Manufacturing-process patents can protect valuable improvements in the way medicines are made. However, they can also present challenges for enforcement. Unlike a product patent, where the patented features may be apparent from the product itself, a process patent requires the patentee to establish what occurs behind a competitor’s closed doors. The patentee may require access to confidential manufacturing information and must prove that the competing process includes every requirement of the asserted claims. Even after obtaining court orders for inspection enforcement may present substantial evidentiary hurdles. Inevitably, enforcement also exposes the patent to validity challenges.
The Federal Court decision in Pfizer Ireland Pharmaceuticals Unlimited Company v Samsung Bioepis Co., Ltd (Infringement and Validity) [2026] FCA 1244 highlights many of the practical challenges associated with enforcing manufacturing-process patents, particularly in the biologics context where competition often takes the form of biosimilar products manufactured using complex and confidential production processes.
This decision represents the culmination of a long-running dispute. In 2017, Pfizer successfully obtained preliminary discovery to investigate whether Samsung’s confidential manufacturing process infringed the patent. In 2024, the Court addressed further discovery disputes in the substantive proceedings, underscoring the need for discovery requests to remain focused on the issues in dispute. The 2026 judgment ultimately determined the infringement and validity issues that had given rise to those earlier procedural battles.
Notably, Pfizer’s infringement case failed despite years of Court ordered discovery and access to extensive evidence concerning Samsung’s manufacturing processes. The Court found that Pfizer had not established that the relevant processes satisfied critical claim requirements, and most of the asserted claims were found invalid for lack of inventive step.
The dispute
Pfizer markets etanercept under the name ENBREL. Etanercept is a fusion protein used to treat autoimmune disorders, including rheumatoid arthritis. Pfizer alleged that eight respondents, including Samsung Bioepis, infringed Australian Patent No. 2005280034 through activities relating to BRENZYS, a biosimilar version of etanercept. A biosimilar is a biological medicine shown to be highly similar to an approved biological medicine, without clinically meaningful differences in safety or effectiveness.
The patent did not claim etanercept itself. It claimed methods for producing therapeutic proteins on a large-scale using mammalian cell culture. The asserted claims focused on aspects of the production process, including the composition of the culture medium and the timing of a change from an initial growth phase to a later production phase.
BRENZYS had been manufactured using two processes, referred to as Process A and Process B. Both used genetically engineered Chinese hamster ovary cells in large-scale bioreactors. However, relevant to the present decision, the processes differed in their use of glutamine. The starting medium used in Process A contained glutamine, whereas Process B deliberately excluded glutamine from both the starting or feed medium, instead using glutamine alternatives hypoxanthine and thymidine.
Pfizer needed to prove that each process included every requirement of the asserted claims. Four respondents cross-claimed that the asserted claims were invalid, including for lack of inventive step.
The outcome
The Court found that none of the asserted claims were infringed.
First, Process B did not meet the requirement for a “medium containing glutamine”. The Court held that this requirement referred to glutamine being deliberately included in the formulated culture medium for both processes. Because Process B intentionally excluded glutamine from both the starting and feed media, instead relying on alternative components, it fell outside the claims.
Secondly, the claims required the cells to have reached between 20% and 80% of the maximum viable cell density that would be achieved had the original culture conditions continued. That maximum was not measured for the relevant manufacturing batches, leaving Pfizer to rely on modelling and reconstruction by an expert of what would have occurred under different operating conditions.
The Court was not persuaded by the expert evidence and concluded that Pfizer had failed to establish the viable cell density requirement for either process. Process A therefore failed to infringe because that requirement was not proven. Process B failed for the same reason, and, independently, because it did not satisfy the separate requirement for a “medium containing glutamine”.
The respondents established that all but one of the asserted claims lacked inventive step. The Court concluded that many of the claimed medium characteristics reflected routine process-development and optimisation work in the field. Claim 2 survived because the respondents could not show that its more specific combination of medium characteristics would have been reached through routine work. However, claim 2 included the viable cell density requirement and was not infringed.
The remaining validity challenges were unsuccessful. The overall result was decisive: Pfizer obtained no relief: none of the asserted claims was infringed, and all but claim 2 were found invalid for lack of inventive step.
Drafting lesson 1: define process components with enforcement in mind
Claim 1 required a medium containing glutamine and having at least one characteristic expressed using cumulative amounts or ratios of components in the medium. The patent defined “cumulative” as the total amount of a component added during the culture process, including amounts in the starting medium and later feeds.
