What is the Starting Point for Inventive Step?

AstraZeneca AB v Apotex Pty Ltd [2014] FCAFC 99 (12 August 2014)

An expanded panel of five judges of the Federal Court of Australia has held that two patents owned by AstraZeneca relating to the cholesterol-lowering drug marketed as CRESTOR (active ingredient rosuvastatin) are invalid.  Subject to the possibility of a High Court appeal, the decision opens the way for generic competition to CRESTOR in Australia, which has been prevented by interlocutory injunctions since 2011.

A particular point of interest in this case lies in the reason for the rare expansion of the Full Court panel from the usual three judges to five.  The judge at first instance (in Apotex Pty Ltd v AstraZeneca AB (No 4) [2013] FCA 162) had found that the AstraZeneca inventions claimed in the patents at issue lacked an inventive step based, in part, on the reasoning of an earlier Full Court panel in Apotex Pty Ltd v Sanofi-Aventis [2009] FCAFC 134.  AstraZeneca contended that the approach taken by the primary judge was wrong, and requested a panel of five judges to enable the earlier Full Court decision to be overruled, if necessary.

So, five judges is better than three, and this decision is inherently important because anything it stands for has precedence over any ‘normal’ judgment issued by a three-judge Full Court panel.  In the event, the appeal court did overrule the reasoning of the primary judge on the approach to the inventive step question at issue.  However, it also found a way to avoid actually overruling the earlier Full Court decision in Sanofi-Aventis.

The key question addressed by the court in this regard is: what is the appropriate ‘starting point’ for assessing whether or not a claimed invention involves an inventive step or, conversely, is obvious? 

The answer to this question is that inventiveness must be assessed by reference to the relevant prior art information, in light of the ‘common general knowledge’ of the person skilled in the art.  The assessment should not be influenced by the description of the invention provided in the patent specification, or by consideration of any problem that the invention is explicitly or implicitly directed at solving, except to the extent that these are, in fact, part of the prior art or common knowledge in the field.

Different ‘Starting Points’?

Taking AstraZeneca’s Australian patent no. 2000023051 (also called ‘the 051 patent’ or ‘the low-dose patent’ in the judgment) as an example, the claims are directed to a dosage regime for treating high cholesterol (i.e. ‘hypercholesterolemia’) with rosuvastatin.  Since it is commonly (though not always) possible to regard an invention as providing a solution to a problem, it makes some sense to ask: what was the problem that was solved by the inventor in this case?

The difficulty that arises with this question is that it may be possible to define the problem is several different ways, each of which may suggest a different starting point for making the claimed invention.  As the court noted in this case (at [208]), a fair reading of the patent specification might lead to a definition of the problem in any one of three different forms:

1. Is it possible to devise a new method for the treatment of hypercholesterolemia?
2. Is it possible to devise a new method for the treatment of hypercholesterolemia using a statin?
3. Is it possible to devise a new method for the treatment of hypercholesterolemia using rosuvastatin?

The primary judge had essentially addressed the question in the third form, and on the basis that the ‘starting point’ for devising the claimed dosage regime was the use of rosuvastatin, because that is what she understood to have been the ‘problem’ addressed by the inventor from a reading of the patent specification.  Thus developing the dosage regime was merely a matter of following standard clinical procedures, which could not involve any inventive step.

A similar situation had arisen in the earlier Sanofi-Aventis case, in which the patent specification claimed a particular structural form of a compound which existed as a ‘racemic mixture’ of two structural forms, known at the time as ‘PCR 4099’.  Although PCR 4099 was not part of the common general knowledge in the field at the relevant date, the court nonetheless found, on the basis of the way the invention was described in the specification, that the racemic mixture was the starting point for assessing inventive step.  Thus, any inventive step would have to reside in separating out the claimed structural form and determining that it was the one with the desired activity.  Since this was obviously desirable, and a matter of routine, it could not involve any inventive step.

The problem with this approach is that it makes inventiveness dependent upon how the invention is described in a patent specification, rather than solely by reference to the objective state of the art at the time the invention was made.  The ‘journey’ taken by a particular inventor, to arrive at an invention, may bear little relationship to the position of a non-inventive skilled worker in the art at the relevant time.

Assessing Inventiveness Objectively

Any test for inventive step, or obviousness, should be, as far as possible, objective and based upon a factual inquiry as to the state of the art.

