What is a “Scintilla” of Utility?

Eli Lilly & Co v Teva Canada Ltd / atomoxetine 2011 FCA 220 Evans JA: Noël, Dawson JJA aff’g Novopharm Ltd v Eli Lilly & Co 2010 FC 915 Barnes J

2,209,735 – atomoxetine – STRATTERA

The factual background to this decision and the sound prediction argument are discussed in my last post. The main issue, both at trial and on appeal, was whether the patent was invalid for lack of demonstrated utility. The accepted position in Canadian law is that “[w]here the specification does not promise a specific result, no particular level of utility is required; a ‘mere scintilla’ of utility will suffice. However, where the specification sets out an explicit ‘promise’, utility will be measured against that promise” 2010 FCA 197 [76]. In practice, the patent can always be construed as making some kind of promise, and utility therefore is almost invariably measured against that promise. In this case the patent was construed as promising that atomoxetine would be a clinically effective treatment for ADHD [29]. The problem for the patentee was that as of the filing date, the sole study supporting the claimed utility of atomoxetine for treating ADHD, the MGH Study, was small scale and preliminary, with significant methodological shortcomings. Barnes J held that it was not adequate to establish clinical efficacy. On appeal, Lilly argued that only a scintilla of utility is necessary [31], but the FCA affirmed the standard doctrine that the patentee will be held to the promise of the patent. Whether that standard has been met is largely a factual question, and the FCA held that it was open to Barnes J on the facts to conclude that it was not.

There is nothing legally novel in this. What is more interesting, in my view, is to ask what the result would have been had the patentee been held only to the “scintilla” standard. I have argued in a previous post that the false promise doctrine, which holds the patentee to the utility specified in the patent, should be abandoned because it means that the validity of a patent may depend on the particular wording of the disclosure rather than the substantive usefulness of what has been disclosed. This is a trap for the unwary drafter – so long as an invention has the minimum required utility, there is no need to promise more – so a patentee who speaks too glowingly of its invention may thereby invalidate what would otherwise be a good patent. On its face this problem appears to be illustrated by the atomoxetine decision; if only the patent description had avoided words like “effective” in favour of perhaps “useful,” the lower “scintilla” standard would have applied, rather than the higher “effectiveness” standard.

However, in arguing that the false promise doctrine should be abandoned, I did not mean to argue that there should be no utility requirement at all. The role of the utility requirement in modern law is to police how far upstream in the development process a patent may be obtained. As the SCC said in Wellcome / AZT 2002 SCC 77 [69], a patent may not be granted for “mere speculation”. See also the famous passage from Brenner v Manson (1966) 383 U.S. 519 at 536: “a patent is not a hunting license. It is not a reward for the search, but compensation for its successful conclusion.” Barnes J made the same point in his atomoxetine decision when he said “[u]tility is not established on the basis of a mere hypothesis, an unproven idea or sheer speculation, even if later established” [FC 92]. Initially, Lilly had a theory that atomoxetine might be worth exploring because it had a selectivity profile arguably similar to several medications that were established to be effective. It then had the MGH Study, and later it had other studies that were sufficient to get regulatory approval. At the theory stage the utility of atomoxetine in treating ADHD is not sufficiently established for it to be patentable, and at the regulatory approval stage it clearly is. Somewhere in between a line must be drawn.

I suggest that this view of utility, as defining where in the development process a patent may be obtained, may be reconciled with the maxim that only a scintilla of utility is required. We should distinguish between the degree of utility, and the degree of proof of that utility. A scintilla of utility is sufficient, but the invention must be known to have that scintilla of utility. Evidence that shows the certainty of a small utility is not the same as evidence that shows the possibility of great utility: the former will support a patent, but the latter will not.

