The decision will be significant in clarifying the law relating to obviousness in Australia. There are potentially two questions at stake in this regard:
- What is the appropriate ‘starting point’ for assessing whether or not a claimed invention is obvious?
- Does the Australian law limit or prohibit ‘selection’ of a favourable prior art document from a potentially large field of candidates as a basis for finding an invention obvious?
Background to the Obviousness IssuesThe original trial judge found (in Apotex Pty Ltd v AstraZeneca AB (No 4)  FCA 162) that the AstraZeneca inventions claimed in the patents at issue were obvious (i.e. lacked an inventive step) based on a ‘starting point’ determined from a reading of the patent specifications. Put simply, the judge accepted that the problem to be solved by the invention was set out in the specifications, and that the solution to that problem was then obvious.
This reasoning was overturned by the five-judge panel of the court, which determined (in AstraZeneca AB v Apotex Pty Ltd  FCAFC 99) that the evidence in the case established that the problem as described in the patent specification was not known to ordinary skilled workers in the field. Thus the correct ‘starting point’ was not the knowledge that might be attributed to the inventor, as disclosed in the patent specification. Instead, the correct perspective from which to assess the existence of an inventive contribution was that of the ordinary worker in the field. (For more on this, see my original article on the decision: What is the Starting Point for Inventive Step?)
Although the five-judge panel rejected the trial judge’s reasons on the ‘starting point’ question, it nonetheless found AstraZeneca’s claims to be obvious. Viewed from the perspective of the ordinary skilled worker, the court found that there was one particular document in the prior art which – so it is said – clearly taught the skilled person that rosuvastatin was a very promising candidate for an improved cholesterol treatment, and that the relevant dosages would then be arrived at by non-inventive routine steps.
Arguing before the High Court back in May, AstraZeneca submitted, in essence, that the Federal Court panel was wrong to accept, as a given, the selection of the one document that could be used to establish obviousness. Its argument was that identifying this particular document involved a process through which a total of about 4500 documents identified in an initial search was reduced to a short-list of 19, and then down to two documents disclosing ‘promising’ candidates, of which rosuvastatin was one.
Under the law of inventive step that applies to the AstraZeneca patents, the court is permitted to have regard to only one prior art document, in combination with the common general knowledge of the ordinary skilled worker. AstraZeneca says that while the five-judge panel only required the one document to reach its conclusion of obviousness, it should have considered that the process by which that document was identified was comparative, and involved consideration of numerous documents in arriving at a final selection.
The High Court is likely to rule on both the ‘starting point’ and ‘selection’ questions.
Options for the High Court on ‘Starting Point’With regard to the ‘starting point’ question, I believe that the trial judge got it wrong, and the five-judge panel was correct. Starting from the information disclosed in the patent specification, unless it can be shown to be consistent with the information available to all skilled workers in the field, is likely to lead to subjective and inappropriate outcomes. Inventors and patent applicants can write whatever they wish in their patent specifications, and objectivity therefore demands that courts, and other decision-making authorities, look to independent sources of information to determine the actual contribution made by a claimed invention.
Nonetheless, the generic pharmaceutical manufacturers opposing AstraZeneca’s appeal have argued that the trial judge’s decision was correct, and is justified on policy grounds. In this, they have the support of the Commonwealth of Australia, which intervened to make submissions in the High Court appeal. Their argument is, in essence, that if there are two steps required to get to a complete invention (say, ‘A+B’), the first of which (to ‘A’) is inventive, and the second (from ‘A’ to ‘A+B’) is non-inventive, then two (or more) patents might be granted where there is, in truth, only one invention, i.e. ‘A’. In the pharmaceutical field, in particular, this creates a risk of ‘evergreening’, whereby additional patents are granted for non-inventive improvements on an initial drug. If these further patents have a later expiry date than the primary patent, then the effective term of the monopoly granted on the basis of the original inventive contribution may be extended.
I actually think this argument is valid, and that this kind of ‘piggy-back’ patenting can, and does, sometimes occur. But I do not think that the issue is specific to Australian patent law, nor is it specific to pharmaceuticals. Until a new development becomes recognised as a sound basis for incremental improvements, it might well form the basis for a number of variations on the inventive ‘theme’. These variations might be developed by a single inventor or company, or they might be independently developed by unrelated inventors. Either way, in practice this period of incremental ‘invention’ is likely to be limited, because once the cat is out-of-the-bag about the original development, all of the variations will rapidly become widely regarded as obvious. To my knowledge, this issue has not proven to be a significant problem in Australia, or in any other jurisdiction.
Options for the High Court on ‘Selection’On the ‘selection’ question, a literal interpretation of the relevant Australian law on inventive step is not necessarily favourable to AstraZeneca’s argument. The law permits the use of a prior art document as the basis for assessing obviousness, and has nothing to say about the process by which that document is identified, other than that it must be information that would be ‘ascertained, understood and regarded as relevant’. The fact that a selected document might be just one of many that would meet this requirement, and that all of the others would result in the skilled person heading off down a different path, does not expressly enter into the inquiry. To find in favour of AstraZeneca, the High Court will need to take a broader and more ‘purposive’ view of the law.
Personally, I hope that the High Court follows a broader approach, as urged by AstraZeneca. The Australian law on inventive step has, for too long, suffered from the unduly prescriptive approach set out in the legislation. (For more on this, see my earlier article on the High Court appeal: Can the High Court Fix Australia’s Obviousness Problem?) A more flexible interpretation that takes into account the actual practices of workers in a field of endeavour would bring the Australian law into closer alignment with our major trading partners, particularly if the same principles could then be applied to the law as recently revised under the Raising the Bar reforms.
This would not necessarily mean victory for AstraZeneca. The High Court could accept AstraZeneca’s arguments, but still find that the prior art document relied upon by the Full Federal Court would have presented itself as so obviously the best candidate for further investigation that it should be used as the basis for assessing inventive step.
I suspect, however, that the High Court will stick quite closely to the wording of the Patents Act, and will substantially uphold the decision of the five-judge Federal Court panel on the ‘selection’ question.