NAPP Pharmaceutical Holdings Ltd v Dr Reddy's Laboratories (UK) Ltd & Anor

Patent claims expressed by reference to numerical composition ranges generally mean the values stated, interpreted in light of the common general knowledge and specification. Rounding or approximation should not ordinarily enlarge a specified range. The word “about” may permit a small degree of imprecision, but its scope must still provide reasonable certainty for third parties.

For a quia timet patent claim, the claimant must prove on the balance of probabilities that threatened acts will infringe on a scale exceeding de minimis. Statistical uncertainty caused by the claimant’s choice of limited data cannot simply be exploited. Even infringement above a de minimis level may warrant only financial relief where an injunction would be disproportionate and impede legitimate trade.

Napp, proprietor of a patent for a seven-day buprenorphine transdermal patch, claimed that Dr Reddy’s and Sandoz threatened to market products infringing claim 1. The claim specified output composition ranges for buprenorphine, levulinic acid, oleyl oleate, polyacrylate and polyvinylpyrrolidone.

The defendants disputed the construction of the claim and contended that their proposed products would fall outside it. Napp’s claim was principally quia timet. The court therefore had to determine the proper construction of the composition limits, the application of the de minimis principle, the burden and standard of proof for future variable production, and the appropriate statistical approach.

1. Construction. The composition percentages referred to the finished product, not the manufacturing inputs. The numerical limits were expressed to the nearest whole number. Thus “10 %-wt” extended from 9.5 to less than 10.5 %-wt, and “10 to 15 %-wt” from 9.5 to less than 15.5 %-wt. The word “about” in “about 10 %-wt oleyloleate” conveyed a small degree of permitted imprecision, but not a range extending beyond 9.0 to less than 11.0 %-wt. The claim lacked clarity in that respect if “about” were given no defined content, but was not invalid on that ground.
2. The skilled person would understand that formulation components, including levulinic acid and oleyl oleate, could influence the formation of a skin depot. The skin-depot hypothesis in the specification was explanatory only and did not limit the claims.
3. De minimis and quia timet. The de minimis principle applied to patent infringement. A threatened act which would infringe only on a de minimis scale did not justify proceedings. Where there was a clear threat to perform acts which would infringe sufficiently often to exceed that threshold, proceedings could be commenced. The question whether an injunction should be granted was separate.
4. Napp bore the burden of proving on the balance of probabilities that the threatened production before patent expiry would result in infringement on a scale exceeding de minimis. The inquiry was: what proportion of patches was proved on that standard to fall within the claim, and was that proportion more than de minimis? A 50% confidence level was appropriate because it reflected the balance of probabilities. Napp could not rely on uncertainty resulting from its decision to accept the defendants’ analytical data and proceed on a small sample.
5. Remedy. Remedies had to be effective, proportionate and dissuasive, while avoiding barriers to legitimate trade and safeguards against abuse. An injunction could be disproportionate where only a very small, randomly distributed proportion of products infringed. In that event financial relief would be appropriate.
6. On the evidence, Sandoz had not shown a threat to produce infringing patches other than, at most, to a de minimis extent. Dr Reddy’s proposed testing regime, using a 50% confidence level and a threshold of one patch in 10,000, was suitable to ensure that marketed patches would not infringe other than to a de minimis extent.
7. Neither defendant’s product fell within claim 1. There was no threat by either defendant to infringe the Patent.