A breast cancer cell. (Natl Cancer Inst)
In the usual trajectory of passing on genetic information, the older tell the younger, when the time is right. Typically, a patient has a genetic test because family history, ethnic group, or some other clue suggests to an astute practitioner an increased risk of something specific.
If a test reveals a mutation that could cause a disease, then the patient and perhaps her partner discuss how, when and what to tell their children – in the best of circumstances, with the help of a genetic counselor.
Direct-to-consumer (dtc) genetic testing has flipped the normal order of things. Young adults are taking dtc tests, some “just for fun,” and through their results, their parents – demographically at the cusp of beginning to fall apart – may learn just how they may do so.
I’m watching this generation reversal in genetic testing right now – and confusion began right away, with misuse of a single word.
'CARRIER' HAS MULTIPLE MEANINGS
A friend of a friend I’ll call Lisa called me last week with a question: her 22-year-old son Justin had driven to nearby Vermont to mail spit to the dtc company 23andMe (it’s illegal in New York, not to spit [except for NYC], but to use the company’s tests). He’d just showed her his DNA test results, with a surprise: he’s a “carrier” for one of two mutations in the BRCA1 cancer gene found mostly in people of Ashkenazi Jewish descent.
Lisa’s Ashkenazi; her husband’s not.
“But I don’t have to worry, right, because he’s only a carrier,” she said.
I hesitated. “Well, not exactly.”
“But I discussed it with three people and they all said it couldn’t affect me because I could just be a carrier, like Justin. Carriers of cystic fibrosis don’t actually get sick. So I won’t. Right?”
“In cancer it doesn’t usually work that way. Why don’t you come over and we can discuss it?”
I was worried, fearing not just high risk of cancer, but the company’s mixing up technical genetics terms with plain English. At fault: the word carrier. Forgive my foray into textbookese, but sloppy use of terminology requires explanation.
A carrier can be a healthy person spreading an infectious disease, like the cook Typhoid Mary, who spread typhoid fever to members of the unfortunate families she served in the New York City area in the early years of the twentieth century.
In genetics, in autosomal recessive inheritance, a carrier has one copy of a mutation and a normal (wild type) copy, and doesn’t have the associated condition. Someone gets the disease by inheriting two recessive mutations from two carrier parents. Lisa was right, carriers of cystic fibrosis don’t inherit full-blown disease (although rarely they can have mild symptoms).
In autosomal dominant inheritance, in contrast, just one copy of a mutant gene causes disease. There are no carriers. If you get the mutation, you get the disease. A person too young to have recognizable symptoms of Huntington disease (HD), for example, isn’t, technically, a carrier – instead he or she is “pre-manifest” or “pre-symptomatic,” in genetics jargon. Early on, HD researchers and clinicians agreed that people under age 18 shouldn't be tested, precisely because of the inevitability of the currently untreatable disease in "carriers."
BRCA mutations are also inherited in an autosomal dominant pattern, but not quite so straightforward as is the case for HD. Inheriting a BRCA mutation gives a person elevated risk, not a certainty of cancer. The classic "two-hit" hypothesis of cancer explains the situation: Justin inherited a susceptibility allele (gene variant) in all his cells, and cancer might develop if/when and where a mutation in the second copy of that gene occurs. That could be in his breast, testes, or prostate – or it might never happen.
But the inheritance of BRCA-related disease is even more complicated, because the risks of developing BRCA1 cancers vary in different population groups, due to the actions of other genes and environmental factors. And the numbers are all over the place, which is driving Lisa crazy. The chance of either breast or ovarian cancer happening by age 70 for a person with a BRCA1 mutation ranges from 40% to 86%, depending on how a particular study selected its families.
The information that Justin received from 23andMe on BRCA cancer is accurate: “Carrier for the 5382insC BRCA1 mutation. Lifetime risk of breast cancer for women is increased from 12% to about 60% and risk of ovarian cancer is increased from less than 2% to about 40%. May significantly increase risk of prostate cancer in men. There is also an increased risk for breast cancer in men.”
MENDEL’S LAWS STILL MATTER
When Lisa came over, I sketched a pedigree as she talked, filling in the circles and squares that represented family members with cancer. Unfortunately cancer has so many guises, and is so common, that its appearance in a pedigree can be coincidence, as it is in my own family.
In Lisa’s family, two cases of ovarian and breast cancer stood out, but the women were older than is typical for inherited cancers. And they weren’t primary relatives (parent, sib, or kid). “Incomplete penetrance” could explain the pattern of filled-in and blank symbols – Lisa and her mom silently passed the mutation on.
If Justin hadn’t sent his spit to 23andMe, likely no one would have suggested Lisa be tested for BRCA genes based on her family history.
As I filled in the square symbol at the bottom of the family tree representing Justin, the implications became clear. Lisa grew teary.
