Hospira UK Ltd v Genentech Inc

For a therapeutic-use claim, a proposed treatment may be novel even though the proposal itself is disclosed. However, the proposal may nevertheless render the claim obvious where the skilled team would undertake a clinical trial with a reasonable expectation that the treatment would work.

Where therapeutic efficacy is a claimed feature, sufficiency and priority require a plausible or credible basis for predicting that the treatment will work. A priority document must also disclose the claimed subject matter directly and unambiguously. A product claim defined by characteristics is not limited to products made on a particular manufacturing scale unless the claim says so.

Hospira sought to invalidate two European patents held by Genentech concerning trastuzumab. The first patent claimed an intravenous 8 mg/kg loading dose followed by 6 mg/kg doses every three weeks for HER2-positive breast cancer. The second claimed trastuzumab compositions containing specified levels of acidic variants.

The issues included obviousness, sufficiency, priority, novelty and a declaration of non-infringement. Hospira relied principally on the FDA label, the Andya application, Waterside conference slides and common general knowledge. The court also considered whether the claimed dosing regimen was plausibly disclosed for priority and sufficiency purposes.

1. 115 patent—obviousness. The skilled team comprised an oncologist and a pharmacokinetics expert. A three-weekly regimen would naturally occur to the clinician because it would provide convenience and quality-of-life benefits when combined with three-weekly paclitaxel. The clinician would consult the pharmacokinetics expert rather than dismiss the idea because of the reported 5.8-day half-life.

2. The pharmacokinetics expert could use the FDA label and common general knowledge to assess higher-dose trastuzumab. The label disclosed dose-dependent half-life, safe administration of a 500 mg weekly dose and a 12-day half-life at that dose. Despite modelling uncertainties, the expert would advise that there was no pharmacokinetic reason not to conduct a small trial of three-weekly dosing. Selecting an 8 mg/kg loading dose and 6 mg/kg maintenance dose was not inventive. Claim 1 was therefore obvious.

3. Sufficiency. On the hypothesis that the claim was inventive, the patent would be insufficient. The skilled team would not have sufficient confidence from the patent and common general knowledge to conduct the clinical trial. The requirement of plausibility was therefore not met.

4. Priority. The requirement for an enabling disclosure, including plausibility of a claimed therapeutic effect, applies to priority as well as sufficiency. The court disagreed with the contrary approach in Gemvax. The priority document disclosed the claimed combination directly and unambiguously, but the claimed invention was not plausibly enabled. Since the claim was obvious, the issue did not affect the final result.

5. 455 patent—claim construction and novelty. Claim 1 was a product claim to a composition suitable for therapeutic use. It was not limited to material made at full manufacturing scale. Andya made an enabling disclosure of a trastuzumab composition containing 82% native protein and no more than 18% acidic variants. Claims 1, 2 and 4 therefore lacked novelty.

6. 455 patent—obviousness. Waterside disclosed the relevant deamidated Asn30 variant and its separation from native trastuzumab at analytical scale. The skilled team would regard minimising acidic variants as desirable and would expect to achieve a specified purity by trading yield against purity. Claims 1 to 4 lacked inventive step over Waterside. Claim 3 was not shown to be obvious over Andya alone.

7. Declaration. Formulations containing 25% to 29% acidic variants fell outside the claims, whose upper limit was 24.5%. The declaration of non-infringement was therefore granted.

8. Both the 115 patent and the 455 patent were invalid.