April 3, 2012 (Chicago, Illinois) — When it comes to predicting the risk for common diseases, including cancer, genome sequencing is not a magic bullet. It might be a valuable tool for people with a strong family history of a disease, but not for the vast majority of people, researchers report.
Genomic sequencing will never be a crystal ball that can reliably predict future health issues, explained researcher Bert Vogelstein, MD, Clayton Professor of Oncology and Pathology at the Johns Hopkins Kimmel Cancer Center in Baltimore, Maryland.
"It cannot substitute for conventional risk-management strategies, including routine check-ups and lifestyle optimization," he said at a press briefing here at the American Association for Cancer Research 103rd Annual Meeting.
Dr. Vogelstein was summarizing the results of a study presented at the meeting and simultaneously published online April 2 in Science Translational Medicine.
The researchers analyzed data collected from thousands of twin-pair groups on the incidence of 24 diseases, including cancer and autoimmune, cardiovascular, genitourinary, neurologic, and obesity-associated conditions. They used mathematical models to predict disease risk.
For the majority of tested individuals, the results would be negative for most diseases. In addition, the predictive value of these negative tests would generally be quite modest, because "the total risk for acquiring the disease in an individual testing negative would be similar to that of the general population," according to the researchers.
Conversely, in the best-case scenario, the results show that the majority of people tested might be alerted to a clinically meaningful risk for at least 1 disease with whole-genome sequencing.
"We stand on the verge of a revolution, and advances in technology and sequencing that have immense implications for many fields of science," said Dr. Vogelstein. "But, as we all know from the recent revolutions in the Middle East, we can't always predict the final outcomes of revolutions."
He added that in genetics, and specifically in personalized medicine, many of the predictions have been based on qualitative arguments and anecdotal reports.
Positive and Negative Tests
A positive test result should indicate that a person has at least a 10% risk for disease. "That means 1 in 10 would develop the disease from all factors combined," he explained.
The usefulness of a negative test result "is in the eye of the beholder," Dr. Vogelstein noted. To be medically useful, the risk would have to be much lower than in the general population.
As an example, Dr. Vogelstein explained that 2% of those taking the test would get positive results for ovarian cancer. "That is 1 in 50 women, and that is the maximum — the best-case scenario," he said. "That can be useful for those women so they can have closer surveillance."
On the flipside, the other 98% of women would get a negative test. "Unfortunately, the negative test is not that informative because it only shows that they have a risk that is slightly lower than the general population," Dr. Vogelstein said.
These results were similar for the other diseases that the researchers looked at, although there were a few "outliers," Dr. Vogelstein explained. "In theory, with coronary heart disease — at least in males — it might be possible that many individuals in the population would have a positive test; this might put them on the alert for heart disease."
Cancer risk is influenced by both environmental and stochastic factors, which further dilutes the ability of whole-genome sequencing to predict disease risk.
To illustrate the limits of genetic testing, Dr. Vogelstein noted that currently, men have a 45% lifetime risk for cancer and women have a 38% lifetime risk. Having a negative test result would lower the risk to 32% to 42% in men and 27% to 36% in women, which is only a slight difference from that of the general population.
Dr. Vogelstein emphasized that information about the genome will not change these estimates, which "are made under the assumption that we are omniscient and understand the effects of every variant and their interactions with one another."
Benefit Seen for Some Conditions
Dr. Vogelstein and his team derived their estimates from 53,666 monozygotic twin pairs and clinical data from registries all over the world. Their analyses suggest that for 23 of the 24 diseases studied, the majority of individuals will receive negative test results, which will probably not be very informative.
With a negative test result, they estimate that the risk of developing 19 of the 24 diseases would be 50% to 80% of that in the general population, at a minimum.
For 13 of 27 disease categories, the researchers note that the majority of patients who would ultimately develop these diseases would not test positive, even in the best-case scenario. For 4 of the disease categories — thyroid autoimmunity, type 1 diabetes, Alzheimer's disease, and coronary heart disease deaths in men — genetic testing might be able to identify more than three quarters of people who subsequently will develop the disease.
Not Ready for Prime Time
A panel of discussants agreed with Dr. Vogelstein's conclusions and pointed out the implications of the study.
Timothy Rebbeck, PhD, professor of epidemiology at the University of Pennsylvania Perelman School of Medicine in Philadelphia, and editor-in-chief of Cancer Epidemiology, Biomarkers & Prevention, noted that "we are going to have to reconsider the value of genetic information and rethink new models and when this information is valuable and when it may not be."
He added that "what we are learning" from this study and previous research is that genetics might not be "the magic cure-all" for all things.
Thomas Sellers, PhD, MPH, executive vice president and director at the H. Lee Moffitt Cancer Center & Research Institute in Tampa, Florida, agreed "with the primary conclusion of this report," adding that this is a very "provocative" study that puts very important issues into perspective.
"Genome sequencing is not going away; there are questions that we have to look at," he said.
The third discussant, Olufunmilayo I. Olopade, MD, professor of medicine and human genetics and director of the cancer risk clinic at the University of Chicago School of Medicine in Illinois, pointed out how many researchers said the same thing about BRCA testing.
"I remember the argument we had almost 20 years ago about BRCA testing," she said. "Some thought nothing good could come out of that research..., now it has been adopted," Dr. Olopade said. "Many women died from ovarian cancer, and we could have prevented it if we had known."
She emphasized that "we are now just beginning our understanding," and that to have an impact on prevention, "we need to have a more elaborate approach."
"I think that genome sequencing can improve public health, but we need to know how we are we going to do it," she said. "We are not there yet, it's not ready for prime time.
American Association for Cancer Research (AACR) 103rd Annual Meeting. Presented April 2, 2012.
Sci Transl Med. Published online April 2, 2012.http://stm.sciencemag.org/content/early ... ed.3003380