“It looks like if Justin has it, I would too. Unless my husband does – but he’s not Jewish,” she said quietly.
“Right.”
“Oh. So, if I’m not a carrier like in cystic fibrosis, that means … do I have the high risk of Ashkenazi Jews of getting cancer?”
“If you have the mutation, I’m afraid so. But the risk may be a little less because the cases in your family are older, and you and your mom don’t have cancer.”
"Some other gene is protecting me?"
"Possibly." She really got it, given the background that some who try to digest the voluminous material on the 23andme website don't have.
Alternatively, Justin could be a "new mutation," but that's unlikely. His DNA would have had to have mutated by chance in the exact same way that the ancestral Ashkenazi mutation did millennia ago.
I was struck by the irony, given recent debate over how we will handle the dense, overwhelming information that is pouring out from exome and genome sequencing. Lisa's case reveals the repercussions of a test on a single, intensely-studied gene on a family. Her distress derives from applying the work of a researcher who conducted his experiments in the 1880s – Gregor Mendel.
ACTION OR ANXIETY?
“Actionable” is a word we’ll be hearing more of as genome sequencing becomes routine. Knowing that her son has a BRCA1 mutation and she probably does too, Lisa’s having a breast exam, physical, and mammogram. She’ll have a BRCA1 test from Myriad Genetics, so insurance can cover the test and follow-up, because 23andMe is not (yet) a medical testing company and licenses their test from Myriad.
If Lisa has her son’s mutation, as her pedigree suggests, she’ll be followed closely. The American College of Obstetrics and Gynecology (ACOG) advises breast exams twice yearly, and mammograms and MRIs annually, to detect cancer when it's treatable. (Even the ACOG news release quotes a physician using “carrier” inappropriately, reflecting a persistent disconnect between genetics and medicine.) And I keep telling Lisa, as did her primary care provider, that in the absence of symptoms or other evidence, a genetic test does not a diagnosis make. She can have the genotype and never develop the phenotype.
But Justin and his sister Maya don’t want to think about, let alone act on, the BRCA1 mutation (that Maya has a 50% chance of having), or anything else lurking in their genomes. After all, Justin just took the tests for fun. “Not so much fun for me,” said his mother.
Justin’s attitude is understandable. Many young adults who’ve grown up in healthy families can’t imagine that one day they might have cancer, or heart disease, or depression, or have a child with a metabolic disease.
So once again I have mixed feelings about the ease with which people can learn their DNA sequences. If common terms like “carrier” are used in ambiguous ways, even by the test providers and the physicians they quote on their websites, and risk statistics vary depending on who makes up the experimental group -- for delivery of information on just one well known gene –- how is the public to fare when a genome sequence can be had for a few thousand dollars, in two days? I hear through the grapevine that this is already possible. And I’d like to know why my husband can take hundreds of genetic tests from 23andMe, without any contact with a health care provider, yet not be allowed to learn his PSA test results without permission from his physician. It's a crazy inconsistency.
I’m reminded of a situation that affected a large family group who had DNA marker tests from a direct-to-consumer genetic ancestry company. Tests revealed that the 96-year-old patriarch was not actually related to any of the others. The old man must have been an “unofficial adoption” – an orphan picked up by a caring couple who never told him -- the heads of the family deduced. And it had never mattered. The elders elected not to give the DNA news to the man, because his entire identity was based on his position in his family.
Yes, this is extreme paternalism. But perhaps it is prescient. Will one of my daughters one day tell me something’s in my immediate future that I might not want to know? And this isn’t just me having my head in the sand, as responses to my blog about not wanting to know my genome sequence suggested a few months ago.
I think from years of writing a human genetics textbook, I'm perhaps more aware than many people of just how complex our genomes are. And I know from years of science writing that we can think we know something -- and then a discovery changes everything. So I think at least some of the information being imparted to people now and in the near future will not have been validated – that is, one mutation might be counteracted by another, yet to be discovered. Some mutations or lesser variants will have effects in one genetic background but not another – like BRCA1 mutations posing different cancer risks in different populations.
There’s still a lot we don’t know, and we can't know what it is we don't yet know. So I’m not ready to be a guinea pig. Yet. But genome sequencing is (for now) a choice.
We can, though, perhaps prevent genetic information disasters. Caveats from genomics companies, conversations among family members so each can know what the others want to know, and more genetics education at all levels can do a lot to stave off upsetting use of genetic information.
Some of my fear of rash decisions based on seeing sequences of A, C, T and G is based partly on response to my book about gene therapy – middle and high-schoolers have no problems at all getting the science, but some of their parents and grandparents struggle. They’ve never learned what genes are and do. That leaves a lot of older people vulnerable to giving too much power to genetic information, of not understanding nuances like incomplete penetrance and one gene's expression affecting another's.
