Generics take Sandoz Rivaroxaban Loss to Heart – Bayer’s Rivaroxaban Patents Valid & Sandoz AU Launch Delayed by Injunction (Appeal Pending)

Sandoz AG v Bayer Intellectual Property GmbH [2023] FCA 1321

Subject to appeal, two of Bayer’s Australian Xarelto^® (rivaroxaban) patents were found valid (one also infringed), whilst the approach taken by Sandoz in the preparation of evidence and at trial was criticised by Justice Rofe.

Introduction

With 158 rivaroxaban products approved at the time the first instance decision was handed down, 149 of them generic products, the result of the Sandoz challenge to the validity of the Australian Bayer (Xarelto^®) patents may be the case with the broadest commercial interest this decade.

At the time of the decision, nine generic companies had multiple products approved (including Accord (32); Alembic (28), Teva (24), Arrotex (23), Cipla (12), DRL (12), Torrent (12), Alphapharm (3), Sandoz (3)), all of whom will be disappointed by the impact of this first instance decision of Rofe J.  Sandoz appealed to the Full Federal Court on 21 November 2023.  Bayer did not cross-appeal.

Subject to the Sandoz appeal, the validity of Bayer’s Australian process/formulation patent 2004305226 (‘226, expiring 13 Nov 2024) and method of treatment (MOT) patent 2006208613 (‘613, expiry 19 Jan 2026) has been upheld, despite the Sandoz multi-pronged invalidity challenge on each.  In brief:

• ’226 is directed to “orally administrable pharmaceutical compositions comprising rivaroxaban ‘in hydrophilized form’ and the use of those pharmaceutical compositions for the prophylaxis and/or treatment of a thromboembolic disease”

First instance finding (under appeal): valid, but Bayer conceded this was not infringed based on the Court’s construction of the phrase “in hydrophilized form”;

• ’613 is directed to “a method of treatment of thromboembolic disorders by once daily administration of a rapid-release tablet comprising rivaroxaban”;

First instance finding (under appeal): valid and infringed; Sandoz injuncted (pending resolution of the appeal).  Bayer’s amendment application to correct an incomplete PCT reference in ‘613 was allowed.

The decision is noteworthy because her Honour:

1. 1. offered a timely reminder to patent revokers to:

      a) avoid taking short-cuts in the preparation of expert evidence;

      b) think twice before opposing minor amendment applications; and

      c) beware adopting a strategy that involves a documentary ‘ambush’ at trial.

   2. provided insights on the success required under the modified “Cripps question” for obviousness.

Summary

Sandoz did not challenge the validity of Bayer’s rivaroxaban API patent (Australian Patent No. 775126), which expired on 24 November 2023 (with PTE).  Demonstrating her commercial awareness, her Honour accepted Sandoz’s requests for delivery of judgment just prior to API patent expiry to accommodate Sandoz’s intention to launch the Sandoz rivaroxaban products after API expiry.  Unfortunately, the outcome was not as Sandoz had hoped, with its commercial activities subject to an injunction pending resolution of its appeal.

Sandoz challenged the validity of patents in suit on the following grounds:

• ‘226 FORM (2024): novelty, inventive step, fair basis, clarity; and
• ‘613 MOT (2026): inventive step, clarity, insufficiency, best method.

Bayer defended validity and cross-claimed for infringement on both patents, seeking permanent injunctive relief.  It also sought to amend the ‘613 patent to correct an incorrect PCT reference.  Bayer conceded that its infringement case on the ‘226 patent depended on the Court’s construction of the phrase ‘in hydrophilized form’.

Bayer succeeded on all points other than construction with the result that (subject to appeal) the ‘613 is valid and infringed, and the ‘226 patent is valid but not infringed.

Key Lessons from Bayer v Sandoz

In round 1 of this dispute (which is under appeal), the Court delivered some hard lessons for Sandoz, namely:

1. Avoid taking short-cuts in the preparation of expert evidence

Sandoz submitted that all the claims of both pre-RTB patents¹ were obvious in light of common general knowledge (CGK) plus one, including WO 01/47919 (WO919).  To use WO919 in this manner, pursuant to s 7(3) of the pre-RTB Act, Sandoz must establish that the PSA could be reasonably expected to have ascertained, understood and regarded WO919 as relevant.

