Wednesday, July 15, 2020

The underlying facts were NOT in fact disclosed

Apotex Inc v Wellcome Foundation Ltd / AZT 2002 SCC 77, aff’g [2001] 1 FC 495 (FCA), var’g T-3197-90, 79 CPR (3d) 193 (FC) Wetston J

            1,238,277

The law of utility has become much quieter since AstraZeneca 2017 SCC 36, so I’ve been spending some time reviewing the case law to see where we have ended up. One issue that remains contentious relates to the enhanced disclosure requirement that arises when utility is established on the basis of a sound prediction. In Wellcome / AZT 2002 SCC 77 at paragraph 70 the SCC remarked that:

Normally, it is sufficient if the specification provides a full, clear and exact description of the nature of the invention and the manner in which it can be practised. It is generally not necessary for an inventor to provide a theory of why the invention works. Practical readers merely want to know that it does work and how to work it. In this sort of case, however, the sound prediction is to some extent the quid pro quo the applicant offers in exchange for the patent monopoly.

This suggests that something other than the normal disclosure is required, though the SCC did not go on to explain what that might be. This passage has been interpreted as requiring a heightened disclosure requirement in the case of a sound prediction, such that the factual basis for the prediction must be disclosed, or at least referred to, in the patent itself: Raloxifene 2011 FCA 220 [46]-[47] aff’g 2010 FC 915 [116]-[17]. This enhanced disclosure requirement is substantively controversial: see Gauthier JA’s remarks in Clopidogrel 2013 FCA 186 [132] and Rennie J’s remarks in Nexium 2014 FC 638 [158-59]. Even parsing the SCC’s statement is not easy: see Rennie J’s careful discussion in Nexium [139-60] arguing that “this sort of case” refers to a new use, not sound prediction generally.

In this post, I am not going to tackle either the interpretation of paragraph 70 or the substantive merits of an enhanced disclosure requirement. Instead, I will focus on a purely factual point. The reason the SCC gave for not elaborating further is that the precise disclosure requirements “do not arise for decision in this case because both the underlying facts (the test data) and the line of reasoning (the chain terminator effect) were in fact disclosed, and disclosure in this respect did not become an issue between the parties” [70, my emphasis]. Similarly, the SCC said that “[t]he trial judge has found that the inventors possessed and disclosed in the patent both the factual data on which to base a prediction,” as well as the line of reasoning necessary to support a sound prediction [75, my emphasis]. In this post I will show that the SCC’s statement that the underlying facts supporting the sound prediction “were in fact disclosed,” is, in fact, false.

This is a purely factual point, involving a comparison between what was disclosed in the patent with the evidence that was found to support a sound prediction. The bulk of this post is devoted to supporting my claim that the factual basis was not disclosed by going through the decision and the patent in rather pedantic detail. It may seem at times that I am belabouring the obvious, but since I am arguing that the SCC made a factual error, I would like to be clear as to the basis for my assertion. I’ll finish by considering the doctrinal implications of this observation.

I should stress that I am not disagreeing with any of the key holdings in AZT. I am arguing that the SCC’s statement at [70] was based on a factually erroneous premise; but that statement was expressly obiter. Given that the remarks were obiter, it is understandable that the SCC was perhaps not as careful in dealing with the issue as it might have been. An obiter comment of this type is clearly not binding in any event: R v Henry 2005 SCC 76 [57]. My argument is that the underlying factual error is one more reason for not placing significant weight on this obiter comment in the future development of the law on this point.

