SELECTION INVENTIONS IN THE PHARMACEUTICAL FIELD

Selection Inventions In The Pharmaceutical Field – Developing Law And Policy TIM POWELL1

I. INTRODUCTION A selection invention involves the selection of individual elements or a sub-set of elements from a broader class which has been defined only in general terms. Selection inventions are typically made in the chemical field where a general formula or list may encompass a large number of individual compounds. Other types of selection may be from a numerical range. Selection inventions have recently been in the spotlight in several European jurisdictions and at first sight there seem to be significant differences in approach to the apparently complex issues which they pose. However, a closer scrutiny of the development of the law reveals deeper similarities than are first apparent but also highlights the dangers of adopting a ‘pick and mix’ approach to patentability. The justification for granting a state monopoly in respect of a selection invention is no different from that for any other type of patent. The long established patent bargain requires the public disclosure of the invention in return for a period of protection during which time the patentee may exclusively exploit the invention. This provides an opportunity for the patentee to obtain a return on the investment put into the making of the invention while ensuring that details which would otherwise remain secret become publicly available. This bargain is even reflected in the original name: ‘Letters Patent for an invention’ derived from the Latin literae patentes meaning ‘open letter’. The public policy objectives behind the bargain are essentially the stimulation of innovation and development. A careful balance which aims at providing sufficient reward to incentivise potential inventors and investors without stifling future progress by unduly restricting the field to others is required. In a leading UK case on selection inventions, E.I. du Pont de Nemours (Witsiepe’s) Application2, Lord Simon specifically commented on this: “Before I turn to consideration of the law relating to selection patents, I venture to remark that Du Pont’s patent appears to satisfy the general ends to which the patent law is directed – encouragement of invention and of the publicising of invention so as to promote technological advance by granting a temporary monopoly, on the one hand, without, on the other,

1 The author would like to thank Claire Robinson of Powell Gilbert for her assistance. 2 E.I. Du Pont & Co. (Witsiepes) Application [1982] F.S.R. 303.

341

TIM POWELL stultifying endeavour at improvement by perpetual or prolonged monopoly or (in historical origin) giving unearned reward to favoured persons.”

Nevertheless, the dilemma posed by selection inventions concerns one of the most fundamental requirements of patentability: that an invention must be new over the state of the art. Almost by definition a selection implies a choice made from preexisting alternatives and indeed the term ‘selection invention’ has come to be used for inventions which lie within parameters previously marked out but for which a new and unexpected advantage has been discovered. The difficulty is therefore in deciding how those pre-existing parameters impact upon novelty. The approach to this issue has varied as the approach to patentability has historically across Europe. The real question is whether this has led to the establishment of different legal principles and whether there is now a conformity of approach following the adoption of the European Patent Convention.

II. UK APPROACH TO SELECTION INVENTIONS In the UK, the first case to directly address the question of the validity of a “chemical selection patent” was that of I.G. Farbenindustrie A.G.’s Patents3. This case concerned a patent for the manufacture of azo dyes. Azo dyes were known but the particular selected dyes were alleged to show excellent colour fastness. In his judgment Maugham J. contrasted “originating patents” concerning the discovery of a new reaction or compound with “selection patents” based upon “the selection of related compounds such as homologues and substitution derivatives of the original compounds which presumably have been described in general terms and claimed in the originating patent”. Maugham J. was clear that a selection invention should not benefit from any exception to the rules governing patentability: “a selection patent does not differ in nature from any other patent and is open to attack on the usual grounds of want of subjectmatter, want of utility, want of novelty and so forth.” Accordingly, this first case was concerned with the question of whether it is possible to show “subject matter” (i.e. inventiveness) in respect of such a selection. Maugham J.’s answer to this was undoubtedly yes. Although reluctant to state definitively the conditions upon which a selection patent depends, Maugham J. did make three general propositions: (i) “First, a selection patent to be valid must be based on some substantial advantage to be secured by the use of the selected members. (The phrase will be understood to include the case of substantial disadvantage to be thereby avoided.)” (ii) “Secondly, the whole of the selected members must possess the advantage in question.”

3 I.G. Farbenindustrie A.G.’s Patents (1930) 47 R.P.C. 289.

342

SELECTION INVENTIONS IN THE PHARMACEUTICAL FIELD

(iii) “Thirdly, the selection must be in respect of a quality of a special character which can fairly be said to be peculiar to the selected group.” In relation to the first proposition Maugham J. noted that a »substantial advantage attributable to the use of the selected members” is inherent in a selection invention and should be differentiated from the requirement of utility. He explained that the second and third propositions also follow from the basis upon which the grant of a patent is justified: “… if the selection embraces selected members which do not possess the alleged advantages, the selection is defective and the patent would be misleading and would also fail for insufficiency and non-utility.”

