Myriad Sticks to Its Business Model

    By Kevin E. Noonan --

Myriad Genetics, one of the preeminent gene-based diagnostics companies, has been targeted by the American Civil Liberties Union, Dan Ravicher’s PUBPAT Organization, a raft of medical researchers and their associations, and individual patients in a lawsuit filed in the Southern District of New York; the avowed aim of the plaintiffs is to get the issue of "gene patenting" before the U.S. Supreme Court (see "Association for Molecular Pathology v. U.S. Patent and Trademark Office").  While the case wends its way through the Federal judicial system, Myriad is continuing its pursuit of diagnostically-relevant cancer markers, the most recent of which is a gene related to susceptibility to pancreatic cancer licensed from the Johns Hopkins University.

The gene encodes PALB2, a tumor suppressor gene reported to be involved in susceptibility to pancreatic cancer in the April 10, 2009 issue of Science (Jones et al., 2009, "Exomic Sequencing Identifies PALB2 as a Pancreatic Cancer Susceptibility Gene," Science 324: 217).  The product of the PALB2 gene is a binding partner of BRCA2, originally identified by its association with breast cancer susceptibility.  The two gene products bind to DNA to repair damage, wherein PALB2 localizes and anchors BRCA2 to damaged DNA in the cell nucleus.  More than 3% of familial pancreatic cancer patients had mutations in PALB2 resulting in the production of a truncated version of the encoded protein.  In contrast, no mutations in the PALB2 gene were found in more than 1,000 normal DNA samples.

Myriad contends that the predictive value of detecting such mutations are due to a 10- to 20-fold increase in the likelihood of developing pancreatic cancer by age 70 for individuals bearing the PALB2 mutation (or mutations in BRCA2 or another tumor suppressor gene, p16).  Early detection is particularly important for pancreatic cancer, which is typically detected only after it has progressed past the stage where intervention has any chance of significantly prolonging life.  Pancreatic cancer is diagnosed in about 200,000 patients worldwide each year, and is fatal almost without exception.

The scope of the difficulty in identifying reliable diagnostic genes is reflected in the following synopsis of the extent of mutations found by the Johns Hopkins researchers in tumor DNA from one pancreatic cancer patient.

Among the 20,661 coding genes analyzed, we identified 15,461 germline variants in Pa10 not found in the reference human genome.  Of these, 7318 were synonymous, 7721 were missense, 64 were nonsense, 108 were at splice sites, and 250 were small deletions or insertions (54% in-frame).

The PALB2 mutation found in tumor DNA from this patient exhibited a germline deletion of four basepairs (TTGT) that produced a frameshift mutation at codon 58, resulting in production of a truncated gene product from this gene.  Additional mutations found associated with pancreatic cancer in the PALB2 gene are a G > T transversion at the 5' splice site of exon 6, a deletion of an A residue in exon 11 (at nucleotide 3116) and a C > T transition mutation in exon 12 (at nucleotide 3256).  Fine structure mapping of mutations found in the PALB2 gene in pancreatic cancer, breast cancer, and Fanconi anemia patients was also reported.

The company estimates that a diagnostic genetic test might be on the market by 2010.

Comments

[BCD]

This is a really intriguing finding. Great science with clinical relevance.

PALB2, the gene in question, is subject of a 2006 PCT application from Mike Stratton's Group in the UK for Fanconi's anemia and for some other cancers.

Ambry already offers PALB2 sequence testing based on PCR amplification of all the exons (see genetests.org). Presumably such sequencing would uncover sequence variations, including those found by the Hopkins researchers to be associated with pancreatic cancer risk. It appears there is already a service on the market that could accommodate this clinical need. Or perhaps not, but why not?

Myriad and Hopkins announced exclusive licensing of patents (by which they meant PCT patent applications, not yet published). How exactly will this play out? Surely Myriad will not stop Ambry from offering a PALB2 sequencing test already on the market? Would they? Cross license?

I'm trying to understand exactly what the business model is. Clearly PALB2 could be a valuable addition to Myriad's cancer-gene-susceptibility testing. But what role will patent exclusivity play? What happens if another lab comes across one of these mutations? This is not a rhetorical question. I really am a bit confused by what the business models are here.

[Kevin E. Noonan]

Dear BCD:

I think the business model will be claims in the "determine and infer" class identified by Professor Collins at Indiana Law School. Since the gene was known, and mutants in the gene were known associated with FA, the claims Hopkins might get and Myriad might have licensed would be something like:

A method for identifying a risk in a human for pancreatic cancer, the method comprising the steps of obtaining a biological sample [could be a pancreatic biopsy but not necessary for germline mutants] from the human and assaying genomic DNA from the sample for a mutation comprising deletion of at least the nucleotide residues TTGT in an exon of human PALB2 gene encoding codon 58, wherein the human is identified if said deletion is detected in the sample.

