Ivy Tsui is a JD candidate at Osgoode Hall Law School and is enrolled in Professor Mgbeoji’s Patents class in Fall 2011. As part of the course requirements, students are asked to write a blog on a topic of their choice.
In the genomic era, the flood of computationally predicted genes has introduced a new challenge for the patent system: How much does a gene patent have to disclose to satisfy the “useful” requirement for patentability? In Human Genome Sciences Inc v Eli Lilly & Co,  UKSC 51, the United Kingdom Supreme Court (UKSC) unanimously held that a patent for a gene sequence, without having done any concrete functional experiments, is a valid patent. The decision has significant implications on the scope of disclosure for gene patents and gives rise to far-reaching consequences in the scientific and biotechnology industry.
Background and Facts
Based on bioinformatics, scientists at the Human Genome Sciences (“HGS”) discovered a novel gene called Neutrokine-α and predicted it to be a member of the TNF ligand superfamily. The patent, which was granted by the European Patent Office (EPO) in 2005, claims the nucleotide of Neutrokine-α, the amino acid sequence and its antibodies, which might be useful for the diagnosis or treatment of an extraordinarily large range of immune disorders. However, there is no experimental data to support these suggestions.
Eli Lilly opposed the patent, arguing that the patent did not disclose a practical way of exploiting the gene. Reportedly, Eli Lilly has already spent $50 million on the development of an antibody to Neutrokine-α and plan on spending another $250 million in clinical trials.
The Opposition Division of the EPO revoked the patent on the basis that the invention did not have a known function. The Technical Board of Appeal held that the patent was valid because a person skilled in the art would have appreciated the general knowledge of the TNF ligand superfamily.
Eli Lilly brought parallel proceedings in the UK Patents Court for revocation of the patent. At the Patents Court, Mr. Justice Kitchin stated that the specification “contains extravagant and sometimes contradictory claims” and struck down the patent. The Court of Appeal also invalidated the patent, asserting that “however clever and inventive you may have been in discovering a gene sequence, you cannot have a patent for it or for the protein for which it encodes if you do not disclose how it can be used.”
Issue on Appeal
The primary issue is whether the requirement of industrial applicability in Articles 52 and 57 of the European Patent Convention (“EPC”) extends to a patent for biological material. To satisfy the requirement of industrial applicability, the invention must be made or used in an industry.
UK Supreme Court Decision
The UKSC unanimously held that the patent satisfied the requirement of industrial application. Lord Neuberger, penning the leading decision, based his reasons on the policy argument advanced by the BioIndustry Association (BIA) and a desire to be consistent with the EPO jurisprudence.
The BIA, which is an association representing the bioscience sector with an aggregate turnover of £5.5 billion in 2010, intervened as a neutral party. It submitted that following the Court of Appeal decision will “make it appreciably harder” to satisfy the patentability requirement in the future, and will cause the “UK bioscience companies great difficulty in attracting investment at an early stage in the research and development process.”
Even though Lord Neuberger recognized that “it would be undesirable to have a monopoly over a particular biological molecular too early, because it risks closing down competition”, His Lordship ultimately agreed with BIA’s submission and said that “it would be wrong to set the hurdle for patentability too high.”
Next, Lord Neuberger made an effort to align UK patent law “so far as possible” with the EPO jurisprudence. Even though the UK Court of Appeal held, and Eli Lilly argued, that the Board’s decisions should carry less weight because the proceedings were unopposed and without cross-examination, Lord Neuberger rejected the argument and stated that the Board has applied its principle consistently. After reviewing the Board’s approach to Article 57 in relation to biological material, Lord Neuberger stated, among other things, that a “plausible” or “reasonably credible” claimed use, or an “educated guess” on biological function can suffice even without any experimental or wet lab evidence. Since the newly discovered gene belongs to the well-known TNF ligand superfamily, where its members share many common features, finding this new member will have great value and the “educated guess” made in the patent application is plausible. Therefore, the gene patent is held to be valid.
Whole genome sequencing projects have inundated the bioinformatics world with a vast amount of data. It is expected that the rate of gene discovery will continue to accelerate; yet these newly discovered genes would not be very useful until further functional or therapeutic research is performed. However, allowing a monopoly on gene sequence will reserve this unexplored territory for the patentee, thereby hindering research in the community. The anti-competitive effect may also undermine the quality of experimental data. Without good biomedical research, human health may suffer.
Further, the patent gives the patentee unjustified control over others who have been actively investigating in the field. For instance, researchers who have worked on the TNF ligand superfamily for years may now have to pay a licensing fee for including the patented gene in a routine experiment. Rather than promoting innovation, this unfair arrangement may become an impediment to scientific progress.
A policy goal of the patent system is to encourage others to perform fair research and make improvements on the patented invention during the monopoly period. However, given the broad patent claims in a gene patent, the line between fair research and patent infringement is blurred. For example, if others discover a genetic variant of the patented gene, it is unclear whether the finding is a patent infringement or not. Such uncertainty in law will inevitably lead to an increased number of patent infringement litigations in the future.
Given that bioinformatics is becoming an increasingly popular method to draw predictions on the function of a gene, and that this is a UKSC decision that many will follow, the trend of gene patenting based on computational predictions will probably continue. Therefore, it is imperative to devise a solution to preserve the quality of granted patents and to prevent clogging up the patent office with unmerited applications.
One possible solution is to set a high threshold of reliability for computationally derived data and grant patents only to those with sufficient accuracy. If computational methods become highly accurate and predicted the function of the gene with 99% accuracy, such gene patents should be allowed. On the other hand, if the prediction has 50% accuracy, then the gene should not deserve patent protection. Setting a clear benchmark will prevent patent applicants from wasting effort to file for patent protection. Until such highly accurate bioinformatics methods become available, a higher standard of disclosure should be required for computationally derived gene patents.