The selection would be similarly defective if it did not encompass a significant proportion of the compounds possessing the special character. Consequently, it is essential that the nature of the characteristics possessed by the selection for which a monopoly is claimed is clearly defined: “… in a selection patent the inventive step lies in the selection for a useful and special property or characteristic adequately defined …”

The patents being litigated were ultimately held to be invalid as the promised advantages of the selection were either not borne out or were too vague. Further, amendments which would have “… greatly diminished the number of the selected compounds described …” were disallowed on the grounds that “… the Patent can only be defended as a selection of compounds possessing special properties or advantages …”. Selection inventions continued to be recognised after the introduction of the 1949 Patents Act. In Beecham Group v. Bristol Laboratories4 Maugham J. was cited by Lord Diplock who summarised the nature of a selection invention: “… The inventive step in a selection patent lies in the discovery that one or more members of a previously known class of products possess some special advantage for a particular purpose, which could not be predicted before the discovery was made (In Re I. G. Farbenindustrie A.G.’s Patents (1930) 47 R.P.C. 283 per Maugham J. at pp. 322/3). The quid pro quo for the monopoly granted to the inventor is the public disclosure by him in his specification of the special advantages that the selected members of the class possess.”

In the House of Lords Du Pont2 case referred to above Lord Wilberforce described this excerpt as a compendious statement of the present position, adding: “The general principle is now securely part of the law and needs no fresh discussion …”. Lord Simon concurred: “That the concept of selection patents is now firmly established in our law appears not only from Lord Diplock’s speech (in which the other members of the Appellate Committee concurred) but also from the numerous authorities reviewed in the judgments appealed from, to which must now be added that selection patents are recognised by the European law devel-

4 Beecham Group Ltd v Bristol Laboratories International S.A. [1978] RPC 521.

343

TIM POWELL oped under the Treaty of Rome: see Bayer A.G. (Baatz & Al.’s) Application [1981] O.J./E.P.O. 206.”

Selection inventions were therefore undoubtedly accepted as being capable of forming adequate subject matter for a patent by the time the Patents Act 1977 was enacted. However, the Du Pont case was primarily concerned with anticipation rather then inventive step as it was alleged that Du Pont’s 1975 patent application was anticipated by a prior I.C.I. specification. This earlier specification disclosed a new class of chemical products, copolyesters, obtained by heating a preferred glycol with other specified ingredients. The resulting copolyesters were said to have enhanced adsorptive capacity for water enabling them to be more readily dyed in the form of fibres for textile materials. The specification also suggested that using any one of another eight glycols would produce copolyesters in the same class with the same properties but none of these compounds was exemplified. The Du Pont application included claims to the copolyester which would result from using a particular one of these eight glycols. However, the invention was directed to a very different use: the copolyesters were said to be especially effective in injection moulding and high speed extrusion applications because of their rapid hardening rates. Lord Wilberforce summarised the issue before the court: “First, it may be true to say as a general rule that where an invention for a substance is specifically disclosed, with a claim for particular advantages, or where a substance is already known, a discovery that the disclosed or known substance has some advantage or useful quality not previously recognised does not give a right to a patent. The difficulty arises when disclosure is made of a group or class of substances for which some advantage is claimed, and later it is found that one or more of this group or class possesses special advantages not belonging to the rest of the group or class, and not previously identified.”

Was Lord Wilberforce thereby suggesting that there may be some exception to the strict rule of novelty? The situation referred to is particularly likely to arise in relation to chemical inventions as the attributes of newly discovered chemical combinations are also likely to be found in homologues or other variants which can easily be described, and therefore claimed, using a generic formula. The crux of the matter is therefore whether such a disclosure is to be regarded as an anticipation: “Is, then, the mere fact that he has disclosed or published in general terms the possibility of these combinations, in such a way that they can be made, a disclosure or publication of unrecognised advantages which may be found to be possessed by one or some of them?”