Background

The key claim in the 277 patent was to the use of AZT for the treatment or prophylaxis of AIDS (claim 22). This claim was held to be useful, at all levels of court, on the basis of a sound prediction of utility. The invention was developed by researchers at Glaxo Wellcome, who made the inventive decision to focus on AZT as a treatment for AIDS [FC 265]. The Glaxo researchers carried out their tests on murine retroviruses, not HIV itself, because Glaxo did not have the facilities required to test compounds against HIV [15]. The key tests carried out by the Glaxo researchers were in vitro “plaque reduction assays” using mouse cells and two different murine retroviruses, the Friend Leukemia Virus (FLV) and the Harvey Sarcoma Virus (HSV) [FC 113-14]. The results of these tests were received in November and December of 1984 [FC 115-16]. The assays showed that AZT “completely eradicated the virus” in the cells [FC 18-19], [FCA 10-11]. It was on the basis of these tests, in combination with existing toxicological and pharmacokinetic information on AZT, that the Glaxo researchers formed the belief that they had discovered a treatment for HIV [FC 116, 124], [FCA 11]. The researcher who carried out the test on the murine retroviruses, Martha “Marty” St Clair, was named as one of the inventors: [FC 113-116]. Glaxo then sent the compound to outside laboratories for further screening [FC 119]. In particular, they sent AZT to researchers at the NIH for in vitro testing against HIV itself, using a human cell line, known as ATH8, that had been developed to be particularly sensitive to HIV [FC 126], [17]. The results of the ATH8 test, communicated to Glaxo in February 1985, confirmed AZT’s efficacy against HIV in a human cell line [FC 126, 169, 170, 185], [73(iv)]. A primary focus of the litigation was whether the NIH researchers should be considered co-inventors. (Ultimately the holding was that the invention was not complete until the ATH8 tests were done, but the NIH researchers were nonetheless not co-inventors.)

The “mouse tests”

The SCC set out the findings which it held supported a finding of a sound prediciton of utility at [72] in narrative form, and at [73] under the heading “The Trial Judge’s Key Findings,” which itemized the key findings with reference to Wetston J’s reasons.

The SCC stated at [72] that the “results of the in vitro test of AZT against the HIV in a human cell line” — that is, the ATH8 tests — “taken together with Glaxo/Wellcome’s own data on AZT, including the mouse tests, provided a factual foundation” [72, my emphasis]. Thus, the SCC considered the factual foundation for the sound prediction to be “the mouse tests”; the ATH8 tests; and Glaxo’s other data on AZT.

What are “the mouse tests”? The only prior reference to “mouse cells” in the SCC reasons is at [11-12], where the Court describes testing using a plaque reduction assay, on “two retroviruses found in mice,” with results in “November 1984"; this undoubtedly refers to the murine assays carried out by St Clair using FLV and HSV. It follows that [72] also refers to the same assays, as no other mouse tests are referred to in the decision.

This is confirmed at [73]. At [73(iii)] the SCC set out four numbered findings, citing [FC 249], which are quoted verbatim from the FC decision at [249]. At [73(iii)(4)] the SCC stated that “Glaxo knew that AZT would inhibit the replication of [two strains of retrovirus in mouse cells] in November-December 1984,” (square brackets in SCC decision), citing [FC 249]. Referring to [FC 249], we see that the bracketed reference to “[two strains of retrovirus in mouse cells]” is a substitute for “MLV and HSV” in Wetston J’s reasons. “MLV” is the abbreviation for the retrovirus “murine leukemia virus” [FC 133]. At [FC 115] Wetston J had described St Clair’s tests as being carried out using the Friend Leukemia Virus, but FLV is a particular strain of murine leukemia virus. While the Glaxo tests were carried out on FLV in particular, much of the evidence turned on whether results from murine retroviruses could be extrapolated to human retroviruses, which is why, by [FC 249], Wetston J was referring to MLV rather than FLV specifically. In describing St Clair’s work, Wetston J abbreviated the Harvey Sarcoma Virus as "HaSV" (which was also used at [FC 163] in describing the same assay), while the abbreviation “HSV” is found only at [249] and is not explicitly defined elsewhere. While the terminology might have been more uniform, it is clear, especially considering the coincidence of dates, that Wetston J’s statement at [249] that “Glaxo knew that AZT would inhibit the replication of MLV and HSV in November-December 1984,” referred to the assays carried out by St Clair. Thus, the reference at [73(iii)(4)] to “two strains of retrovirus in mouse cells” is undoubtedly a reference to the murine assays carried out by St Clair using FLV and HSV. This is further confirmed by the FCA discussion of the facts at [FCA 8-11]. The only mouse / murine tests discussed by the FCA are the plaque reduction assays carried out by St Clair (see [FCA 8 fn2]). It is natural that the SCC would refer to the “mouse tests” as meaning the tests of that description discussed by the FCA.

