Originally published Nov. 1.
Citing Myriad Genetics' proprietary database of gene variants as the primary example of an industry practice that is harmful to patients and the advancement of genomic medicine, policy and legal experts have put forth a number of proposals for ensuring that life science companies share data on the association between genes and diseases.
Myriad operates an internal database containing information on thousands of BRCA variants that it has associated with risk of breast or ovarian cancer. Although the company contributed such variant data to public resources until late 2004, it decided the following year to amass this information as a trade secret.
In a paper published this week in the European Journal of Human Genetics, Robert Cook-Deegan of Duke University; James Evans of the University of North Carolina, Chapel Hill; and lawyers John Conley and Daniel Vorhaus of Robinson, Bradshaw & Hinson assert that by operating a proprietary database of gene-disease associations, Myriad is denying researchers, physicians and patients access to "basic scientific and medical information" that would improve their knowledge of variants linked to cancer and enable more accurate diagnoses.
Myriad is the sole provider of BRCA testing for breast and ovarian cancer susceptibility in the US, largely due to its exclusive patent position on testing methods and the disease-linked BRCA mutations themselves. In other parts of the world, such as in the EU, the company's patent position is not as strong as in the US, but its BRACAnalysis test is still a top selling product.
Particularly for markets where it is lacking patent protection, Myriad touts its proprietary variant database as an advantage over competing BRCA tests. Myriad presented data earlier this year at a major European conference highlighting that the rate of variants of unknown significance reported by its lab is around 3 percent, while other European BRCA testing labs have a VUS rate of up to 30 percent (PGx Reporter 6/27/2012). VUS are makers that have an unclear association with disease.
Myriad's proprietary database "is probably the major factor in explaining the company's ability to interpret VUS result more successfully than others," the authors of the EJHG paper assert. Given that Myriad is presently trying to penetrate further into European territories, Cook-Deegan and colleagues believe that the time is ripe for payors and regulatory bodies in those countries to adopt policies that require companies to submit variant data and any interpretive algorithms to public databases.
"I'm not a Myriad-hater. I think they do good testing at prices that are in the ballpark, and they do a good job of getting third-party reimbursement, write clear reports, and have a solid clinical staff. I also don't think they're doing anything illegal," Cook-Deegan told PGx Reporter via email.
"What I do think is disturbing is that most physicians and patients are unaware that Myriad is keeping their patient data and not sharing it publicly. It was sharing through late 2004, but then changed policy. But it did not tell docs or their patients about the change," he continued. "Their public stance is deliberately disingenuous, in my view, and I expect this article will really be the first time they've been called out on it specifically."
The authors of the EJHG paper recommend that scientific journals; government-run databases such as the National Institutes of Health's Genetic Testing Registry; scientific organizations; medical groups; and particularly payors and regulatory agencies that are part of national health systems should enact new policies or enforce the regulations they already have in the books to require companies to report any data that would enable independent validation of their tests. "First, [national health systems] could ask testing firms to voluntarily adopt policies to share mutation data publicly," the authors suggest. "Second, payors could refuse payment unless clinically relevant data are shared and subject to independent verification for both accuracy and validity of interpretation."
Comparatively, in the US, the authors allege that "hundreds of [coverage] agreements have been signed" between payors and test developers, under which insurers haven't asked companies to disclose key genetic interpretation data that would benefit public health. "Payors and regulators in Europe, South America, Asia, and other markets need not be so passive," they stated.
Another, more expensive, option proposed in the paper would have national and international institutions recreate data in proprietary databases by establishing large-scale registries that collect genomic and clinical data. Lastly, national payors could put forth incentives — such as payment codes that companies could submit upon depositing information on genomic markers to public databases — for encouraging such behavior among industry.
These types of policy changes, the authors feel, are necessary to foster the type of collaborative environment necessary to speed interpretation of molecular data from whole-genome sequencing studies, which will uncover reams of variants with uncertain or unknown disease associations.
