DALLAS
— ON Aug. 6, researchers announced in The New England Journal of
Medicine that they had found that mutations in a gene called PALB2
greatly increase the risk of breast cancer.
This is one of the biggest developments since the discovery in the ’90s
of the role of mutations in the BRCA1 and BRCA2 genes in breast and ovarian cancer.
The
response among patients has been predictable. One woman’s email to me
summed it up: “I’d like to get an entire genome scan to rule out a
hidden cancer diagnosis.”
Genetic
testing has revolutionized how we think about cancer, allowing us to
make some decent predictions about who might get certain cancers and who
might benefit from preventive treatments. Many know the story of Angelina Jolie, who used her family history to learn she had a BRCA1 mutation. She chose to have a double mastectomy
instead of waiting to see if she developed cancer. We may not envy this
choice, but we do appreciate the power that comes with taking
evidence-based action against a deadly disease. Many of us want that
power for ourselves.
The
problem is that many patients think genetic testing can tell us far
more than it does. Despite the exaggerated claims of some entrepreneurs
and lab owners, we can’t predict patients’ cancer risk and advise them
appropriately just by sequencing their genome. At least not yet.
First,
it’s important to know the difference between sequencing a tumor and
sequencing a germ line (which is what the woman who emailed me wanted).
The
germ line genome is found in every cell of the body and encodes traits
such as eye color and the tendency to develop diseases like cystic fibrosis. It is the blueprint you inherited from your parents.
The
tumor genome is like an ugly addition on top of the original
construction. It’s acquired rather than inherited, and contains a mess
of new mutations, a subset of which cause the cells to proliferate out
of control and transform into cancer.
Sequencing
a portion of a tumor genome can sometimes tell us what went wrong and
how the cancer might be treated more effectively. For instance, testing
people with melanomas for mutations in their BRAF genes can influence
treatment. But it’s unclear how many people would benefit from
sequencing all the genes in a tumor; it, like full germ line sequencing,
remains mainly a research tool.
Both
tumor and full germ line sequencing suffer from the same problem: The
whole genome doesn’t make a lot of sense to us yet. We have a very
limited understanding of which variations in DNA contribute to disease
development. Most of your genome sequence looks the way Shakespeare does
to a toddler — incomprehensible.
Conscientious
doctors won’t order a lab test that they can’t understand, so they’re
unlikely to say yes to a patient’s request for full genome sequencing.
And insurance companies shouldn’t pay for a test if there isn’t research
backing up its use.
The
genetic tests that are clearly valuable at this point for predicting
cancer risk are those for the specific mutations that we understand. But
even then there can be other mutations in the same genes that are of
unknown clinical significance. So the best results come from sequencing
the subset of genes that could potentially explain patterns of disease
observed in the family tree. If your family shows an increased incidence
of certain cancers, it makes sense to look for known mutations that
increase the risk of those cancers.
For example, we recently saw a 25-year-old patient at our clinic with a strong family history of colon cancer.
His father had colon cancer, but had refused any kind of genetic
testing. When the son tested positive for a broken colon cancer gene,
the rest of his family followed. Those who tested positive now know they
need to have frequent colonoscopies.
And most people with strong family histories of breast cancer should
consider getting genetic counseling and, if recommended, testing for
mutations in genes such as PALB2, BRCA1 and BRCA2.
Family
history doesn’t just help us decide where to look for broken genes; if
we find a mutation that could lead to cancer, the number of cancers in
the family helps us determine what kind of risk that mutation might
carry. Some patients with BRCA mutations have a 40 percent chance of
developing breast cancer by age 70, but for others, the chance has been
estimated to be as high as 87 percent. It looks as if this is the case
for PALB2 mutations as well. According to the New England Journal of
Medicine study, by Marc Tischkowitz at the University of Cambridge and
others, patients with a broken PALB2 gene and no family history of
breast cancer have a 35 percent chance of developing breast cancer, but
if they had two or more family members with cancer, the risk rises to 58
percent. This is significant when you’re considering life-altering
preventive surgery or screenings.
Genetic
testing is a process for many patients, not a one-time event. In the
coming weeks, in light of the report on PALB2, as well as the emergence
of sophisticated tests that can now analyze many cancer genes at one
time, we’ll advise former patients who tested negative for the BRCA or
other specific genes to return for more genetic counseling and
potentially further testing.
One
patient, who is now healthy but who dealt with a breast cancer
diagnosis in her early 30s, emailed to ask if she’d been tested for the
PALB2 gene when she had the BRCA test in 2007. We’d found no mutations,
and she seemed to be a genetic mystery. My reply was easy: She hadn’t
been tested for PALB2, but she should be now, along with any other new
cancer genes. If she tests positive, we’ll know more about the risk of
her having a second cancer, and be able to test her family members.
For
any patient who still wants to get his or her full genome sequenced,
there is one good reason to do it: scientific curiosity. Something of
interest could come of it as the technology improves over time. But it’s
worth doing only if a clinical trial is paying for it, or if you can
easily afford it. Getting the data can cost under $2,000, but having the
data interpreted can cost in the tens or hundreds of thousands. It’s a
fine purchase, but definitely a luxury. And bear in mind that the data
that come back will raise more questions than they answer. There is a
good chance that mutations of unknown clinical significance will be
found that will increase anxiety without illuminating any known risks.
We
have to have patience with the pace of research. We can now use genetic
information to prevent breast, ovarian and colon cancer in many
patients, and we will get better at this. In the meantime, most people
should focus less on the high-tech future of genetic testing and more on
the low-tech history of their family trees. Those who don’t know their
own family histories, because of adoption, secrecy, loss or
estrangement, should take comfort in the fact that we are one big
family. Data in aggregate from many families, gathered together
in studies like Dr. Tischkowitz’s, will eventually teach us how to
manage our risks, how to treat disease and how to save lives.
Comments
Post a Comment