Idenix Pharmaceuticals, Inc v Gilead Pharmasset LLC2017 FCA 161

Prospective patentees should exercise caution when beginning the patent process based on what they merely think can be accomplished, as this type of approach may lead to problematic gaps in disclosure.

In this decision, the Federal Court of Appeal (“FCA”) dismissed [16] Idenix Pharmaceuticals, Inc.’s (“Idenix”) appeal to the decision in Gilead Sciences, Inc v Idenix Pharmaceuticals Inc, 2015 FC 1156 (PCK Reporter summary available Here), where the Federal Court (“FC”) found Idenix’s Canadian Patent No. 2,490,191 (“the ‘191 Patent”) invalid, and dismissed Idenix’s counterclaim against Gilead Sciences, Inc., Gilead Sciences Canada, Inc. and Gilead Pharmasset LLC (collectively, “Gilead”). [1] This case raised no new issues of principle or novel application of established principles to the facts, and only sought to deal with the allegations of error put forth by Idenix. [3] The FCA did not find there to be any reversible error in the FC’s conclusion that the ‘191 Patent was invalid. [53]

Background

The ‘191 Patent and Gilead’s Canadian Patent No. 2,527,657 (“the ‘657 Patent”) [1] both claim nucleoside analogues with various bases and a particular structure at the 2’ position on a 5-carbon sugar ring consisting of a methyl group in the “up” position and a fluorine atom in the “down” position (“2’-C-Me/F”). [4] Idenix had previously files two US patent applications for synthesized 2’-C-Me/OH nucleoside analogues that have activity against the Flaviviridae family of viruses, which includes Hepatitis C. [5] Idenix proceeded to attempt synthesis of several similar compounds including 2’-C-Me/F, and, before successful synthesis, filed US patent applications in 2002 and 2003, which established priority for the ‘191 Patent. [5] Gilead’s predecessor succeeded in synthesising 2’-C-Me/F in 2003 and submitted a provisional US patent application in 2004, both detailing the step-by-step synthesis and antiviral activity against Hepatitis C, and laying the foundation for the ‘657 Patent. [6] Gilead brought a challenge to the validity of the ‘191 Patent in 2012, where the FC found in favour of Gilead and noted the ‘191 Patent as being invalid for insufficient disclosure and no sound prediction of utility. [7]

While there were three possible pathways whereby 2’-C-Me/F could be synthesized, the ‘191 Patent did not teach the person of ordinary skill in the art (“POSITA”) of any particular starting point at all. [8] Idenix argued that the ‘191 Patent taught the POSITA to select one of two of the pathways, or that those two pathways would have been apparent. [9] Despite this argument, the FC found that many steps in the synthesis remained undisclosed, particularly when it came to referencing intermediate compounds, or how to install the necessary fluorine group. [10] Common general knowledge and/or experimentation were not found to be adequate substitutes for the gaps in disclosure. [10] The FC also noted that the ‘191 Patent did not soundly predict utility of 2’-C-Me/F since Idenix had filed the patent prior to successful synthesis. [11] The activities of the ‘657 Patent were found to fall within the scope of the ‘191 Patent claims and, as a result of the invalidity of the ‘191 Patent, it was determined that the ‘657 Patent was not anticipated. [12]

Issues

Idenix alleged that the FC had made a number of factual and legal errors in reaching its decision. [13] The FCA distilled the allegations into the following three issues:

  1. Whether there was sufficiency of disclosure in the ‘191 Patent, [14]
  2. Whether there was sound prediction of utility in the ‘191 Patent, [14]
  3. Whether the ‘657 Patent was anticipated/invalid and/or infringing. [13]

Analysis

Insufficiency of Disclosure

Idenix alleged that the FC did not read the disclosure of the ‘191 Patent from the perspective of a POSITA with the benefit of common general knowledge. [20] This resulted in the FC’s finding that a key step in the synthesis was not disclosed. [20] The FCA found that on fair reading of the FC’s reasons, it was evident that the concern was with how the POSITA would have understood the patent. [27] Idenix’s argument here was found to be lacking based on this fair reading. [27] The FCA did, however, note that the FC’s choice of the term “express written disclosure” was not appropriate. [27]

Idenix also argued that the FC had used the improper candidate as an example of the POSITA, but the FCA found this argument to be of no consequence [28] based on the principle that courts must be informed by the evidence as to what knowledge a POSITA possesses and what they would have been able to do with it. [29]

With regards to the treatment of the expert and witness evidence, the FCA did not agree with the arguments Idenix put forth pertaining to any of the four witnesses. [31-41] In particular, the FCA rejected the arguments that the FC misunderstood comments made by expert witnesses, [33] that the FC made improper assessments of the expert evidence, [36] that certain transcripts were inadmissible, [38] and that the FC made unreasonable inferences regarding the conveyance of information between witnesses. [41]

Finally, Idenix submitted that the FC mistakenly understood that specific reaction conditions were required for certain steps, which led to the belief that the steps required more experimentation than the evidence showed. [43] The FC had found Idenix attempting to read something into the ‘191 Patent that simply was not there. [45] The FCA agreed with the FC’s determination that the disclosure of the ‘191 Patent was insufficient. [50]

No Sound Prediction of Utility

As no reversible error in the FC’s conclusion was found that the ‘191 Patent was invalid for insufficiency of disclosure, it was not necessary for the FCA to deal with Idenix’s submissions on utility. [52] The FCA made is clear however that this did not mean it was endorsing the FC’s approach to sound prediction. [52]

‘657 Patent Not Anticipated/Invalid Nor Infringing

Similarly, as no reversible error in the FC’s conclusion was found, the ‘191 Patent was held to remain invalid for insufficiency of disclosure. [53] The ‘657 Patent was not found to be anticipated nor did it infringe the ‘191 Patent. [53]

Conclusion

The appeal was dismissed with costs. [54] An application for leave to appeal at the Supreme Court of Canada has been filed.

Commentary

As in this decision, many patentees, particularly in the pharmaceutical industry, often begin the patent process without even having successfully synthesized the compound they seek to protect. In addition to the obvious issue of a lack of testing present, there is also the strong possibility of insufficient disclosure in the patent application, since all of the steps may not have even been documented yet. As this decision illustrates again, this is a profound gamble for such patentees. While Gilead first successfully synthesized 2’-C-Me/F and then submitted a provisional patent with step-by-step instructions along with the effects of the drug, Idenix chose to begin the patent process with what they thought they might eventually be able to accomplish. The result of Idenix’s approach was the presence of undeniable gaps in the disclosure – gaps that were not found by the courts to be fillable by common general knowledge.