Justice Rofe was critical of Sandoz’s truncated approach to preparing expert evidence for its inventive step case.  The ‘gold standard’ for evidence to establish that a document satisfies s 7(3) is the search process set out in AstraZeneca AB v Apotex Pty Ltd (2014) 107 IPR 177 and endorsed by the High Court in AstraZeneca AB v Apotex Pty Ltd (2015) 257 CLR 356.  This typically involves the expert first engaging in routine and conventional literature (and perhaps patent) searches to ascertain relevant art, and then selecting and reviewing relevant papers for relevance and understanding.  Justice Rofe was critical of the short-cuts Sandoz took in the preparation of its evidence, finding that Sandoz did not establish that WO919 was part of the prior art base under s 7(3).  Her honour noted that:

• the complete results of the expert’s literature searches were not put into evidence;
• the expert was provided a copy of WO919 before reviewing redacted and limited search results; and
• the expert was given two hours to review a redacted, 71 page spreadsheet containing only patent literature search results, rather than the complete literature results of his hypothetical search (which would include journal articles).

Rofe J reiterated that the absence of evidence concerning hypothetical literature searches is not always fatal to a s 7(3) case, citing the recent decision in Hanwha Solutions Corporation v REC Pte Ltd [2023] FCA 1017 (Hanwha) which we have reported here.  Her Honour distinguished WO919 from the relevant information in Hanwha, finding that merely demonstrating that WO919 appeared amongst thousands of search results was insufficient to establish that a PSA would be reasonably expected to have ascertained, understood and regarded WO919 as relevant for the purposes of s 7(3).

Justice Rofe reminded the parties that “to properly test the proposition of ascertainment and relevance, the most prudent approach… would have been for Sandoz’s lawyers to test whether Professor Roberts would have actually identified WO919 as a document he would read from the results of his own search strategy”.  Her Honour found that Sandoz’s failure to do so could not be rectified.

Patent challengers ought to be aware of the potential pitfalls in taking short-cuts with expert evidence, as it may prove fatal to their invalidity arguments.

2. ‘Think twice’ before opposing minor typographical amendment applications

Sandoz challenged the ‘613 MOT patent alleging that Bayer sought to maintain an unfair advantage by hiding the best method in plain sight via an incomplete reference to a PCT which contained the best method.  Both Bayer and Sandoz’s evidence demonstrated that the correct PCT number would have been found in 15 minutes or less by the skilled team (inclusive of a patent searcher).

In response to this best method challenge, Bayer sought to amend the ’613 to correct the PCT number.  Her Honour concluded that the public were not sent off on a “wild goose chase” whilst Bayer kept the best method of performing the invention a secret.  In respect of Sandoz’s arguments of unreasonable delay by Bayer, Rofe J considered the practicality of amending a typo-style mistake in a global portfolio of patents, highlighting the quirk of the Australian best method invalidity ground at [784]:

“…the best method requirement is a peculiarly Australian ground of invalidity… I consider that it would impose an unrealistic burden to expect a patentee to circulate to its patent attorneys in every jurisdiction each and every issue that arises in the course of the prosecution of any application in a patent family around the world in case it raises a jurisdiction specific problem.”

Bayer’s delay in making the amendment application was not a culpable delay and her Honour exercised her discretion to allow the amendment.

Her Honour was critical of the Sandoz challenge (and parties opposing patent amendments more broadly, where the amendments sought are not made in bad faith, are not directed to claims, and do not help the patentee avoid prior art (see [797])) encouraging the hypothetical opponent to “think twice” before doing so.

3. Beware adopting a strategy that involves a documentary ‘ambush’ at trial

Late in the trial Sandoz sought to tender a document which it had been aware of since before the hearing, that allegedly demonstrated that the chemical name and structure of rivaroxaban was published before the priority date.  Bayer objected.  The parties agreed to make submissions, and Sandoz withdrew its tender.  Sandoz then sought to re-tender the same document during closing submissions, which was rejected by Rofe J.  Her Honour was critical of the Sandoz’s failure to raise the document earlier, referring to it as an ‘ambush’ at trial.  Justice Rofe sent a clear warning to parties that good case management and conduct is imperative.