"Whole-genome analysis stands poised to have a major impact on medical care if it can be harnessed appropriately. But the biggest challenge to its implementation is properly interpreting the variants found upon analyzing any individual’s genome," the authors wrote. "As whole-genome and whole-exome sequencing become commonplace, the rate of truly novel mutations will eventually decline. For the foreseeable future, however, each individual whose genome is sequenced will have vast numbers of variants of uncertain clinical signiﬁcance."
The proposals in the paper, while embraced by some, will likely be viewed by many industry players as unfriendly to their competitive interests. Felix Frueh, former associate director for genomics at the US Food and Drug Administration, told PGx Reporter that the value of molecular diagnostics lies in how well they can inform disease and guide treatment. Recognizing this, Frueh suggested that many of the recommendations in the EJHG paper would eliminate the incentives for industry to invest in complex genomic tests.
"[T]here is little revenue to be made with just a diagnostic test alone, and … the real value is in the interpretation of the test result fed by the data that accumulates over time," Frueh said via email. "What if the only incentive for a company is to capture, [for instance,] the sequence data? Would there be any such companies and would we get the necessary accompanying phenotypic information that makes this (sequencing) data actionable and valuable to patients? Probably not."
The Risk of Indefinite Monopoly
The authors of the paper, as well as many researchers, healthcare providers, and patients believe that the lack of policies encouraging submissions to public databases has enabled Myriad to monopolize a significant segment of the breast and ovarian cancer genetic testing market.
Currently, researchers and patients represented by the American Civil Liberties Union and the Public Patent Foundation are challenging patents held by the University of Utah and exclusively licensed to Myriad. The plaintiffs allege that Myriad's patents are invalid because they claim gene sequences in the body, which are naturally occurring substances and cannot be patented under US law (PGx Reporter 8/22/2012).
In the EJHG paper, Cook-Deegan and colleagues point out that even if Myriad loses this legal battle and many of its patents underlying the BRACAnalysis test are invalidated, it won't keep the company from offering the test or remaining competitive in the diagnostics space, largely due to all the other patents that would remain valid and its proprietary database. "In an environment in which new technologies, including whole-genome and whole-exome sequencing, are already beginning to change clinical practices in genetic testing, a proprietary database gives Myriad indeﬁnite exclusivity independent of patent protection," they wrote.
While Myriad was still submitting data to public resources, researchers from the company and several universities published an article in the American Journal of Human Genetics on the clinical assessment of more than 1,400 variants of unknown significance. The paper reports 18 deleterious and 100 neutral markers, but does not detail the rest of the other variants or the interpretive algorithms used, which means that others can't validate the findings. "This contrasts with recommendations of the National Academies in two reports that call for depositing data and methods sufficient for replication and interpretation," the study authors point out.
Until the variant data and interpretive algorithms are publicly available, other labs conducting BRCA testing will continue to provide less robust tests, the authors noted. "This asymmetry has clinical impact: a woman might be able to receive BRCA testing from another laboratory in Malawi or Malta, where Myriad’s BRCA patent rights are not in force and testing is perfectly legal, but that laboratory will have no access to Myriad’s data and will thus be unable to interpret many VUS results."
Although geographic inequities in access exist for all medical products, many patients and healthcare providers have taken particular issue in the case of BRCA testing because of the way Myriad has exercised its IP rights while benefiting from information available in public databases and research funded with taxpayer dollars.
"Trade secret protection is used in all manner of research," Vorhaus, one of the authors of the EJHG paper, said via email. "Think about, for instance, pharma conducting pre-approval studies and research on drug candidates, biomarkers, etc. Much of that never sees the light of day." As a lawyer for Robinson, Bradshaw & Hinson, Vorhaus represents the interests of many life sciences companies, and has knowledge of how firms protect their competitive assets. In his view, Myriad stands out in the industry in terms of how its business strategy has impacted patients' access to genetic tests.
