Proving What the POSITA Would Have Thought

Apotex Inc. v. H. Lundbeck A/S / escitalopram 2013 FC 192 Harrington J
            CIPRALEX / escitalopram / 1,339,452

The primary issue in this escitalopram decision was obviousness, and “the prime issue, really, is what two over-qualified experts in 2012 would think an ordinary pharmaceutical chemist would have thought of doing in 1988 [the claim date]" [161]. It is well-recognized that the obviousness determination must guard against the hindsight bias, which means that something may have seemed difficult at the time, but obvious in hindsight. But the problem in this case, while related, is even more basic: what kind of proof is need to today, to establish what a hypothetical person counter-factually would have thought 25 years ago?

This was an action for impeachment of the ‘452 patent, and a counterclaim for infringement. Apotex was previously unsuccessful in NOC proceeding in which Harrington J was also the judge at first instance: 2010 FCA 320 aff’g 2009 FC 146. Harrington J noted that because of the long learning curve necessary to understand pharmaceutical patents, it is the practice of the Court to assign the infringement action to the same trial judge who heard the NOC application. As a matter of law, this does not give rise to an apprehension of bias [66], and the parties were agreeable; the party which prevailed in the application has the comfort of facing the same judge, while the losing party has a second chance, with better evidence, to address the judge’s concerns [67].

Infringement was admitted. While novelty and lack of utility were argued by Apotex, Harrington J quickly disposed of these arguments. The main battleground was obviousness, though Harrington J’s holding relating to Sildenafil-type insufficiency is also of interest, as are some of the remedial points. I will deal with these in subsequent posts.

The ‘452 claims escitalopram, which is the (+) enantiomer of citalopram, and related compounds and methods. At the claim date, citalopram was a known compound which was known to be racemic and known to be useful for treating depression [10]. It was also common general knowledge that one enantiomer might have a better therapeutic effect than the other [33], and for this reason there was a motivation to resolve the racemate [100]. The challenge was to actually resolve the enantiomers out of the racemic compound.

The evidence indicated that at the claim date, in 1988, there were many techniques available for resolving enantiomers generally, though some were more popular than others, but there was no single technique that was guaranteed to work. Consequently, the question of obviousness “boils down to whether or not it was ‘obvious to try’ certain techniques and whether at one point one might give up in frustration” [93].

The answer to this question turned on the expert evidence. Between the NOC proceedings and this action, both parties attempted to impugn the other side’s witnesses on the basis that they were hired guns, who had appeared many times for that party [111-112]. Harrington J refused to accept this position [112], on the basis that the ultimate question was whether their evidence was accurate: [115], [116]. It strikes me that the fact that a witness has testified many times for one party in itself says nothing as to the objectivity of their evidence; if a world-renowned expert honestly and firms believes a certain view to be true, and that view supports one party, it is not surprising that that expert would be called many times by that party. However, it is significant that in this case none of the witnesses had changed their evidence in any important respect. (Harrington J noted that the evidence one of one expert was not “on all fours” with his evidence in other cases, but he noted that the context was different and “[t]here was no need for [the expert] to deal with what had not been alleged” [114]. If any expert had substantially changed his testimony, this might be a different matter.

The more important problem with the experts is that they were experts.

[122] [Apotex’s experts] each came up with a resolution method without even having the patent at hand, a method very similar to the patent. Apotex says that since they could do it, having put themselves in the shoes of the “ordinary” medicinal chemist, it was therefore obvious to those to whom the patent was addressed. Lundbeck submits that this simply proves their point that the two doctors were so overskilled they could not put themselves in the shoes of the “ordinary” medicinal chemist in 1988. I agree with Lundbeck.

The fact that a world-leader in the field can explain how the problem could have been solved using the technology of the day, does not establish that an ordinary skilled person would have arrived at that solution [213], [215].

This does not necessarily imply that world-renowned experts should not be called, but it does mean that they should remain focused on the task, which is not to explain how they could have solved the problem at hand, but rather to explain to the court how a person of ordinary skill could have solved it. Perhaps in some circumstances, a top-flight teaching professor, who has one teaching awards rather than research recognition, might be more helpful in an obviousness attack?

Harrington J also remarked that “the party claiming obviousness must not only be able to demonstrate that the prior art exists, but to show how the person of ordinary skill in the art would have been led to combine the relevant components from the mosaic of prior art” [211]. The echos, to some extent, the “suggestion” aspect of the US TSM test, which is intended to address the question of why an ordinary skilled person would have combined the elements of the prior art. I would not suggest that we adopt such a test in Canada, which was of course rejected by the USSC in KSR v Teleflex 550 US 398 as being overly rigid. But even if there should not be any requirement of an explicit suggestion in the prior art to combine elements of the prior art, an obviousness attack will be strengthened if the experts can explain explicitly why an ordinary person would have been led to the solution in question.

Apotex also conducted experiments to show that the racemate could be resolved using 1988 techniques. Even these experiments foundered on the technical difficulty of recreating the 1988 environment. For example, the HPLC columns that were used in the experiments were not 1988-vintage columns, and the silica gel used in those columns was not the same as that which was available at the time [196]. In a mature technology this might not matter, but the HPLC equipment was evolving rapidly at the time [202].

Establishing obviousness requires a counter-factual inquiry based on hindsight. As this case illustrates, this is inherently a difficult undertaking, both for the court and for the parties.