Three strikes, you’re out: Rosuvastatin generics in the clear

The High Court of Australia handed down its decision yesterday in the appeal by AstraZeneca (AstraZeneca AB v Apotex Pty Ltd (2015) HCA 30), against last year’s ruling by the Full Court of the Federal Court,  affirming the Full Court’s decision that AstraZeneca’s low dose rosuvastatin patent was obvious.


The High Court has unanimously sided with the Full Court, finding that there was no error in its reasoning in applying the test for inventive step.  In particular, the Full Court’s construction of section 7(3) of the Patents Act 1990 (Cth) was affirmed.

The High Court side-stepped any decision on the ‘starting point’ approach, on the basis that it was not necessary to consider the issue in light of its finding on the construction of s7(3). This approach was previously adopted by an earlier Full Court in Apotex Pty Ltd v Sanofi-Aventis [2009] FCAFC 134 under the Patents Act 1952. It allowed information contained in the relevant patent specification, but which did not form part of the common general knowledge, to be taken into account in an obviousness analysis. In Rosuvastatin, the Full Court found that the ‘starting point’ approach is not applicable under the current Patents Act. As the High Court did not consider this issue, that decision now stands.


AstraZeneca appealed to the High Court after being unsuccessful both at first instance (Apotex v AstraZeneca (No 4)) and before the Full Federal Court (AstraZeneca AB v Apotex Pty Ltd [2014] 312 ALR 1). Although three patents were considered at first instance, and the appeal before the Full Court concerned two patents, the only patent in suit in the High Court appeal was AU 200023051 (“051 Patent”).  The ‘051 Patent claimed methods of treating a patient suffering from hypercholesterolemia comprising administration of 5 or 10 mg of the pharmaceutical compound, rosuvastatin (sold under the brand name CRESTOR) as a starting dose for the treatment of hypercholesterolemia.

The primary judge found the ‘051 Patent invalid on three grounds: lack of entitlement; lack of novelty in light of two prior art publications; and obviousness within the meaning of s 7(2) of the Act. The Full Court of the Federal Court of Australia overturned only the finding of lack of novelty and otherwise dismissed the appeals from the primary judge’s decision.

Issues of interpretation of the Patents Act

The primary issue considered by the High Court was the meaning of ss 7(2) and 7(3) of the Patents Act. Subsection 7(2) requires inventive step to be assessed by reference to common general knowledge, considered separately or together with prior art information falling within s 7(3).  The “threshold requirement” under s 7(3) is that the prior art information could, before the priority date of the relevant claim, be reasonably expected to have been ascertained, understood and regarded as relevant to work in the relevant art in Australia.

Importantly, under the Patents Act as it stood at the time relevant to the ‘051 Patent, s 7(3) only permitted combining documents which would be treated as a single source of information. Following amendments in 2001 this requirement was relaxed to allow combination of documents for obviousness purposes where the skilled person could be reasonably expected to do so. Mosaicing of prior art information which is not common general knowledge is not permitted unless this requirement is met.

The expert evidence in the case identified literature and patent searches which the experts carried out when posed with the problem of finding a more effective alternative statin. These searches led to the identification of the prior art information invoked against the ‘051 Patent, in particular the Watanabe article.  This article disclosed the compound rosuvastatin and the fact that it was in clinical trials. However the experts also identified another document, the Aoki article, which disclosed another promising compound.

AstraZeneca asserted firstly that in identifying the Watanabe article as relevant for the purposes of s7(3), the experts had had regard to other prior art publications disclosed by their searches, by way of a process of comparison.  AstraZeneca argued that a process of satisfying the ‘regarded as relevant’ requirement of s 7(3) which involves comparing various non-“cgk” publications, amounted to the impermissible combining of such documents for the purpose of s7(3).

Secondly, it argued that the Full Court had taken a “single avenue approach”. By this it argued that, having identified the Watanabe article as ascertained and relevant, the Full Court had decided the question of inventive step on the basis that the only course available to the skilled person was the course identified in that document i.e. the use of rosuvastatin.  In light of the existence of the Aoki article, it argued that it had not been shown that the skilled person would have chosen rosuvastatin rather than the other promising new compound identified in Aoki.

 The High Court Judgment

 The High Court had no difficulty in finding that there was no impermissible combination of documents in order to find the Watanabe article relevant.  The words “considered separately” in s7(2) qualify the way in which prior art information complying with s 7(3) must be used, but they say nothing about how the relevance requirement of s7(3) is to be satisfied. Accordingly, the finding of Jessup J in the Full Court was correct: it was wholly within the scheme of the subsection that the skilled person might well sort through all manner of information with a view to finding something that is regarded as relevant.

On the second argument, AstraZeneca seemed destined to fail in light of the evidence of at least one of the expert witnesses that he would have selected rosuvastatin as the most promising candidate, on the basis that the Watanabe article suggested that it had high potency with reduced side-effects and it had progressed to clinical trials.

However this begs the question of what the position might be in a situation where a document falling within s 7(3), considered alone, would directly lead the skilled person to the invention, but where there might be other documents (also falling within s 7(3)) which might lead the skilled person in a different direction.

French CJ considered that it sufficed to say that the text of s7(2) “simply did not offer a constructional choice imposing the limitation for which AstraZeneca contends”, that is, the limitation that the Court was not permitted to decide the question of obviousness on the basis that the only course available to the skilled person was that identified in the relevant s7(3) document.

Gageler and Keane JJ considered that (at paragraph [115]):

“Section 7(2) does not contemplate that a choice between apparently effective solutions must be attributed to the notional skilled addressee, much less that the notional skilled addressee might be so befuddled by an embarrassment of choices as to cease pursuit of the solution.”

 While understandable as a construction of ss 7(2) and (3) (as they stood at the relevant time for this case), it does appear that this construction could lead to an obviousness analysis, divorced from the actual reality that in order to arrive at an invention the skilled person may be required to choose between a number of promising options. On the High Court’s reasoning it appears that a s 7(3) document could form the basis for an obviousness finding, even if the expert gave evidence that they would, in practice, have preferred a different solution contained in another s7(3) document.

Interestingly Kiefel J acknowledged that it was necessary for the expert to select the best candidate because the clinical trials which would follow are extremely expensive. However since this comparison was not done with a view to determining the relevance of each document per se, it could not affect the relevance of any given document under s7(3).

A possible alternative construction of s7(3) could perhaps arise under the Patents Act as it now stands (both pre- and post- Raising the Bar). As mentioned above, s7(3) now permits combination of two or more pieces of s7(3) information, where the skilled person could be reasonably expected to do so.  A more restrictive version of s7(3) was in place at the relevant time of this case. Arguably, when confronted with two documents pointing to two different, and both potentially useful, compounds, it might be said that the skilled person could reasonably be expected to combine the information in both documents to reach a decision as to which path to take.

It is important to note that post – the Raising the Bar amendments, the requirement that relevant prior art information could reasonably have been ascertained, understood and regarded as relevant has been removed from s 7(3). As a result, any publically available information can be taken into account under s 7(3). Applying the High Court’s reasoning, this would appear to open the door even further for argument that if any one document considered on its own (even if not likely to be ascertained by the skilled addressee) would directly lead the skilled addressee to the invention, the invention will be obvious. That conclusion could be reached regardless of whether the skilled person would likely, in reality, choose a different path.