Finally, a Truthful
Assessment of Cancer Genomics: For Now, Everyone is An N of 1
Nothing, and I mean nothing, in all of modern science has
more infuriated me as a former medical oncologist and former cancer
pharmacologist than the horse hockey that I have heard about the power of
cancer genomics. I have written about it in the lay press and in this space
Let’s summarize briefly some points I have made in the past:
proclivities for the development of cancer is a fine research concept as long
as it is done under the supervision of an IRB with full informed consent and authorization. To do
this clinically on a routine basis is also fine as long as the patient (or
subject for these people are often among the worried well with no symptoms or
signs of cancer) has a clear understanding of the risks vs. benefits of both
the testing, the value or harm of the test results to family members and the
possible clinical responses available if unusual risk of malignancy is identified
(e.g., bilateral mastectomy of healthy breast tissue or total colectomy).
Counseling is imperative and each patient/subject is an N of 1. (Shouldn’t all
patients be? Don’t YOU want to be? Hasn’t that always been the essence of
medicine—to be a ‘personalized’ relationship between one physician and one patient and his/her family?)
testing should be very limited to those clinical situations where the testing
might lead to an intervention that might alter the tested individual’s life.
definite mutation associated with a successful intervention is identified and
then it should be fully covered as all such lab tests are.
the benefits of doing these expensive tests on all cancer patients and having
either insurers or the patients themselves paying for it, don’t be surprised
that the FDA wants in on oversight of the validity, safety, risk and
therapeutic response to this new technology. I hate it, but don’t be surprised
by it. In my opinion, genomics has been overhyped as was tumor stem cell assays to assess
chemosensitivity and high dose chemotherapy for primary breast cancer. The
cancer community has not always behaved well tripping over itself for more
money from the NIH or private sources. This FDA response should be
expected, albeit unfortunate.)
genetic profile data of tumor genomic analysis (e.g., in osteosarcoma samples)
is the result of the carcinogenic process not its cause? This would surely be
consistent with the Fidler/Kripke model of cancer heterogeneity and genetic instability that has been
widely accepted for decades.
I have been writing this stuff and sure a lot of people resent it because they
have a lot of eggs in the genomic basket. Some truly believe in the curative power of this stuff and
some are afraid to buck the trend that has led to the raising of millions of
dollars to do this work and to erect institutes in which to do it. In fact, the genomic apologists really want their cut.
genomics is good business.
like a refreshing breeze from Dallas comes the Opinion piece in today’s NY
Times by one our UT colleagues Theodora Ross at UT SW.
cannot say it any better though I have tried.
urge instant memorization of the details here and handing it out to all of your
patients. Here is a real deal director of a cancer genetics program at a sister
institution giving us her obviously way-better-than-mine insight into the state
of the art.
the belief, particularly by the government, that you can treat every patient
with algorithms, guidelines and protocols, the results of genomic testing is
actually showing us, that every single cancer patient may be different than
every other one. Individual risk of developing cancer varies. The cancer
developed may look the same under the microscope but be driven by or
characterized by completely different genetic, epigenetic, and biochemical etiologies. And
we really don’t know in most cases if the aberrant genetics is cause or effect.
As Dr. Ross says: “We have to have patience with the pace of research…we will
get better at this.”
only further comment: READ AND THEN READ AGAIN!
job Dr. Ross!!!!