Must the Factual Basis for Utility Be Disclosed in the Patent?

Pfizer Ireland Pharmaceuticals v Apotex Inc 2014 FCA 13 Sharlow J: Dawson, Mainville JJ aff’g 2012 FC 1339 Zinn J
            2,163,446 / VIAGRA / sildenafil

I hesitate to post this comment, because if I have understood the Apotex / Viagra decision correctly, the FCA, in a couple of short paragraphs which cite no authority, has wrought a major change in the law which is arguably inconsistent with SCC precedent, by requiring disclosure of the factual basis for demonstrated utility in the patent itself. This change is so radical and thinly reasoned that I can hardly believe that I have interpreted it correctly, hence my hesitation in posting. [UPDATE: On reconsideration, I've realized that my initial reading was indeed wrong: please see the update at the end of this post.] The decision is very short, and deserves to be read in full by every patent practitioner, so I hope that readers with a different interpretation of the holding will correct me.*

As is well known, in the NOC proceeding in Teva / Viagra 2012 SCC 60, the SCC held Pfizer’s Viagra patent invalid for failure to disclose the invention. In a subsequent impeachment action, Apotex / Viagra 2012 FC 1339 Zinn J granted summary judgment in favour of Apotex, holding the ‘446 patent invalid, on the basis that “the determination by the Supreme Court in its Reasons that the `446 Patent fails to meet the requirement of sufficient disclosure pursuant to subsection 27(3) of the Patent Act is a legal determination binding on this Court” [30] (see blog here). On appeal, Pfizer successfully argued that this holding was wrong in light of the well established jurisprudence to the effect that NOC determinations are not binding in subsequent infringement or impeachment: [24]. See my post here making the same argument

But it might have been better for Pfizer, and certainly for patentees generally, if it had lost on this point, as the FCA nonetheless affirmed Zinn J’s decision holding the patent invalid, and its reasoning was even more problematic. Recall that the basic problem with the ‘446 patent, which led the SCC to hold that the disclosure was insufficient, was that there were two claims to individual compounds: Claim 6 to a compound referred in this decision as UK-114,542; and Claim 7 to sildenafil. The FCA accurately summarized the SCC decision:

[15j] Since Pfizer had conducted tests that demonstrated that sildenafil was effective in treating erectile dysfunction, and none of the other compounds in the 446 patent had been shown to be effective in doing so, the invention was the use of sildenafil for the treatment of erectile dysfunction. That is what had to be disclosed in order to meet the requirements of subsection 27(3) of the Patent Act.

To counter this point, before the FCA in this case, “Pfizer points to evidence that UK-114,542, the compound named in Claim 6, was in fact tested before the Canadian filing date and found to be useful for treating erectile dysfunction” [26]. In other words, Pfizer’s argument is that in fact, the true invention was that both UK-114,542 and sildenafil had been shown to be effective in treating erectile dysfunction, and consequently the true invention was the use of both for treating ED. If that is right, then it would seem that the true invention was indeed disclosed.

The FCA rejected this argument:

[25] However, to establish that there is a triable issue with respect to the construction of the 446 patent, it was incumbent on Pfizer to adduce evidence or to refer to evidence already adduced in this matter. In particular, Pfizer should have adduced or referred to evidence that addresses how the skilled reader would construe the specification, and why that construction casts doubt on the correctness of the construction adopted in Teva 2012. The record contains no such evidence.

This implies that the evidence might be found in the record generally, but the FCA immediately went on to say:

[26] Pfizer points to evidence that UK-114,542, the compound named in Claim 6, was in fact tested before the Canadian filing date and found to be useful for treating erectile dysfunction. However, what the patent discloses is that as of the filing date studies had confirmed that one of the compounds in the especially preferred group induces penile erection in important males. There is no disclosure that UK-114,542 was tested, and Pfizer has adduced no evidence that is capable of establishing that the skilled reader should have discerned from the specification that it was tested. Therefore, evidence that UK-114,542 was tested is not capable of establishing a genuine issue for trial on the question of the sufficiency of the disclosure.