I have gone on at some length to show that the “mouse tests” that the SCC considered to be part of the factual basis for the sound prediction is certainly a reference to the FLV and HSV tests carried out by Glaxo in the fall of 1984. This is significant because those tests are not disclosed in the patent. The only reference to any mouse tests in the patent is a disclosure of tests using the Rauscher Murine Leukaemia Virus (p3, 14-15). Even allowing for the possibility that the nomenclature is not consistent, this is clearly not a reference to the Glaxo assays of 1984 because the patent discloses an in vivo test in which AZT was administered in the drinking water to “female BALB/c mice infected with [the] Rauscher Murine Leukemia Virus.” This is clearly not the same as the in vitro plaque reduction assays carried out by St Clair.

Was the factual basis disclosed?

We are now in a position to summarize the key data and whether it was disclosed. For data that was disclosed, I have provided a page number referring to the specification of the 277 patent, dated 19th November 1987, as downloaded from the CIPO website. (I also reviewed the file wrapper and the disclosure dated 26th February 1986. It contains the same information on the points below as does the 19th November specification.)

The SCC set out the trial judge’s key findings at [73], which began by stating that “The trial judge upheld the claim in suit on the following bases.” After summarizing Wetston J’s findings at [73-74], at [75] the SCC said “These conclusions support a finding of sound prediction.” This implies that the evidence supporting a finding of sound prediction are those set out at [73-74].

● “information regarding toxicity, pharmacokinetics and the pharmacology of AZT in an internal document called the ‘Wise-Burchall report’” [73(i)].

The Wise-Burchall report was a compilation of Glaxo’s existing information on the effects of AZT derived from its earlier efforts to develop AZT as an antibiotic [FC 117]. It contained the following information:

• Toxicity in mice and rats:

• Disclosed (p18 “Toxicity Assay”)

• Pharmacokinetics in rats, pigs, chickens, sheep and calves, including an oral uptake in a rat, pig, chicken and calf and a half-life study in rats and sheep [FC 118, 177]:

Not disclosed.

• Information regarding anti-bacterial action [FC 118, 177]:

Not disclosed, though perhaps not relevant

●“some preliminary testing on HeLa Alpha-DNA polymerase” ([73(i)], citing FC [118])

Not disclosed.

● Glaxo/Wellcome's experience with other nucleoside analogues [73(ii)] citing FC [177]):

Not disclosed.

●Nucleoside analogues could act as chain terminators; effect of inhibiting reverse transcriptase; that MLV and HIV are retroviruses [73(iii)(1)-(3)], citing [FC 249]

• Common general knowledge [FC 249].

● AZT inhibits replication of FLV and HSV: [73(iii)(4)], citing [FC 249]

Not disclosed

● The results of the NIH / ATH8 test: [73(iv)], citing [FC 179]

• Disclosed. The ATH8 tests are probably the tests disclosed at p3, 17 “(iii) Preventing Infection by AIDV,” as relating to ‘cloned T4, tetanus- specific, T-helper lymphocytes,” because the ATH8 cell line is in fact such a cell: see US Patent 4,704,357, the patent for the ATH8 cell line, at col 2 of the disclosure, describing the invention as ‘A tetanus toxoid-specific T-cell line.” It is possible that the ATH8 tests are actually referred to by the patent as the tests on the H9 cell line, referred to at p3, 17-18, as ATH8 is an improved version of H9 [FC 205].

