Genomic Testing: The Risk of Knowing Too Much

CONSIDER THE FOLLOWING: A healthy woman in her early 20s has a family history of breast cancer. She decides to undergo genetic testing to determine her own breast cancer risk. But instead of simply testing for the BRCA1 and BRCA2 gene mutations, her doctor offers her a new kind of test that looks at more than a dozen genes associated with a wide range of tumor types. The results come back. The good news is she doesn’t have any mutations in the two breast cancer genes. The bad news is she has a mutation in a gene called TP53 that is associated with a rare disorder called Li-Fraumeni syndrome (LFS). Women with the inherited disorder have more than a 90 percent chance of developing some kind of cancer—breast cancer, brain cancer, sarcoma, leukemia or another cancer type—during their lifetime.

Kenneth Offit, a cancer geneticist and medical oncologist at Memorial Sloan Kettering Cancer Center in New York City, has seen two cases like this in his clinic. Explaining these kinds of test results to patients is not easy, he says. Although patients with LFS can undergo intensive screening with regular blood tests, MRIs and physical examinations, determining the best screening methods can be a challenge because of the wide range of tumors that can develop. What’s more, there are no treatments for LFS.

While genetic testing for inherited cancer mutations has become more common as an increasing number of people seek to understand their cancer risk, advances in genome sequencing—the process of determining the precise order of the four chemical bases A, T, C and G that make up a person’s DNA—have taken testing to a new level. Instead of testing just one or two genes, companies now offer gene panels that analyze dozens of cancer-related genes simultaneously from a single patient sample. Some tests sequence hundreds of genes, and even entire genomes. The hope is that more comprehensive genomic testing will lead to a more precise understanding of a person’s cancer risk, allowing patients to take preventive measures like routine screening or prophylactic surgery to protect themselves. The additional genetic knowledge could also help protect their family members.

But sequencing that much DNA presents oncologists and patients with a conundrum: The tests generate an enormous amount of information, some of which isn’t yet clearly understood. And even if the findings are understood, they can blindside a person by uncovering an unexpected disease risk or discovering a cancer-associated mutation for which there are no preventive measures or treatments.

“It’s really important for people to know what they’re getting into when they undergo testing with these multigene panels,” says Susan Domchek, a medical oncologist at the University of Pennsylvania’s Abramson Cancer Center in Philadelphia. “They need to talk to their doctor about the pros and cons, because I think there is the assumption that more is better. But like everything, that’s not always true.”

Multigene Panel Testing on the Rise

More than a half dozen major commercial laboratories provide multigene cancer panel tests. Ambry Genetics offers a test called CancerNext that looks at 28 genes associated with breast, ovarian, colorectal, uterine and several other cancers. Myriad Genetics offers Myriad myRisk, a test that looks at 25 different genes and identifies mutations associated with an increased risk for eight types of cancer. “Gene panels are essentially replacing single gene testing,” says Offit, who anticipates that soon physicians will be unable to order some single gene tests from large commercial companies.

The reasons are partly economic. It’s cheaper and more efficient to test multiple genes at once than to test those same genes one by one in individual tests. It can also be easier on patients. For example, if a person with a family history of melanoma and pancreatic cancer tests negative for a CDKN2A gene mutation associated with a high risk of these cancers, it can be reassuring. But a doctor trying to understand the reason for the family history may order more genetic tests to look for other mutations that predispose to these cancers.

“That means more visits to the clinic, more waiting for test results, more time off work,” says oncologist and cancer geneticist Theodora Ross of the University of Texas Southwestern Medical Center in Dallas. “So there are a lot of reasons to do a panel up front.”

On the other hand, because some multigene panels cover a wide range of cancers, people may not be prepared to deal with all of a test’s findings. A patient might have a panel test to determine her risk of ovarian cancer and instead discover she has a mutation in a gene called CDH1, associated with a very high risk of developing stomach cancer. Such instances are rare but have big consequences. Doctors routinely recommend that patients with this mutation have their stomachs removed as a preventive measure.

“That is a huge deal and has major issues about quality of life,” says Domchek. And although patients can choose not to receive certain findings, they need to be aware of the consequences of not knowing that information, says Offit, such as the effect on relatives who might want to be warned of an inherited mutation or the impact on any future children.

