Decoding variants: Analyzing genetic test results
A patient’s genetic test reports variants. How do you dig deeper? What’s next?
This is the second in a multi-part podcast with my collaborator Dr. Jonathan Golob. In the first podcast, we discussed the basics of genetic screening tests for cancer.
In this episode we look at the actual results that you’d get from a genetic cancer screening test. We discuss how a physician or geneticist goes about investigating a particular variant. We use the NCBI’s website to explore two specific variants:
- BRCA2 c.3865_3868del (p.Lys 128(Alafs*3)) (pathogenic)
- ATM c.7502A>G (p.Asn2501Ser) (variant of unknown significance, VUS)
A link to the full podcast, summary of what we discuss, as well as a transcript of highlights from the podcast follow. We also include short videos explaining our exploration of each of these two variants.
This podcast assumes a basic knowledge of genetics.
What we discuss:
- Variants are initially flagged based on deviations from reference genomes. They may or may not affect the protein coding sequence.
- Nomenclature of variants: For e.g. What do “BRCA2 c.3865_3868del (p.Lys 128(Alafs*3))” and “ATM c.7502A>G (p.Asn2501Ser)” mean?
- How this information is used to guide clinical care (using the ClinVar tool). Case where a known pathogenic variant has not yet guided clinical practice.
- A more detailed look at both the BRCA2 and ATM variants of interest with three minute video walk-throughs on the trusted NCBI website.
- A discussion of population analysis. The need for diverse and representative genetic data in databases to accurately assess variants across different populations.
Highlights from the podcast:
What you get back from a genetic test result:
“What you end up with is sort of a list, and then a shorthand that is used describe when there is a variant ... So there’s three letter codons that each correspond to an amino acid, but most amino acids can have more than one codon all going to the same place. So there’s some tolerance if you have a mutation in there without a change in the protein coding sequence”.
“You can also describe this in the level of DNA the DNA at this location is no longer an A, it is a T and that may or may not have an effect on the protein coding sequence.”
BRCA2 mutation: Nomenclature and clinical significance
Nomenclature:
“The words of DNA that go into making an amino acid the building blocks of a protein are they are three letter words. So if you’re missing four what this causes after that point in the gene is what’s called a frame shift mutation”.
“Because all of the words are different there, so this results in a massive change in the gene at least the protein coding level. And so if the gene is important, that gene is a very likely to not function at least in the same way. If not be completely unmakeable at all by the cell.”
Clinical significance of this mutation:
“That is just so absolutely in gene BRCA2. That is very important in breast cancer and so that’s why this is one of the best studied genetics susceptibilities to answer. Even here you can see (on clinical significance) it’s not all the way to four stars which is fascinating in and up itself “
“So the difference between the three star is that this has been reviewed by an expert panel versus four, has been integrated as part of practice guidelines such as a change in screening or prospective mastectomy. In this case in this particular variant has not reached that level of certitude, it is clearly pathogenic but has not yet had the sort of studies done to support it driving by default, a change in a care.
That does not eliminate a good discussion with someone’s provider about doing something about it.
What this means for a patient with this result:
“If the motivation for ordering the test was say a family history, a close family member who had breast cancer either aggressive or discovered early or both and then one orders a genetic test and this result comes, it all is telling a very same story. That’s where having a discussion with a trusted physician to really interpret it and then add in the person’s age, their risk tolerance and their desires.”
Walk-through on the NCBI website, of the BRCA2 mutation (3 min):
ATM mutation: Nomenclature and clinical significance:
Nomenclature:
“It’s a single nucleotide variant. It’s only one base pair long and it is a missense. So, this does affect the amino acid in this protein. The protein will still be made, but you’re going from an Asparagine to Serine”
Clinical significance of this mutation:
“People had different arguments about what disease are we even considering? So there’s some broadly cancer diagnosis and the general consensus is that some people are saying that it’s probably benign some are making no arguments about it. .This is a nice summary of it.”
What this means for a patient:
“When this variant that is observed in that particular patient has been observed multiple times in the past makes us think a little bit more that at least it is real, and it’s significance is very much up in the air here.”
Walk-through on the NCBI website, of the ATM mutation (3 min):
On population frequency analysis:
“So if you have a big room of people some of whom have increased risk and some of whom do not have increased risk for breast cancer and in this case, if there’s a gene or a variant of a gene that is assisted with breast cancer risk..
One can look for the prevalence of a variant in cases of people with breast cancer and controls matched as well as you can without, or you can do it as a cohort you take the room and then divide that room in half where the room is selected and by definition you’re not doing any matching.”
To conclude:
“I think bringing back to the focus of an individual patient. This is both the excitement and the challenge of this kind of study or these kind of tests is that bad news is the most interpretable.”
“Unfortunately, variants that have clear pathogenic potential are the ones that are probably going to be the most actionable for an individual patient. Whereas for all the complexities that we’ve just mentioned there’s lots of different ways for a gene to vary. It’s pretty hard to do the study to accumulate enough patients to understand the significance particularly of rare variants that I would expect the default answer for most variants that come out of a genetic test to be unknown significance.”
