In the much anticipated AstraZeneca AB v Apotex Pty Ltd1, the High Court of Australia unanimously upheld a decision by an enlarged Full Federal Court (that is, 5 judges rather than the usual 3), that an AstraZeneca patent relating to the blockbuster statin drug CRESTOR (rosuvastatin) is invalid. The AstraZeneca patent is directed to methods of using low doses of rosuvastatin to treat high cholesterol (the Low Dose Patent) which the High Court found lacking in inventiveness (that is, "obvious") when compared with the "prior art base".

The High Court's decision2 clarifies how expert witnesses may undertake searches to locate "prior art information" which may be taken into account when addressing the issue of inventive step and the proper treatment of this "prior art information" in the inventive step analysis.

The High Court's decision also leaves untouched the Full Court's decision on "starting point" – that is, information against which the question of inventiveness is to be determined (common general knowledge (CGK) and any relevant prior art) cannot be enlarged by reference to the description in the body of the patent specification.


  • An expert witness tasked with identifying relevant "prior art information" for the purpose of assessing inventive step is permitted to refer to and consider other pieces of "prior art information" (as well as CGK) in their search process, in order to determine which pieces of "prior art information" they consider relevant.
  • For patents filed prior to April 2002 (as the Low Dose Patent was), each piece of "prior art information" ultimately identified by an expert must be considered one at a time together with the CGK in determining whether a claimed invention contains an inventive step. If the invention is obvious in light of the CGK plus any one of the ascertained pieces of "prior art information" then the patent is invalid for lack of inventive step.
  • In industries (such as the pharmaceutical industry) where animal and human clinical studies are routinely performed, absence of animal or human safety data in a piece of "prior art information" does not mean that it cannot be used to challenge the inventiveness of a claimed treatment regimen.


At first instance, Justice Jagot found that the claims of the Low Dose Patent were invalid on three grounds. Firstly, AstraZeneca was not entitled to the Low Dose Patent. Secondly, the invention disclosed in the Low Dose Patent lacked novelty in light of two pieces of prior art – the Watanabe Article and the 471 Patent, both of which disclosed the compound rosuvastatin. Thirdly, the invention disclosed in the Low Dose Patent did not involve an inventive step.

Justice Jagot's decision was appealed to an enlarged bench of the Full Federal Court, constituted primarily to determine the key issue of the correct "starting point" for assessing inventive step.3 Although it rejected Justice Jagot's finding on the "starting point" question and held that the problem addressed in a patent specification is not necessarily part of the CGK for assessing obviousness, the Full Federal Court upheld the conclusion that the invention claimed in the Low Dose Patent was obvious, in light of the CGK plus either of the Watanabe Article and the 471 Patent. The reasoning is explained below. Justice Jagot's finding that AstraZeneca was not entitled to the Low Dose Patent was also upheld, but the other findings were overturned.

While another AstraZeneca patent (which related to particular compositions of rosuvastatin) was revoked in the first instance and Full Federal Court decisions, it was not pressed in the High Court.


On further appeal to the High Court, AstraZeneca sought to reverse the Full Federal Court's findings of obviousness and lack of entitlement. As the High Court ultimately upheld the Full Federal Court's finding of obviousness it did not consider the entitlement issue.

Argument in the High Court appeal centred on the construction of the words of section 7(2) and 7(3) of the Patents Act 1990 (Cth) (the Act), with the High Court ultimately holding that the positions put forward by AstraZeneca were "constructions of ss 7(2) and 7(3) which their text will not bear".4


In order to be patentable, under section 18(1)(b)(ii) of the Act, a claimed invention must involve an inventive step when compared with the "prior art base" as it existed before the priority date of the claim.

The test for determining inventive step is set out in section 7(2) of the Act (emphasis added):

" invention is to be taken to involve an inventive step when compared with the prior art base unless the invention would have been obvious to a person skilled in the relevant art in the light of the common general knowledge as it existed in the patent area before the priority date of the relevant claim, whether that knowledge is considered separately or together with either of the kinds of information mentioned in subsection (3), each of which must be considered separately."

Section 7(3)(a) further provided:

"For the purposes of subsection (2), the kinds of information are:

(a) prior art information made publicly available in a single document ...

being information that the skilled person mentioned in subsection (2) could, before the priority date of the relevant claim, be reasonably expected to have ascertained, understood and regarded as relevant to work in the relevant art in the patent area."

The parties' arguments related to whether "prior art information" must be "considered separately" at both stages of the inquiry, that is, when searching for relevant "prior art information" and when assessing inventive step.


  1. Undertaking searches - the threshold "relevance requirement"

First, AstraZeneca argued that the phrase "considered separately" in section 7(2) required "prior art information" in a non-CGK source to be reviewed in isolation from other non-CGK sources in undertaking searches, that is, to determine whether the source of information satisfies the threshold "regarded as relevant" requirement in section 7(3). On AstraZeneca's view, the approach taken by the expert witnesses in this case5 - to perform a literature search and then compare the documents which resulted from the search to determine which they considered most relevant - was not allowed.