"What Myriad is doing — taking information of clear clinical relevance for a product already on the market, and holding that as a trade secret — may be somewhat unique given Myriad’s unique and patent-enabled market position," he said. "What Myriad is doing could potentially represent a dangerous precedent that is worth discussing, [for example] in the academic and policy literature, … potentially heading off through one of the pathways we suggest."
In defense arguments against the ACLU and PUBPAT, Myriad has said that it hasn't enforced its patent rights to restrict research, highlighting that 9,000 papers have been published by 18,000 authors on BRCA 1/2 mutations. However, according to Cook-Deegan's analysis, Myriad has contributed very little to the published knowledge around these mutations. "[Myriad's] publications include 440 authors (2.5 percent) and constitute a very tiny fraction of those 9,000 articles," he noted. "Many or most of those articles are not about clinical testing and don't help inform it, but even of those that are about genetic testing for BRCA, Myriad is not the dominant [research] force by any means. So yes, they benefit enormously from publicly funded research."
From its stringent enforcement of its exclusive licenses and its patent rights to its decision to keep gene association data as a trade secret, Myriad has often been portrayed as the villain of the genetic testing industry. When the ACLU was looking to launch a case against patenting genes, it could have challenged IP held by Myriad or Athena Diagnostics (now a Quest subsidiary), but decided to go with Myriad because of its more egregious and hotly criticized patent enforcement practices.
In the EJHG paper, Cook-Deegan and colleagues point out that Myriad is one of the few labs that has declined to submit gene variant information to the public MutaDatabase resource. Meanwhile, around 100 labs — both academic and commercial — have contributed data to the database. "Most companies, as well as research laboratories, are very, very supportive of contributing data and curating it for use," said Sherri Bale, managing director of GeneDx, a molecular diagnostics firm that is not only contributing data to MutaDatabase, but is helping build its infrastructure.
"I think Myriad is a stand-out in its position," Bale said. "There are one or two others … who have said to us that they believe that the information from testing patients is proprietary. But the vast majority of people realize that if we don't share this stuff, all of us are at a disadvantage." GeneDx has nearly all its data ready for submission to MutaDatabase and plans to do a bulk upload in the next two to three months if the database infrastructure is ready.
If only a few companies are keeping gene association data as a trade secret while the majority of firms are sharing this type of data publicly, then it raises the question as to whether regulatory and reimbursement policy changes of the kind proposed in the EJHG paper are necessary. Bale believes, however, that policy changes are needed "to help loosen up the funds" needed to build these types of public databases. "[When] someone says, 'Unless you send that data, Blue Cross isn't going to pay for that test,' suddenly there will be money to get that data out," she said.
Cook-Deegan, too, acknowledged that Myriad represents an "extreme" case and noted that the company's ability to hoard its genetic association data is "as much the fault of payors and health plans and regulators who have not demanded data access."
The Value of Interpretation
In the genomic testing industry at large, there are a number of firms keeping the interpretation portion of their tests a trade secret. For example, the algorithms underlying multi-analyte tests such as Genomic Health's Oncotype DX and Agendia's MammaPrint are not known to the public and are considered "black boxes."
In Cook-Deegan's view, however, the difference between Myriad's practices and companies marketing multi-analyte algorithm-based tests is that the latter have made a more significant contribution to research.
Developers of multi-analyte algorithm-based tests "have built their black boxes. They select what to measure and methods to measure it. Then they pay for clinical research to verify that it works," Cook-Deegan said. "Myriad's target is found in nature. They have done relatively little clinical research. They're sitting on data needed to interpret not their particular cluster of expression profiles, but the mutations that predispose to cancer that exist in nature. I concede it's not a sharp distinction, but I think it's a distinction that matters for policy."
As a general principle, Cook-Deegan believes that no company's business strategy should restrict independent validation of its test methods. "Frankly, if the other companies' proprietary strategies mean their results can't be independently verified, then their data and algorithms should not [emphasis in original] be proprietary either; they should be public."
