Photo by Testalize.me on UnsplashDeveloping and manufacturing biologic medicines can involve years of research, experimentation and process optimisation. However, as a recent Federal Court decision demonstrates, the technical complexity and commercial significance of that work do not necessarily translate into patentable inventions or enforceable patent rights.
In Pfizer Ireland Pharmaceuticals Unlimited Company v Samsung Bioepis Co., Ltd (Infringement and Validity) [2026] FCA 1244, the Federal Court considered both the validity and alleged infringement of Pfizer’s patent directed to the large-scale production oftherapeutic proteins using mammalian cell cultures
The Court found that Pfizer had failed to establish infringement and that all but one of the asserted claims lacked an inventive step.
The decision provides important lessons for businesses developing and protecting biologic manufacturing technologies, particularly concerning how inventions are defined and claimed, the distinction between routine optimisation and inventive ingenuity, and the importance of claim construction and technical evidence in establishing infringement.
The proceedings concerned Australian Patent No. 2005280034 (the Patent), entitled Production of polypeptides, owned by Pfizer Ireland Pharmaceuticals Unlimited Company and exclusively licensed to Pfizer Australia Pty Ltd (together, Pfizer).
The Patent relates to methods of producing therapeutic proteins using large-scale mammalian cell cultures. In particular, it describes cell culture methods involving specific nutrient compositions and culture conditions to promote cell growth and protein production.
The dispute concerned etanercept, a therapeutic fusion protein used to treat autoimmune conditions, including rheumatoid arthritis. Pfizer markets etanercept under the brand name ENBREL.
Samsung Bioepis Co., Ltd and the other respondents (together, the respondents) were involved in manufacturing and supplying a biosimilar etanercept product marketed as BRENZYS.
Pfizer alleged that the manufacturing processes used to produce BRENZYS infringed the claims of the Patent. The respondents denied infringement and challenged the validity of the claims on several grounds, including lack of inventive step.
Claim 1 was central to the infringement dispute. It claimed a method of producing a polypeptide in a large-scale mammalian cell culture, comprising the following four main steps:
1. Providing a cell culture containing mammalian cells genetically modified to produce the desired polypeptide, together with “a medium containing glutamine and having a medium characteristic selected from the group consisting of:
(i) a cumulative amino acid amount per unit volume greater than 70 mM,
(ii) a molar cumulative glutamine to cumulative asparagine ratio of less than 2,
(iii) a molar cumulative glutamine to cumulative total amino acid ratio of less than 0.2,
(iv) a molar cumulative inorganic ion to cumulative total amino acid ratio between 0.4 to 1,
(v) a combined cumulative amount of glutamine and asparagine per unit volume of greater than 16mM, and combinations thereof;”
2. maintaining said culture in an initial growth phase under a first set of culture conditions for a first period of time sufficient to allow said cells to reproduce to a viable cell density within a range of 20% - 80% of the maximal possible viable cell density if said culture were maintained under the first set of culture conditions;
3. changing at least one of the culture conditions, so that a second set of culture conditions is applied;
4. maintaining said culture for a second period of time under the second set of conditions and for a second period of time so that the polypeptide accumulates in the cell culture.
The infringement dispute ultimately turned on two requirements of claim 1:
Two manufacturing processes were used to produce etanercept for BRENZYS, referred to as Process A and Process B.
Both processes involved genetically modified Chinese hamster ovary (CHO) cells cultivated in progressively larger vessels before transfer to a 15,000 L production bioreactor. The cells were cultivated for approximately 14 days, with the temperature reduced from 37°C to 30°C on day 6. Nutrient feeds were also added during cultivation.
Process A used formulated media containing glutamine. It was used to manufacture nine regulatory batches in 2013 and two commercial batches in 2016, the latter being incorporated into BRENZYS products imported into Australia.
Process B was developed by optimising Process A. Unlike Process A, its formulated media did not contain glutamine and instead contained hypoxanthine and thymidine. The optimised formulation achieved comparable cell growth and product quality, and Process B subsequently replaced Process A for commercial production.
