DNA sequencing of tumour enables ultimate personalised cancer
10 April 2012
DNA sequencing technology has been used to not only identify
mutations at the root of a patient’s tumour but to map the genetic
evolution of disease and monitor response to treatment at Washington
University School of Medicine in St Louis.
“We’re finding clinically relevant information in the tumour
samples we’re sequencing for discovery-oriented research studies,”
says Elaine Mardis, PhD, co-director of The Genome Institute at the
School of Medicine. “Genome analysis can play a role at multiple
time points during a patient’s treatment, to identify ‘driver’
mutations in the tumour genome and to determine whether cells
carrying those mutations have been eliminated by treatment.”
This work is helping to guide the design of future cancer
clinical trials in which treatment decisions are based on results of
sequencing, says Mardis, who is speaking April 1 at the opening
plenary session of the American Association for Cancer Research
annual meeting in Chicago. She also is affiliated with the Siteman
Cancer Center at the School of Medicine and Barnes-Jewish Hospital.
To date, Mardis and her colleagues have sequenced all the DNA of
tumour cells from more than 700 cancer patients. By comparing the
genetic sequences in the tumour cells to healthy cells from the same
patient, they can identify mutations underlying each patient’s
Already, information gleaned through whole-genome sequencing is
pushing researchers to reclassify tumours based on their genetic
makeup rather than their location in the body. In patients with
breast cancer, for example, Mardis and her colleagues have found
numerous driver mutations in genes that have not previously been
associated with breast tumours.
A number of these genes have been identified in prostate,
colorectal, lung or skin cancer, as well as leukemia and other
cancers. Drugs that target mutations in these genes, including
imatinib, ruxolitinib and sunitinib, while not approved for breast
cancer, are already on the market for other cancers.
“We are finding genetic mutations in multiple tumour types that
could potentially be targeted with drugs that are already
available,” Mardis says.
She predicts, however, that it may require a paradigm change for
oncologists to evaluate the potential benefits of individualized
cancer therapy. While clinical trials typically involve randomly
assigning patients to a particular treatment regimen, a personalized
medicine approach calls for choosing drugs based on the underlying
mutations in each patient’s tumour.
“Having all treatment options available for every patient doesn’t
fit neatly into the confines of a carefully designed clinical
trial,” Mardis acknowledges. “We’re going to need more flexibility.”
When during the course of cancer mutations develop also is likely
to be important in decisions about treatment. In a recent study,
Mardis and her team mapped the genetic evolution of leukemia and
found clues to suggest that targeted cancer drugs should be aimed at
mutations that develop early in the course of the disease.
Using “deep digital sequencing,” a technique developed at The
Genome Institute, they sequenced individual mutations in patients’
tumour samples more than 1,000 times each. This provides a read-out
of the frequency of each mutation in a patient’s tumour genome and
allowed the researchers to map the genetic evolution of cancer cells
as the disease progressed.
They found that as cancer evolves, tumours acquire new mutations
but always retain the original cluster of mutations that made the
cells cancerous in the first place. Their discovery suggests that
drugs targeted to cancer may be more effective if they are directed
toward genetic changes that occur early in the course of cancer.
Drugs that target mutations found exclusively in later-evolving
cancer cells likely may not have much effect on the disease because
they would not kill all the tumour cells.
Mardis says that sequencing the entire genome of cancer cells is
essential to piecing together an accurate picture of the way cancer
cells evolve. If the researchers had sequenced only the small
portion of the genome that involves genes, they would not have had
the statistical power to track the frequency of mutations over time.
(Only 1 to 2 percent of the genome consists of genes.)
In another study, a phase III clinical trial of post-menopausal
women with estrogen-receptor positive breast cancer, the Washington
University researchers have shown that sequencing can help to
predict which women will respond to treatment with aromatase
inhibitors. These estrogen-lowering drugs are often prescribed to
shrink breast tumours before surgery. But only about half of women
with estrogen-receptor positive breast cancer respond to these
drugs, and doctors have not been able to predict which patients will
Interestingly, by sequencing patients’ breast tumours before and
after aromatase inhibitor therapy, the researchers identified
substantive genomic changes that had occurred in responsive
patients, whereas the genomes of unresponsive patients remained
largely unchanged by the therapy.
“No one has ever looked at treatment response at this level of
resolution,” Mardis says. “It’s so obvious who is responding.”
In addition, the researchers have identified a series of
mutations in the breast tumours that have corresponding
small-molecule inhibitor drugs that target defective proteins. This
finding indicates that for women who are not responding to aromatase
inhibitors, treatment options may include combining conventional
chemotherapy with the indicated small-molecule inhibitor.
“We felt it was important to show there could be therapeutic
options available to patients who are resistant to aromatase
inhibitors,” Mardis says. “As we move forward, we think sequencing
will contribute crucial information to determining the best
treatment options for patients.”