Eli Lilly & Company v Janssen Alzheimer Immunotherapy

A patent claim for a product for use in preventing or treating a disease requires the product to be suitable for achieving the claimed therapeutic effect. Animal evidence may establish plausibility, but it need not establish efficacy in humans. A claim remains insufficient if the specification does not make it plausible that the invention works across its scope, or if performance across that scope requires an undue research burden. Later evidence may be used to show that the invention does not work across the claim.

The patent was construed broadly to cover antibodies to monomeric and aggregated Aβ. That breadth was not supported by a plausible teaching that every human IgG1 antibody to Aβ would prevent or treat amyloid disease.

JAI owned a European patent concerning pharmaceutical compositions comprising human IgG1 antibodies to Aβ for preventing or treating diseases characterised by amyloid deposition. Lilly sought revocation for added matter, lack of novelty, obviousness and insufficiency, together with a declaration concerning solanezumab.

The principal issues were the construction of “antibody to Aβ”, “for use in preventing or treating”, “aggregated Aβ” and “dissociated Aβ”; whether the claims were novel and inventive over Konig and Becker; whether the specification plausibly disclosed the claimed therapeutic application across its scope; and whether solanezumab fell within the claims.

1. Construction. “An antibody to Aβ” included antibodies binding monomeric as well as aggregated Aβ. “Aggregated Aβ” was not confined to insoluble plaques. “Dissociated Aβ” was synonymous with, or at least included, soluble monomeric Aβ. The claim was not implicitly limited to a particular mechanism of action or to antibodies inducing the patent’s defined immune response.
2. The words “for use in preventing or treating” imposed a functional limitation. The composition had to be suitable for achieving a beneficial effect on the amyloidogenic component of the disease. The primary criterion indicated by the specification was success in a Phase 2 trial; available Phase 3 results would be the best guide.
3. Validity. The patent was not invalid for added matter. Claim 1 was novel over Konig because Konig disclosed neither human IgG1 nor the required therapeutic efficacy. It was not obvious over Konig, Becker or on Agrevo grounds. The skilled team would have been motivated to investigate the proposal, but Konig and Becker provided no real expectation of success.
4. Insufficiency. The court applied a two-stage inquiry. First, the specification and common general knowledge had to make it plausible that the invention worked across the claim. Secondly, later evidence could establish that the invention could not in fact be performed across that scope without undue burden. The patent made it plausible that passive immunisation with a suitable antibody could be effective, and that N-terminal antibodies might work. It did not make it plausible that any human IgG1 antibody to Aβ would be effective. The claim therefore lacked sufficiency. In any event, selecting and validating an effective antibody required a lengthy, costly research programme with a high prospect of failure.
5. Infringement. Solanezumab was specific for monomeric Aβ, did not appreciably bind plaques, did not induce downstream Fc-mediated effects, and was not shown to induce FcRn-mediated clearance. If valid, claims 1 and 5 would have been infringed. The patent was nevertheless invalid for insufficiency.