In a unanimous decision released last Thursday (Apotex Inc. v. Sanofi‑Synthelabo Canada Inc., 2008 SCC 61), the Supreme Court of Canada confirmed that selection patents are, in principle, valid under the Canadian Patent Act.  In making their pronouncement, the Supreme Court polished other legal requirements for patentability along the way.

The case began 6 years ago, when Apotex sought marketing approval for a generic version of Plavix, a drug prescribed to prevent the risk of heart attacks or strokes.  In order to gain regulatory approval, Apotex alleged that the patent protecting the active ingredient in Plavix was invalid.  Although that patent was narrow in scope, relating to the active ingredient in Plavix, an earlier patent had disclosed that same ingredient as a member of over 250,000 possible compounds. 

In patent terms, the earlier patent is known as the “genus” patent while the latter is termed the “species” (or “selection”) patent.  Selection patents are granted in many fields of research when an inventor discovers the “needle in a haystack” with improved properties relative to the others in its class.  Innovative pharmaceutical manufacturers had argued that selection patents are necessary to protect useful drugs, since seven to eight of every ten drugs on the market are protected by a selection patent.

In denying Apotex’ suggestion that selection patents are invalid in principle, the Court held that although a genus patent may soundly predict a set of compounds and their properties, if a particular compound in that set was not made and its special advantages were not known, a patent should not be denied to the inventor who uncovers that compound and its special advantages for the first time.

The Court then reassessed the patentability requirements of novelty and obviousness.  In refining the previous patent novelty requirement that the exact invention be already made and publicly disclosed, the Supreme Court held that an approach should be adopted that inquires into what special advantages are taught by the prior disclosure and whether a skilled person is able to perform the subject matter of a prior disclosure without undue burden.  Under the “prior disclosure” requirement, the court must first look to the allegedly novelty-destroying document and ascertain whether there is disclosure of subject matter that would necessarily result in infringement of the selection patent.  No experimentation is allowed at this stage – the genus patent must disclose the special advantages of the selection patent invention to satisfy this portion of the test.  Next, the court must ask whether a person skilled in the art would have been able to perform the subject matter of the prior art without applying any inventive step.  A patent may fail for want of novelty if these two requirements are met.

This is a reasonable requirement.  Where a genus patent discloses a series of thousands of compounds which are said to contain a specific beneficial property, a new patent should not be issued simply for picking out a previously-unprepared compound from that series and confirming the predicted properties.  The current decision appears to reward research that produces unforeseen results.  In the case of Plavix, the court relied on evidence that showed that millions of dollars had been spent developing a form of the drug from the genus patent before channeling efforts towards chemical refinement to isolate a compound with decreased toxicity and increased tolerability.  It was apparently those efforts that uncovered the beneficial properties of the active ingredient in Plavix.

In addressing the obviousness requirement, the Supreme Court noted that both the United States and the United Kingdom have accepted that it is appropriate here to as ask whether an invention was “obvious to try.”  Although lower courts in Canada have rejected this approach, the Supreme Court held that in certain narrow circumstances an “obvious to try” test may be warranted.  This advance in the law allows a court to consider the relevant state of the art in the field of an invention: if there are a “finite number of identified solutions,” if routine trials will answer the question, and if there is a motive to research in a certain area, a patent may now fail by reason of obviousness.

The decision from Canada’s highest court is especially interesting in light of parallel cases litigated by the same parties in the United States and Australia.  In the US, the District Court for the Southern District of New York (492 F. Supp. 2d 353) held that the corresponding US patent was valid since the genus patent did not direct a skilled person towards the Plavix ingredient, and since that ingredient contained the unexpected combination of beneficial therapeutic properties and decreased toxicity.

In contrast, the Federal Court of Australia ([2008] FCA 1194) held that the Plavix active ingredient, although providing an advantage, did not provide one that was not already disclosed in the genus patent.  In other words, the Australian court was convinced that the selection patent was simply confirming properties that were already known from the genus patent.

he broader implications of selection patents are that they arguably extend a patent monopoly past the statutory 20-year limit.  While Supreme Court acknowledged that this is a legitimate concern that may run contrary to the objectives of the Patent Act, they were not willing to abolish selection patents on this basis.  At the end of the day, the court emphasized that selection patents encourage inventors to go back to the pool of potential candidates and uncover the hidden gem – the one that “does something better than and different from what was claimed in the genus patent.”  So long as a selection patent achieves this goal, it will stand on good footing.