This may be narrower than what they are after, but you wouldn't have assayed or detected this mutant in the prior art unless you knew it was there, and so Hopkins should have novelty.

Thanks for the comment.

[Bob Cook-Deegan]

I think they identified at least two mutations that would not have been previously associated with familial pancreatic cancer (presumably that's why Science published the paper), and I would certainly think that might be patentable, as it's novel and useful and probably nonobvious. Whether they describe their invention in sufficient detail to enable the claims and provide sufficient written description will depend on the precise wording of their patents.

But my question was not whether it is patentable. My question addresses the business model, and how exclusivity will work in the market for diagnostic tests. That's what I find puzzling about exclusively licensing such rights, given that there are already clinical services offering a test that would uncover these mutations, and could continue to do so without the added transaction costs of licensing and enforcement. There is nothing stopping Myriad from doing such testing too. But what is the likely use of exclusive rights in this case in the future when there are granted patents? I can imagine it's a way to force cross-licensing, but I worry it could be used as a club to shut down clinical testing elsewhere.

The only use of patents in genetic diagnostics of which I am aware has been"cease and desist" actions against labs already offering a competitive test. This suggests that in genetic testing (in contrast to therapeutics), there are already products/services on the market, and the role of patents has not been to encourage their availability but to change the business model from competition to monopoly. This is what Myriad did with BRCA in the US (but did not succeed in any other jurisdiction), Athena does with many tests, and PGxHealth did with long-QT testing until GeneDx forced a deal by obtaining countervailing exclusive rights.

[Kevin E. Noonan]

Dear Bob:

I agree that the business model is exclusivity ("monopoly" is like "troll" - even when accurate it rarely advances the debate).

But I think the way you frame the question (which is the way many do) is to look at individual cases (where there may be an "anticompetitive effect") instead of the bigger picture. It is true that under the current system, there are two different models: Myriad's, which is for exclusive access to patented technology (we may call this the "traditional approach") and Stanford's non-exclusive licensing (for Cohen & Boyer's restriction enzymes, for example), which we can call the "enlightened approach." Remember, the companies are trying to make money, so from their perspective either business model may work, depending on the circumstances.

However, there are societal and political costs involved in Myriad's approach that have implications for the entire industry. The ACLU lawsuit is just one example; others would be petitioning HHS to exercise its "march-in" rights under Bayh-Dole regarding the NIH grants used to support the work done initially at Utah (and, for pancreatic cancer, at Johns Hopkins).

What is not going to be an effective way is some sort of ban. Bans on gene patenting are a bad idea for lots of reasons, such as inhibiting the development of therapeutics (which are frankly and necessarily exclusionary, given the development and regulatory costs). It is also unnecessary, since these patents are rarely "blocking" (according to a study from Catholic University in Belgium; see today's post).

Bans on diagnostic testing would likely be hard to craft narrowly against those involving genes, and if successful are fraught with the risk of unintended consequences. One ready example is trade secret protection as an alternative, for example. Consider the following hypothetical.

Right now, most disease-related polymorphisms are disclosed by university researchers who patent (as required by Bayh-Dole) and (for most) are more focused on disclosure for advancing science/and their own careers. They are limited by grant money and the criteria for getting it. If there is a gene patenting ban what we can expect is that diagnostic companies will use the brute-force power of sequencing, coupled with banks of frozen tumor samples available from many sources (commercial or otherwise), to develop other, maybe better, more reliable, diagnostic markers. This is the kind of work unlikely to be done in academia (recognizing that this is an overbroad statement).

A company would have no incentive to disclose this information since its test could not be patented. Moreover, there would be no reason to disclose since it could be commercialized in a way very difficult to reverse-engineer. For example, a company could make a chip with 10,000 gene sequences on it, that would light up when contacted with a sample containing a gene or polymorphism diagnostic for a disease like cancer. But that wouldn’t be the only thing that would light up, since there would be controls and other things on the chip (including decoy sequences). The company could have a bar code on each chip that would be read by a reader that would interpret the data, but the company would never have to disclose the polymorphism or where it was on each chip (it would be different each time) – the bar code could be encrypted, for example. And that technology would be protected forever with no disclosure and no expiration.

I think the answer in the short term is for universities to offer non-exclusive licenses, maybe with field of use or other restrictions, that would prevent the most dire consequences of patent exclusivity. And see what happens - I think the hypo above will arise once the costs of development are justified by the exclusivity profits.

Thanks for the comment.