In a widely anticipated decision, the High Court has upheld a decision of the Full Federal Court (which was made by five, rather than the usual three, Federal Court judges) that AstraZeneca’s patent relating to low dosages of rosuvastatin is invalid for want of inventive step.
Rosuvastatin is used in the treatment of high cholesterol and is marketed by AstraZeneca as Crestor. The decision will be a relief for companies that have already launched generic versions of rosuvastatin.
While AstraZeneca appealed the Full Court’s findings on inventive step, novelty and entitlement, it was only necessary for the High Court to consider inventive step.
The High Court’s decision is important because:
- it does not disturb the Full Court’s decision below that when assessing inventive step, the invention as claimed must be compared against the common general knowledge (CGK) and any prior art information, irrespective of the description of the inventive step in the specification; and
- it clarifies that an invention will be obvious in light of one prior art document irrespective of the fact that other prior art documents which have also been shown to be relevant may suggest an alternative course.
The starting point for assessing inventive step
The High Court did not disturb the Full Court’s finding below that when assessing inventive step, the comparison must focus on the claimed invention, which must be compared against the CGK and any prior art information. The Full Court’s decision effectively overturned an earlier Full Court decision in Apotex Pty Ltd v Sanofi-Aventis in which the Court held that when assessing inventiveness, it is necessary to determine the nature of the claimed invention and the inventive step as described in the specification and then assess that inventiveness by reference to the CGK and any prior art information. According to the Full Court in this proceeding at least, information against which the question of inventiveness is to be determined cannot be enlarged by reference to the description in the body of the specification.
At first instance, the primary judge relied on the Apotex v Sanofi decision to conclude that the description of the invention was such that rosuvastatin could be taken into account as part of the “starting point” when considering whether the claimed invention was obvious. On appeal, the Full Court overturned this aspect of the primary judge’s decision and found that whether the patentee knew of the existence of rosuvastatin was not relevant to considering the question of inventive step – the question must always be whether the claimed invention is obvious in light of the CGK and any prior art information.
This important aspect of the Full Court’s decision did not upset the finding of obviousness as rosuvastatin was nonetheless disclosed in the prior art and the invention was found to be obvious in light of this prior art. However, the Full Court’s decision does represent a shift in approach to obviousness which the High Court has not disturbed.
The existence of more than one prior art document
In the High Court, AstraZeneca argued that an invention should not be found to lack an inventive step where one prior art document is selected from a group of documents which have all been identified as potentially relevant. That is, to determine the question of obviousness on the basis that the only course available to the person skilled in the art is the one identified in one of multiple prior art documents ignores false routes suggested by other prior art documents which also meet the relevance requirement. Such an approach, it was argued, gives the skilled person the benefit of hindsight.
The High Court rejected this. The Court confirmed that when assessing inventive step, it is permissible to consider different prior art documents separately (and each in light of the CGK), provided it is established that the skilled person could be reasonably expected to have ascertained, understood and regarded each as relevant. If the invention as claimed lacks inventive step having regard to either, the invention will be invalid for obviousness.
The High Court also rejected AstraZeneca’s argument that in ascertaining a relevant single prior art document, sections 7(2) and 7(3) of the Patents Act preclude the use of other information that is not part of the CGK. Astra Zeneca argued that when establishing that a publication “could be reasonably expected to have ascertained, understood and regarded as relevant”, it is impermissible to look to publications which are not part of the CGK.
The High Court upheld the Full Court’s finding that the skilled person may well sort through all manner of information in order to find something that is regarded as relevant. There is nothing in the Act which places an embargo on using combinations of sources of information along the road to that destination. While section 7(2) (at the relevant time) precludes the combining of information in separate publications for the purposes of assessing obviousness, it does not preclude a combination of documents being considered for the purposes of assessing each document’s relevance.
It is important to note that the priority date of the patent in suit was 6 February 1999. Since that time the Patents Act has been amended such that combining information from multiple prior art documents is permissible. Similarly, following the IP Laws Amendment (Raising the Bar) Act 2012, it is no longer necessary to establish that prior art information is information which the skilled person could be reasonably expected to have ascertained, understood and regarded as relevant.
The test for inventive step – the Cripps question
The High Court agreed with the Full Court below that the claimed invention lacked inventive step in light of the CGK considered with either of two prior art documents. The Court applied the Cripps question from Olin Mathieson Chemical Corporation v Biorex Laboratories Ltd and found that the skilled person would have been led directly as a matter of course to try rosuvastatin in the expectation that it might well be an efficacious treatment. Further, it would be a routine step to test rosuvastatin in the claimed dosage range in clinical trials which, while expensive, could be conventionally carried out.