And similarly,

Therefore, for a considerable time before the summary judgment motion was made, it should have been apparent to Pfizer that it could be a problem that the disclosure in the 446 patent does not mention the testing of UK-114,542.

I have reproduced these crucial passages in full, because they seem to me to be saying that the fact that UK-114,542 was tested must be disclosed in the patent itself; even if there was evidence in fact “that UK-114,542 was tested” this is irrelevant, if this fact of the testing was not disclosed in the patent itself.

This seems to me to be tantamount to saying that the factual basis for utility must be disclosed in the patent. If this is the correct interpretation of the decision, it is a major change in Canadian patent law. As outlined in my article “Must the Factual Basis for Sound Prediction Be Disclosed in the Patent?” (2012) 28 CIPR 39, current case law requires that the factual basis for sound prediction must be disclosed in the patent, but this disclosure is not required when utility is demonstrated. As Barnes J put it in Lilly / atomoxetine 2010 FC 915, summarizing the authorities:

[116] It seems to me that it is beyond debate in Canada that where a patentee asserts that the utility of its invention has been demonstrated, it need not assert its supporting evidence in the patent.

Thus the FCA’s holding in this case is a major change in the law. The FCA has framed this issue as a matter of sufficiency of disclosure, rather than as a matter of utility, but this does not distinguish the prior case law. The principle that the factual basis for utility need not be disclosed in the specification is usually traced back to Consolboard [1981] 1 SCR 504, which expressly turned on sufficiency of disclosure. (And of course, functionally, for patentees, it does not matter where the disclosure requirement is found.)

Moreover, the FCA’s holding that the testing of UK-114,542 had to be disclosed in the patent seems to me to be inconsistent with Teva / Viagra. There, the SCC, after quoting its own decision in Wellcome / AZT 2002 SCC 77, stated that

[40] Nothing in this passage suggests that utility is a disclosure requirement; all it says is that "the utility required for patentability (s. 2) must, as of the priority date, either be demonstrated or be a sound prediction". Utility can be demonstrated by, for example, conducting tests, but this does not mean that there is a separate requirement for the disclosure of utility.

The SCC did then go on to say that “In any event, Pfizer disclosed the utility of sildenafil by disclosing that tests had been conducted” [41], which perhaps makes the SCC’s statement that disclosure is not required strictly obiter. But while lower courts are not strictly bound by SCC obiter, cogent reasons should nonetheless be given before departing from obiter which is obviously intended for guidance: R v Henry, 2005 SCC 76 [57]. At the very least, the FCA should have engaged with this passage and explained why that guidance should not be followed.

The FCA’s holding in Apotex / Viagra raises all the problems that are familiar from the requirement that the factual basis for sound prediction must be disclosed in the patent. There is the obvious and important fact that some patents will now be invalid which would have clearly been valid under the law as understood since Consolboard. But this new doctrine will also create uncertainty and arbitrariness as the parties seek to “discern[] from the specification” whether the compounds had been tested. For example, what exactly does it mean to “disclose” the testing? From Teva / Viagra it is clear that simple reference to the testing, without any details, is sufficient. Indeed, the decision strongly implies that a skilled person would know that the individually claimed compounds would work; in effect, the fact of testing would be implicitly disclosed by the fact that the compounds were individually claimed . Suppose then, that in fact, as Pfizer submits, both UK-114,542 and sildenafil had been tested and found to be effective, but the patent had failed to include the sentence stating that “one of the especially preferred compounds induces penile erection in impotent males”? Would the patent be valid, because the implied disclosure is true, or invalid, because of the lack of reference to testing in the description? If the former, why is the implicit disclosure insufficient in this case?

In summary, as I understand it, the FCA decision in Apotex / Viagra is a radical and thinly reasoned departure from existing law. I hope that I have not understood it correctly.

*UPDATE: I believe the hesitation I expressed at the outset of this point was warranted, as I did indeed interpret the decision incorrectly. On re-reading this passage, I now see the FCA's discussion as relating only to disclosure that sildenafil was the tested compound — the point at the heart of the SCC decision — and does not relate to disclosure of the evidence as such. While I now think my analysis of the passage is wrong, it is my policy not to edit my blog substantively after posting; hence this update.