In its judgment, given on 18 December 2009 in Dr Reddy's Laboratories (UK) Limited v Eli Lilly and Company Limited ([2009] EWCA Civ 1362) the Court of Appeal unanimously upheld the judgment of the Patent Court finding Eli Lilly’s patent for an individual compound selected from a class which had been previously disclosed valid.
The Court of Appeal rejected Dr Reddy’s argument that the validity of the patent, as a selection patent, should be assessed in the light of the earlier patent, in accordance with the rules set out in re IG Farbenindustrie AG's Patents (1930) 47 RPC 289, which was decided prior to the 1977 Act. Lord Justice Jacob, giving the leading judgment, clarified that the approach to be taken to "selection inventions" was that adopted by the EPO.
Claim 3 of the patent concerns the chemical entity olanzapine, an anti-psychotic widely used to treat schizophrenia. Dr Reddy’s attacked the validity of the patent on the grounds that it was not novel over a prior Lilly patent application published in 1978 (the “235 Application”). Dr Reddy’s also argued that the patent was obvious over the 235 Application and also a paper by a Lilly scientist, Dr Chakrabarti published in 1980.
Novelty
The 235 application claimed a wide class of thieno-benzodiazepines by way of a general chemical class formula. This general formula (also known as a “Markush” group) was broad and extended to more than 1019 compounds, of which olanzapine was one. The 235 application also identifies a “preferred” class, comprising 86,000 compounds, although no explanation was provided for this preference.
Dr Reddy’s argued that the 235 Application disclosure of the chemical class made olanzapine “available to the public my means of a written…description”. The Court rejected this contention, which amounted to every chemical class disclosure disclosing each and every member of the class, even if simply written down without any utility. The Court’s reasons were based on both EPO case-law and also as a matter of logic. The idea that it was possible to make available a particular leaf by stating that it was in a forest was ridiculous. To accept such a contention would make any generalised prior disclosure in this case e.g. “3-ring organic compounds” novelty destroying, which makes no sense, such a description does not disclose particular matter which falls within it.
The Court followed EPO case law, as summarised in Hoescht Enantiomers (T 0296/87) that for the prior art to anticipate there must be an “individualised description” of the compound or class of compounds claimed. The Court noted that the EPO approach was consistent and clear and that there was no good reason for not following it. The Court observed that the findings were also consistent with the approach adopted by the German Courts (STADApharm No.1-2W 47/07).
Obviousness over the 235 Application
The Court summarised the pre-1977 case law in relation to “selection patents”. The “individualised description” approach did not form part of the UK approach to anticipation, therefore any general disclosure would be prima facie novelty destroying. The approach to avoid such anticipation in relation to a patent for a compound or subclass was set out in I. G. Farbenindustrie’s Patent (1930) RPC 289). In that case Maugham J stated that the requirements of a valid selection patent are: (i) the selected class must have a substantial advantage over the prior disclosure; (ii) the advantage must be possessed by the entire selected class; and (iii) the advantage should be unique to the class so selected. Furthermore, the advantage claimed must be clearly defined. The Court considered that in reality it would be difficult to comply with these rules without carrying out an enormous range of experiments and that they are best considered to be legal history and not part of the current law.
The correct approach is that taken by the EPO, as applied in AgrEvo (T 0939/92), to consider whether the selection from a class has made a technical contribution. If the selection was merely arbitrary it would be regarded as obvious because it cannot involve an inventive step.
The technical problem to be solved by the patent did not relate to making a selection from the general class disclosed in the 235 Application, but to find a compound similar to the existing treatments for schizophrenia which avoided the side effects associated with them. The patent also disclosed the superior properties of olanzapine. Lilly’s selection of olanzapine made a technical contribution and could not be regarded as arbitrary. The Court also accepted the judge’s finding of non-obviousness because the skilled team would have been sceptical about the vague promises of the 235 Application in light of the general unpredictability in the field rendering the inclusion of the class of compounds in research as a step which would not be obvious to take.
Obviousness over Chakrabarti
Chakrabarti synthesised and tested the activity of 59 compounds, based on the results it concluded that a particular class of compounds would merit further development, and highlighted certain members of the class (which did not include olanzapine). Dr Reddy’s argued that based on this disclosure the skilled team would have synthesised all of the compounds in the disclosed class, one of which would have been olanzapine. The judge found that the approach of the skilled team would have been to appreciate the systematic investigation conducted by Chakrabarti and select one of the highlighted compounds as the basis for future testing. The suggestion that all of the compounds would have been synthesised was a hindsight argument. The Court found that the judge was correct in his approach and that there were no grounds for overturning this finding of fact by the judge.
Conclusion
The judgment of the Court of Appeal was made by the two most senior patent judges, Jacob LJ and Lord Neuberger. The decision provides clarity on the law applicable to selection patents and confirms that the pre-1977 approach to selection patents in the UK is no longer good law. The judgment also reflects the Court’s attempts to be consistent with EPO authority so far as possible which accords with the House of Lords in Merrel Dow v Norton ([1996] RPC 76) and Conor v Angiotech ([2008] UKHL 49).