Let’s hope that we can enter and embrace this age of genomics while preserving individuals’ right to determine what they do and do not want to know.
(This blog is also posted at PLOS DNA Science blog.
If a test reveals a mutation that could cause a disease, then the patient and perhaps her partner discuss how, when and what to tell their children – in the best of circumstances, with the help of a genetic counselor.
Direct-to-consumer (dtc) genetic testing has flipped the normal order of things. Young adults are taking dtc tests, some “just for fun,” and through their results, their parents – demographically at the cusp of beginning to fall apart – may learn just how they may do so.
I’m watching this generation reversal in genetic testing right now – and confusion began right away, with misuse of a single word.
'CARRIER' HAS MULTIPLE MEANINGS
A friend of a friend I’ll call Lisa called me last week with a question: her 22-year-old son Justin had driven to nearby Vermont to mail spit to the dtc company 23andMe (it’s illegal in New York, not to spit [except for NYC], but to use the company’s tests). He’d just showed her his DNA test results, with a surprise: he’s a “carrier” for one of two mutations in the BRCA1 cancer gene found mostly in people of Ashkenazi Jewish descent.
Lisa’s Ashkenazi; her husband’s not.
“But I don’t have to worry, right, because he’s only a carrier,” she said.
I hesitated. “Well, not exactly.”
“But I discussed it with three people and they all said it couldn’t affect me because I could just be a carrier, like Justin. Carriers of cystic fibrosis don’t actually get sick. So I won’t. Right?”
“In cancer it doesn’t usually work that way. Why don’t you come over and we can discuss it?”
I was worried, fearing not just high risk of cancer, but the company’s mixing up technical genetics terms with plain English. At fault: the word carrier. Forgive my foray into textbookese, but sloppy use of terminology requires explanation.
A carrier can be a healthy person spreading an infectious disease, like the cook Typhoid Mary, who spread typhoid fever to members of the unfortunate families she served in the New York City area in the early years of the twentieth century.
In genetics, in autosomal recessive inheritance, a carrier has one copy of a mutation and a normal (wild type) copy, and doesn’t have the associated condition. Someone gets the disease by inheriting two recessive mutations from two carrier parents. Lisa was right, carriers of cystic fibrosis don’t inherit full-blown disease (although rarely they can have mild symptoms).
In autosomal dominant inheritance, in contrast, just one copy of a mutant gene causes disease. There are no carriers. If you get the mutation, you get the disease. A person too young to have recognizable symptoms of Huntington disease (HD), for example, isn’t, technically, a carrier – instead he or she is “pre-manifest” or “pre-symptomatic,” in genetics jargon. Early on, HD researchers and clinicians agreed that people under age 18 shouldn't be tested, precisely because of the inevitability of the currently untreatable disease in "carriers."
BRCA mutations are also inherited in an autosomal dominant pattern, but not quite so straightforward as is the case for HD. Inheriting a BRCA mutation gives a person elevated risk, not a certainty of cancer. The classic "two-hit" hypothesis of cancer explains the situation: Justin inherited a susceptibility allele (gene variant) in all his cells, and cancer might develop if/when and where a mutation in the second copy of that gene occurs. That could be in his breast, testes, or prostate – or it might never happen.
But the inheritance of BRCA-related disease is even more complicated, because the risks of developing BRCA1 cancers vary in different population groups, due to the actions of other genes and environmental factors. And the numbers are all over the place, which is driving Lisa crazy. The chance of either breast or ovarian cancer happening by age 70 for a person with a BRCA1 mutation ranges from 40% to 86%, depending on how a particular study selected its families.
The information that Justin received from 23andMe on BRCA cancer is accurate: “Carrier for the 5382insC BRCA1 mutation. Lifetime risk of breast cancer for women is increased from 12% to about 60% and risk of ovarian cancer is increased from less than 2% to about 40%. May significantly increase risk of prostate cancer in men. There is also an increased risk for breast cancer in men.”
MENDEL’S LAWS STILL MATTER
When Lisa came over, I sketched a pedigree as she talked, filling in the circles and squares that represented family members with cancer. Unfortunately cancer has so many guises, and is so common, that its appearance in a pedigree can be coincidence, as it is in my own family.
In Lisa’s family, two cases of ovarian and breast cancer stood out, but the women were older than is typical for inherited cancers. And they weren’t primary relatives (parent, sib, or kid). “Incomplete penetrance” could explain the pattern of filled-in and blank symbols – Lisa and her mom silently passed the mutation on.
If Justin hadn’t sent his spit to 23andMe, likely no one would have suggested Lisa be tested for BRCA genes based on her family history.
As I filled in the square symbol at the bottom of the family tree representing Justin, the implications became clear. Lisa grew teary.
“It looks like if Justin has it, I would too. Unless my husband does – but he’s not Jewish,” she said quietly.
“Right.”