Lord Wilberforce considered it necessary for anticipation that the prior disclosure clearly indicated that use of the selected material would result in a product having advantages predicted by the class. It is interesting (and in contrast to the approach of the EPO) that Lord Wilberforce considered it irrelevant whether the prior class was described by a generic formula or by a list of compounds including the later selected compounds. He observed that the skilled person could readily transform a generic formula into a list. Similarly he did not consider the size of the prior disclosed class to be relevant to the issue of anticipation although it may be relevant to obviousness. 344

SELECTION INVENTIONS IN THE PHARMACEUTICAL FIELD

Shortly after the Du Pont case the Patents Act 1977 became law and for the first time the principles of patent law across Europe were brought into conformity in accordance with the European Patent Convention. The case of Hallen v Brabantia5 in 1990 confirmed that the concept of selection inventions had survived the introduction of the new law. In one of the only cases outside the field of chemistry, it was held that a patent for self-pulling corkscrews coated with PTFE was not a valid selection invention as the specification had failed to disclose the advantages which were peculiar to such devices over corkscrew coated with PTFE generally6. In Ranbaxy v Warner-Lambert7 the court returned its focus to novelty even though the issue of selection inventions had been raised in relation to a lack of inventive step reference. Pumfrey J. confirmed that as a matter of English law it is possible to make a selection invention where there was a prior disclosure of a class of compounds “rather than a disclosure of the individual members of that class as distinct entities”. Pumfrey J. considered that if the later patent did not state the advantage possessed by the selected class, it is merely an arbitrary selection among things already disclosed, and therefore lacking novelty. This set the scene for the most recent UK case on selection inventions, Dr Reddy’s v Eli Lilly8 but before examining that case it is convenient to consider the EPO approach to selection inventions.

III. EPO APPROACH TO SELECTION INVENTIONS The EPO’s approval to novelty has in some ways been stricter than that of the UK courts. Unlike the approach in DuPont the discovery of a new technical effect can never add novelty to a class of compound that is otherwise old. However, as a balance to this the EPO had developed rules by which novelty can be conferred on the selection of compounds per se as discussed below. The EPO’s Examination Guidelines explain that selection inventions “deal with the selection of individual elements, sub-sets, or ranges, which have not been explicitly mentioned, from within a larger known set or range.” The Technical Board of Appeal’s decision in T12/819 is described in the fifth edition of the Case Law of the Boards of Appeal of the European Patent Office as “a decision of fundamental importance with regard to novelty in the field of chemistry”. It has long been established law that substances which are the inevitable product of a process described in a prior document cannot be novel. This case established when a substance

5 6 7 8 9

Hallen Co. and Anr. v Brabantia (UK) Ltd. [1991] R.P.C. 195. Apparently the coating unexpectedly ease extraction as well as insertion. Ranbaxy UK Ltd v Warner-Lambert Co [2006] F.S.R. 14. Dr Reddy’s Laboratories (UK) Ltd v Eli Lilly and Co Ltd [2008] EWHC 2345 (Pat). T 12/81 (BAYER/Diastereoisomers) OJ 8/1982, 296.

345

TIM POWELL

is to be regarded as being an inevitable product in circumstances where the earlier disclosure is framed in terms of classes of starting compounds and/or classes of processes. In T12/81 the prior art listed 20 starting compounds and gave a choice between 5 different processes for reducing ketones to their corresponding secondary alcohols which could take two diastereomeric forms. The patent application was for the product resulting from the use of a particular starting compound and a particular process as it had been discovered that the resulting compound comprised a very high ratio of the preferred diastereomer. The Board differentiated circumstances where a choice is left open to the skilled person in implementing a prior art teaching, highlighting that the concept of substance selection presupposes that a choice has been made from a group of substances: “If on the other hand two classes of starting substances are required to prepare the end products and examples of individual entities in each class are given in two lists of some length, then a substance resulting from the reaction of a specific pair from the two lists can nevertheless be regarded for patent purposes as a selection and hence as new.”

At first sight this may seem to be somewhat arbitrary but the Board’s reasoning is that where two starting compounds must be selected the end product is the result of two variables parameters: “The new element – indispensable if a substance selection is to be recognised as new for patent law purposes – is not attributable to the absence of a reference to the end product but to the fact that the combination actually selected from the wide range of possibilities has not been disclosed to the public.”

However, where only one starting compound must be selected (and these are all specifically disclosed) the variable process parameter does not introduce a new element. The selection of a sub-range from a broader numerical range is another important category of selection invention. In T198/8410 it was established that such inventions are novel if the selected sub-range (a) is narrow compared to the known range (b) is sufficiently far removed from specifically disclosed examples and from the end points of the known range; and (c) importantly is not an arbitrary selection. This means that it must not simply be an embodiment of the prior art but another invention i.e. it must be a purposive selection resulting from new technical teaching. The Board clarified that the sub-range singled out must be new per se, a newly discovered effect occurring only in the sub-range cannot confer novelty but can confirm that the selection is not arbitrary. The Board reiterated that the purpose of Art.54(1) of the EPC is to prevent the state of the art being patented again. This is perhaps the key to understanding the EPO approach (at least to inventions relating to claimed sub-ranges). The subject matter of the new invention is the selection of a narrower class or range of compounds. This selection must be purposive and confer an advantage over the other, non selected members of the prior disclosed class, otherwise it is not in fact a new invention at all. 10 T 198/84 (Hoechst) OJ 1985, 209.

