[Ip-health] New Yorker: Open-Sourcing a Treatment for Cancer

Thiru Balasubramaniam thiru at keionline.org
Fri Feb 28 08:15:43 PST 2014


FEBRUARY 27, 2014
POSTED BY GARY MARCUS<http://www.newyorker.com/magazine/bios/gary_marcus/search?contributorName=Gary%20Marcus>

Elana Simon was given a diagnosis of a rare form of liver cancer at the age
of twelve. Six years later, a few months shy of her high-school graduation,
she is not only a survivor but a certified cancer researcher: today, she
published an article<http://www.sciencemag.org/lookup/doi/10.1126/science.1249484>
her disease, fibrolamellar hepatocellular
in *Science*, one of the world's most important scientific journals.

One of the unique issues that Simon and others with extremely rare
illnesses face is that there's often not enough data to know exactly how to
treat them. There are approximately two hundred and thirty thousand new
cases of prostate cancer each
the United States, compared to roughly two hundred new cases of
fibrolamellar. This means that doctors know far less about the progression
of rare diseases--and, because there are so few cases, they are often
misdiagnosed. Meanwhile, drugs for rare diseases are often more difficult
to develop, because sample sizes for experimental trials are necessarily

The paucity of data that could help treat rare diseases is exacerbated by
doctors' reluctance to share what little information does exist, because
hospitals tend to compete with one another rather than collaborate. For
instance, if a hospital treats a young child with an uncommon ailment, the
data it collects in the process often goes unseen and unheard of by doctors
at other hospitals. Hospitals often keep data and cell culture to
themselves, as anyone who has read Rebecca Skloot's "The Immortal Life of
Henrietta Lacks <http://rebeccaskloot.com/the-immortal-life/>," which
documents how cells from the tumor of a poor farmer were used extensively
in medical research over the past half century, might realize.

As a high-school student, Simon worked to understand her own disease and
volunteered at Mount Sinai Medical Center. Before long, she realized that
simply finding and aggregating data was deeply important. With the help of
a friend and the surgeon who treated her, Simon began collecting cancerous
tissue from several other patients facing fibrolamellar for the study
published today.

The sample size was small--just fifteen subjects--but the data was
astonishingly clear and consistent: a genomic analysis revealed a single
common gene mutation in the tumor of every patient. Importantly, the
mutation was found only in the tumor, not in the surrounding tissue,
potentially suggesting a precise diagnostic for fibrolamellar cancer. That
analysis has also helped identify specific genes that are turned on in
fibrolamellar, which could be targets for treatment.

Simon's experience also led her to spend the past year developing the
Fibrolamellar Registry, a Web site where fibrolamellar patients can share
their medical data <http://www.fibroregistry.org/>. The service acts as
both an electronic medical-record repository and a tool for collecting data
from patients in different hospitals, cities, and countries. A number of
parties, including clinicians, computer scientists, and the National
Institutes of Health's office of rare diseases have agreed to join the
project as collaborators. Ideally, the registry could serve as a model for
larger, more general systems that track all diseases, where individuals can
share anonymized data for clinicians from any institution to analyze.

As Simon put it to me over the phone earlier today, patients usually freely
surrender their tissue and information to hospitals, but why should they?
"It's their information, it's their health, so they should have the right
to do with it what they want," she said. Perhaps the best thing patients
can do is to give their data away for free to work, bit by bit, toward a
cure. The open-access movement has worked for software and genetics; it can
work for medicine, too.

*Photograph: Jim Varney/Science Source*

More information about the Ip-health mailing list