Process A satisfied the “medium containing glutamine” requirement because its formulated starting medium contained glutamine. Process B, however, intentionally excluded glutamine from both its formulated starting and feed media. To establish that Process B nevertheless satisfied the requirement, Pfizer argued that the relevant medium could include glutamine carried into the production bioreactor from an earlier bioreactor, produced by the cells or present from another source.
The Court disagreed, holding that the relevant medium was the formulated solution containing selected nutrients and did not include glutamine produced by the cells, carried into the production bioreactor with the inoculum or present as an incidental contaminant. As a result, Process B, which intentionally excluded glutamine from both the starting and feed media, fell outside the claims.
Definitions may both clarify and confine claim scope. If a process claim is intended to cover components from different sources, the specification should expressly address whether it includes material carried over from an earlier step, produced during the process or present as an impurity or residue. It should also address how those sources are treated in any claimed concentration, ratio or cumulative amount.
Drafting lesson 2: draft claims that can be proven
In respect of the viable cell density requirement the claims required a temperature shift to occur when the cells had reached between 20% and 80% of the maximum density they would have achieved if the original culture conditions had continued. In practice, the temperature was reduced on day six. Pfizer therefore had to prove a hyopothetical: what occurred during the manufacturing process, and also what would have occurred had the temperature reduction on day six never been made.
The Court was not persuaded by the expert evidence, expressing concerns about the limited data, assumptions about nutrient supply, evidence of declining cell growth before the temperature change, and evidence that the cells were already moving towards the stationary phase.
Applicants should consider not only whether a claim feature distinguishes over the prior art, but whether it can realistically be proven in an infringement action. Features that depend on hypothetical outcomes, counterfactual scenarios or information that is unlikely to be measured or retained may create significant enforcement difficulties. Claims directed to observable materials, steps or operating conditions may be easier to enforce in practice.
Drafting lesson 3: ensure the invention lies beyond routine optimisation
The Court found that many features of the claimed method formed part of routine bioprocess development and were within the common general knowledge. These included the use of CHO cells, serum-free enriched media, fed-batch processes, media optimisation and a temperature shift from a growth phase to a production phase.
The respondents also relied on an earlier patent publication, DelaCruz, which taught reduced glutamine concentrations and preferably no glutamine added during culturing. The Court found that, applying the common general knowledge to that teaching, the skilled team would have arrived at the relevant claim features through routine work.
Pfizer argued that developing the process was arduous and could take one to two years and involve many experiments. The Court did not consider the duration of the work sufficient to establish an inventive step. Skilled teams may routinely undertake substantial and time-consuming development programmes. However, the key question was whether the claimed features reflected inventive ingenuity or the results of routine optimisation.
Several dependent claims added minimum quantities of amino acids and vitamins. The patent did not explain the technical benefit of many of those limits, and the evidence indicated that suitable amounts could be identified through routine optimisation. The Court described them as “mere bells and whistles”, rather than features supporting an inventive step.
The practical lesson is that numerical ranges, concentration limits and optimisation parameters are more likely to withstand validity challenges where the specification explains why they matter. Before asserting a patent, the owner must assess validity. That review should identify what distinguishes each claim from the prior art, whether those features reflect routine work, whether their technical significance is explained and whether a narrower but stronger claim would still cover the competing process.
Strategic takeaways
The decision highlights the need to:
- draft process claims with future enforcement in mind;
- define components and their possible sources clearly;
- include claims directed to measurable steps and operating conditions;
- explain the technical significance of numerical limits;
- link claimed features to unexpected or valuable results; and
- assess both infringement and validity before commencing proceedings.
Process patents remain an important part of pharmaceutical protection and freedom-to-operate strategy, particularly in the biologics and biosimilars context. Their value, however, depends on claims that capture a meaningful technical contribution and can realistically be enforced.
For patent owners, these issues should be addressed when the application is drafted and revisited before enforcement. For companies planning market entry, freedom-to-operate analysis should consider process patents early enough to determine whether changes to the manufacturing process are required before regulatory submissions, process validation and commercial supply arrangements are finalised.
Effective patent protection depends on more than securing grant. Griffith Hack can assist with patent drafting, portfolio strategy, infringement and validity assessments, and freedom-to-operate reviews. Please contact us here for further information.