Although doctors routinely talk to their patients in advance about what the tests could reveal, and genetic counselors are frequently involved, how to best communicate all that information to patients is still a work in progress. “I think it’s a fascinating and exciting time, but it’s also a time where we need to focus carefully on both physician and patient education efforts,” says Stacy W. Gray, a medical oncologist at the Dana-Farber Cancer Institute in Boston. “Patients need to understand why they’re being tested and what the implications are, and physicians need to understand enough to know when to offer the different kinds of tests.”

Dealing With Uncertainty

Interpreting the results from a panel test is no easy feat. Some genes on the panels are well-understood in terms of their role in cancer and the amount of risk associated with them. Medical options—screening tests, preventive surgery or drug therapies—are available to patients with mutations in these genes. Other genes on the panel are less understood.

Physicians may know there is some degree of risk associated with mutations in these genes, “but they don’t exactly know what the risk is, how high the risk is, or if this is a risk for people in their 60s or in their 30s,” says Domchek. Finally, some genes on those panels may have mutations that flag clear risks, even though a patient’s family may have no history of that disease, but there are no clear guidelines for how to screen or treat patients with those mutations.

Even more challenging, for every gene in a person’s genome, there can be small variations in the DNA—changes in the normal sequence of A, T, C and G. Most misspellings of genes are harmless, but a small percentage of them can cause disease. When researchers find a new misspelling in a genetic sequence, they call it a variant of unknown significance (VUS), because its role in cancer risk is unknown. “Sometimes a single change in [a DNA letter] is extraordinarily important,” says Domchek, and can increase a person’s cancer risk substantially. “And sometimes it’s no big deal at all and is just a benign chance. How you determine whether it’s important or no big deal isn’t trivial.”

To figure out if a VUS may be harmful or not, and to better understand some of the genes on the panels, Domchek and her colleagues in the fall of 2014 launched a national online registry called PROMPT, which stands for Prospective Registry of MultiPlex Testing. The registry is a collaboration of the Abramson Cancer Center of the University of Pennsylvania in Philadelphia, the Dana-Farber Cancer Institute in Boston, the Mayo Clinic in Rochester, Minnesota, and the Memorial Sloan Kettering Cancer Center in New York City. Together, these institutions have established partnerships with some of the major diagnostic companies in the country, including Ambry Genetics and Myriad Genetics.

The goal is to collect data from thousands of patients nationwide who have had multigene testing for cancer risk and to follow them over many years to learn how their genetic alterations affect their health. Patients who have multigene testing offered by one of the diagnostic companies involved are invited to enter their test results, as well as their personal and family health histories, into the registry. They can choose to identify themselves by name or enter anonymously, and they can invite their family members to participate as well. The researchers are also planning to conduct molecular experiments in the lab to see how different VUS affect gene function.

Fergus Couch, a cancer geneticist at the Mayo Clinic and one of the collaborators on the project, hopes that as his team analyzes the data, it will be able to report back to participating patients with a more precise assessment of their cancer risk. Learning more about a particular VUS, even if the information is provided many years after testing, could offer patients some comfort, especially if the VUS turns out to be harmless. Or, says Couch, he and his colleagues might find that the VUS confers, for instance, a 20 percent increased risk for cancer by age 50. “That patient might say, ‘I’m only 30 years old right now. I can go ahead and have kids and I don’t have to worry too much. I can deal with it later,’ ” says Couch. “Or the patient might do watchful waiting all her life with MRIs and not think too much about surgeries. These are the kinds of things that can be useful for patients and allow patients to better decide how they want to move forward.”

Incidental Findings

Following on the heels of cancer panel tests are even more sophisticated tests that could raise additional questions for people undergoing genomic testing. Advances in genome sequencing have made it cost-effective to sequence not just a dozen genes at once, but a person’s entire exome—the parts of the genome that code for proteins essential to the body’s functioning. But sequencing more DNA inevitably results in finding more mutations. What’s more, the tests can reveal incidental findings, results unrelated to the purpose of the test, in genes involved in a range of illnesses such as heart disease, nerve diseases and inherited eye disorders.

