Be Careful what you Promise: Selection Patents are Held by the Federal Court of Canada to a Higher Standard of Utility than their Corresponding Genus Patents

Published December 12th, 2011

In Eli Lilly v. Novopharm, 2011 FC 1288, the Federal Court held that Eli Lilly’s patent was an invalid selection patent on the ground that there was no sound line of reasoning as of the Canadian filing date to predict that olanzapine would have the advantages promised by the patent.

It is well-established law in Canada that a system of genus and selection patents is acceptable in principle.  However, in order to be valid the selection patent must, in addition to meeting all of the normal requirements for patentability, set out a substantial advantage to be secured or disadvantage to be avoided by using the selected species.

It is also well-established law in Canada that an invention must have utility.  When no specific utility is promised by the patent, only a “scintilla of utility” is required.  However, when specific promises are made by the patent, the claims are held to those promises.  The utility of a claimed invention may be demonstrated by examples within the specification or it may be based upon a sound prediction.  A sound prediction requires a factual basis, a sound line of reasoning, and proper disclosure of both the factual basis and the line of reasoning as of the Canadian filing date.

In the present case, Eli Lilly had been issued a selection patent to olanzapine (Canadian Patent No. 2,041,113; “the ’113 patent”), which was encompassed, but not specifically claimed, by an earlier genus patent (Canadian Patent No. 1,075,687; “the ’687 patent”).  The ’113 patent stated that olanzapine had several advantages, such as a “marked superiority and a better side effects profile than prior known antipsychotic agents, and…a highly advantageous activity level.”  

Novopharm argued that this list of advantages constituted a specific promise of utility and that there was no sound line of reasoning in the ’113 patent to support the prediction that olanzapine would have these advantages.  Eli Lilly argued that the ’113 patent simply promised that olanzapine is a relatively safe and effective anti-psychotic, which is the same utility that was promised by the ’687 patent.  The Trial Judge agreed with Novopharm on this issue, stating that “The promise of the ’113 patent must be greater than that of the ’687 patent…. It is simply not enough for a selected compound to achieve what was promised in the genus patent.” (paragraphs [110]-[111]).

In relying on the recent decision of the Federal Court of Appeal in Pfizer Canada Inc. et al. v. Apotex Inc., 2011 FCA 236, with respect to the medicine latanoprost, the Trial Judge stated that “where a patented compound is claimed to be safe and effective in the treatment of a chronic condition, utility will be demonstrated if the patent discloses studies showing that the patented compound, when administered over a long term, meets that promise.”  Thus, based on the wording of the ‘113 patent, the evidence and the fact that schizophrenia is a condition of a chronic nature, the Trial Judge determined the promise of the ’113 patent to be that olanzapine, when administered over the long term, treats schizophrenia patients in the clinic in a markedly superior fashion with a better side-effects profile than other known antipsychotics. (paragraphs [209-210]).Having construed the promise of the ’113 patent, the Trial Judge then applied the promise to both the compound claims and the use claims and had to determine whether, as of the Canadian filing date, there was a sound line of reasoning to support the prediction that olanzapine would have the promised utility.  The Trial Judge stated that:

“[w]hile the ‘113 patent contains some information about the testing of olanzapine, it does not set out a line of reasoning to support its alleged clinical superiority in the treatment of schizophrenia and better side effects profile. It does set out a rational basis for making a sound prediction that olanzapine would be useful in the treatment of schizophrenia, but not grounds for a sound prediction that olanzapine would treat schizophrenia in a markedly superior fashion, with a better side-effects profile than other known antipsychotics.” (paragraph [255])

In other words, the Trial Judge found that the ’113 patent provided a sound line of reasoning for some utility for olanzapine, but not for the specific utility of the advantages promised in the patent.

From a practical standpoint, this decision highlights the importance of careful patent drafting when it comes to a species selected from a previously disclosed broad genus.  While advantages of the selected species must be referred to in the patent in order to overcome double patenting and novelty issues, the language used to describe the advantages must be carefully chosen so as to avoid making promises that cannot be backed up by data and/or a sound line of reasoning.