Genetics may soon take much of the guesswork out of treating cancer. As this
ScienCentral News video reports, breast cancer tumors have genetic signatures
that could tell doctors how to treat them.
Personalized Medicine Goes Clinical
Oncologists are beginning to use powerful new genomics tools to customize cancer
treatment. Breast cancer patients at the Netherlands Cancer Institute (NKI)
will be the first in the world to have their tumor cells genetically profiled
to decide whether they should have aggressive chemotherapy.
"The mere fact that patients having the same disease respond very differently
to the same treatment tells us that these cancers are not the same and require
individualized treatment, based on knowledge of the pathways that are defective
in those tumors," says Rene
Bernards, head of NKI's Division of Molecular Carcinogenesis.
"There may be some patients who today are being treated because people
are unsure whether or not their disease is going to relapse, whether it's
going to metastasize," explains Steve
Friend, President of Rosetta Inpharmatics and a Senior Vice President
at Merck & Company.
"Maybe there are some patients who actually don't need that treatment.
In turn there may be some patients who appear to have early disease but in
whom one's not certain whether to give therapy."
The profiling at NKI in Amsterdam is based on research the Dutch and American
scientists published in the journals Nature
and the New England
Journal of Medicine. The oncologists collected breast tumor samples
from 78 patients a decade ago, and kept careful records of the womenâ€™s
outcomes. At Rosetta, in Kirkland, WA, geneticists used DNA microarrays to
measure the activity of all the genes in the tumor cells.
"That allowed us to find out, of all the 25,000 genes that might change
in the cells, a very small subset… just 70 genes… which are actually
the core signature that allowed us to say there's a tumor that's likely to
go on and metastasize and cause aggressive disease," Friend says.
Based on the tumor signatures, the researchers were able to group patients
into two groups: one with a good prognosis likely to do well without chemotherapy,
and another with a poor prognosis indicating aggressive treatment. When they
compared those predictions with records of patient outcomes, they found the
test's prognosis was more accurate than all other current clinical markers
combined. "So the real hope is being able to take the current markers
and complement them with the signature," Friend says. “[That] will
really increase the information that will be available to the patient, and
to the doctor, to make these decisions.”
Bernards says that in the U.S., introduction of the microarray test on a large
scale won't be done without validation in clinical trials, which will take
two years to complete. But, he adds, "we will discuss with the relevant
authorities in the U.S. the possibility to use the chip test for patients
that are considered to be low risk by the current criteria. We estimate that
some 30 percent of those patients are in fact at high risk for recurrent disease,
and it would be useful to be able to identify those patients as soon as possible."
"When you're making that decision whether you want to undergo chemotherapy,
you don't want to be thrown in among a group of 10,000 women" with a
good or poor prognosis, says Erich
Huang of Duke University School of Medicine. "You want to know what
your personal prognosis is."
Huang, his father Andrew Huang and colleagues at Duke used gene chips to profile
breast tumors from 89 patients of the Koo
Foundation Sun Yat-Sen Cancer Center in Taipei, Taiwan. They reported
in The Lancet
that a more complex profile comprising nearly 500 related clusters of genes
could narrow risk predictions down to the individual patient level.
In a second paper in the journal Nature
Genetics, the group reported on research in mice that combines clinical
markers with microarray analysis, giving even better accuracy. Clinical trials
to validate their method will likely take place in Taiwan before it can be
introduced in the U.S.
Other research groups are developing ways to use microarrays to guide treatment
of other cancers, including prostate cancer, lymphoma and lung
"This is personalized medicine in its earliest form and just the tip of
the iceberg," says NKI's Bernards. "I am convinced that in 10 years'
time nearly every treatment of cancer will start with some form of profiling."
"I think over the next decade we'll begin to find markers for other diseases,"
says Friend, "whether it's diseases such as cancer, but more importantly
diseases such as obesity
and other diseases that we've really had no ways of tracking how a patient
was going to respond to a drug or to a treatment."