A Conversation With
A Conversation about Personal Genomic Tests with Dr. Muin Khoury
Last December, NIH and the CDC convened a workshop to discuss the science behind personal genomic tests. These tests, which include “genetic risk profiles” for cancer and other common diseases, are sold directly to consumers, who provide DNA through saliva or a cheek swab. At the workshop, representatives from academia, personal genomics companies, government, and policy groups discussed the scientific basis of the tests, what needs to happen before results can be used in the clinic, and recommendations for advancing the field.
Dr. Muin Khoury is the first author of a report on the workshop in the August Genetics in Medicine. He directs the National Office of Public Health Genomics at the CDC and is a senior consultant in public health genomics in NCI’s Division of Cancer Control and Population Sciences.
What are some examples of personal genomic tests?
Over the last few years, more and more of these products have become available, and they come in a variety of flavors, such as tests for people interested in ancestry. Companies are also selling people genetic risk profiles to assess the risk of common diseases like cancer and diabetes. You can buy these services directly from the companies without interference from your health care provider. The business world seems to be trying to capitalize on recent advances in the science of gene discovery, and as the science continues to mature we are likely to see cheaper and cheaper products out there.
What are your concerns about some of these products?
I am most concerned about the tests that offer advice on health-related issues, such as screening and risk assessment that may lead to clinical interventions. These tests are not ready for prime time—that’s the bottom line; I’ve said this many times and in many articles in the past few years. The currently available tests are not predictive enough to be useful for screening. By and large, they don’t add much value to current risk factors used in risk assessment and targeted therapies.
The field of personal genomics is evolving, but it’s still in its infancy. The promise of personalized medicine is not what’s being offered today to consumers, and the reality is that we may have to wait a while longer before that promise gets fulfilled.
How are risk profiles for cancer determined?
For common cancers such as breast and prostate, researchers have used genome-wide association studies to identify anywhere from 5 to 30 robust genetic variants for each cancer that are associated with the risk of developing these diseases. These variants tend to contribute modestly to a person’s overall risk of disease. But they can be assembled together to create some kind of a test to profile a person’s risk. The problem with these tests is that, taken alone, they provide incomplete and potentially misleading information about a person’s overall risk for these cancers. We do not yet know all of the genetic factors that contribute to risk or how these factors interact with environmental risk factors.
What is known about the usefulness of these tests in the clinic?
If you look at the clinical validity and utility of these tests, it’s clear that we haven’t crossed a certain threshold of evidence that would allow these tests to be used in the context of primary care. At present, these tests are not reimbursed or covered because of the lack of documented clinical utility.
What do you know about how these tests are being used?
In 2008, our office conducted national surveys of health care providers and consumers about their perceptions and use of personal genomic tests. We found that there was very little use of these direct-to-consumer tests in the community, but awareness of the tests was increasing. More people were asking questions about them. The survey looked at a cross-section of the population in 2008, and of course no one knows how these things will track over time.
Might personal genomic information lead people to change their behavior?
If you are not following the recommended health guidelines and practices, such as eating well and exercising, we do not know whether personal genomic information will cause you to change your behavior. Finding out, for instance, that you have a slightly increased or slightly decreased risk of developing disease “x” may or may not cause you to alter your behavior. At the end of the day, you should still follow the recommendations for screening and detection, and these recommendations currently do not make use of genetic risk information. This could change in a few years if we have additional information and interventions that are based on specific genetic profiles. But it’s not the case now.
Do you think that talking about these tests may inadvertently draw people to them?
People need to be savvy consumers when it comes to their own health and health care. They should take a hard and a skeptical look at these products. I’m all for educating the public and the providers, and doing so can lead to teachable moments between providers and patients.
What would you like to see happen next in the field?
We need to discover more genes and understand what these genes do in relation to health outcomes, especially in the context of environmental and social factors. We need to figure out how people respond to genomic information and whether the public benefits from—or is hurt by—having incomplete information that may lead to unnecessary medical procedures, follow-up tests, or even surgical interventions.
In another area, there also needs to be more standardization within the industry. When you buy whole-genome tests that different companies put out, you should come up roughly with the same results. It is one genome after all. But there’s been quite a bit of heterogeneity, or lack of consistency, in the laboratory processes and clinical interpretation. There are many kinds of questions, which are laid out in the workshop proceedings we just published in the August 2009 issue of Genetics in Medicine.
Are you curious about your own genomic profile?
If I had a few hundred dollars to spare, I probably wouldn’t buy information about my own genome. Some people do find these tests interesting, amusing, and informative, perhaps from a recreational or some other perspective. But from the perspective of promoting health and preventing disease, I don’t think we’re there yet.
Is there a disease for which personal genomic information may be relevant soon?
Not yet. I don’t want to be dogmatic because the science is moving so quickly that things could change in a hurry in this field. But as of today, we’re not there.