Boehringer v. JAMP continued: the Federal Court’s look at anticipation and sound prediction of utility

The Federal Court’s recent decision in Boehringer v JAMP, 2024 FC 1198 concerned two patents relating to nintedanib capsules (which Boehringer markets as OFEV). Our previous post looked at the Court’s discussion of claim elements as essential or non-essential. This post focuses on anticipation and utility of the 083 Patent, a patent relating to a new use: the treatment or prevention of idiopathic pulmonary fibrosis (IPF).

Anticipation

Invalidating a patent for anticipation has two requirements: disclosure and enablement. The dispute in this case focused on the former: whether the prior art patent application (WO 948) disclosed the invention.

WO 948 related to the use of Lck inhibitors for the treatment of immunologic diseases. It claimed treatment of “immunologic disease or pathological condition”, including “lung fibrosis”. JAMP asserted the skilled person would understand “lung fibrosis” in WO 948 to include IPF.

The Court held that lung fibrosis must be considered in context of the disclosure of WO 948. It found the skilled person would understand lung fibrosis as referring to lung fibrosis arising from autoimmune diseases believed to be associated with inappropriate T-cell activation. In short, since WO 948 focused on treatment of autoimmune or T-cell mediated diseases, the references therein to lung fibrosis were limited to lung fibrosis caused by an autoimmune disease. The evidence was that IPF was not an autoimmune or T-cell mediated disease and WO 948 contained no reference to IPF.

The Court relied on Gilead Science Inc v Canada (Health), 2013 FC 1270, which held: “[i]f there is doubt about what the prior art reference includes, it cannot be taken to meet the definition of anticipation”. Given the focus in WO 948 on Lck and T-cell mediated inhibition, the Court was unable to conclude that the skilled person would necessarily consider the treatment of “lung fibrosis” in WO 948 to encompass IPF.

The Court went on to conclude that, even if the skilled person would view IPF as a form of “lung fibrosis” contemplated by WO 948, the disclosure of “lung fibrosis” as one of the many disease conditions within WO 948 would be insufficient to constitute disclosure of the use of nintedanib to treat IPF. JAMP argued that prior disclosure of a medicine that can treat an umbrella of diseases can constitute disclosure of treating individual diseases within that umbrella, even if the specific disease is not explicitly identified. The Court found that in this case a number of choices would be required to arrive at the invention claimed in the 083 Patent. Thus, WO 948 did not provide clear and unmistakable directions to the use of nintedanib for the treatment of IPF. For instance, there was no prior association of nintedanib with IPF, “lung fibrosis” was not a homogeneous condition but could vary in severity depending on its cause, and not all lung fibrosis was associated with IPF.

The prior art did not disclose nintedanib for the treatment of IPF and thus did not anticipate the claims of the 083 Patent.

Sound Prediction of Utility

The parties agreed that the utility of the invention was the use in treatment of IPF in human patients. Utility was not demonstrated at the time of patent filing but was premised on a sound prediction. For a sound prediction to be established, there must be a factual basis for the prediction, the inventor must have an articulable line of reasoning from which the desired result could be inferred from the factual basis, and there must be proper disclosure.

The decision contains much discussion on whether the Court can consider only the disclosure in the patent or also the inventors’ work. The Court held that the knowledge, activities and endeavours of the inventors must be considered. While an understanding of the disclosure and the prediction as disclosed therein is ultimately assessed from the perspective of the skilled person, this does not shift the focus of the factual inquiry away from the inventors.

The Court stated that, if the full factual basis and line of reasoning was disclosed in the patent, then the issue would have been whether the skilled person would find this work was sufficient for a sound prediction. But JAMP took issue with the fulfillment of the disclosure requirement, asserting that the inventors did not disclose foundational work in the patent. The Court stated that the analysis must therefore also focus on the inventors’ work and what made the prediction sound.

However, the decision did not seem to turn on this. The Court walked through the inventors’ work, but ultimately focused on the patent disclosure itself and whether the information therein was sufficient for a sound prediction.

A patent need not disclose all test results obtained by the inventors, but must disclose the information that made the prediction sound. Boehringer’s position was that the inventors’ additional work was corroborative rather than grounding. The Court effectively agreed, as it ultimately found the disclosure in the patent sufficient for a sound prediction of utility.

It was not in dispute that the patent did not disclose all work conducted by the inventors. The only elements of the factual basis disclosed in 083 Patent were: a) a general discussion of certain in vitro testing (i.e., lab tests not in a living organism); and b) an evaluation of nintedanib in an in vivo bleomycin-induced rat model (i.e., testing in rats) relating to a preventative regimen.

To be sound, the prediction must be more than “a lucky guess”, “mere speculation”, belief, or a hypothesis that might later prove useful. The public is entitled to a teaching that is solid, accurate and meaningful and one that is based on exact science. But it is not a regulatory standard and the threshold is not high. A patent may be sustained provided a prima facie reasonable inference of utility exists.

The parties agreed the in vitro information would be insufficient on its own. JAMP argued that limitations in the in vivo data, such as the fact that data from a preventative regimen would not translate to efficacious results in a treatment regimen, rendered it incapable of supporting a sound prediction of efficacy in humans.

The Court agreed there were known limitations and criticisms of the in vivo model used but found the model was still regarded as the “gold standard” at the time and was the best available option for evaluating compounds for treating IPF. Disclosure of the positive results of testing using the preventative regimen provided the skilled person with data from the leading in vivo model, indicating that nintedanib had an inhibitory effect on fibrosis and therefore establishing nintedanib as a compound of interest.

Proof of clinical efficacy is not the standard for sound prediction. It is understood that further work may need to be done. The disclosure of the 083 Patent and its line of reasoning was sufficient as “it provided the PSA with a sound prediction of utility of nintedanib in the treatment of IPF that was more than mere speculation or hypothesis, but one grounded in the accepted in vivo model for evaluating potential efficacy.”

The Court held the 083 Patent was not invalid for lack of sound prediction of utility.

A copy of the decision may be found here.