Ascertained and obvious – Full Court declots Australia’s law on routine steps

In this regard, the Full Court has confirmed that inherent risks in conventional drug development work are not sufficient to provide an inventive step, rejecting the approach of the primary judge who found two Bayer patents inventive because (she said) the person skilled in the art would not have the relevant expectation of success at the pre-clinical phase of the drug development process, particularly in view of evidence that at least 90% of new chemical entity development never progresses the entire way along the drug development path.   

The Full Court found that her Honour erred in finding that the inventions claimed in both Bayer patents involved an inventive step, in addition to erring in finding that the skilled person could not be reasonably expected to have ascertained the key prior art document. 

Key takeaways

• An expectation of success is inherent in a person skilled in the art choosing to go down the conventional development route.
• Even if inherently risky, such risk will not always be sufficient to render something inventive.
• The decision provides greater scope for inventive step challenges to the validity of similar granted patents. 

Background

In 2021, Sandoz brought proceedings seeking to clear the way for its generic rivaroxaban product following the expiry of the compound patent.  

Sandoz contended that both Australian Patent No. 2004305226 (226 Patent) and Australian Patent No. 2006208613 (613 Patent) were invalid for lack of inventive step on the basis of the common general knowledge (as it existed at the relevant time) together with either International Patent Publication No. WO 01/47919 (WO 919), or the journal Blood.

The 226 Patent asserts a “special treatment” of rivaroxaban to improve its bioavailability – namely, the preparation of solid oral compositions comprising rivaroxaban in hydrophilized form. 
The 613 Patent asserts an inventive step in the administration of rivaroxaban orally once daily for at least five consecutive days, where the rivaroxaban has a plasma concentration half-life of 10 hours or less.  The administration of the tablet once a day is described as inventive because the half-life of rivaroxaban suggests more frequent dosing. 

Obviousness and WO 919

At first instance

In Australia, the test for determining whether an invention involves an inventive step requires a consideration of “whether the person skilled in the art at the relevant date with the relevant knowledge would directly be led, as a matter of course, to try the claimed invention in the expectation that it might well produce a useful alternative to, or a better drug than, the existing compound”. 

At first instance, Rofe J made extensive findings about the drug development process, and the likelihood of success in the drug development process both generally, and by reference to what was taught by WO 919. 

The findings of her Honour at first instance included that:

(a) it was not uncommon for the drug development process to fail; 

(b) the majority of lead drug candidates that make it to pre-clinical studies never advance to clinical trials, and most drugs that enter clinical trials do not get regulatory approval – with only a 10% chance of this occurring;

(c) the risk of failure for New Chemical Entities (NCE) is particularly acute because the development team is starting with a “blank slate”, so there is no way of anticipating what will be encountered in the development process;

(d) the substantial risks and uncertainties associated with drug development are particularly problematic in the high risk field of thrombosis involving a novel, first in class drug. 

Against this background, she found that WO 919 would be of interest to the skilled person, and that reading WO 919 that they would likely select rivaroxaban as a lead candidate to take into development work, including the full suite of pre-clinical tests.  

However, her Honour’s view was that the identification of rivaroxaban as a lead, or preferred candidate was a very early step along the drug development pathway, and that each further step presents an opportunity for “failure or change in direction” – such risk being compounded “when the compound is a first in class NCE, the side effects are entirely unknown, [and] other unrelated compounds … are known to have failed in early phase studies”. 

“The steps foreshadowed by [the experts] form part of a new, arduous, comprehensive, risky and unpredictable research project for a first class NCE”, she said, and so “the pre-clinical tests are not undertaken with the requisite expectation of developing a safe and efficacious drug”. 

Rather, such steps, she said, were a mere information gathering exercise along the path to development, and were “more akin to the voyage of discovery”.  

Accordingly, she found that the inventions claimed were not obvious. 

Full Court

On appeal, the Full Court rejected this approach. 

There was no evidence before the Court that any “failure or change in direction” would have arisen in this case in the course of routine development work that would have conventionally been carried out following the identification of rivaroxaban as a lead candidate to take forward – or that the dosage regimen, or the formulation covered by the two Bayer patents would not have been identified. 

The evidence of the experts before the Court was clear that they would conduct pre-clinical and clinical tests in taking rivaroxaban forward, and that this was routine and conventional work along a well-established drug development pathway. 

In the circumstances, the Full Court said that Rofe J placed too much emphasis on the risks and unknowns associated with tests that would have been undertaken as a matter of course, and that this was an error in circumstances where her Honour had found that a person reading WO 919 would likely select rivaroxaban as a candidate, if not the candidate, for development. 

An expectation of success is implicit in such a finding, as “one would not expect the person skilled in the art to even commence pre-clinical trials if such an expectation was absent”, they said. 

In this regard, the person skilled in the art does not need to have an expectation that the chosen compound would pass through all stages of development. 

“The relevant expectation is to be measured against the ordinary level of expectation and risk inherent in routine work in the particular field… the relevant test is merely expecting that the steps may well work”. 

Accordingly, the Full Court found that the invention claimed in both the 613 Patent and the 226 Patent lacked an inventive step because:

• the skilled team would have been lead from the disclosure in WO 919 to take rivaroxaban into drug development work;
• there was no evidence of any issues or failures that would occur in that process that would lead to a change in direction; and
• where there was no evidence that the formulation or dosage regiment involved an inventive step, these matters would have been identified during the course of conventional trials. 

Ascertained

In rejecting Sandoz’s invalidity challenge at first instance, Rofe J also made a finding that the skilled person could not be reasonably expected to have ascertained WO 919 for the purpose of section 7(3) of the Patents Act 1990 (Cth) (Act). 

Under the version of the Act at the time, only information that the skilled person could, before the priority date of the patent, be “reasonably expected to have ascertained, understood and regarded as relevant” can be used in the assessment of whether an invention is obvious. 

For patents to which this test applies, the pool of information that can be used to assess whether an invention involves an inventive step becomes restricted.  It can also impose an additional evidentiary burden on a party seeking to make an obviousness challenge because it may be necessary to prepare evidence that retrospectively proves that certain information could have been found by the skilled person at an earlier date. 

This standard was amended as part of the Raising the Bar amendments introduced in 2013 in order to align the approach to obviousness under Australian law with other key jurisdictions, including Europe.  However, the test remains relevant to patent applications examined prior to the introduction of these amendments. 

In this case, in the course of preparing his evidence, Sandoz’s expert witness was asked to assume that he was tasked with developing a formulation for an immediate release solid oral dosage form that contained a factor Xa inhibitor. The expert then described the searches that he would have undertaken in order to develop this formulation. He was subsequently asked to review search results provided to him and identified WO 919. 

However, the primary judge found that the search results that were provided to the expert by Sandoz’s lawyers were narrower than the results of the searches that would have been obtained, if the searches had been conducted in accordance with the expert’s usual practices. In particular, the focus on factor Xa inhibitors (to the exclusion of other inhibitors) resulted in fewer results. 

On appeal the Full Court accepted that the standard under section 7(3) of the Act does not require proof that the prior art information would have been ascertained, rather it is only necessary to demonstrate a reasonable expectation that it would have been ascertained on the balance of probabilities.  

In this case, the evidence was sufficient to establish that the skilled person would have conducted searches of patent databases using search terms that included factor Xa inhibitors and subsequently identified WO 919. It was immaterial that additional searches could have been performed and that additional documents may have been identified as relevant. 

The Full Court confirmed that a party does not need to adduce evidence that the skilled person preferred or prioritised one piece of prior art over all other information which they could have found out.