The absence of added glutamine in Process B was significant because claim 1 required a "medium containing glutamine", raising a dispute over whether glutamine originating from other sources could satisfy this requirement.
Accordingly, a central infringement issue was whether the expression "medium containing glutamine" in claim 1 was limited to glutamine deliberately included in the formulated media or extended to glutamine present in the culture from other sources.
Pfizer argued for a broader construction, under which glutamine could originate from the formulated basal or feed media, be carried into the production bioreactor with the inoculum from the preceding N-1 bioreactor, or be produced and secreted by the cells during cultivation ([177]).
Pfizer also relied on dependent claim 8, which required glutamine to be provided only in the initial medium, arguing that this included glutamine transferred with the inoculum. Pfizer contended that this supported a broader interpretation of "medium containing glutamine" in claim 1, extending beyond deliberately added medium components ([178]–[180]).
The respondents, by contrast, argued that the expression referred only to glutamine deliberately added as part of the formulated basal or feed media, excluding glutamine originating from inoculum carry-over, cell secretion or incidental contamination. They emphasised that the Patent concerned carefully formulated culture media with specified nutrient amounts and ratios, rather than all substances present in the production bioreactor ([181]–[182]).
Justice Burley preferred the respondents' construction, finding that the claimed "medium" referred to the formulated basal and feed media, rather than everything present in the production bioreactor ([183]–[214]).
His Honour's reasoning centred on the language and technical context of the specification.
First, the Patent described culture media in terms of formulated nutrient compositions, with specified amounts and ratios intended to support cell growth and polypeptide production.
Second, the Patent defined cumulative amounts by reference to nutrients added during cultivation, rather than their actual concentrations at a particular time. Justice Burley considered it inconsistent to include glutamine secreted by cells when determining whether the medium contained glutamine, but exclude that glutamine when calculating the claimed cumulative amounts and ratios. His Honour considered that both requirements should be interpreted consistently by reference to the same medium ([189]–[191]).
Third, Justice Burley observed that the specification did not refer to the inoculum or appear to account for it in describing the invention. Instead, the specification focused on the precise formulation of culture media using quantified components added through the basal and feed media. This supported the conclusion that glutamine incidentally present through inoculum carry-over or cell secretion was not contemplated as part of the claimed medium ([193]–[214]).
Accordingly, Process A satisfied the glutamine requirement, because its formulated growth medium contained glutamine. Process B did not, because glutamine was not deliberately included in its formulated basal or feed media ([252], [254]).
The second major infringement issue concerned the timing of a temperature shift in the culture.
Claim 1 required the cells to be maintained under an initial set of culture conditions until they reached a viable cell density (VCD) within 20%-80% of the maximum possible viable cell density (mpVCD) that would have been achieved if those initial conditions had continued.
Both Process A and Process B involved reducing the culture temperature from 37°C to 30°C on day 6. The parties agreed that this constituted a change in culture conditions. The dispute was whether, at the time of the temperature shift, the cells had reached a VCD within the claimed range.
The claimed range of 20%-80% meant that the mpVCD had to be between 1.25 and 5 times the VCD at the time of the temperature shift. As the lower threshold of 20% was not disputed, Pfizer needed to establish that, had the initial culture conditions continued, the cells could have achieved a maximum viable cell density of at least 1.25 times their VCD on day 6 ([333]).
This required the Court to consider the competing expert evidence concerning the likely growth trajectory of the cells had the temperature remained at 37°C, including whether the cells would have continued growing or were already approaching their maximum viable cell density when the temperature was reduced.
Pfizer relied on the evidence of Dr Croughan, who estimated the mpVCD by extrapolating cell growth data beyond day 6 using a logarithmic curve. His modelling accounted for potential growth-limiting factors, including osmolality, and assumed that nutrient feeding would continue at the day 5 level. On this basis, Pfizer argued that the mpVCD would have been at least 1.25 times the day 6 VCD for all relevant batches except one Process B batch, satisfying the claimed requirement ([334]–[338]).
