Why Genetic Test Results Require So Much Interpretation

Anyone who’s been to a doctor or even an urgent care clinic lately has probably received a test result from a diagnostic laboratory. From cholesterol levels to infectious disease detection to blood cell counts, these results dictate many of our medical decisions. Unfortunately, the certainty of those types of results has left us unprepared for the complexity and probabilities inherent in many genetic test results. This is a real problem, especially as more and more people are encouraged to get genetic testing for an increasingly broad range of healthcare uses.

The diagnostic tests we’re used to are mostly black-and-white, reporting simple numerical values or yes-or-no answers. Genetic testing, on the other hand, includes many shades of gray. While some results are straightforward — yes, we found this genetic variant, or no, we didn’t — a lot of them are based on subjective judgment and interpretation.

This is a reflection of where we are in the genomic era: very much still learning. At the annual meeting of the American College of Medical Genetics and Genomics in Baltimore last week, presentations from a number of speakers highlighted the careful attention — and angst — that goes into decisions about which results are reported and how they are reported back to patients and physicians.

To understand the challenge, consider how genetic variants are classified. The clinical and research communities have done heavy lifting in the past few decades to determine whether genetic variants are disease-causing (pathogenic) or harmless (benign). That has led to four broad categories of variants: pathogenic and likely pathogenic, and benign and likely benign. But there’s a chasm in between these ends of the spectrum populated by variants of unknown significance. (To be clear: this is not the case for genetic tests that simply look for the presence or absence of a known variant.)

These unclassified variants are the ones that keep genetic professionals up at night because we don’t yet know whether they’re harmful or harmless. Each of us has a genome that harbors millions of genetic variants, including hundreds that might be unique to us. Even with all of the genomes that have been sequenced so far, each new genome analyzed still presents plenty of opportunity to find variants that have never been seen before. Between new variants and variants that simply haven’t been studied enough yet for scientists to understand their function, there’s a giant pool of these “unknown significance” findings.

When a genetic test is run — particularly one where researchers are trying to find the cause of a disease — the data is generated and then analyzed by a team of specialists. Variants that have already been classified can be sorted fairly easily. For the rest, analysts will comb through scientific studies looking for any mention of each variant, carefully weighing the confidence of each report and its relevance to the case at hand. They’ll also consult databases of genetic variants, discuss options among the team and possibly with outside experts, and maybe even run follow-up tests to try to learn more.

At the end, though, the team still has to decide which variants to report — and which to leave out. In many clinical labs, only the classified variants are reported back to patients and physicians; variants of unknown significance are left out. If you know me personally, you know that I advocate for broad information sharing in healthcare. The medical field as a whole has an uncomfortably paternalistic view about what patients should and shouldn’t know, and I think it’s high time that changed. I believe we are all entitled to any health-related information generated about us.

However, the presentations at the ACMG meeting had me questioning that position. Consider the case of a patient with an undiagnosed rare disease, where the genetic analysis team finds a variant that looks like it might be related, but there is no clear scientific evidence or precedent to back it up. That variant may have nothing to do with the disease at hand, but if the team reports it — even noting that the variant’s significance is unclear — it could bias the physicians treating the patient. They might change treatment course as a result, potentially harming the patient, or fail to consider other diagnoses due to a belief that this variant is the answer. In a field where “do no harm” is the primary mandate, releasing information in a formal clinical report might actually do more harm than good.

There is still a case to be made that patients deserve access to their data, but what these talks really illustrated for me is how early we are in our understanding of genomic medicine. Reports from genetic testing can make it seem like our knowledge is more advanced or more certain than it actually is, and that’s something patients should be aware of when they consider their own genetic analysis results.

Short Takes

Earlier this week, a federal judge blocked many of the changes to childhood vaccine recommendations announced by the U.S. Department of Health and Human Services back in January. Even though the block is temporary, it’s good news for people pushing to keep kids as safe as possible — and to ensure that future vaccine recommendations are based firmly on scientific evidence.

In a first, the rate of cigarette smoking among adults in the U.S. dropped below 10%. The new study that analyzed the data found that use of tobacco products overall is still nearly double that, but lowering cigarette use is at least some progress in efforts to reduce the incidence of lung cancer.

A second astronomy mission has confirmed that the basic DNA and RNA building blocks of life have been found on asteroids, supporting the idea that life on Earth may have been seeded from space.