An Australian Federal Court judge has found that the specification of Australian patent no. 2003200700 does not disclose the ‘best method’ of performing the claimed invention (in this case, the compound perindopril arginine, which is sold under the name COVERSYL as a drug to treat hypertension). The patent is owned by French pharmaceutical company Les Laboratoires Servier, and was challenged by generic drug manufacturer Apotex, which wishes to bring its own version of the drug to market.
Not surprisingly, Apotex applied to the court for revocation of the Servier patent on a number of additional grounds: lack of novelty; lack of inventive step; lack of fair basis for the claims in the description; and that the patent was obtained by false suggestion or misrepresentation. Apotex failed on all of these grounds, succeeding only in persuading the court that the patent did not describe the best method known to Servier of performing the invention.
Lack of novelty, inventive step or fair basis is assessed on a claim-by-claim basis, i.e. for any of these grounds the patent might have been only partially invalid. Failure to disclose the best method, on the other hand, would usually be expected to result in the patent being declared wholly invalid (I shall explain why below), making Apotex entirely successful in its application for revocation of the Servier patent.
Interestingly, however, in this case the court has stopped short of immediately ordering revocation of the patent. The judge (Justice Rares) has instead indicated that he will hear further submissions from the parties on the question of appropriate relief, as well as on the matter of an award of costs.
This strikes me as odd. It is difficult to conceive of any reason to maintain a patent where the applicant failed to disclose the best method known to it at the time of filing – this being one of the fundamental requirements of the Australian patent law. Nonetheless, the Patents Act 1990 allows the court a discretion not to revoke a patent, even if a ground of revocation is made out.
If the matter proceeds to a hearing on the question of whether or not the patent should be revoked, it will be interesting to see the arguments presented on behalf of Apotex and Servier, and the decision-making process of the judge. However, this is perhaps not very likely, at least not in the near future, since (as in most pharmaceutical patent cases) an appeal seems almost certain to be filed before any further hearing can take place.
The Role of ‘Best Method’Section 40(2)(aa) stipulates that a complete patent specification must ‘disclose the best method known to the applicant of performing the invention’. (Note, however, that in this case the former version of s 40 applies, in which exactly the same requirement was set out in paragraph (2)(a).)
The obligation to disclose the best method known to the applicant has long been a feature of the patent law in many countries, and particularly those (such as Australia) whose laws are derived ultimately from earlier British statutes.
The reason for the obligation is simple – it is a pillar of the ‘bargain’ that was traditionally viewed as being at the heart of the patent system: the applicant receives a limited-term monopoly to exploit the invention, in exchange for a full disclosure enabling others to learn from it, improve it and/or to exploit it freely once the monopoly expires.
An applicant’s failure to disclose the best method known for performing the invention, at least if this can then be maintained as a trade secret, potentially provides an advantage over its competitors which would extend beyond the expiry of the patent. The ‘best method’ requirement has therefore often been regarded as essential to the patent bargain, and a failure to comply as a form of ‘cheating the system’.
Having said that, however, many modern patent laws have diluted the importance of disclosing the best method, or even eliminated it completely. The European Patent Convention, for example, contains no corresponding provision – Article 83 requires only that ‘[t]he European patent application shall disclose the invention in a manner sufficiently clear and complete for it to be carried out by a person skilled in the art.’ And while the US patent law retains a ‘best mode’ requirement in 35 USC 112(a), since commencement of the America Invents Act on 16 September 2011, 35 USC 282 (which provides defences in a patent validity or infringement proceeding) no longer includes the failure to disclose the best mode as a basis on which any claim of a patent may be cancelled or held invalid or otherwise unenforceable. In other words, an applicant is still expected to disclose the best method in the application, but once a patent has been granted it cannot be invalidated only because of a failure to do so.
In Australia, however, a failure to disclose the best method known for performing the invention remains a ground for revocation of a patent, under section 138(3)(f) of the Patents Act.
A few other points are worth noting about the best method requirement in Australia:
- the relevant date for assessing the ‘best method’ is the date of filing the application (Rescare Ltd v Anaesthetic Supplies Pty Ltd  FCA 537), i.e. if the applicant identifies a better method at a time subsequent to filing, there is no obligation to amend the specification to include that method;
- if the best method is not disclosed at filing, the specification may be amended (if the amendment is otherwise allowable), at least up until the date of grant, in order to rectify this deficiency (Pfizer Overseas Pharmaceuticals v Eli Lilly  FCAFC 224);
- while the best method must be disclosed in the specification, it need not be identified, i.e. if multiple methods are described, the requirement is satified so long as one of them is the best known to the applicant, even if it is not expressly identified as such (also Pfizer v Eli Lilly).