“Oh. So, if I’m not a carrier like in cystic fibrosis, that means … do I have the high risk of Ashkenazi Jews of getting cancer?”
“If you have the mutation, I’m afraid so. But the risk may be a little less because the cases in your family are older, and you and your mom don’t have cancer.”
"Some other gene is protecting me?"
"Possibly." She really got it, given the background that some who try to digest the voluminous material on the 23andme website don't have.
Alternatively, Justin could be a "new mutation," but that's unlikely. His DNA would have had to have mutated by chance in the exact same way that the ancestral Ashkenazi mutation did millennia ago.
I was struck by the irony, given recent debate over how we will handle the dense, overwhelming information that is pouring out from exome and genome sequencing. Lisa's case reveals the repercussions of a test on a single, intensely-studied gene on a family. Her distress derives from applying the work of a researcher who conducted his experiments in the 1880s – Gregor Mendel.
ACTION OR ANXIETY?
“Actionable” is a word we’ll be hearing more of as genome sequencing becomes routine. Knowing that her son has a BRCA1 mutation and she probably does too, Lisa’s having a breast exam, physical, and mammogram. She’ll have a BRCA1 test from Myriad Genetics, so insurance can cover the test and follow-up, because 23andMe is not (yet) a medical testing company and licenses their test from Myriad.
If Lisa has her son’s mutation, as her pedigree suggests, she’ll be followed closely. The American College of Obstetrics and Gynecology (ACOG) advises breast exams twice yearly, and mammograms and MRIs annually, to detect cancer when it's treatable. (Even the ACOG news release quotes a physician using “carrier” inappropriately, reflecting a persistent disconnect between genetics and medicine.) And I keep telling Lisa, as did her primary care provider, that in the absence of symptoms or other evidence, a genetic test does not a diagnosis make. She can have the genotype and never develop the phenotype.
But Justin and his sister Maya don’t want to think about, let alone act on, the BRCA1 mutation (that Maya has a 50% chance of having), or anything else lurking in their genomes. After all, Justin just took the tests for fun. “Not so much fun for me,” said his mother.
Justin’s attitude is understandable. Many young adults who’ve grown up in healthy families can’t imagine that one day they might have cancer, or heart disease, or depression, or have a child with a metabolic disease.
So once again I have mixed feelings about the ease with which people can learn their DNA sequences. If common terms like “carrier” are used in ambiguous ways, even by the test providers and the physicians they quote on their websites, and risk statistics vary depending on who makes up the experimental group -- for delivery of information on just one well known gene –- how is the public to fare when a genome sequence can be had for a few thousand dollars, in two days? I hear through the grapevine that this is already possible. And I’d like to know why my husband can take hundreds of genetic tests from 23andMe, without any contact with a health care provider, yet not be allowed to learn his PSA test results without permission from his physician. It's a crazy inconsistency.
I’m reminded of a situation that affected a large family group who had DNA marker tests from a direct-to-consumer genetic ancestry company. Tests revealed that the 96-year-old patriarch was not actually related to any of the others. The old man must have been an “unofficial adoption” – an orphan picked up by a caring couple who never told him -- the heads of the family deduced. And it had never mattered. The elders elected not to give the DNA news to the man, because his entire identity was based on his position in his family.
Yes, this is extreme paternalism. But perhaps it is prescient. Will one of my daughters one day tell me something’s in my immediate future that I might not want to know? And this isn’t just me having my head in the sand, as responses to my blog about not wanting to know my genome sequence suggested a few months ago.
I think from years of writing a human genetics textbook, I'm perhaps more aware than many people of just how complex our genomes are. And I know from years of science writing that we can think we know something -- and then a discovery changes everything. So I think at least some of the information being imparted to people now and in the near future will not have been validated – that is, one mutation might be counteracted by another, yet to be discovered. Some mutations or lesser variants will have effects in one genetic background but not another – like BRCA1 mutations posing different cancer risks in different populations.
There’s still a lot we don’t know, and we can't know what it is we don't yet know. So I’m not ready to be a guinea pig. Yet. But genome sequencing is (for now) a choice.
We can, though, perhaps prevent genetic information disasters. Caveats from genomics companies, conversations among family members so each can know what the others want to know, and more genetics education at all levels can do a lot to stave off upsetting use of genetic information.
Some of my fear of rash decisions based on seeing sequences of A, C, T and G is based partly on response to my book about gene therapy – middle and high-schoolers have no problems at all getting the science, but some of their parents and grandparents struggle. They’ve never learned what genes are and do. That leaves a lot of older people vulnerable to giving too much power to genetic information, of not understanding nuances like incomplete penetrance and one gene's expression affecting another's.
Let’s hope that we can enter and embrace this age of genomics while preserving individuals’ right to determine what they do and do not want to know.
(This blog is also posted at PLOS DNA Science blog.