‘Affymetrix’ is usually the first word that comes to mind when the subject of microarray-related intellectual property is raised. After all, the company has more than 170 issued and 370 pending US patents.
But discussions with several leading patent lawyers who practice in the genomics and biotechnology field revealed that other patent issues may come into play more strongly as microarrays move downstream in the drug discovery and development process, and ultimately into the clinic.
Among these issues are the need for array companies to navigate around patents that could cover antibody, antigens, and protein fragment content on microarrays as well as the polymorphisms on SNP arrays. And from the point of view of an inventor seeking a patent, there is an urgent need to keep in mind recent case law that can weaken a patent’s scope of protection if the patent’s scope has been narrowed during the application process.
With this first issue, an emerging protein microarray company needs to look into patent issues before jumping out on the market, said Beth Arnold, a partner at Foley Hoag’s Boston office who specializes in biotechnology patent law. “With these kinds of arrays you need to clear every component that’s on the array as well as every component and process used in generating the array,” she said, just the way a pharmaceutical company would clear any candidate compound that it develops with its patent lawyers. The rights to certain genes, proteins, or antibodies “may in certain instances be very well blocked. If there’s not a license available, you may not want to include that antibody or protein on your array, since you may be stepping into a patent infringement lawsuit,” Arnold said.
This rule applies not only to patents that cover proteins and antibodies — which proteomics companies have been busy filing lately — but also to ones that cover genes. In fact, gene patents typically include claims to the protein and to the antibodies that recognize the protein encoded by the genes, Arnold said.
This concern with content in the gene, protein, and antibody array arena contrasts with the preoccupation with patent issues for oligonucleotide arrays. Since an EST or other oligo’s use as a probe is not considered a specific, substantial, and credible utility by the US Patent Office — a requirement that any invention must meet to be patented — there really aren’t many gene fragment patents out there and the ones that currently exist may not be valid, according to Arnold.
Polymorphisms present a different story. With SNP patents proliferating, SNP arrays present similar content-related IP issues to antibody and other protein arrays, said Elizabeth Howard, an intellectual property partner at Orrick, Herrington & Sutcliffe’s Silicon Valley office in Menlo Park, Calif. As SNP arrays are eventually going to be used to identify genetic predispositions to disease, “the companies that identify the SNPs have every interest in realizing some revenue from the use of their SNPs in this process,” said Howard. “The bioarrays are almost certainly what’s going to facilitate that.”
So does this mean that a company making SNP-based arrays has to take 1,000 licenses for a 1,000-SNP chip?
Theoretically, yes, but there will be a lot of room to work around this problem, Howard said. “There’s not one single SNP for a disease, so there’s always going to be another way.” Even though some forward-thinking companies are now patenting haplotypes, or blocks of co-variant SNPs that can serve as more robust biomarkers of disease, array companies going for SNP-based diagnostic applications may even be able to come up with alternative haplotypes in their products. And even if the companies need to take out licenses to SNPs, “the universe of patents is not going to be inordinately large,” said attorney Laura Coruzzi, the head of the biotechnology group at Pennie and Edmonds in New York.
Seeking Shelter in Hatch-Waxman
Companies marketing SNP or protein arrays to pharmaceutical companies may find respite from intellectual property wars under the umbrella of the so-called Hatch-Waxman Amendments passed by the US Congress in 1984. This law, while best known for giving birth to the generic drug industry, also protects from patent infringement suits substances or devices used in research that is performed in order to submit data for FDA approval of a new drug or diagnostic test.
In a recent lawsuit where Rhône-Poulenc Rhorer pharmaceuticals sued Bristol-Myers Squibb for infringing its patent covering the breast cancer drug Taxol, the US District Court for the Southern District of New York delivered a broad interpretation of this research exemption — known as 271E, said Coruzzi. Bristol-Myers Squibb was using compounds involved in the synthesis of Taxol to discover an alternative Taxol-like drug for cancer, and even though there was no evidence the company was applying for FDA approval of these compounds, Bristol-Myers Squibb argued that, “everything we do we hope will be a drug, and therefore every piece of data we generate we intend to submit to the FDA,” Coruzzi said. The court agreed. Although the case is going to trial and the law could take three years to be resolved after appeal to the Federal Circuit, this interpretation of the law “virtually immunizes” from infringement compounds or diagnostics involved in the drug discovery process from infringement claims in the mean time. In other words, companies making SNP or antibody arrays who follow the guidance of this caselaw could market their products to pharmaceutical companies for drug discovery and development without having to worry too much about patent infringement claims.
Would-be infringers may have an additional powerful defensive weapon if the US Supreme Court decides to uphold the Federal Circuit’s decision in the Festo v. Shoketsu Kinzoku Kogyo Kabushiki case. The Federal Circuit held in Festo that when a patent is rewritten to narrow the scope of the claims during the application process, this bars the patent holder from prosecuting the patent under the doctrine of equivalents. (This doctrine allows infringement claims when another invention is not literally identical to the patented invention, but is basically equivalent, whereas literal infringement requires the patent holder to show that the infringing invention is identical.) So if a patent examiner initially rejects a patent application because the patent is too broadly framed, then the patent applicant rewrites it more narrowly. This weakens the patent quite a bit because the patent can only be enforced in cases where it is being literally infringed.
The Supreme Court is expected to hand down its decision in this case before it recesses in early July.
Meanwhile, patent attorneys are operating defensively, said Corruzi. “If you file narrow claims, then broaden them during prosecution, nothing will happen. But if you are not doing careful analysis and end up narrowing your claims, the price to be paid could be bad. From the freedom-to-operate point of view it is always best to be conservative,” Coruzzi said.