Formycon AG & Anor v Regeneron Pharmaceuticals Inc. & Anor

Infringement by equivalence must be assessed by reference to the invention at the level of generality of the claims. A court must identify the inventive concept and apply the three Actavis questions, while treating them as guidelines rather than rigid rules. In an empirical formulation field, routine screening methods do not make every resulting formulation obvious. The court must assess the overall expectation of success, the number of alternatives, the technical context and any specific risks. A patent may validly claim a narrow, specifically tested formulation even though its specification discloses broader concepts. The claim scope is not enlarged merely because later workers know that a stable formulation is possible. Experiments cannot ordinarily be used to discover how a variant works for the second equivalence question.

Two actions concerned European Patents (UK) Nos 2 364 691 and 2 944 306, relating to formulations of aflibercept for intravitreal administration. Formycon, Klinge and Samsung sought declarations of invalidity and non-infringement. Regeneron and Bayer counterclaimed for infringement by equivalence.

The products used different buffers, concentrations of sodium phosphate, sodium chloride and sucrose from the claims. The issues included the skilled team, common general knowledge, infringement by equivalence, priority and added matter, obviousness over Wiegand II, insufficiency and technical contribution. The central questions were whether the variants achieved the inventive concept in substantially the same way, whether that would be obvious on the assumption that they worked, and whether strict compliance with the claim language was intended.

1. Outcome. Neither product infringed either patent. The ’306 patent lacked priority and was invalid for added matter. The ’691 patent was entitled to priority and valid over the prior art. Amendments to ’691 were allowed; amendments to ’306 were not.
2. Equivalence. The court applied the approach in [2017] UKSC 48. The inventive concept had to be identified at the level of generality of the claims, having regard to the specification and common general knowledge. The specific formulations corresponding to the tested examples were narrow inventions, distinct from the broader disclosure of stable high-concentration formulations.
3. The Formycon product, using histidine instead of sodium phosphate, achieved stability through a different particular combination of excipients and therefore failed Actavis Q1. The Samsung product, using less sodium phosphate, no sodium chloride and more sucrose, likewise differed in the way the claimed formulation achieved osmolality and stability. Even on Regeneron’s broader inventive concept, the dependent claim, its detailed formulation, and its correspondence with specific examples showed that strict compliance was intended, so Actavis Q3 was not satisfied.
4. For Actavis Q2, experiments could not be used to discover how an otherwise working variant achieved stability. Allowing such experiments would remove the practical operation of the second question and undermine legal certainty.
5. Validity. The ’306 claim impermissibly generalised the specific formulation of the priority examples to a broader class of fusion proteins introduced by “comprises”. The ’691 claims were adequately disclosed. The 4 mg dose was not obvious from Wiegand II. Although formulation methods and individual excipients were common general knowledge, the particular pH and combination were not obvious, especially given the calculated isoelectric point and the oxidation risk associated with polysorbate 20.
6. The 40 mg/ml concentration made a technical contribution because it enabled a clinically useful dose to be administered within the practical intravitreal volume. The applications to add Nguyen 2006 and the Regeneron press release were refused.