The five judges of the Full Court agreed.  The Australian Patents Act 1990, in section 7, requires that inventive step be assessed by reference to the common general knowledge of a person skilled in the art, with regard to prior art information as set out in subsections 7(2) and (3).  As the court stated (Besanko, Foster, Nicholas and Yates JJ, at [202-3]):

…whether a claim of a patent is invalid for lack of inventive step is to be determined by comparing the invention, so far as claimed, against the common general knowledge and any s 7(3) information. The question is then whether the invention would have been obvious to the hypothetical person skilled in the art in light of that knowledge considered separately from, or together with, the s 7(3) information. So understood, it is apparent that the relevant provisions of the Act do not expressly or impliedly contemplate that the body of knowledge and information against which the question whether or not an invention, so far as claimed, involves an inventive step is to be determined may be enlarged by reference to the inventor’s (or patent applicant’s) description in the complete specification of the invention including, in particular, any problem that the invention is explicitly or implicitly directed at solving.

If the problem addressed by a patent specification is itself common general knowledge, or if knowledge of the problem is s 7(3) information, then such knowledge or information will be attributed to the hypothetical person skilled in the art for the purpose of assessing obviousness. But if the problem cannot be attributed to the hypothetical person skilled in the art in either of these ways then it is not permissible to attribute a knowledge of the problem on the basis of the inventor’s “starting point” such as might be gleaned from a reading of the complete specification as a whole.

The claim, in the case of the 051 patent, is directed to the use of rosuvastatin, according to a particular dosage regime, for the treatment of hypercholesterolemia.  The specification makes it clear that no claim is made to the invention of rosuvastatin itself, however, as Jessup J put the point, at [502]:

The question implicitly posed by the Patents Act is not “given rosuvastatin, would the administration of it at the claimed doses be obvious?” Rather, the question is “given a patient suffering from hypercholesterolemia, would the treatment of him or her with rosuvastatin at the claimed doses be obvious?”

In the case of the 051 patent, the rejection of the approach taken by the primary judge did not actually result in any better outcome for AstraZeneca.  In this case, the court found that rosuvastatin was known in the prior art and that, following the approach required by section 7 of the Patents Act, it would be obvious to treat a patient suffering from hypercholesterolemia with rosuvastatin at the claimed doses.

The Overruling Issue: Five Trumps Three

As for overruling the Full Court judgment in Sanofi-Aventis, the expanded panel dodged that issue on the basis that the earlier case had been decided under the provisions of the Patents Act 1952.  The earlier legislation had no equivalent of section 7 of the 1990 Act, instead requiring that inventive step be assessed by reference to what was ‘known or used’ – a phrase interpreted to encompass only the common general knowledge in the relevant art (see Minnesota Mining & Manufacturing Co v Beiersdorf (Australia) Ltd [1980] HCA 9, and earlier Patentology article ‘Inventiveness – It’s Just Not Obvious!’).

There are fewer than 50 patent remaining to which the 1952 Act provisions apply, and the last of these will expire by 2016.  It is perhaps unsurprising, then, that the court here elected not to upset the status quo at this late stage.  The majority stated, at [217]:

It is not strictly necessary to express a view as to the correctness of the Full Court’s decision in Sanofi-Aventis (2009). Whatever may be the correct position under the 1952 Act, the position under the [1990] Act is, in our view, as previously described.

Justice Jessup was, perhaps, a little more forthright, stating, at [500]:

To the extent that Bennett and Middleton JJ held in Apotex that the “starting point” must always be “a determination of the invention, as described in the specification”, I take the view, with respect, that their Honours’ test would misdirect the focus of the relevant inquiry away from the claim and thus cannot stand alongside Insta Image.

Conclusion – A Good Result for Consistency and Objectivity

Notwithstanding the doubt now attending the approach to inventive step under the (almost defunct) 1952 Act, this decision of five Federal Court judges makes clear that the ‘starting point’ approach is not the law, going forward, under the 1990 Act.  This is a positive outcome, because otherwise it is likely that single judges sitting at first instance would have continued to consider themselves bound by Sanofi-Aventis.

The kind of subjective and drafting-dependent approach employed in Sanofi-Aventis is, in my view, no way to go about assessing inventive step.  While total objectivity may be impossible when dealing with the tricky question of whether or not something is ‘obvious’, at the very least the process by which the question is approached should be objective and consistent.  In this respect, it is interesting to compare the approach of the Australian Full Court in the AstraZeneca case with this statement of the US Supreme Court in KSR Int'l Co. v Teleflex Inc. 550 US ___ (2007):

The question is not whether the combination was obvious to the patentee but whether the combination was obvious to a person with ordinary skill in the art. Under the correct analysis, any need or problem known in the field of endeavor at the time of invention and addressed by the patent can provide a reason for combining the elements in the manner claimed.

Thus, looking to the specification to assess obviousness is likely to mislead.  Of course, this cuts both ways.  The inventor, being inventive and perhaps possessing above average knowledge and skill, may describe the invention in a way that makes it appear obvious, even though it would not be apparent to the ordinary skilled worker.  On the other hand, it should not be possible to make an obvious advance appear inventive merely by ‘talking up’ the inventor’s contribution in the specification!

Image Copyright (c) 123RF Stock Photos

Tags: Appeal, Australia, Obviousness