Late in 2013, many Canadian news outlets blared a headline similar to the one reported on the CBC’s website, “Eli Lilly Files $500M NAFTA Suit Against Canada Over Drug Patents.” The Indianapolis-based pharmaceutical giant’s lawsuit against Canada followed two court rulings in which the Canadian patents protecting the drugs Straterra and Zyprexa were invalidated in court after being challenged by a generic drug manufacturer. (Strattera (atomoxetine) is used as an attention deficit-hyperactivity disorder therapy and Zyprexa (olanzapine) is an antipsychotic drug.)

Eli Lilly’s suit against Canada was filed under the NAFTA provisions and alleges that Canada is in violation of its obligations to foreign investors as a result of the court rulings. The lawsuit has brought to the fore a very contentious issue in Canadian patent law known as the promise doctrine. This column discusses how the promise doctrine involves an analysis of the patent to determine how much experimental data is required to support the utility of an invention.

As I have mentioned in the past, there are three main criteria that a purported invention must meet in order for a patent to be obtained: the invention must be novel, inventive and have utility. With respect to utility, the question that must be asked is whether an invention does what that patent states it will do. This tenet of patent law seems simple enough. If a patent discloses that a compound inhibits an enzyme, it should be fairly easy to determine whether the compound possesses that utility and whether such experimental data is disclosed in the application — either the compound inhibits the enzyme or it doesn’t. As often is the case however, the reality is much more complicated, as evidenced by Eli Lilly’s lawsuit.

Over the past decade, Canadian courts have developed the promise doctrine with respect to the issue of utility. Under this doctrine, it is not sufficient to ask whether a compound, for example, simply inhibits an enzyme. The question that a court now attempts to determine is whether the inventors have made a promise in the patent, explicit or implied, which raises the threshold required to support the utility of the invention. This is a complex issue, so let me provide you with a straightforward example that demonstrates how even simple language in a patent can have significant consequences. A patent discloses a compound that inhibits Enzyme A, which may be involved in the etiology of breast cancer. The patent contains data that demonstrates that the compound inhibits Enzyme A but, like most pharmaceutical patents, does not contain in vivo testing. If the patent protecting the compound were to state that the “compound inhibits Enzyme A and may be useful for the treatment of breast cancer,” a court may find that the promise of the patent simply relates to compounds that inhibit Enzyme A.

In the eyes of the court, the treatment of breast cancer may simply be a goal, which may or may not be attained through further research and development. Some patents, however, use superfluous language, possibly in an attempt to bolster the patentability of the invention. For example, were the patent to state that the “compound is an outstanding inhibitor of Enzyme A and is useful for the treatment of breast cancer in humans,” a court may find that the inventor has promised that the compound will treat breast cancer. However, data from a simple in vitro assay may not be sufficient to support a promise that the compound will, indeed, treat breast cancer. Both of the patents covering Straterra and Zyprexa were invalidated on the basis that the data in the patent did not support the promise as construed by the court.

While the promise doctrine is unique to Canadian patent law, the principles that flow from the doctrine are widely applicable: include as much data as you can in a patent application and avoid flowery language to bolster patentability. In the end, only two questions need to be answered when drafting a patent application: what is the invention and what does it do?