DNA Project Aims to Make Public a Company’s Data on Cancer Genes
(ED: The Supreme Court will hear arguments this week on this case, and it could have deep implications for genomics)
By GINA KOLATA
Published: April 12, 2013 44 Comments
Anyone in the United States who wants to know if she has mutations in two breast cancer genes has little choice of where to be tested. One company alone has patents on the genes, and that company pretty much controls the market.
On Monday, the Supreme Court will take up the issue of whether companies can own patents on genes. But there is another issue, often overlooked, that might make the patent question beside the point. No matter which way the patent decision goes, the company, Myriad Genetics, will still own the largest database that tells patients what various mutations mean.
With 17 years of experience, millions of tests looking for thousands of mutations in the genes, and a $500 million investment, the company was able to amass a huge database that tells which DNA changes increase cancer risk and by how much, and which are inconsequential blips in DNA. And it is keeping that data to itself.
Some genetics researchers are furious and have now figured out a way to get the data anyway. Every time Myriad sends out a report on a gene test, it specifies not just the mutations it found but also what they mean. As a result, Myriad’s data on each of the mutations is scattered in millions of reports in the hands of doctors and patients. If the geneticists could just gather those reports, they say, they can recreate Myriad’s database.
So they started a grass-roots project, Sharing Clinical Reports, and are asking cancer clinics and doctors to provide them with all the Myriad data they have from patients who have been tested.
None of the data have names of patients or other identifiers, so confidentiality is not an issue, advocates say. But their task is huge because the amount of data needed is vast. The project’s leader, Dr. Robert L. Nussbaum, chief of the division of genomic medicine at the University of California, San Francisco, estimates that with about 1,000 mutations collected so far, he has only about 1.5 percent of what Myriad has.
“Myriad is probably laughing at me, saying, ‘Here is this little flea,’ ” Dr. Nussbaum said.
The story began in 1996, when Myriad got patents on the two isolated DNA molecules known as the BRCA1 and BRCA2 genes and provided a test to determine if the genes carried mutations. The company realized, though, that it would be crucial to figure out how risky each mutation was.
That sort of analysis requires linking the mutations to people’s cancer history. Obtaining that data is a momentous project, said Mark Capone, president of Myriad Genetic Laboratories, a wholly owned Myriad subsidiary. In 1996, the company classified 40 percent of mutations as being of uncertain significance because it did not have enough information to know what they meant. By 2004, the figure was down to 20 percent. Now it is just 3 percent.
Mr. Capone explained how the work is done. The company finds a mutation and is not certain what it means. To find out, it needs to see the same mutation in at least 20 other people, asking whether they had cancer and, if so, what type. So Myriad offers free testing to other family members to get more information.
Until 2004, Myriad posted its data on a site for researchers. But, Mr. Capone said, the company became aware of problems with the way its data were being used. For example, he said, the person running the database part time updated it only every couple of months. And the database included risk estimates submitted by laboratories all over the world, not just by Myriad.
“We might classify a mutation one way, and someone else might call it something different,” Mr. Capone said. That is fine if the data were being used as intended — for research purposes only. But instead, they were being used to tell patients their cancer risk and to make major medical decisions.
Myriad’s database, Mr. Capone said, is highly regulated. “We can only use our database to provide clinical results for patients who had their genes sequenced in our lab,” he said. So, he said, when the company became aware of how the research database was being used, “we had to act — we didn’t have any choice.”
Myriad stopped posting its data.
Dr. Nussbaum does not buy that argument. “The Myriad approach is a big black box,” he said. “It’s a ‘trust us, we know best’ approach.” And, he said, “it is contrary to the public health.”
One thing it does is preclude independent second opinions, said Sherri Bale, managing director of GeneDX, a gene testing company working with Dr. Nussbaum. Yet the consequences of some mutations are so dire that women may have their breasts and ovaries removed to protect themselves from cancer. “You are going to remove my breasts, you are going to remove my ovaries? Let me ask one other person,” Dr. Bale said.
Myriad disagrees, though, saying it has licensed to LabCorp the ability to independently use its technology to search for mutations. Myriad then uses its data to say what the mutations mean.
But having one company control the data for genes is contrary to the way medicine is developing, said Heidi Rehm, a Harvard geneticist who also is working with Dr. Nussbaum.
Today, doctors and researchers are scanning all of a patient’s genes and the old system is crumbling. No one can know enough to interpret the results without a public database.
“Now it is, ‘I have 22,000 genes and have to stay on top of 50 million variants,’ ” Dr. Rehm said. “There is just no way a single laboratory can manage that.” And if companies control the data on the interpretation of results, doctors, clinical laboratories and researchers are in a quandary.
Dr. Rehm, Dr. Nussbaum, Dr. Bale and others now are working with the National Institutes of Health to start a public database of variants in all the genes that have been studied. But first, they are working on the Myriad issue.
The Myriad project began about a year ago when Dr. Nussbaum decided he and others could take matters into their own hands.
Working from home at night and on weekends, with the volunteer help of two genetics counselors, he began contacting geneticist friends and people he knew at large clinics, asking them to send in Myriad reports.
“I would say, ‘How about pulling your results?’ They would say, ‘It’s a big pain.’ So I would tell them, ‘I will pay you 33 cents for each variant,’ ” Dr. Nussbaum said, explaining that a clinic would usually end up getting about $50 if it sent in all of its patients’ mutations and their interpretations. Most then agreed to do it, assigning a student who needed extra money.
Clinics that sent in at least 200 unique genetic variants would get an iPad mini.
The funds for these inducements were supplied by Peter Kolchinsky, managing director of RA Capital Management in Boston.
Dr. Kolchinsky, a scientist by training, said he would like to see many gene testing labs compete on the basis of cost, speed and customer service. But they would all share data on interpreting alterations in genes rather than creating what he called “gene- or disease-specific trade secret monopolies.”
“That works for Coke, not for cancer,” Dr. Kolchinsky said.