[MUSIC] GRAY: Cancer is one of the most complex
diseases that mankind has to face. We used to think that cancers were more or
less similar as long as you stayed within one anatomic site, like breast cancers tended to
be similar to one another and colon cancers tended to be similar to one another.
What we now know is that’s not at all true. HAYES: If you look at lung cancer, the number
one cancer killer in the world by a mile, the number one variant of this tumor is called
non-small cell lung cancer. If you think about that for a second, that means “we don’t know what it is,
but it’s not small-cell.” That’s the diagnosis that most people in the world carry today when they’re
diagnosed with the number one cancer killer. So, it’s a disease, which is not further classified
in most cases and I will acknowledge that there have been some very interesting developments
over the last five years. For most patients, those developments still are not part of the treatment paradigm.
So, I’m looking for rethinking the number one cancer killer in the world. LEVINE: If we can figure out what type of genomic
defects exist in a given tumor for an individual person, we can then try to tailor this therapy
towards that individual person. HAYES: In glioblastoma we were able to generate
an amazing data set of integrated data with gene expression and copy number mutation. One of the
things that we noticed was groups of patients with glioblastoma who had very different patterns of disease.
We called these subtypes and this type of subtype discoveries something we do a lot in cancer research.
One of the important things is that the subtypes look like they differed by one of our most exciting therapeutic
targets and that’s the epidermal growth factor receptor. LEVINE: In the past, many targeted trials have
failed because they’ve taken one drug to treat a whole population of patients. But if you can
identify the best population to match a given drug with, you’re more likely to achieve success,
and even great success, in a smaller fraction of patients who have this disease. CHIN: That type of success example is driving
the effort to count in parallel characterize the changes as we’re doing in TCGA, but in parallel,
also influencing a shift in clinical practice, where more and more physicians and more and more
hospitals are beginning to think about how they can profile their patients coming in so that they
have the information and match them up to the right therapy. LINEHAN: I can see a day not too far from now,
when physicians like me would be doing that routinely and that would then let us pick a directed
therapy, a personalized therapy for that patient’s cancer. We set out nearly thirty years ago to identify the genes
that cause cancer of the kidney. It took us ten years to find our first gene. Now, we’re very happy about that.
We can use that gene to make the diagnosis earlier in a whole lot of families and in patients with this disorder,
with this cancer. But also, now the FDA has approved six drugs, six drugs for patients with advanced
cancer based on that first cancer gene. What it took us ten years to do with the Cancer Genome
Atlas, we could probably now do in six months. It’s just so much faster. Even though we’ve
come a long way, we have a long way to go. [MUSIC]