346

SELECTION INVENTIONS IN THE PHARMACEUTICAL FIELD

IV. OLANZAPINE IN THE UK – THE DECISION OF THE HIGH COURT The principles established by the EPO were considered in Dr Reddy’s v Eli Lilly8, the latest UK case on selection inventions. Eli Lily had claimed the anti-psychotic agent, olanzapine. The prior art included a patent application for a wide class of compounds defined by reference to a general formula. This formula had four variable substituents 19 and was estimated to cover up to 10 compounds. Preferred values for the substituents narrowed the class to 86,000 compounds. Olanzapine was included within this preferred class but was not amongst the approximately 100 compounds listed in the specification by name and nor was it one of the 15 synthetic examples provided. The validity of the patent was challenged on several grounds, including an allegation that it did not fulfil the criteria for a valid selection invention. In assessing the case the judge, Floyd J., first considered novelty, referring extensively to EPO case law in the field of selection inventions. In particular, Floyd J. applied case T12/81 (discussed above) finding that there are two ways in which a novel selection can be made. Firstly, the selection may be of an “unmentioned” compound from territory marked out by the state of the art and secondly, a valid selection may be made in circumstances where, in order to arrive at the compound, a combination of starting materials has to be chosen from “two lists of some length”. He inferred that it is the need to make a choice which prevents the disclosure from “unalterably establishing” the claimed compound. In case T7/8611 Floyd J. found support for applying the principles of T12/81 to polysubstituted chemical compounds where the individual substituents have to be selected from two or more lists of some length. Floyd J. also highlighted T181/8212 in which the Board distinguished between a compound which is merely covered by a definition and one which is expressly taught or, in other words, the “purely intellectual content” of the definition and “the information content” in the sense of “a specific teaching with regard to technical action”. The Board held that a specifically mentioned compound is disclosed whereas a group of compounds in which the substituent is characterised by a range only teaches the skilled person about individuals specifically designated from the group. Following this analysis Floyd J. found that “a general formula with multiple substituents chosen from lists of some length will not normally take away the novelty of a subsequent claim to an individual compound”. He was also persuaded that “a prior disclosure does not take away the novelty of a claim to a specific compound unless the compound is disclosed in ‘individualised form’”. He arrived at the following conclusions: “i) In relation to lack of novelty, it is doubtful in the light of the EPO jurisprudence whether a newly discovered effect complying with Maugham J’s principles could overcome a finding that a compound was specifically disclosed in a prior document.

11 T 7/86 (Draco/Xanthines). 12 T 181/82 (Ciba/Spiro compounds) OJ EPO 1984, 401.

347

TIM POWELL ii) Whether or not that is so, provided there is novelty on conventional grounds, obviousness is to be decided according to ordinary principles. iii) The existence of an advantage possessed by the selected compound will be relevant to the overall assessment of obviousness, but is not an essential pre-requisite. iv) Compliance with Maugham J’s principles in IG Farbenindustrie’s Patent is equally not an essential requirement for inventive step to be found.”

Applying his conclusions, Floyd J. found olanzapine to be novel using the EPO’s approach. The earlier patent application did not make an “individualised disclosure” of olanzapine as the skilled person would have to make a number of choices of substituents from the general formula. Inventiveness was found because there was no clear direction in the earlier document to make the particular selection of olanzapine based on structure-activity considerations. In essence, this reasoning completely side-steps the requirement to investigate as part of the analysis of validity whether the narrower class is merely an arbitrary selection from the prior, wider disclosure. It is a peculiarity of this approach that the prior generic disclosure does not appear to have been considered at all in assessing whether a new invention has been made. Once the generic disclosure was dismissed for novelty purposes, given the breadth of the disclosure it formed no part of the obviousness analysis that the class of compounds including olanzapine had previously been claimed as useful for the very same purpose (as anti-psychotic drugs) as that for which olanzapine is used. It is interesting that (in case he was found to be wrong on appeal) Floyd J. also applied the Maugham J. principles in IG Farbenindustrie discussed above. He found that the patent did not meet the “classic” requirements for a selection invention over the prior generic disclosure. The patent claimed only two advantages over individual compounds specifically described in the prior art rather than any quality which was of special character over the prior disclosed class as a whole. Nor was there any assertion that olanzapine was special and that the advantages were not shared by the class as a whole.

