According to the Canadian Patent Act a patent will only be granted if an invention is novel, non-obvious and useful, assuming that the claims define statutory (i.e., patentable) subject matter. The "utility" requirement has become a major challenge for patent applications in one area of BIO-IT that is developing particularly fast, namely, "personalized medicine". It has become increasingly difficult to convince Canadian patent examiners, especially in the case of "method of diagnostic" inventions, that a description demonstrates that the invention is useful, meaning that it is controllable and reproducible so that its objectives are predictably achieved.
In Apotex Inc. v. Wellcome Foundation Ltd.[i] the Supreme Court of Canada confirmed that utility of a patent must either be demonstrated or soundly predicted as of the Canadian filing date of the patent. The Court enunciated three requirements of the test for sound prediction, namely, that there must be:
- a factual basis for the prediction;
- an articulable and "sound" line of reasoning from which the desired result can be inferred from the factual basis; and
- proper disclosure.
In a recent decision Eli Lilly Canada Inc. v. Apotex Inc. ("Lilly")[ii] under the Patented Medicines (Notice of Compliance) Regulations, the Federal Court of Appeal was faced with an allegation of patent invalidity on the basis of lack of sound prediction. The FCA upheld the trial judge's decision which found that although the first two requirements of sound prediction had been satisfied, the patent was invalid for failure to meet the third element, proper disclosure. The FCA found that the underlying data upon which the inventors' relied for their sound prediction should have been disclosed in the patent.
The patent at issue, Canadian Patent No. 2,101,356, was directed to the use of a group of compounds in the treatment or prevention of osteoporosis by inhibiting bone loss in humans. The patent specification disclosed results of animal model studies and referred to an ongoing study on a group of human subjects where certain results were expected. According to the FCA, when a patent is based on a sound prediction, the disclosure must include the prediction. The FCA reasoned that "where the claimed invention had not yet actually been reduced to practice, the patent must provide a disclosure such that a person skilled in the art, given that disclosure, could have as the inventors did, soundly predicted that the invention would work once reduced to practice." In this particular case, the appellant accepted for purposes of the appeal that a particular human study, which was published after the priority date but before the Canadian filing date, was required in order to turn the prediction on which the patent was predicted into a sound one. The FCA concluded that the human study had to bedisclosed in this case in order to substantiate the inventor's prediction.
The FCA has thus articulated a position regarding inventions based on sound prediction that may have a direct impact on patents related to personalized medicine and other types of diagnostic and theranostic methods.
SATISFYING THE TEST OF "PROPER DISCLOSURE"
The Lilly decision has arguably raised the standard for the proper disclosure of sound prediction. It may be prudent for patentees to consider including as much data as possible to provide an adequate factual basis and to clearly enunciate the sound line of reasoning for the prediction in the patent. It should be noted that the law in Canadacontinues to be that establishment of utility through clinical trials is not required for patentability. The Supreme Court of Canada in Apotex, supra, stated that clinical testing was a regulatory standard and not required for patent utility and that "the doctrine of sound prediction, in its nature, presupposes that further work remains to be done."
A POSTERIORIDATA - RELEVANT FOR ASSESSMENT OF ACTUAL UTILITY
The capacity of genetic analyses, which form the basis of many BIO-IT patent applications, to predict behaviours or results has also been challenged by a posteriori data published in the literature. One example is the correlation between stress-induced depression and the short promoter of the human HTT gene. Risch[iii] demonstrated that, contrary to what was previously believed, there is no statistically significant link between depression and the presence of the short promoter of the human HTT gene. These a posteriori results question the utility of "method of diagnostic" claims that are based on assessing the predisposition to depression of a subject by detecting the presence of copies of a short promoter allele of an HTT gene.
Not only is it already difficult to translate the promise of personalized medicine into clinically validated diagnostic tests, but now it seems that the bar has been raised regarding the proper disclosure requirement in Canada. As a result BIO-IT applications may be filed at a more advanced stage of research and development in an attempt to satisfy the disclosure requirement. Though a posteriori data cannot be used to invalidate a patent on the basis of sound prediction, they may be relevant to an assessment of the actual utility of a patent. Although increasing numbers of diagnostic tools are being developed in the field of BIO-IT, obtaining patent protection for personalized medicine inventions has become more challenging, making it more important than ever to involve BIO-IT patent practitioners with in-depth knowledge of the industry at an early stage in the process in order to overcome the obstacles.