The Court also:

4. Held there may be clarity in a phrase that is unknown to the person skilled in the art (PSA)

Sandoz’s clarity attack and Bayer’s counterclaim for infringement of the ‘226 turned on the construction of ‘in hydrophilized form’, a phrase not defined in the specification, and one that neither party’s experts had heard before.

Sandoz argued that ‘in hydrophilized form’ was unclear, and must refer to a composition that is produced using wet granulation, a process Sandoz does not use.  Bayer argued that meaning could be gleaned from prior art referred to in the ‘226 specification (but not incorporated by reference).

Justice Rofe highlighted that the phrase did “not have an ordinary meeting”, neither of the “very experienced formulation experts were familiar with the term”, and that “both experts agreed that a formulator would probably not have heard the term before reading the ‘226”.  Despite this, her Honour concluded that a PSA would have recourse to refer to the prior art referenced in the ‘226 patent, as both experts confirmed they would do so to understand the meaning of the phrase.

Rofe J concluded that ‘in hydrophilized form’ would be understood by a PSA in light of the ‘226 patent and the prior art to necessarily include the use of a liquid.  Therefore both Sandoz’s clarity attack and Bayer’s infringement claim of ‘226 were unsuccessful.

5. Addressed the level of success under the reformulated ‘Cripps question’

Justice Rofe set out the well-established principles for assessing obviousness, namely whether the PSA would (our emphasis) “directly be led, as a matter of course, to try the claimed invention in the expectation that it might well produce a useful alterative to, or a better drug than, the existing compound.”² But in her analysis Rofe J appears to have considered that the PSA would have to of had an expectation of success that the drug would (a) pass through all drug development stages to successful completion of phase III trials, and (b) be approved for use in human as a safe and effective.  It is unclear whether her Honour intended to lift the standard of the success required under the reformulated ’Cripps question’ to now include the successful completion of phase III trials and regulatory approval for use in humans as a safe and effective.  In our view, this would be the most interesting aspect of the appeal.

Conclusion

Subject to the pending appeal, the take home message to future patent revokers are clear:

• avoid taking short-cuts in the preparation of expert evidence;
• beware the documentary ‘ambush’ at trial;
• ‘think twice’ before opposing minor amendment applications; and
• consider the success relevant to the modified ‘Cripps question’, as the bar may have been raised.

___
¹ Pre-Raising the Bar (i.e. prior to the 2012 changes introduced by the Intellectual Property Laws Amendment (Raising the Bar) Act 2012 (Cth)
² See [294], referencing Aktiebolaget Hassle v Alphapharm Pty Ltd (2002) 212 CLR 411, and Graham J in Olin Mathieson Chemical Corporation v Biotex Laboratories Ltd [1970] RPC 157 at 187 – 188

About Pearce IP

Pearce IP is a boutique firm offering intellectual property specialist lawyers, patent attorneys and trade mark attorneys to the life sciences industries (in particular, pharmaceutical, biopharmaceutical, biotech, ag-tech and food tech).  Pearce IP is the 2021 ‘Intellectual Property Team of the Year’ (Lawyers Weekly Australian Law Awards) and was shortlisted for the same award in 2022.  Pearce IP is ranked in IAM Patent 1000 and Managing IP (MIP) IP Stars, in Australasian Lawyer 5 Star Awards as a ‘5 Star’ firm, and the Legal 500 APAC Guide for Intellectual Property.

Our leaders have been recognised in virtually every notable IP listing for their legal, patent and trade mark excellence including: IAM Patent 1000, IAM Strategy 300, MIP IP Stars, Doyles Guide, WIPR Leaders, 5 Star IP Lawyers, Best Lawyers, and Australasian Lawyer 5 Star Awards, and have been honoured with many awards including Australian Law Awards – IP Partner of the Year, Women in Law Awards – Partner of the Year, Women in Business Law Awards – Patent Lawyer of the Year (Asia Pacific), Most Influential Lawyers (Changemaker), among other awards.