In the atomoxetine case, the MGH Study suggested a very substantial utility: over half of the patients in the study showed a substantial reduction in symptoms, compared with only 10% in the placebo group. The problem was not that the degree of utility implied by this study fell below the scintilla threshold; the problem was that the methodological limitations of the study meant it was not sufficiently well established that atomoxetine actually had that degree of utility. If a large scale study with no significant methodological limitations had confirmed exactly the same results, there is no doubt that the utility thereby established would easily have been sufficient to sustain a patent. On the other hand, if a large controlled study with no significant methodological limitations had demonstrated that atomoxetine had a small but statistically significant effect in reducing ADHD, this would also sustain a patent, illustrating the “scintilla of utility” maxim.

This view is consistent with the leading cases. There is very little in the modern case-law defining what is meant by a “scintilla” of utility because, as noted, the patentee is almost always held to the promise of the patent instead. The SCC decision in Consolboard is sometimes cited for this proposition (see e.g. 2009 FC 638 at [77]), but it is not in itself significant authority on this point: the paragraph in question expressly stands for the proposition the utility need not be specified in the disclosure, not for what the threshold is. The scintilla maxim is really established by the cases cited in Halsbury’s 3^rd, which is cited by the SCC in Consolboard, and by Fox, 4^th ed at p.159, which is also often cited for this point (see e.g. 2010 FC 1265 at [482]).

It is not possible here to go through all of those cases in detail, so a couple of examples will have to suffice. As a preliminary point it should be noted that the distinction between the degree of utility and the evidence of utility is most acute in pharmaceutical cases. In the case of mechanical inventions, it can normally be established exactly how well the invention worked at the relevant date, and the only question is whether the established utility is sufficient. It is perhaps for that reason that the older cases do not clearly make the distinction between the degree of utility and proof of that utility. Those cases nonetheless illustrate the point that the role of the utility requirement is control how far upstream in the development process a patent may be granted. Wandscheer v Sicard Ltd [1948] SCR 1 (cited by Fox), is a good illustration; the patent in question was for a snowblower with a rotating ejector pipe. The patentee had the idea of a curved neck to guide the snow, but used a sharp angled pipe that clogged with snow, rather than the parabolic curve which ultimately proved effective. The SCC held that utility was not established. Badische Anilin und Soda Fabrik AG v Levinstein (1887) 12 AC 710 (HL) (also cited by Fox), illustrates the scintilla maxim. The patentee had claimed a class of dyes, only one of which had achieved commercial success. The patent was attacked on the basis that the other dyes lacked utility, but Lord Halsbury disagreed, saying “[e]ach season very often has its own fashionable colour.” The other dyes had only a scintilla of practical utility, in the sense that they were not actually used and offered at most a choice, but there was no question that they were indeed effective in dying fabric, albeit in shades that no one wanted to purchase.

Lord Halsbury did also discuss what was meant by lack of utility: “I think what is so often referred to in Patent Cases as a laboratory experiment and as not admitting of the quality of utility is intended to be applied to cases where really there is no product at all. Such an experiment proves the possibility of a product but does not really practically produce one in such quantities and under such circumstances as really to make it a product at all.” That was really the conclusion of Barnes J in the atomoxetine case; the results of the MGH Study were “promising but only preliminary” – in other words, it established only the possibility of a product, albeit it a very good one.

If this point is accepted, the practical question remains: what evidence is needed to establish utility. As the Lilly pointed out on appeal, Barnes J’s decision in the atomoxetine case is somewhat at odds with the Viagra case, 2010 FCA 242 aff’g 2009 FC 638 in which it was held that utility was established on the basis of a pilot study on a similarly small scale. One might also add 2011 FC 239, in which a primary screen was held sufficient to establish utility. However, as the FCA noted at [42], “utility is largely a question of fact that is decided in each case on the basis of the evidence and the judge’s assessment of it.” This is not simply taking the easy way out; all of these cases were no doubt close to the line, and we cannot expect each court, on somewhat different facts, to draw the line in exactly the same place. While we cannot ask for certainty on this issue, I do suggest that the law would be clarified if we could focus directly on this issue, rather than obscuring it with the current focus on what exactly is promised by the patent.