●The inventors possessed and disclosed in the patent “a line of reasoning (chain terminator effect)” [75]

• Disclosed: (p4, p16 (b)(i) “AIDV Reverse Transcriptase Activity”)

It is therefore perfectly clear that what the SCC itself described at [73] as the “key findings” on which the sound prediction was based were not all disclosed in the patent. Even if we suppose the crucial findings were only those mentioned by the SCC at [72], namely the ATH8 tests and the “mouse tests” (ie the in vitro FLV and HSV assays), the full factual basis for the sound prediction was not disclosed in the patent.

Why wasn’t the factual basis disclosed in the patent?

Why wasn’t that evidence included in the patent? I’ve gone on about all the data that was not disclosed, because that is the evidence identified by the SCC as supporting a sound prediction of utility, but that’s not to say that the specification didn’t have any data at all. The specification did have quite a bit of data relating to antiviral activity (p14-17),* but it almost all related to HIV; no doubt this is because the drafters believed that the best data to support claims to the use of AZT to treat HIV/AIDS would be testing of AZT on HIV itself. Recall that Glaxo did not have the facilities necessary to deal with HIV itself, so all of Glaxo’s testing was done on other retroviruses, including FLV and HSV. The testing on HIV that was disclosed in the patent would all have been done after the mouse tests which Glaxo relied on to establish sole inventorship. Because the utility argument was driven by the underlying issue of inventorship, the utility argument turned on evidence that was, for the most part, not disclosed in the patent.

The trial judge did not find that inventors disclosed the factual basis in the patent

There is another problem. The SCC stated that “[t]he trial judge has found that the inventors possessed and disclosed in the patent” the factual data and the line of reasoning to support a sound prediction [75, emphasis added]. This statement is wrong. There was never any finding by the trial judge that the factual basis was disclosed. On the contrary, Wetston J never distinguished between evidence that was or was not disclosed in the patent itself. That is why it is necessary to scour the patent itself to determine whether the mouse tests were disclosed; Wetston J discussed the tests at some length, but never mentioned whether or not they were disclosed. The discussion of utility and related issues is at [77-186], and Wetston J simply does not distinguish between information that is or is not disclosed in the patent. For example, at [129], Wetston J asked “Did the utility claimed at February 6, 1985, exceed the invention?” and he began his analysis by simply noting that at that date the inventors “had” the various items of information — without specifying whether or not they had disclosed it in the patent. The only discussion of what was disclosed in the patent was in the context of overbreadth, e.g. in the context of discussing whether claims for prophylaxis and for treatment of all human retroviruses were “not invented and. . . not disclosed in the patent” at [285]-[293], [294]-[303]. And here the argument regarding disclosure was that “there is no information in the disclosure on dosages, timing, duration, or method of treatment with respect to prophylaxis” [286], which is to say, disclosure of how to use, not data supporting a sound prediction. Even in that context, at [300] Wetston J considered evidence that was clearly not disclosed, and while he was not persuaded by it, he considered it rather than dismissing it on the basis that it had not been disclosed.

Why didn’t the trial judge discuss whether the factual basis was disclosed?

Why did Wetston J not make any finding that the factual basis was disclosed in the patent? Why did he not even mention whether the data he discussed was disclosed in the patent? Why wasn’t the relevant evidence included in the patent? The answer is simple: neither the drafter of the patent, nor the parties to the litigation, nor the trial judge, thought that it mattered whether the evidence was disclosed in the patent. While Wellcome / AZT is now known primarily as the leading case on sound prediction, the primary issue was whether the NIH researchers who carried out the ATH8 tests were co-inventors. Glaxo argued (successfully), that they were not, while Apotex and Novopharm argued they were. Because of this underlying issue, the focus of the sound prediction argument at trial was whether the invention was useful, and therefore complete, before the samples were sent to NIH for testing on the ATH8 cell line. This is why the argument focused on the information available up until, and including, the ATH8 testing.

Why did the SCC say the factual basis was disclosed?

