Eli Lilly’s '356 patent claims raloxifene and other benzothiophenes for use in treating or preventing osteoporosis or bone loss, particularly in post-menopausal females, and particularly without eliciting significant estrogenic responses in primary sex tissues.  

For a patent to be valid in Canada, a patent must have “utility.” The inventor must be in a position to establish utility as of the date the patent is applied for, based on either demonstrated utility or a sound prediction.  

Not surprisingly, at the time Eli Lilly filed its patent, it did not yet have data demonstrating the utility of the claimed use. Rather (and as the court observed), the invention was based on a prediction.  

The court noted that the '356 patent disclosed that:  

  1. raloxifene is a known compound having certain known medical uses in estrogen treatment, and it is known how to make it;  
  2. studies on rats with raloxifene showed that bone loss is prevented with minimal increases in uterine weight; and  
  3. studies on post-menopausal women were contemplated and expected to show that raloxifene is effective in inhibiting bone loss.  

The rat studies were positive, but the claim that raloxifene could have the same effect on women — in particular, estrogen-deficient, post-menopausal women who suffered from bone loss — was necessarily based on a prediction.  

Eli Lilly conceded that the basis for the soundness of the prediction was a study it had conducted on 251 post-menopausal women in Hong Kong; this study showed that raloxifene held promise as a skeletal antiresorptive.  

The Test for Sound Prediction

The court referred to the decision of the Supreme Court of Canada in Apotex Inc. v. Wellcome Foundation Ltd [2002] SCC 77, which provided a three-fold test for the requirement of sound prediction:  

  1. There must be a factual basis for the prediction.  
  2. The inventor must have, at the date of the patent application, an articulable and sound line of reasoning from which the derived result can be inferred from the factual basis.
  3. There must be proper disclosure.

The Court of Appeal held that when a patent is based on sound prediction, the disclosure must include the underlying factual basis for the prediction and the sound line of reasoning that grounded the inventors’ prediction. Because the prediction in this case was made sound by Eli Lilly’s “Hong Kong Study,” the court held that the results of this study should have been disclosed in the patent itself for there to have been sufficient disclosure. Eli Lilly’s failure to do so was fatal to its patent.

McCarthy Tétrault Analysis

This is a watershed decision that is particularly relevant to the filing of patent applications henceforth.

It is not unusual for claimed utility to be based not on demonstrated utility but rather prediction. This decision now requires, for the first time, that all data and studies that constitute the factual basis upon which the prediction is made should be disclosed clearly in the patent specification itself. In addition to including this data, patent applications should also articulate the science, logic and reasoning behind the prediction as it relates to the data disclosed.

Without both items being clearly set out in the patent application (i.e., the disclosure and prediction), the patentee is at risk of having its patent found invalid.