V. OLANZAPINE IN THE UK – THE DECISION OF THE COURT OF APPEAL In December 2009, the Court of Appeal confirmed the decision of the High Court. The Court of Appeal dealt briefly with novelty, rejecting the notion that every chemical class disclosure (for example, a Markush formula) discloses each and every member of the class. Instead, the Court held that anticipation requires an “individualised description” of the later claimed compound or class of compounds and simply noted that “This case is miles from that”. The appeal focused on the arguments made in relation to obviousness. Most notably the English common law rules developed in I.G. Farbenindustrie A.G.’s Patents have been rejected as “part of legal history, not … living law” in favour of the approach of the EPO. The Court observed that no technical contribution 348

SELECTION INVENTIONS IN THE PHARMACEUTICAL FIELD

is provided by an arbitrary selection and that the EPO Boards “deploy the objection of obviousness where the patentee has in truth made no real technical advance”. Following consideration of EPO jurisprudence, the Court held that the correct question to ask is “whether the selection … was ‘arbitrary’, or whether the teaching of the Patent established that the selected compound achieved ‘a particular technical result’ and, in answering that question, one must bear in mind that it arises in the context of the broader proposition that ‘the extent of a patent monopoly should correspond to and be justified by the technical contribution to the art’”. For now, it seems “English” selection inventions have been relegated to legal history.

VI. OLANZAPINE IN GERMANY – THE DECISION OF THE GERMAN SUPREME COURT Eli Lilly’s patent on olanzapine had an eventful progress through the German court system. As is well known, the patent was revoked by the Federal Patents Court but while the appeal was pending before the German Supreme Court, the Dusseldorf Appellate Court granted preliminary injunctions to Lilly, casting severe doubts on the Federal Patent Court’s decision. The German Supreme Court subsequently overturned the decision of the Federal Patents Court and declared the patent valid. The majority of the discussion as to the validity of the olanzapine patents centred not on the prior patent disclosure discussed above but on a separate piece of prior art by Lilly scientists, Chakrabarti 1980, which also disclosed a generic formula that could encompass olanzapine. The German Supreme Court, following EPO jurisprudence, held that the prior disclosure did not disclose olanzapine in individualised form. The German Supreme Court also noted that this was consistent with the approach of Floyd J. in the parallel UK proceedings. The Supreme Court also found that olanzapine was not obvious over the disclosure in Chakrabarti, based on similar reasoning to that of Floyd J. (there were not sufficient directions in Chakrabarti to focus on olanzapine based on the disclosed structure activity relationship). The prior patent that disclosed the generic class including olanzapine was dealt with quite quickly by the Supreme Court. Following the same reasoning as above, it did not deprive the later patent to olanzapine of novelty. The closest specifically exemplified compound in the earlier patent was ethyl-olanzapine and it was held that it would not be obvious to replace the earlier ethyl substituent with a methyl substituent to form the later patented compound. In summary, the German Supreme Court appears to have taken a very similar approach to that of Floyd J. – an approach which does not appear to give any consideration to the particular issues raised by selection patents.

349

TIM POWELL

VII. CONCLUSION Although patent law in the UK has been harmonised with the European Patent Convention, the EPO’s approach to the examination of patentability differs slightly to the way in which the UK courts approach validity. The UK courts have increasingly sought to bring UK law into conformity with the approach of the EPO but are also bound by precedent. The result can some times be a “crashing of gears” as established UK principles give way to a more European approach. An example of this does appear to be the case of selection inventions. As has been seen, the law on selection inventions in the UK was well understood and consistently applied in the UK for some 80 years but is now in disarray. In fact it is doubtful whether there is any remaining application for “classic” selection inventions in the UK except perhaps in the case of numerical sub-ranges. The UK has moved towards the EPO approach to novelty of chemical compounds. A prior disclosure of a generic class encompassing a specific compound does not anticipate a later individualised disclosure of that compound. There are good policy reasons for this approach but also dangers. Is it right for a patentee to obtain a 20 year monopoly for a generic class of compounds and then obtain a further 20 year monopoly simply by picking one compound out of the general formula? It is submitted that this should only be allowed where the selection itself is a new invention, which should be defined in distinction to the wider class of compounds from which the selection was made. Assessing whether such an invention had been made would in practice involve the application of Maugham J.’s selection criteria discussed above. Perhaps the selection criteria can be resurrected in the assessment of obviousness, taking the prior generic disclosure as the relevant prior art for this purpose. There should be a requirement that the “objective technical problem” associated with the wider class of compounds is identified in the later patent specification and it should be established that the selected compounds solve this problem in contradistinction to the compounds which do not form part of the selection. In this way, the classic UK approach to selection inventions and the EPO approach to novelty and obviousness may be brought into harmony.

350