Genetic tests won’t give flawless disease risk predictions

Mike Zacharias

Genetic testing might allow doctors to predict the likelihood of a patient contracting a specific disease. But professor Neil Holtzman of Johns Hopkins University said the quality of these predictions, and of some laboratories themselves, can be less than ideal.

“We have the expectation that in the next 10 to 20 years, medicine will be able to predict whether we are going to get a particular common disease in the future,” said Holtzman, former chairman of the federal Task Force on Genetic Testing. “That is not going to happen, with very rare exceptions.”

Holtzman spoke on the topic Wednesday at the Weisman Art Museum.

Currently, laboratories are not required to provide information on the clinical validity, or accuracy, of a genetic test. While Holtzman said accurate tests can be conducted for diseases like breast and colon cancer, it is more difficult to decipher the genetic indicators for common diseases such as high blood pressure, diabetes or manic depression.

“The reason for this is not a technological limitation, it’s that the genetics are too complicated,” Holtzman said. “There are many different diseases as well as environmental factors that have to be present together to explain a large proportion of people who are going to get the disease.”

Holtzman said some diseases can be predicted through genetic testing because scientists have identified the locations of the few indicative genes of that disease. Unknown numbers of indicative genes and their locations presents a difficulty of identifying the disease. Also, there can be a multitude of gene combinations that can cause these diseases.

For example, Holtzman said, some people with diabetes might have identifiable genes at locations W, X, Y and Z, while other people with the same type of diabetes can have genes at locations A, B, C and D. With more possible combinations, in addition to undiscovered genes, it is extremely difficult to accurately predict the likelihood of disease.

“People have looked very hard for genes for these other disorders, and there are some weak predictors that have been found,” Holtzman said.

These predictors’ lack of strength, Holtzman said, should be known to the people receiving the results, because some might test as likely to be affected by a disease – and never get it.

There are also problems with the regulation of genetic testing in clinical laboratories, Holtzman said, because federal guidelines were drafted before genetic testing advanced to the level it is today.

“So we don’t have special requirements for looking at the lab quality of genetics,” Holtzman said.

Jeffrey Kahn, director for the University’s Center for Bioethics, said the current problems with lab regulation mean labs do not have to live up to the same standards as other government-regulated health products.

“Maybe most problematic is that labs now market directly to consumers, sometimes through the Internet,” Kahn said. “So all you have to do is send them your credit card number, and they’ll send you a kit to swab the inside of your cheek and send back the cells.”

After the lab does the test and sends back its findings, the problem arises as to what to do with the test results.

“I think it’s hard for professionals, let alone for the public, to understand the quality of that information and what to do with it,” Kahn said.

But both Holtzman and Kahn said they expect the quality of genetic testing and the validity of the tests to improve soon because of government intervention.

Due to former Secretary of Health and Human Services Donna Shalala’s advisory committee on genetic testing, the Food and Drug Administration is already working to find a way to enact the committee’s recommendations, said Susanne Haga, policy analyst for the committee.

The committee recommended labs be held accountable for the predictability of their tests, and regulatory measures be enacted for genetic testing.


Mike Zacharias welcomes comments at [email protected]