Servier’s ShortcomingServier’s patent specification described a method for making the claimed invention, i.e. the compound perindopril arginine, in sufficient detail to enable the relevant skilled person to put the invention into effect. In particular, a passage covering about a page of the specification describes ‘preparing perindopril arginine from L-arginine and perindopril according to a classical method of salification of organic chemistry’ (see paragraph  of the judgment). It was not in dispute that this generally described the methods actually used by Servier’s chemists to prepare the component, nor that the description was sufficient to enable other chemists to achieve the same result.
I do not propose to get into the details – if you are interested in the chemistry, you can read the judgment! Broadly speaking, the problem the court found with Servier’s disclosure was that it left critical parameters of the preparation to be discovered by any person attempting to reproduce the method. There does not appear to have been any question of this being anything more than a matter of routine, perhaps involving some trial and error. However, the fact was that Servier chemists had, on at least two separate occasions prior to filing of the patent application, successfully performed a’ classical method of salification’, and therefore were clearly in a position to provide the additional details necessary to save the reader from performing this additional investigative work.
As the judge said (at ):
I am of opinion that the 1986 and 1991 methods are better than that disclosed in the patent because they eliminated the risks as to time and resources needed in the trial and error experimentation potentially involved in a skilled addressee following the variety of possible, but unspecified, methods of classical salification.
The court’s full reasoning on this point may be found at paragraphs - of the judgment.
Will Servier’s Patent Now Be Revoked?As I indicated at the outset, my expectation would be that a failure to disclose the best method should result in the patent being wholly revoked. In such a case, the patentee has – either intentionally, or through an unfortunate lack of diligence – failed to hold up its side of the patent bargain, i.e. to disclose everything it relevantly knows about performing the invention in exchange for a term of monopoly.
Section 138(3) of the Patents Act unquestionably authorises revocation, although it does afford the court a discretion in whether or not to do so (the specific wording that ‘the court may, by order, revoke the patent…’). Apotex applied for the patent to be revoked on a number of grounds, and has been successful on one of those grounds. There does not appear to be any reason why it should not obtain the relief it sought, i.e. an order that the patent be revoked.
Yet this is what the court concluded on this occasion (at -):
Apotex has succeeded only in establishing that the specification in the patent did not describe the best method known to the applicant of performing the invention and so did not comply with s 40(2)(a). Accordingly, that raises the question of what relief should be granted under s 138(3) of the Act. Apotex’ many other challenges to the validity of the patent have all failed and those took up the bulk of the evidence and hearing. That failure should be reflected in costs. …
I will hear the parties on the question of relief and on costs.
I really cannot see how any ‘question’ is raised in relation to relief, or why the word ‘only’ should be used in relation to a failure to disclose the best method. The only relief authorised by section 138(3) is revocation, and if the court were to decline to make such an order this would leave Apotex (and, presumably, other generic manufacturers) exposed to an infringement action, despite Servier’s failure to satisfy the ‘best method’ requirement.
As for costs, I can see the court’s point regarding the time and expense incurred on grounds that were not successful. On the other hand, however, it does not appear that those grounds were manifestly baseless, and ultimately Apotex has succeeded in establishing what it set out to prove, i.e. that the patent is invalid. If the court takes a piecemeal approach to costs, then it seems that Apotex could end up having to pay a majority of Servier’s costs of the proceeding, despite having been successful in its application.
This all strikes me as rather odd, although it may not have the opportunity to play out since an appeal to a Full Bench of the Federal Court is a practical certainty. Justice Rares has already foreshadowed this, requesting that the parties ‘identify any errors or oversights in these reasons that may readily be corrected, including any argument or issue of substance that may not have been addressed so as to avert any unnecessary issue in any appeal’ (at ).
Conclusion – What to Make of This?Failure to disclose the best method of performing an invention is not often raised as a ground of invalidity – presumably due to the difficulty of proving the subjective state of mind of the applicant at the time of filing of the application. It is even more rarely successful. In 2011, Justice Bennett considered the criteria by which to assess whether there has been a failure to describe the best method in refusing an application for summary dismissal of this ground of invalidity. However, that dispute subsequently settled out of court, and the issue never proceeded to a full trial.
In this case, it seems almost as though Justice Rares considers a failure to disclose the best method as some form of ‘second rate’ ground of invalidity, which should perhaps be subject to discretion if the patent is valid on all other – apparently more substantive – grounds. But that cannot be right, in view of the centrality of the ‘best method’ requirement to the traditional notion of the ‘patent bargain’.
Maybe, however, I am reading too much into this. It is possible that the judge is simply so certain that an appeal will follow that he did not want to waste his time, or that of the parties, dealing with orders for relief and costs before the decision has been reviewed by a Full Court. An order that the patent be revoked would clearly need to be stayed pending a decision in the appeal (although there is not a great deal of work or administration involved in issuing, or staying, such an order).
If anybody has any further thoughts or theories, please share in the comments below.
The proceedings have been stood over for a further appearance on 14 February 2014. Any appeal will need to be filed in the meantime, so it is unlikely that this date will be kept!