Glaxosmithkline UK Ltd v Wyeth Holdings LLC

In assessing patent validity, plausibility is a threshold requirement. The disclosure must provide a credible reason for supposing that the invention works across the claim’s scope; it need not prove the invention conclusively. The same threshold applies to AgrEvo-type obviousness. A claimed combination must be construed in light of the patent as a whole, including its disclosed technical contribution. Here, the contribution was a vaccine composition in which 2086 protein provided protection across diverse meningococcal strains and one or more PorA proteins extended coverage. Anticipation by prior use requires an enabling disclosure: the skilled person must be able to identify and reproduce the claimed subject matter without undue burden. A broad list or cross-reference does not necessarily provide an individualised disclosure. The patent was valid, and infringement was established for all material claims except claim 3.

GSK sought revocation of European Patent (UK) 2,343,308 and a declaration of non-infringement. Wyeth, the registered proprietor, counterclaimed for infringement by GSK’s Bexsero meningococcal vaccine. The patent concerned compositions containing 2086 protein, now known as factor H binding protein, together with at least one PorA protein.

The principal issues were construction, infringement, added matter, priority, plausibility, insufficiency, novelty and inventive step. The court considered anticipation by several patent documents, vaccines and conference materials, and obviousness over WO 01/64922 A2. The central questions were whether the patent disclosed a credible technical contribution, whether the priority documents enabled the claimed invention, and whether the prior art disclosed or made it obvious.

1. Construction and infringement. The expression “at least one PorA protein” was not limited to free or isolated PorA and could include membrane-associated PorA in OMVs. Claim 3 required 100% sequence identity because the claims formed a cascading series of narrowing back-ups, and a different percentage identity was specified wherever intended. Claims 1, 2, 5, 6, 10 and 18–20 were infringed by Bexsero; claim 3 was not.
2. Functional claims. Claims 18–20 required a discernible therapeutic effect, but the relevant technical contribution was population-level protection against a broad range of diverse strains. The combination was not a mere juxtaposition: 2086 protein provided broad protection and PorA extended protection to additional strains. Synergy was not required.
3. Plausibility and sufficiency. Plausibility was a threshold test requiring a credible, non-speculative basis for the claimed effect. The patent data made it plausible that lipidated and non-lipidated 2086 proteins could produce bactericidal antibodies, that a single PorA could extend protection, and that suitable aluminium adjuvants could be used. The AgrEvo allegation therefore failed.
4. Added matter and priority. The application as filed disclosed the relevant 2086 sequences, the 95% and 97% identity boundaries, and combinations with PorA. Selections from lists were not governed by a rigid rule; the disclosure had to be assessed as a whole. The first priority document directly and unambiguously disclosed the combination and enabled it. The second priority document also enabled the sequence listings through its PCR and sequencing instructions.
5. Novelty. Anticipation required an enabling and sufficiently individualised disclosure. WO 01/52885 A1 did not individually disclose sequence 2536/741 in combination with a PorA-containing meningococcal preparation. The Cuban Vaccine did not anticipate because its pre-priority composition was not proved to contain fHbp, the skilled person could not identify it without undue burden, and any very low abundance would not be shown to make a material contribution to the claimed functional effects. Similar objections failed in relation to Andersen and WO 03/009869 A1.
6. Inventive step. WO 01/64922 A2 concerned expression methodologies, not identification of vaccine candidates. It did not give a fair prospect of success for selecting 2086/741 without hindsight. The patent’s technical contribution was materially different from the prior art. All validity challenges were rejected.

The patent remained valid. Claims 1–2, 5–6, 10 and 18–20 were infringed; claim 3 was not.