The Federal Court of Appeal's (FCA) recent decision in Teva v. Leo Pharma addresses important issues related to the doctrines of sound prediction and sufficiency. At first instance, Locke J granted Leo Pharma's application for an order of prohibition, rejecting Teva's obviousness, utility and sufficiency allegations. Gauthier JA wrote for a unanimous Court of Appeal and dismissed Teva's appeal.
The patent at issue related to ointment compositions comprising two active ingredients, calcipotriol and a corticosteroid, along with a solvent. The prior art taught that calcipotriol and corticosteroids were both useful for the treatment of psoriasis. However, an instability problem had prevented these ingredients from being formulated together. The claimed compositions overcame that compatibility problem.
Sound Prediction - Evidence for Establishing a Sound Prediction
The first issue addressed by the FCA relates to the manner by which patentees can establish the elements of the sound prediction test. Teva argued that the Federal Court had before it no evidence of the factual basis or line of reasoning of the inventors, and thus that it was improper to find a sound prediction.
The FCA rejected this argument, holding that patentees are not limited in terms of the evidence that can be used to show the soundness of a prediction. In this case, there was evidence that the inventors could not participate as witnesses. There was also evidence as to the factual basis that the inventors possessed. Further, the patent itself conveyed the inventors' logic and line of reasoning to the skilled reader. Thus, Teva's argument failed despite the lack of direct evidence from the inventors as the inventors' prediction was established and was conveyed to the skilled person through the language of the patent.
Sufficiency - Trial and Error Experimentation
The most substantive of Teva's insufficiency arguments was that there was a particular order of adding the different components of the composition that must be followed in order to make the desired ointment. Teva argued that the patent was insufficient as the failure to describe this mixing order was a failure to describe an essential element of the invention.
Locke J had disposed of this argument in the Federal Court by finding that processes for making ointments were well-known at the relevant time, and that a skilled person could make the claimed compositions with their common general knowledge and possibly the use of some trial and error experimentation. Teva objected to the use of "trial and error" experimentation in Locke J's analysis on the basis of the Viagra decision's "minor research project" language. Teva argued that the SCC in Viagra overturned the law that non-inventive trial and error experimentation could be used in establishing enablement.
The FCA dismissed Teva's argument, distinguishing Viagra. Viagra was said to relate to a failure to disclose the "invention itself", whereas this case related to the second aspect of sufficiency, namely, the requirement that a skilled person be able to put the invention into practice (enablement). The FCA held that the longstanding law that non-inventive trial and error experimentation is permissible in considering enablement was correct and was properly applied in the Court below
 2017 FCA 50, on appeal from Leo Pharma v Teva, 2015 FC 1237
 Teva v Pfizer, 2012 SCC 60