The SCC twice said that the factual basis for the sound prediction was “in fact” disclosed in the patent, once at [70] and a second time at [75], where the SCC began by saying that “These conclusions support a finding of sound prediction,” clearly referring to the conclusions that had been summarized in the two preceding paragraphs. It is really very difficult to understand why the SCC said the factual data had been disclosed in the patent. Indeed, at the end of the paragraph summarizing the key findings, the SCC at [73(v)] quoted the trial judge [FC 185] as saying that his conclusion was based on the data “considered cumulatively, in conjunction with all of the evidence adduced and considered in this trial” (emphasis added). It is abundantly clear that not all of the evidence adduced at trial was disclosed in the patent. Remarkably, it was only four sentences later that the SCC at [75] went on to say that the factual data on which to base a sound prediction was disclosed “in the patent,” referring to “these conclusions” which it had just finished summarizing.

The SCC’s statement that the factual basis for the sound prediction was “in fact” disclosed, is simply a factual error. Part of the explanation for the error, it must be said, is sloppiness; perhaps we can blame the clerks. But part of the explanation is that the SCC did not have the benefit of argument, or indeed, of any submissions at all, on this point. The issue of the disclosure requirement for sound prediction was not raised at trial, nor was it raised in submissions to the SCC. The submissions by Apotex focused on the inventorship issue, and on the point that patents must not be granted for speculation and that the invention must therefore be complete as of the priority or filing date. The discussion of “Disputes Over Sufficiency of Disclosure” at [56]-[59] (Factum of the Appellant, Apotex Inc, SCC Court File 28287) argued that the invention had not been made because sufficient testing had not been carried out — e.g. AZT had not been tested in humans—and not on the basis that the experiments had not been disclosed in the specification. Novopharm (Appellant’s Factum (Novopharm Ltd.), SCC Court File 28287) had a more extensive discussion of sufficiency of disclosure in the context of “Issue 3: The Specification Must Give Directions as to How to Use the Invention” (p32), but this was directed at the argument that “that there was not enough information in the patent to determine how to use the drug as a prophylactic” (Novopharm factum at [122], quoting Wetston J at [288]). That is, the complaint about the disclosure was directed at disclosure of how to use AZT as a prophylactic, and not about disclosure of the test data. See also Novopharm Factum [138] submitting that the disclosure must “give all information that is necessary for successful operation of use of the invention,” and [140] arguing that “[w]here the invention rests in the therapeutic utility of a known compound, the therapeutic utility must be sufficiently disclosed.” Certainly, none of the submissions by any of the parties discussed whether the key information set out by the SCC at [72-73] was or was not disclosed in the patent itself.

Summary

In summary, the SCC in AZT was factually wrong in stating that the data at issue had been disclosed in the patent; it was also factually wrong in stating that the trial judge had made a finding to that effect. The reason for these errors was, in part, that the SCC did not have the benefit of argument by counsel.

What are the implications of this error? First, it is clear that the suggestion that there might be a heightened disclosure requirement “in this sort of case” was novel. Surely, if Apotex and Novopharm, represented at trial by highly experienced patent litigators, had had any inkling of a requirement that the factual basis for a sound prediction should be disclosed in the patent, the issue would have been raised at trial and at least adverted to by Wetston J.

Perhaps more importantly, the error reinforces the point, noted by Gauthier JA in Clopidogrel [132], that Binnie J’s brief statement at paragraph 70 of Wellcome / AZT may not be a proper basis for a heightened disclosure requirement. The comments are obscure — paragraph 70 implies that the disclosure requirement “in this sort of case” is different from normal, but there is no suggestion whatsoever as to what disclosure requirement might be appropriate, and it is not clear what kind of case is “this sort of case.” The comments were obscure because they were expressly obiter, and not intended for guidance, per R v Henry 2005 SCC 76 [57]. The comments were made without the benefit of any argument by counsel, or any consideration by the Federal Court or the Federal Court of Appeal. And finally, the comments were premised on what is, with respect, a clear factual error. In another part of Wellcome / AZT, the SCC held at [44] that the appropriate “standard of review” of a granted patent is “reasonableness simpliciter.” In Eli Lilly / cefaclor 2009 FC 991, Gauthier J declined to follow this holding by the SCC [350]-[370]. She noted that this holding would have drastically modified existing law, and “[o]ne would expect such a shift to be done clearly and in express terms” [355]. She also noted that the SCC had raised the issue itself, without the benefit of submissions by counsel [356 fn126]. The case for ignoring paragraph [70] is even stronger.

