.1001/jama.285.1.91-JMS0103-3-1

Modern medicine emphasizes treatment of the sick. It is often said that the widespread genetic testing soon to follow the completion of the Human Genome Project will usher in a new era of preventive medicine. Such changes require new ways of thinking, however. For example, there may be nothing clinically wrong with a healthy patient who requests genetic testing, even if the tests reveal disease genes. Since all individuals have genetic skeletons in their closets, it is important to be careful not to confuse having disease genes with having the diseases that they cause. Unfortunately, many in the public have adopted a kind of genetic determinism that sees genes as destiny: for example, having the gene associated with colon cancer means they will develop colon cancer. Physicians tend to be more careful, yet even they are not immune to subtle versions of genetic determinism.

One example of this is the uncritical categorization of certain diseases as "genetic." In fact, an adequate concept of genetic disease is extremely difficult to come by. The simplest notion would require a 1:1 correspondence between a disease and its genes, but this is the exception rather than the rule. For example, cystic fibrosis (CF) is often put forward as a good example of a genetic disease, since it seems to result from mutations in a single gene, CFTR. Even in this case, however, the exact relationship between CFTR mutations and disease is not clear, as virtually every possible combination of sweat chloride test results, genetic test results, and symptoms has been observed.1 If a patient presents with the classic symptoms of CF and is found to have a mutation in the CFTR gene, the physician might understandably infer that the mutation caused the disease. But if an asymptomatic patient is discovered to have a CFTR mutation, it is unclear what this means. The physician might tell the patient the gene is abnormal and that he or she is likely to develop specific symptoms, but it is not really known whether even this qualified prognosis is true. This is because current knowledge of CF is based largely on studies of people who have the disease. We do not yet know how likely it is that someone would test positive for a CFTR mutation and remain healthy all his or her life. The claim that these situations are rare is thus based on an implicit genetic determinism and is not supported by data. Such assumptions may be harmful if, for example, they cause patients to make inappropriate treatment decisions. Nevertheless, because they remain tacit and thus escape critical scrutiny, assumptions like this are quite common.2

Moreover, the causal role of a specific gene is typically less direct than in CF. Suppose a woman from the general population, who is not otherwise at risk for breast cancer, tests positive for the BRCA1 gene. She is likely to be told that she has an 85% lifetime risk of developing breast or ovarian cancer. This is serious news, especially since it is offered in the seemingly unassailable, quantitative language of science. However, the 85% figure is actually based on cancer-gene covariance in families that have an unusually high incidence of early-onset breast cancer.3 It is an open question to what extent such families accurately represent the general population. In these circumstances, should the patient even be given a numerical estimate of risk, knowing that quantitative results relayed by a physician are often taken as absolute? Physicians may have an intuition that information is always harmless and desirable, but this is defensible only if the information can be used to make appropriate decisions. Surely, if the results of a genetic test cause a patient to undergo what later turns out to be an unnecessary radical mastectomy, the test information can not be considered harmless.

One thing physicians can do for patients is direct them to trained genetic counselors who will discuss the desirability and interpretation of genetic tests in detail. Unfortunately, given the dearth of trained counselors and the pressures of managed care, this is not always possible. Primary care physicians forced to deal with these new tests thus have a 2-fold responsibility. First, they must keep themselves up to date concerning the complexities of gene-disease relationships, taking careful note of what is still unknown as well as what is known. Second, they must communicate this information openly and honestly with the patient. In particular, physicians must be prepared to debunk uncritical notions about the causal power of genes and even to argue against the advisability of genetic testing, especially in situations where there is no clear differential treatment based on what the tests reveal. It is better to admit to patients that medical science cannot yet answer their questions than to offer a false sense of certainty through tacit endorsement of information of uncertain quality.

Smith  KC A disease by any other name. Eur J Med Phil Health Care. In press.
Smith  KC Equivocal notions of accuracy and genetic screening of the general population. Mount Sinai J Med. 1998;65178- 183
Hall  JMLee  MKNewman  B  et al.  Linkage of early onset breast cancer to chromosome 17q21. Science. 1990;2501684- 1689Article