Eli Lilly v Canada[1] is the first final patent law decision in international investment arbitration brought under Chapter 11 of the North American Free Trade Agreement (“NAFTA”). In this regard, it provides insight into likely approaches to future arbitrations seeking to challenge national jurisprudence under NAFTA or other trade agreements.

The Tribunal has demonstrated that it will take a conservative approach in adjudicating disputes under the investor-state settlement provisions of NAFTA, according significant deference to the decisions of national courts. Perhaps most notably, in its decision, the Tribunal seems to be guided by the view that, so long as a decision is not entirely inconsistent with the preceding jurisprudence, it cannot be considered a dramatic change from the existing law. This likely means that, at least for countries having common law systems, national court decisions or doctrinal developments will be overturned in only the rarest of cases.


Eli Lilly argued that, during the mid 2000s, Canadian courts adopted a “promise utility doctrine” for patent evaluation distinct from the previous “mere scintilla” utility test and inconsistent with Canada’s patent protection obligations under NAFTA chapter 17. Under the traditional “mere scintilla” test, Eli Lilly argued, so long as an invention was capable of being put to a specific use, it qualified as “useful” under the Patent Act. Under the “promise utility doctrine,” by way of contrast, when a promise is made, the patent examiners look to the promises made in the patent disclosure and measure utility.

The new promise utility doctrine, Eli Lilly argued, had been unfairly and retroactively applied to invalidate its Canadian patents in respect of the pharmaceuticals marked as STRATTERA and ZYPREXA in Canada. This invalidation thereby constituted an unlawful expropriation of Eli Lilly’s investments under NAFTA Article 1110, and a breach of Canada’s obligation to provide the minimum standard of treatment under NAFTA Article 1105, leading it to pursue its claim in arbitration.


The Tribunal, based on its analysis of the Supreme Court of Canada’s 1981 decision in Consolboard Inc. v. MacMillan Bloedel Saskatchewan Ltd. (“Consolboard”), concluded that the “promise utility doctrine” was not entirely novel and was implicitly anticipated by that decision. Justice Dickson, writing for the court, drew on the discussion of the meaning of “not useful” in British patent law as set out in Halsbury’s Laws of England. Quoting from that text, the Consolboard decision notes that “not useful” means “that the invention will not work, either in the sense that it will not operate at all or, more broadly, that it will not do what the specification promises that it will do.”[2] A number of patent decisions after Consolboard followed Justice Dickson in reproducing this excerpt from Halsbury’s, although they did not always do so in the context of a patent utility analysis.[3]

Eli Lilly argued that the mere reproduction of the full quotation did not demonstrate an application of the promise standard, rather than the lower “mere scintilla” operability standard. It pointed out that, in fact, Consolboard had not been cited as a source of the “promise standard” until 2005.

The Tribunal agreed with Eli Lilly that none of the pre-2005 cases citing Consolboard applied the promise standard. Nevertheless, it held that citations of the “not useful definition” as set out in Consolboard “cannot be ignored as indications of similar analysis and policy considerations to that of the promise doctrine.” In other words, the mere reproduction of the full sentence, without discussion or judicial comment, was enough for the Tribunal to infer a “background” consideration of the promise standard.

In particular, the Tribunal held that the “promise standard” was latent in the language of Consolboard (and those cases that cited it), had merely become patent (pun unintended) post-2005, and therefore did not constitute a dramatic change:

While the promise standard may not have played a significant role in the Canadian jurisprudence before 2005, and courts looked to the disclosure for the promise in relatively few cases, the rule was clearly “out there”, to be ignored at a patentee’s peril. Furthermore, the rule has a strong foundation in the language of Consolboard, which Canadian courts still cite for the promise standard today, as already discussed.[4]

Takeaway: NAFTA Chapter 11 claims won’t succeed if a decision can be reconciled with previous jurisprudence.

[1] UNCT/14/2.

[2] Consolboard Inc. v. MacMillan Bloedel (Sask) Ltd., [1981] 1 S.C.R. 504), at 525.

[3] Eli Lilly v Canada, para. 239

[4] [4] Eli Lilly v Canada, para. 324.