The Federal Court of Appeal has dismissed an appeal by Pharmascience from a decision of the Federal Court (for further details, see "Federal Court decision regarding glatiramer acetate finds one patent obvious and another valid and infringed") that found Canadian patent No. 2,760,802 (the 802 patent) valid and infringed by Pharmascience's proposed glatiramer acetate product (Teva's Copaxone and Pharmascience's Glatect).(1) A second patent, Canadian patent No. 2,702,437 (the 437 patent), which was found invalid for obviousness by the Federal Court, was not at issue on the appeal.


The 802 patent claims the use of 40mg of glatiramer acetate (GA) injected three times weekly with at least one day between each injection for use in the treatment of a form of multiple sclerosis called relapsing-remitting multiple sclerosis (RRMS). In comparison with the known RRMS treatment (which employed daily injections of 20 mg of GA), the 802 patent describes the claimed thrice-weekly dose of 40mg of the drug as being an effective low-frequency dosage regimen that increases the tolerability of GA treatment.

The 802 patent discloses details of a phase III clinical trial (the GALA trial), including its expected results. However, the actual results were not included in the patent because the study had not yet concluded when the application for the 802 patent had been filed. The Federal Court found the utility was soundly predicted. The Federal Court also found that the claims of the 802 patent would not have been obvious. Pharmascience appealed both findings.


Pharmascience argued that the trial judge had erred regarding the disclosure requirement for sound prediction because she had not acknowledged the requirement to describe the factual basis and line of reasoning supporting the prediction of utility. The parties did not disagree on whether a "heightened" disclosure requirement for sound prediction exists, and the Federal Court of Appeal, therefore, noted that it was not necessary to comment on that question. The Federal Court of Appeal disagreed with Pharmascience's argument, holding that the trial judge had not misunderstood the disclosure requirement under the sound prediction doctrine.

Pharmascience also argued that the trial judge had erred in relying on a series of small studies not described in the 802 patent to support the soundness of the prediction. The Federal Court of Appeal agreed that the trial judge appeared to have incorrectly believed these studies were referred to in the 802 patent but held that this was not a reviewable error: even if it were a palpable error, it did not go to the very core of the outcome of the case because it was not clear that the trial judge had relied on those studies.

Pharmascience further pointed to a statement in the 802 patent that the complex pharmacokinetic behaviour of GA makes the effect of variation in the frequency of administration unpredictable and argued that this was an acknowledgement that the common general knowledge (CGK) could not provide a sound prediction of utility in varying the frequency of administration of GA. The Federal Court of Appeal disagreed, holding that this was instead an acknowledgement of the need for the GALA trial to test the theory that the lower frequency dosing will be safe and effective; the doctrine of sound prediction presupposes that further work remains to be done.

Finally, the Federal Court of Appeal was hesitant to interfere with the conclusions of the trial judge on utility because Pharmascience had adduced no evidence from its own experts on this issue, despite having the burden of proof.

Pharmascience argued that if the 802 patent does not fail for lack of utility, it must fail for obviousness: as the 802 patent does not provide any results of experiments, the factual basis and sound line of reasoning must come from the CGK, which would make the invention obvious to try. The Federal Court of Appeal disagreed, holding that there is no necessary inconsistency between the finding, on the one hand, that an idea is sufficiently described in the patent disclosure and the CGK to support a sound prediction that it will be useful and, on the other, that the idea is not sufficiently known in the prior art to lead the person skilled in the art directly and without difficulty to the solution taught in the patent. It was open to the trial judge to find enough in the CGK to support a sound prediction of utility, but not enough to find the invention obvious.

The Federal Court of Appeal also rejected arguments from Pharmascience that the trial judge had erred in her formulation of the test for obviousness by discounting or ignoring certain prior art that would not have been located on a reasonably diligent search and in finding that the obviousness argument was based on a mosaic of prior art. There was no error in the trial judge's concern that, given the difficulty in locating certain prior art, the skilled person would not have been led directly and without difficulty to combine these references.

Pharmascience would need leave from the Supreme Court of Canada to appeal this decision.

For further information on this topic please contact Andrea Berenbaum at Smart & Biggar by telephone (+1 416 593 5514) or email ([email protected]). The Smart & Biggar website can be accessed at www.smartbiggar.ca.


(1) Pharmascience Inc v Teva Canada Innovation, 2022 FCA 2.