CRISPR – the application of the wrong inventive step test by the Patent Office means a CRISPR patent application is upheld

Date: 02 June 2021
Forum:  Australian Patent Office
Delegate: Damian Triffett

Background

In June 2021, the Australian Patent Office handed down a much-anticipated CRISPR decision, rejecting an opposition filed by JH Corporate Services¹ (JHCS) against Patent Application 2018229489 (Application) in the name of Sigma-Aldrich Co. LLC (Sigma) on the grounds of manner of manufacture, novelty, inventive step, utility, clear and complete enough disclosure, support, best method and clarity (post-Raising the Bar).

CRISPR-Cas9 gene editing technology (CRISPR) is one of the most important breakthroughs in molecular biology since the elucidation of the double helix structure of DNA. Since CRISPR can efficiently target and modify genes it has broad and game-changing applications in a wide range of research areas including drug discovery, cell development and differentiation, production of transgenic animals, improving crops, diagnosis of diseases, such as HIV and COVID-19, and curing genetic diseases.

Key Issues

A key aspect of the case was the inventive step argument. The test for inventive step in Australia involves consideration of both limbs of the ‘modified Cripps question’, namely:

• whether the prior art directly leads the skilled person as a matter of course to the invention; and

• whether the skilled person would have an expectation that the invention might well work.

It was agreed between the parties that the relevant prior art publication did “directly lead” the skilled person to the claimed invention. The only inventive step issue in dispute therefore was the second part of the test, whether the skilled person would have an expectation that performing the claimed invention might well produce the desired result.

Federal Court case law has emphasised that what is required in this respect is an expectation that the invention may well work, and that it is not necessary to know that steps will or would or even may well work. This was underscored by the Full Court’s judgment in Mylan Health Pty Ltd v Sun Pharma ANZ Pty Ltd,² suggesting that the threshold may be met by something which may be ‘no better than fifty-fifty’.

Sigma’s submissions largely followed those made to the Opposition Division of the EPO in respect of the counterpart patent EP3138910 (rejected there), including that (our emphasis):

• “it was not known whether such a bacterial system would function in eukaryotic cells”. and
• “there is no guarantee that Cas9 will work effectively on a chromatin target”.

Outcome

The Application was upheld, conflicting with the outcome in Europe, where the counterpart patent was overturned for lack of inventive step.

Contrary to the case law referred to above, the Delegate appears to have set the benchmark for establishing obviousness as whether there was a reasonable expectation that the invention “would work”, stating (our emphasis) that he “did not consider the steps to achieve targeted integration in eukaryotes would be a matter of routine, as it is not clear whether there was a reasonable expectation that such steps would work”.

Implications

The test for obviousness was elevated in this case by the Delegate to a level that appears to be much higher than a Court would require. As the decision was not appealed, the application of the “would work” standard rather than the established “may well work” standard will not be redressed by the Federal Court in this instance.

The decision appears to leave room for patent applicants to argue during prosecution that unless an invention would be known to work before it was performed, there cannot be an expectation that it may well work, taking the test closer to novelty than the established test for obviousness.

1. Pearce IP acted for JHCS in this case
2. [2020] FCAFC 116