The respondents relied on the evidence of Dr Mather and Professor Mahler, who considered that the cells were already transitioning towards the stationary phase before the temperature shift on day 6. They identified a significant slowing of cell growth between days 5 and 6, supported by analysis of the VCD data, changes in lactate metabolism and experiments involving delayed temperature shifts ([339]–[345], [437]–[464])
Justice Burley preferred the respondents' expert evidence, identifying several weaknesses in Dr Croughan's modelling. In particular, the modelling did not adequately account for the significant decline in the rate of cell growth between days 5 and 6 ([400]–[406]). His Honour also questioned the reliability of the logarithmic curve, which, for certain batches, would have been based on only three data points ([407]–[413]).
Further, the assumption that maintaining the day 5 nutrient feed would support continued growth was not adequately supported by the evidence, particularly as cell growth had already slowed before the temperature shift ([422]–[436]). Experiments involving delayed temperature shifts provided some support for the respondents' position, while changes in lactate metabolism tended to indicate that the cells were transitioning towards the stationary phase around day 5 ([437]–[464]).
Overall, Justice Burley considered Dr Croughan's extrapolation unpersuasive and found that Dr Mather and Professor Mahler provided a more credible explanation of the observed cell growth trends. Accordingly, Pfizer had failed to establish that the mpVCD requirement was satisfied.
Although experiments specifically designed to determine mpVCD might have assisted, his Honour declined to draw an adverse inference from Pfizer's failure to conduct them. The evidence presented was nevertheless insufficient to establish that the claimed feature was met.
The Court found that neither Process A nor Process B infringed the asserted claims.
For Process A, the glutamine requirement was satisfied, but Pfizer failed to establish that the temperature shift occurred within the claimed mpVCD range.
For Process B, Pfizer failed to establish both the glutamine requirement, under the Court's preferred construction, and the mpVCD requirement.
Accordingly, Pfizer had not established infringement of any asserted claim.
In addition to denying infringement, the respondents challenged the validity of the asserted claims on several grounds. Their challenges based on lack of fair basis, utility, clarity and failure to disclose the best method were unsuccessful. The Court did not consider the insufficiency challenge, which depended on Pfizer's proposed claim construction, which the Court rejected.
The principal issue was inventive step, on which the respondents succeeded in invalidating all but one of the asserted claims.
The respondents primarily relied on common general knowledge (CGK) combined with the prior art document referred to as DelaCruz (US2003/0087372), which disclosed methods of producing recombinant proteins in mammalian cell cultures, including a culture medium known as Medium A and a preference for adding glutamine only to the starting medium.
The respondents contended that a skilled team would routinely adopt DelaCruz's teachings, including testing Medium A containing 4mM glutamine and using feeds without further glutamine. They argued that Medium A satisfied four of the five medium characteristics in claim 1 before feeding, and that at least three would remain satisfied after feeding. They further submitted that the additional limitations of the dependent claims would be met by following DelaCruz or through routine optimisation of nutrient concentrations ([553]–[558], [586]–[598], [659]–[662]).
Pfizer disputed that the claimed invention lacked an inventive step, arguing that DelaCruz's teaching of reduced glutamine was linked to abnormally high glucose levels and that the respondents had not adequately accounted for the effect of glucose-containing feeds on the claimed medium characteristics. Pfizer also disputed that the skilled team would necessarily follow DelaCruz's preference for excluding glutamine from feeds ([589]–[594]).
Pfizer further contended that developing a suitable fed-batch process was not sufficient. Rather, the task involved developing anew method of producing a polypeptide at large scale that achieved better titres. Pfizer characterised this as a one-to-two-year research project involving numerous variables and alternative approaches, without an expectation of success ([566], [640]–[646]).