As whole-genome sequencing technologies are used more readily for discovery research, sharing gene-association data may become more necessary, but the value of that interpretation will increase. In this kind of environment, many companies may choose to take the trade secret route to remain competitive. Policy changes mandating companies to share novel gene associations they've discovered internally, with their own funds, will likely be viewed by industry as a disincentive to invest in a field that already has significant regulatory and reimbursement hurdles.
Currently, there is no compelling reason for companies to submit variant interpretation data to public repositories. According to GeneDx's Bale, the only incentive driving those participating in MutaDatabase, for example, is that they would be seen as good industry players.
Bale believes that in the US, payors would have the most power to demand the kind of data sharing proposed by Cook-Deegan and others in the EJHG paper. "Payors would love to tell us that we'd have to submit [this type] of data or they won't pay," she said. "It just gives us another hoop to jump through."
In fact, the Centers for Medicare and Medicaid Services has instituted a program called MolDx, through which Medicare contractor Palmetto GBA is reviewing the analytical validity, clinical validity, and clinical utility of molecular diagnostics as a condition for coverage. Although any data submitted to Palmetto for this assessment is strictly confidential, labs have expressed significant concern that participating in this program could jeopardize proprietary test data (PGx Reporter 2/29/2012).
Ultimately, the decision to make public data from genetic tests may be outside of the mandates of regulatory and reimbursement bodies, said Frueh, who recently joined the venture capital firm Third Rock Ventures and provides strategic advice to companies on personalized medicine. Data from genetic tests should belong to patients, in Frueh's view, and they should have the final say on how their information is shared.
"The patient — either directly or indirectly via insurance — is paying for a service that generates data, so in essence one can argue that the sequence data generated, as well as the interpretation (even if based on a proprietary and undisclosed algorithm) belongs to the entity who bought this service, so they could decide who can use it," he said. "It’s a privacy question, too, as probably not all patients would agree to having (even if only parts of) their genome made public."
Moreover, Frueh noted that in healthcare, experience is of value, and for a company that experience lies in the patient data they've accumulated over time. "One can certainly argue that not all healthcare professionals, or any other professionals for that matter, are of equal value and expertise, which at least to some extent is based on their experience (i.e. data) accumulated over time," he said. This is "not all that different from the situation described with Myriad and their experience with BRCA1/2 testing.
While the recommendations in the EJHG paper may result in some short-term gains, Frueh believes they are shortsighted from an innovation standpoint. "We would risk losing an industry that is just starting to capture and make accessible to patients the effort we invested in over the last two decades in deciphering and learning from the human genome," he reflected. "What we have to realize and accept is that this part of health care is rapidly becoming an information service business. What we want is the best possible information to make the best possible decisions."
Meanwhile, Cook-Deegan and others would argue that companies hoarding gene interpretation data as trade secrets are the shortsighted ones, since such practices restrict other researchers from ensuring the accuracy of marketed tests or from using the data in new ways and, most importantly, prevent patients from getting the most accurate tests.
"We simply cannot afford to allow 'trust us' tests to become the norm," Cook-Deegan said. "The science needs to be replicable and transparent."
Even if labs continue to keep gene association data as trade secrets, "gaps" in public knowledge of variants of unknown significance could be addressed in other ways. One alternative Cook-Deegan suggested may be research consortia, such as the Evidence-based Network for the Interpretation of Mutant Germline Alleles, or ENIGMA — an effort that draws information from different databases and applies bioinformatics analyses to improve VUS interpretation.
"Admittedly those structures don't exist or are only inchoate, but it's become apparent to everyone doing whole-exome and whole-genome sequencing that such structures are needed," he said. "So, publicly funded research can fill the gaps even if the labs don't close the data loop — so long as someone else does."
Myriad did not respond to requests for comment for this article.