Increasingly, academic centers and commercial labs are offering whole exome sequencing as a diagnostic test for cancer patients. Yet the jury is still out on whether these tests are more useful than panels. Gail Jarvik is a medical geneticist at the University of Washington in Seattle and is conducting a trial of whole exome sequencing in colorectal cancer patients whose cancer is suspected of having a genetic basis. “Our goal really is to see if whole exome sequencing is faster and more efficient than usual-care genetic testing, but also what are the ethical and logistical issues with doing a whole exome and returning genomic results including incidental findings,” she says.

In the trial, doctors or genetic counselors only return to their patients incidental findings that are “medically actionable,” she says. For instance, if a genetic mutation is found that increases cancer risk, doctors can recommend screening or prophylactic surgery. If a variant is related to heart disease, the patient can have imaging tests or get a defibrillator for a heart rhythm abnormality. “In other words, something can be done,” says Jarvik. “There are reasonably useful preventative measures that can be taken.”

She and her colleagues try to set realistic expectations for patients beforehand by telling them that most people do not have genetic changes detected that significantly affect their health. It turns out that only a small percentage of patients will actually have an incidental finding, says Jarvik. “Setting those reasonable expectations has been good for patients, and the general response from them has been good.”

Dana-Farber’s Gray is involved in a similar study looking at whole exome sequencing in patients with advanced lung and colorectal cancer. In this case, the researchers sequence their patients’ tumor genome to help find mutations that can be targeted by existing drugs. But to identify those mutations, the researchers also sequence their patients’ inherited genome for comparison, which can reveal incidental findings. One of her patients is 66-year-old Elizabeth Kenner from Cranston, Rhode Island, who at the age of 62 was diagnosed with stage I lung cancer that eventually progressed to stage IV. She underwent whole exome sequencing and had an incidental finding—a mutation in a colon cancer gene. Because she has no family history of colon cancer and results from her recent colonoscopy were negative, her doctors were reassured—though she’ll continue to get screened. She is happy she had the genomic test. “If they find something and you know about it, you can do something about it,” she says. “If not, I don’t believe you should just close your eyes. I think it’s better to know.”

Researchers are also finding whole exome sequencing to be useful in diagnosing pediatric cancers and in helping to determine which mutations might increase susceptibility to cancer in some children. Sharon Plon, a cancer geneticist at the Baylor College of Medicine in Houston, is investigating whole exome sequencing in children newly diagnose​d with brain tumors and other solid tumors in the body. Although such comprehensive testing can identify mutations linked to rare cancers in children that may not be included on many gene panels, the tests can also reveal mutations associated with adult-onset cancers—findings that have implications not only for the future health of the child, but also for that of parents and other family members.

The value of reporting incidental findings when testing pediatric patients is a controversial subject and is still being debated, says Plon. “If you find a mutation that predisposes the patient to an adult-onset cancer, should you disclose that to a child’s parents?” Plon believes doctors should. “It’s likely they inherited that mutation from their parents and it’s probably to the benefit of the child that the parents be aware of that risk,” she says. At the American Society of Human Genetics annual meeting in San Diego in October 2014, Plon presented preliminary findings from her study. Of the 115 children enrolled, three were found to have mutations in the BRCA1 or BRCA2 gene. Whether these mutations are associated with the children’s tumors is still unclear, but Plon says researchers noted a history of breast and ovarian cancer in the children’s families.

Other physicians are uncomfortable with the idea of reporting incidental findings. “They say you’re doing a test that the patient didn’t ask for,” says James Evans, a medical geneticist at the University of North Carolina School of Medicine in Chapel Hill. Yet patients can choose whether they want all the results. In fact, the American College of Medical Genetics recently updated its guidelines to allow patients to opt out of receiving incidental findings. “People are allowed to make their own decisions,” Evans says. “The trick is to make sure they’re educated about it, that they know what it is they are declining.”

Evans is enthusiastic about these new genomic tests and sees many patients benefiting from them. However, like every new technology, physicians need to be thoughtful in how they apply the tests, he says. “Like an MRI, it’s a miraculous technology that gives us great insight into specific problems that patients have, but we don’t give an MRI to everybody, for very good reasons. We should use these tests where they’ve been shown to have some reasonable chance of helping patients.”

Cancer Today magazine is free to cancer patients, survivors and caregivers who live in the U.S. Subscribe here to receive four issues per year.