Justice Burley rejected Pfizer's argument that the task required better titres, noting that the asserted claims did not require an improved process. His Honour found that the complexity and duration of development did not necessarily establish inventiveness. The relevant question was whether the skilled team would arrive at the claimed invention as a matter of routine and without inventive ingenuity ([640]–[650]).
In relation to claim 1, which required at least one of five specified medium characteristics, Justice Burley found that following DelaCruz would satisfy two characteristics, namely a cumulative glutamine-to-asparagine molar ratio of less than 2 and a cumulative glutamine-to-total-amino-acid molar ratio of less than 0.2. Because DelaCruz taught feeding without further glutamine, these ratios would remain below the claimed limits ([622]–[629]).
However, his Honour was not satisfied that the cumulative amino acid concentration would exceed 70 mM. Although Medium A initially contained 71.01 mM amino acids, glucose-containing feeds could dilute the medium below the claimed threshold ([599]–[621]).
Nevertheless, as claim 1 required only one of the five characteristics, the two established characteristics were sufficient. The remaining process steps were also considered routine. Accordingly, claim 1 lacked an inventive step ([650]–[652]).
The position was different for independent claim 2, which required a cumulative glutamine-to-asparagine ratio of less than 2, together with at least two additional medium characteristics. Although the skilled team would arrive at the glutamine-to-asparagine and glutamine-to-total-amino-acid ratios, the respondents failed to establish that a third characteristic would be met.
In particular, the Court was not satisfied that the cumulative amino acid concentration would remain above 70 mM after feeding, or that the claimed inorganic-ion-to-amino-acid ratio or combined glutamine and asparagine concentration would be achieved. The respondents had also not advanced an alternative case that the skilled team would routinely modify DelaCruz's process to achieve these characteristics. Accordingly, the respondents failed to establish that claim 2 lacked an inventive step([653]–[658]).
Justice Burley also found that dependent claims 5, 7, 8 and 33 lacked an inventive step, as their additional limitations concerning glutamine concentrations, its inclusion only in the initial medium, and starting medium characteristics would be satisfied by following DelaCruz ([632]–[635], [659]).
The remaining asserted dependent claims, claims 37–40, 42,45, 46 and 49, specified minimum concentrations of particular amino acids or vitamins that exceeded those disclosed in Medium A. Justice Burley concluded that these additional limitations did not render the claimed combinations inventive, as the relevant nutrient concentrations could be reached through routine optimisation without inventive ingenuity ([660]–[672]).
Ultimately, the Court found that all asserted claims except independent claim 2 lacked an inventive step in light of DelaCruz and CGK. The respondents' three alternative inventive step challenges were unsuccessful( [673]–[683]).
The decision highlights several important considerations for businesses seeking to protect innovations in biologic manufacturing processes.
The different validity outcomes for claims 1 and 2 demonstrate the importance of how technical features are combined and expressed in patent claims. Broad claims encompassing alternative process conditions may be vulnerable where even one alternative is obvious, while more specifically defined combinations may be better positioned to withstand an inventive step challenge.
The Court's interpretation of "medium containing glutamine" illustrates how the specification and its technical context can influence the scope of patent claims. Patent applicants should carefully consider whether the terminology used in their claims and descriptions adequately captures the variations in manufacturing processes they intend to protect.
Establishing infringement of biologic manufacturing process claims can be particularly challenging where the claims depend on process parameters that are not directly measurable or require assumptions about hypothetical operating conditions. Careful consideration should be given to how infringement of the claimed features could be established in practice.
Ultimately, the decision reinforces the importance of identifying and clearly defining the inventive contribution, drafting claims that provide commercially meaningful protection while remaining defensible against validity challenges, and considering how infringement could be established against competing manufacturing processes.
For more information on the patent issues concerning biologics, biosimilars and associated manufacturing methods, please contact Dr Catrina Olivera or Daniel McKinley.
We are proud members of
Office
Level 1/98 Riley Street
East Sydney NSW 2010
Australia
GPO Box 683
Sydney NSW 2001
Australia
Contact