The High Court rejected AstraZeneca's argument and affirmed the approach taken by Justice Jessup in the Full Federal Court. Specifically, the High Court held that while the phrase "considered separately" prevents the use of a combination of unrelated documents ("mosaicking") to defeat the inventiveness of a claimed invention under section 7(2), it did not restrict how the threshold "regarded as relevant" requirement in section 7(3) is satisfied.

  1. Using prior art information - assessing inventive step

Second, AstraZeneca submitted that the Full Federal Court had applied the test for inventive step in section 7(2) incorrectly by considering each piece of "prior art information" in isolation. AstraZeneca argued that this approach suggested that the "only course available to the skilled person [was] that identified in the s 7(3) document [a]dded to the prior art base" and allowed the expert to ignore potential "false" avenues to solve the problem suggested by other relevant documents; an approach fundamentally affected by hindsight.6 AstraZeneca argued that in reality, the skilled person would have been confronted with a number of choices of compounds and therefore the choice of rosuvastatin was not obvious.

In support of its argument, AstraZeneca referred to an additional document (the Aoki Article) located by the expert witnesses, which identified the compound, NK-104, as a promising cholesterol lowering agent. AstraZeneca argued that once NK-104 was identified, the skilled person would be confronted with a choice as to whether NK-104 or rosuvastatin should be selected and that there was an evidentiary gap in this regard.

Chief Justice French rejected AstraZeneca's argument on the basis that the words of section 7(2) simply did not support such a construction; once identified as relevant in accordance with section 7(3), each piece of "prior art information" must be considered on its own, with the CGK. In any case, Justice Kiefel noted that there was evidence that a skilled addressee would have selected rosuvastatin over NK-104.


In determining whether the invention claimed in the Low Dose Patent was obvious or not, the High Court adopted the 'reformulated Cripps question' considered by Justice Jessup, namely7:

"whether the notional skilled person would have directly been led as a matter of course to try the invention as claimed in the expectation that it might well produce a better method of treating a patient suffering from hypercholesterolemia than existing statins and doses."

The High Court agreed with Justice Jessup's view that the effect of the expert evidence was that a person skilled in the art, having read the Watanabe Article in light of the CGK would have entertained the expectation that rosuvastatin might well be at least as efficacious a treatment as atorvastatin (the leading treatment at the time). It would then have been a routine step to test rosuvastatin at the lowest efficacious dose as part of conventional clinical trial procedure.8 Accordingly, the absence of animal or human trial safety data (which would be useful in determining doses that should be tested in clinical trials) in the Watanabe Article and 471 Patent did not prevent a conclusion that the invention claimed in the Low Dose Patent lacked inventiveness.


The High Court has clarified the approach that may be taken by the hypothetical skilled addressee to identify relevant "prior art information" for the purpose of assessing inventive step.

In the words of Justices Gageler and Keane9:

"Nothing in s 7(3) suggests that the notional skilled addressee may not trawl through multiple documents, discarding the irrelevant and retaining the useful, as part of the ascertainment of relevant information."

In addition, the decision highlights that when considering the question as to whether an invention is obvious in light of the CGK and "prior art information", the skilled person must consider each piece of prior art separately. This is the case even where there may be a number of pieces of prior art, some of which propose different solutions to the problem.

It is important to note that the priority date of the Low Dose Patent was 6 February 1999. Since then amendments have been made to section 7(3) of the Act which permit certain combinations of prior art documents when determining inventiveness. Similarly, the amendments introduced by the Raising the Bar Act10 mean that it is no longer necessary to establish that prior art would have been "ascertained, understood and regarded as relevant" by the skilled person.


1 AstraZeneca AB v Apotex Pty Ltd; AstraZeneca AB v Watson Pharma Pty Ltd; AstraZeneca AB v Ascent Pharma Pty Ltd [2015] HCA 30 (the High Court Appeal).

2 Each High Court judge (with the exception of Justices Gageler and Keane) delivered separate judgments.

3 Our earlier IP Preview Article discussed the Full Federal Court decision in detail.

4 High Court Appeal at [8].

5 The expert evidence relevant to inventive step was that of Professor O'Brien and Dr Reece, both of whom were experts for the generic parties.

6 High Court Appeal at [24].

7 High Court Appeal at [40].

8 See also High Court Appeal at [80].

9 High Court Appeal at [111].

10 Intellectual Property Laws Amendment (Raising the Bar) Act 2012 (Cth)

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.

Most awarded firm and Australian deal of the year
Australasian Legal Business Awards
Employer of Choice for Women
Equal Opportunity for Women
in the Workplace (EOWA)