Routine risks: Full Court invalidates two follow-on pharmaceutical patents

Sandoz AG v Bayer Intellectual Property GmbH [2024] FCAFC 135

What you need to know

• In a significant win in the Full Court of the Federal Court of Australia, Sandoz has revoked two Bayer follow-on patents for a formulation and a method of treatment using rivaroxaban (Xarelto).
• The Full Court held that the two follow-on patents were obvious, in light of an earlier compound patent filed by Bayer that disclosed and claimed rivaroxaban.
• The Full Court clarified the test for inventive step, and confirmed that the risks inherent in routine drug development work are not sufficient, on their own, to render follow-on patents inventive.
• The Full Court reaffirmed that, for prior art to be "ascertained" as part of the inventive step enquiry, there is no need for an "artificial ex post facto literature search", nor evidence that the skilled person would "prefer, prioritise or select" one piece of prior art over others.

What you need to do

• This case highlights the need for follow-on patents for pharmaceutical products to involve inventive work in order to be valid, whether in the formulation, dosage regime or another aspect.
• Generic pharmaceutical companies should keep this decision front of mind when evaluating patents standing in the way of generic entry.
• Originator pharmaceutical companies should be mindful of the decision when drafting follow-on patents and defending their validity, and be sure to highlight any inventive aspects of follow-on work.

Rivaroxaban

Rivaroxaban (brand name, Xarelto) is one of the most significant drugs in the Australian market. It was the 13th most prescribed active ingredient in 2023.¹ Rivaroxaban is Bayer's best selling drug and earned the company over $4B in global sales in 2023.²

Bayer's compound patent for rivaroxaban, international application WO 01/47919 (Compound Patent), expired in Australia in November 2023.

Bayer sought to enforce two follow-on patents:

1. AU 2004305226 (the Formulation Patent), which claims a formulation of rivaroxaban and expires on 13 November 2024; and
2. AU 2006208613 (the Dosage Regimen Patent), which claims a once-daily dosage regimen for administering rivaroxaban and expires on 19 January 2026.

Sandoz clears the way

In November 2021, Sandoz commenced revocation proceedings to "clear the way" for its proposed launch following the expiry of the Compound Patent in November 2023. At first instance, the primary judge held that the Formulation Patent and the Dosage Regimen Patent were valid. Sandoz appealed.

On 23 October 2024, the Full Court of the Federal Court of Australia allowed Sandoz's appeal and held that the Formulation Patent and Dosage Regimen Patent should be revoked.

Ashurst acted for Sandoz. The decision is Sandoz AG v Bayer Intellectual Property GmbH [2024] FCAFC 135.

Inventive step

At first instance the primary judge accepted that, with the Compound Patent in hand, the skilled team would select rivaroxaban as a lead candidate to take into further drug development. The primary judge found that the drug development process involved a well-established series of steps, which included the formulation of the drug and the selection of a dosage regimen.

However, the primary judge held that the inventions claimed in the Formulation Patent and Dosage Regimen Patent were not obvious, because the skilled team would not have had the requisite expectation of success. This was due to the risks and uncertainties inherent in the drug development process, and the limited pre-clinical data about rivaroxaban in the Compound Patent. In particular, the primary judge referred to literature estimating that only 10% of drug candidates successfully proceed from pre-clinical testing to product approval.

As a result of the primary judge's reliance on the general risks and uncertainties in the drug development process, the primary judge did not make any finding that the specific features of the claimed formulation and or dosage regimen were inventive.

On appeal, the Full Court rejected the primary judge's approach. The Full Court held that:

• in drug development, carrying out clinical trials and other tests in order to obtain further data is routine work: at [89];
• while there are multiple steps that follow selection of a lead candidate, each of which may lead to failure, "that risk is common to the field and establishes the baseline or context for the inventive step inquiry; it should not be determinative of the outcome": at [92];
• the primary judge "placed too much emphasis on the risks and unknowns associated with the pre-clinical and clinical tests": at [94];
• it is implicit in selecting a drug as a lead candidate to subject to further drug development work that there is an expectation of success – the skilled person would not proceed into pre-clinical and clinical tests if they did not have a sufficient expectation of success: at [96]; and
• it is not necessary for the skilled person to know a particular outcome at the outset of that work: at [103].

The Full Court held that the inventions claimed in both patents were obvious, because the evidence showed that:

• the skilled team would select rivaroxaban from the Compound Patent as a lead candidate to take forward into drug development;
• once taken forward, there was no evidence that any particular problems would arise during pre-clinical and clinical tests that would require inventive skill to overcome; and
• there was no evidence that the claimed formulation (wet granulation into rapid release tablets) and dosage regimen (once daily) would not have been identified during the course of conventional pharmaceutical development work.

Ascertainment

The primary judge also held that the Compound Patent was not available as prior art under section 7(3) of the Patents Act 1990 (Cth) (the Act), because Sandoz's search evidence did not establish that the skilled person could be reasonably expected to have "ascertained" the document. This issue is relevant for all patents considered under the pre-Raising the Bar form of the Act.

The Full Court overturned this finding, and confirmed that section 7(3) only requires proof of a reasonable expectation that the prior art would be ascertained, on a balance of probabilities. In particular, the Full Court rejected the primary judge's finding that ascertainment required the skilled person to "prefer, prioritise or select" the prior art over other information which might be returned by the searches, or other searches that might be completed.