Eli Lilly & Company v Human Genome Sciences Inc

For an invention consisting of an isolated gene sequence or encoded protein, industrial applicability requires more than proof that the structure can be made or that it probably has some medical value. The application, read with common general knowledge, must disclose a sufficiently precise practical use, supported by a sound and concrete basis and a real likelihood of exploitation.

Sequence homology and family membership may make a function plausible, but they do not by themselves establish industrial application where family members have diverse or poorly understood effects. A use that remains a research programme, or is supplied only by post-published evidence, is insufficient. Industrial applicability is fact-sensitive, and national courts should respect first-instance evaluation of the evidence.

Human Genome Sciences Inc appealed from Kitchin J’s decision in the Patents Court, reported at [2008] EWHC 1903 (Pat), that its patent for Neutrokine-α, related antibodies and encoding sequences was invalid. The judge found lack of industrial applicability, insufficiency and obviousness, with a further insufficiency finding concerning claims 18 and 19.

After the appeal was lodged, the EPO Technical Board of Appeal allowed a separate appeal on more restricted claims. The Court of Appeal heard the appeal on those claims. The central issue was whether the patent disclosed a sufficiently specific and reliable practical use to satisfy Article 57 of the European Patent Convention.

1. Appeal dismissed. The appeal failed on industrial applicability. It was unnecessary to decide the remaining insufficiency and obviousness issues.
2. Under Article 57 of the European Patent Convention, an invention must have a practical industrial use disclosed in the patent, read with the common general knowledge. There must be a sound and concrete basis for recognising a real prospect of exploitation. A vague statement that a protein or its antibodies may have some pharmaceutical use is insufficient if the skilled person must undertake a research programme to discover what that use is.
3. The word plausible requires more than a conclusion that a proposed use is not impossible. The proposed use must have a real reason for being regarded as likely and must be stated with sufficient precision. Sequence homology or membership of a protein family may assist, but the result depends on the facts, including how well the functions of the known family members are established and whether their effects are diverse or contradictory.
4. Post-published evidence cannot supply an industrial application absent from the application as filed. It may, however, assist in assessing whether the patent, together with common general knowledge, already conveyed a practical use. Subsequent research showing that the inventors were still trying to discover the protein’s function may indicate that the original disclosure was insufficient.
5. The national court was not bound by the EPO Board’s factual assessment. The Board and Kitchin J had considered different evidence under different procedures. Questions of degree, including industrial applicability, require evaluation of the evidence, and the Court of Appeal should show substantial caution before reversing the first-instance judge’s findings and value judgments. National courts remain the final arbiters of validity in their own jurisdictions.
6. UK courts should follow clearly established EPO principles of law, subject only to the exceptional right to differ where the principle is plainly wrong. That deference does not extend to factual evaluations or standards applied to particular facts in other EPO cases. On the evidence, Neutrokine-α’s anticipated roles in T-cell and B-cell activity were too general and speculative to establish a practical industrial application.

• Court of Appeal (Civil Division): The appeal was dismissed on 9 February 2010. The Court considered the restricted claims allowed during the related EPO proceedings and upheld the finding that the invention lacked industrial applicability.
• Chancery Division (Patents Court): Kitchin J held the patent invalid on grounds including industrial applicability, insufficiency and obviousness in [2008] EWHC 1903 (Pat).
• European Patent Office Technical Board of Appeal: The Board allowed HGS’s separate appeal on more restricted claims on 21 October 2009 and gave reasons on 1 December 2009. The Court of Appeal did not defer to its factual assessment.