In R v Henry, [57] the SCC remarked that “The notion that each phrase in a judgment of this Court should be treated as if enacted in a statute is not supported by the cases and is inconsistent with the basic fundamental principle that the common law develops by experience.” There is no doubt a discussion to be had as to whether there is or should be an enhanced disclosure requirement for sound prediction. That discussion, in my view, should be based on sound law and policy, without being burdened by reference to paragraph 70 of Wellcome / AZT.

Notes

*In particular, the patent disclosed or referred to the following data that was not referred by Wetston J, or at any other level of court: in vitro tests by AIDV in H9 and U937 human lymphoblastoid cell-line (p3), (p17-18); PHA- stimulated white blood cells (p3)(p17-18); cultured peripheral blood lymphocytes (p3); TM-11 cells co-cultivated with HIV producer MJ-tumour cells(p15); as well as the in vivo murine tests Rauscher Murine Leukaemia Virus (p3, 14-15). The chain terminator effect, which formed the basis for the sound prediction, was disclosed in the patent, as was the ATH8 tests. The ATH8 tests are probably the tests described at p3, 17 as relating to ‘cloned T4, tetanus- specific, T-helper lymphocytes,” because the ATH8 cell line is in fact such a cell: see US Patent 4,704,357, the patent for the ATH8 cell line, at col 2 of the disclosure, describing the invention as ‘A tetanus toxoid-specific T-cell line”; moreover, Wetston J at [126] described the ATH8 assay as using “the HIV virus in a human cell line of immortalized cloned T4 cells.” It is possible that the ATH8 tests are actually referred to by the patent as the tests on the H9 cell line, as ATH8 is an improved version of H9 [FC 205]. The mere fact that it is unclear which tests disclosed in the patent refer to the ATH8 cell line is telling in itself: it is not clear from the decision, precisely because there was not any finding by Wetston J that the relevant data was disclosed in the patent itself, even in respect of the ATH8 cell line, which was disclosed.

4 comments:

  1. Interesting and important point.
    As I am sure you are aware, the stem case for sound prediction is the British case: Olin Mathieson Chemical Corp. v. Biorex Laboratories Ltd., [1970] R.P.C. 157, referred to at paragraphs 60 and 61 of AZT. In that British case there was no suggestion that a heightened disclosure was required to establish sound prediction.

    ReplyDelete
  2. Agreed. I’d go further: in Olin Mathieson it is clear that the sound prediction was based on data that was not in the patent. At 195 Graham J identifies two points “of great importance,” showing that the CF3 substitution has the greatest activity, saying “it seems right, and I so hold, that the plaintiffs are entitled to rely on [that] fact.” He expressly says “they do not say or promise this in their specification,” and he goes on to say that if it is true “and it appears to be so from works....subsequent to the date of the patent” then the patentee has made a substantial contribution. Immediately thereafter he concludes that the claims are “fairly based,” which is the sound prediction point. It couldn’t possibly be more explicit that the key evidence was not in the patent.

    ReplyDelete
  3. You are right! The "heightened disclosure" requirement really was created out of whole cloth.

    ReplyDelete
  4. Also in Monsanto Co. v. Commissioner of Patents, [1979] 2 SCR 1108, sound prediction was established on the basis of affidavits submitted to the Board: 1119. See also the dissent of Martland J at 1132 saying that the Board is entitled to consider the material submitted to it, including the affidavits.

    ReplyDelete