This article was first published by the International Law Office, a premium online legal update service for major companies and law firms worldwide. Register for a free subscription.


Claims are technical solutions seeking protection and should be a generalisation of the content sufficiently disclosed in a patent description. However, poorly drafted claims that include inappropriate generalisations risk being unsupported by their description.

Determining the appropriateness of generalisations remains controversial in practice. Medical use is an essential improvement drug invention. When ascertaining whether a claim of medical use invention is supported, it is crucial to reasonably interpret the scope of the claim.

To shed some light on patent drafting, granting and the affirmation procedure, this article dissects the Re-examination and Invalidation Department's (CNIPA's) examination methodology in an invalidation case concerning a medical use patent.


The invalidation case (Decision 38074) concerned a patent for the medical use of Levo-ornidazole (L-ornidazole). The claimant sought to protect the "use of L-ornidazole in the preparation of an anti-parasitic infection medicament". L-ornidazole is a single isomer of the marketed drug racemic ornidazole.

The patent inventor found that L-ornidazole can dramatically reduce central nervous system (CNS) toxicity while retaining the effect of ornidazole, and thus has a therapeutic advantage. The description's working examples provided experimental data on L-ornidazole's anti-protozoan infection effect and toxicology.

In the invalidation proceedings, the petitioner considered the scope of the term 'parasite' to be overly broad – different parasites have distinct habits and metabolic systems – and thus the claim could not be supported by the description, which provided data for only trichomonas vaginalis and cecal amoeba. According to the petitioner, persons skilled in the art would not be able to predict the efficacy of L-ornidazole against other parasites.


The CNIPA held that the patent's improvement lay in the discovery that L-ornidazole had lower CNS toxicity, rather than new pharmacological activities (eg, activity against a broadened spectrum of parasite). It was known in the prior art that nitroimidazole drugs such as ornidazole had anti-parasitic (especially anti-protozoan) effects.

Due to the structural similarity between L-ornidazole and racemic ornidazole, their efficacies are predictably similar. Based on the disclosure of the description, the CNIPA affirmed that those skilled in the art would be able to reasonably determine the protection scope of the claims, rather than arbitrarily extend it to the extreme case (ie, being effective for all parasites).


Claims should be reasonably interpreted by considering the inventive improvement and prior art, rather than relying solely on the literal meaning of claim language. In judging whether a claim can be supported by a description, it is vital to identify the substantial contribution made by the invention to the prior art and the specific working examples should not be the sole basis for judging support of the claim.